EEG Journal - February 2020 Vol. XXVI, No. I (2) by Association of Environmental & Engineering Geologists (AEG)

EDITORIAL OFFICE: Environmental & Engineering Geoscience journal, Department of Geology, Kent State University, Kent, OH 44242, U.S.A. phone: 330-672-2968, fax: 330-672-7949, ashakoor@kent.edu. CLAIMS: Claims for damaged or not received issues will be honored for 6 months from date of publication. AEG members should contact AEG, 201 East Main St., Suite 1405, Lexington, KY 40507. Phone: 844-331-7867. GSA members who are not members of AEG should contact the GSA Member Service center. All claims must be submitted in writing. POSTMASTER: Send address changes to AEG, 201 East Main St., Suite 1405, Lexington, KY 40507. Phone: 844-331-7867. Include both old and new addresses, with ZIP code. Canada agreement number PM40063731. Return undeliverable Canadian addresses to Station A P.O. Box 54, Windsor, ON N9A 6J5 Email: returnsil@imexpb.com. DISCLAIMER NOTICE: Authors alone are responsible for views expressed inarticles. Advertisers and their agencies are solely responsible for the content of all advertisements printed and also assume responsibility for any claims arising therefrom against the publisher. AEG and Environmental & Engineering Geoscience reserve the right to reject any advertising copy. SUBSCRIPTIONS: Member subscriptions: AEG members automatically receive digital access to the journal as part of their AEG membership dues. Members may order print subscriptions for $75 per year. GSA members who are not members of AEG may order for $60 per year on their annual GSA dues statement or by contacting GSA. Nonmember subscriptions are $310 and may be ordered from the subscription department of either organization. A postage differential of $10 may apply to nonmember subscribers outside the United States, Canada, and Pan America. Contact AEG at 844-331-7867; contact GSA Subscription Services, Geological Society of America, P.O. Box 9140, Boulder, CO 80301. Single copies are $75.00 each. Requests for single copies should be sent to AEG, 201 East Main St., Suite 1405, Lexington, KY 40507. © 2020 by the Association of Environmental and Engineering Geologists

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER Abdul Shakoor Department of Geology Kent State University Kent, OH 44242 330-672-2968 ashakoor@kent.edu

EDITORS

Brian G. Katz Florida Department of Environmental Protection 2600 Blair Stone Rd. Tallahassee, FL 32399 850-245-8233 eegeditorbkatz@gmail.com

Sasowsky, Ira D. University of Akron Katz, Brian G. Florida Department of Environmental Protection Shakoor, Abdul Kent State University

ASSOCIATE EDITORS Brankman, Charles, Consultant Boston MA Bruckno, Brian Virginia Department of Transportation Clague, John J. Simon Fraser University, Canada De Graff, Jerome V. California State University, Fresno Fryar, Alan University of Kentucky Hauser, Ernest Wright State University Hutchinson, Jean Queens University, Canada Keaton, Jeff AMEC Americas Marinos, Vassillis Aristotle University of Thessaloniki, Greece

McBride, John Brigham Young University Mwakanyamale, Kisa Illinois State Geological Survey Santi, Paul Colorado School of Mines Dee, Seth University of Nevada, Reno Shlemon, Roy R.J. Shlemon & Associates, Inc. Stephenson, William U.S. Geological Survey Stock, Greg National Park Service Sukop, Michael Florida International University Ulusay, Resat Hacettepe University, Turkey Watts, Chester F. “Skip,” Radford University West, Terry Purdue University

Environmental & Engineering Geoscience February 2020 VOLUME XXVI, NUMBER 1 SPECIAL ISSUE ON NATURALLY OCCURRING ASBESTOS (NOA) GUEST EDITORS: R. MARK BAILEY AND SARAH KALIKA

SUBMISSION OF MANUSCRIPTS Environmental & Engineering Geoscience (E&EG), is a quarterly journal devoted to the publication of original papers that are of potential interest to hydrogeologists, environmental and engineering geologists, and geological engineers working in site selection, feasibility studies, investigations, design or construction of civil engineering projects or in waste management, groundwater, and related environmental fields. All papers are peer reviewed. The editors invite contributions concerning all aspects of environmental and engineering geology and related disciplines. Recent abstracts can be viewed under “Archive” at the web site, “http://eeg.geoscienceworld.org”. Articles that report on research, case histories and new methods, and book reviews are welcome. Discussion papers, which are critiques of printed articles and are technical in nature, may be published with replies from the original author(s). Discussion papers and replies should be concise. To submit a manuscript go to http://eeg.allentrack.net. If you have not used the system before, follow the link at the bottom of the page that says New users should register for an account. Choose your own login and password. Further instructions will be available upon logging into the system. Please carefully read the “Instructions for Authors”. Authors do not pay any charge for color figures that are essential to the manuscript. Manuscripts of fewer than 10 pages may be published as Technical Notes. For further information, you may contact Dr. Abdul Shakoor at the editorial office.

Cover photo Fibrous tremolite/actinolite in talc tremolite schist, Franciscan Complex, Richmond, CA, USA imaged using field emission scanning electron microscopy. Photo courtesy of Bradley Erskine, Ph.D., PG of Erskine Environmental Consulting.

Volume XXVI, Number 1, February 2020

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from AEG.

EDITORIAL BOARD Watts, Chester “Skip” F. Radford University, Chair Hasan, Syed University of Missouri, Kansas City Nandi, Arpita East Tennessee State University Oommen, Thomas Michigan Technological University

ENVIRONMENTAL & ENGINEERING GEOSCIENCE

Environmental & Engineering Geoscience (ISSN 1078-7275) is published quarterly by the Association of Environmental & Engineering Geologists (AEG) and the Geological Society of America (GSA). Periodicals postage paid at AEG, 201 East Main St., Suite 1405, Lexington, KY 40507 and additional mailing offices.

THE JOINT PUBLICATION OF THE ASSOCIATION OF ENVIRONMENTAL AND ENGINEERING GEOLOGISTS AND THE GEOLOGICAL SOCIETY OF AMERICA SERVING PROFESSIONALS IN ENGINEERING GEOLOGY, ENVIRONMENTAL GEOLOGY, AND HYDROGEOLOGY

EDITORIAL BOARD Watts, Chester “Skip” F. Radford University, Chair Hasan, Syed University of Missouri, Kansas City Nandi, Arpita East Tennessee State University

Oommen, Thomas Michigan Technological University Sasowsky, Ira D. University of Akron

SUBSCRIPTIONS:

Environmental & Engineering Geoscience February 2020 VOLUME XXVI, NUMBER 1 SPECIAL ISSUE ON NATURALLY OCCURRING ASBESTOS (NOA) GUEST EDITORS: R. MARK BAILEY AND SARAH KALIKA

SUBMISSION OF MANUSCRIPTS

Member subscriptions: AEG members automatically receive digital access to the journal as part of their AEG membership dues. Members may order print subscriptions for $75 per year. GSA members who are not members of AEG may order for $60 per year on their annual GSA dues statement or by contacting GSA.

Environmental & Engineering Geoscience (E&EG), is a quarterly journal devoted to the publication of original papers that are of potential interest to hydrogeologists, environmental and engineering geologists, and geological engineers working in site selection, feasibility studies, investigations, design or construction of civil engineering projects or in waste management, groundwater, and related environmental fields. All papers are peer reviewed.

Nonmember subscriptions are $310 and may be ordered from the subscription department of either organization. A postage differential of $10 may apply to nonmember subscribers outside the United States, Canada, and Pan America. Contact AEG at 844-331-7867; contact GSA Subscription Services, Geological Society of America, P.O. Box 9140, Boulder, CO 80301.

The editors invite contributions concerning all aspects of environmental and engineering geology and related disciplines. Recent abstracts can be viewed under “Archive” at the web site, “http://eeg.geoscienceworld.org”. Articles that report on research, case histories and new methods, and book reviews are welcome. Discussion papers, which are critiques of printed articles and are technical in nature, may be published with replies from the original author(s). Discussion papers and replies should be concise.

Single copies are $75.00 each. Requests for single copies should be sent to AEG, 201 East Main St., Suite 1405, Lexington, KY 40507.

To submit a manuscript go to https://www.editorialmanager.com/EEG/ default.aspx. If you have not used the system before, follow the link at the bottom of the page that says New users should register for an account. Choose your own login and password. Further instructions will be available upon logging into the system. Please carefully read the “Instructions for Authors”.

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER EDITORS

Brian G. Katz Environmental Consultant Tallahassee, FL 32309 eegeditorbkatz@gmail.com

Authors do not pay any charge for color figures that are essential to the manuscript. Manuscripts of fewer than 10 pages may be published as Technical Notes. For further information, you may contact Dr. Abdul Shakoor at the editorial office.

Volume XXVI, Number 1, February 2020

Abdul Shakoor Department of Geology Kent State University Kent, OH 44242 330-672-2968 ashakoor@kent.edu

ENVIRONMENTAL & ENGINEERING GEOSCIENCE

Environmental & Engineering Geoscience Volume 26, Number 1, February 2020 Table of Contents 1

Foreword to the Environmental & Engineering Geoscience Special Edition on Naturally Occurring Asbestos R. Mark Bailey and Sarah Kalika

Naturally Occurring Asbestos: A Global Health Concern? State of the Art and Open Issues Alessandro F. Gualtieri

Overview of Naturally Occurring Asbestos in California and Southwestern Nevada R. Mark Bailey

Clastic Sedimentary Rocks and Sedimentary Mélanges: Potential Naturally Occurring Asbestos Occurrences (Amphibole and Serpentine) John Wakabayashi

Asbestiform Minerals of the Franciscan Assemblage in California with a Focus on the Calaveras Dam Replacement Project R. Mark Bailey

Does Exposure to Naturally Occurring Asbestos (NOA) During Dam Construction Increase Mesothelioma Risk? Daniel W. Hernandez

NOA Air-Quality Lessons Learned during Calaveras Dam Replacement Project Bart Eklund, John Roadifer, Noel Wong, and Michael Forrest

Naturally Occurring Asbestiform Minerals in Italian Western Alps and in Other Italian Sites Elena Belluso, Alain Baronnet, and Silvana Capella

Naturally Occurring Asbestos in Valmalenco (Central Alps, Northern Italy): From Quarries and Mines to Stream Sediments Alessandro Cavallo and Jasmine Rita Petriglieri

Naturally Occurring Asbestos in France: Geological Mapping, Mineral Characterization, and Technical Developments Florence Cagnard and Didier Lahondère

Naturally Occurring Asbestos in France: a Technical and Regulatory Review Erell Léocat

Regulations Concerning Naturally Occurring Asbestos (NOA) in Germany—Testing Procedures for Asbestos Stefan Pierdzig

Fibrous Tremolite in Central New South Wales, Australia Marc Hendrickx

Management of Naturally Occurring Asbestos Area in Republic of Korea Sungjun Yoon, Kyubong Yeom, Yongun Kim, Byungno Park, Jaebong Park, Hyesu Kim, Hyeonyi Jeong, and Yul Roh

Identification and Preliminary Toxicological Assessment of a Non-Regulated Mineral Fiber: Fibrous Antigorite from New Caledonia Jasmine Rita Petriglieri, Christine LaPorte-Magoni, Emma Salvioli-Mariani, Maura Tomatis, Elena Gazzano, Francesco Turci, Alessandro Cavallo, and Bice Fubini

Geologic Investigations for Compliance with the CARB Asbestos ATCM Bradley G. Erskine

107

Geological Model for Naturally Occurring Asbestos Content Prediction in the Rock Excavation of a Long Tunnel (Gronda di Genova Project, NW Italy) Luca Barale, Fabrizio Piana, Sergio Tallone, Roberto Compagnoni, Chiara Avataneo, Serena Botta, Igor Marcelli, Andrea Irace, Pietro Mosca, Roberto Cossio, and Francesco Turci

113

New Tools for the Evaluation of Asbestos-Related Risk during Excavation in an NOA-Rich Geological Setting Francesco Turci, Chiara Avataneo, Serena Botta, Igor Marcelli, Luca Barale, Maura Tomatis, Roberto Cossio, Sergio Tallone, Fabrizio Piana, and Roberto Compagnoni

121

Sampling, Analysis, and Risk Assessment for Asbestos and Other Mineral Fibers in Soil Ed Cahill

129

Refinement of Sampling and Analysis Techniques for Asbestos in Soil Julie Wroble, Tim Frederick, and Daniel Vallero

133

Discerning Erionite from Other Zeolite Minerals during Analysis Robyn Ray

Foreword to the Environmental & Engineering Geoscience Special Edition on Naturally Occurring Asbestos R. MARK BAILEY Asbestos TEM Laboratories, 600 Bancroft Way, Suite A, Berkeley, CA 94710

SARAH KALIKA Cornerstone Earth Group, 1220 Oakland Boulevard, Suite 220, Walnut Creek, CA 94596

This special issue of Environmental and Engineering Geoscience (E&EG) is an outgrowth of the Naturally Occurring Asbestos (NOA) Symposium held as part of the combined XIII Congress of the International Association for Engineering Geology and the Environment (IAEG) and the Annual Meeting of the Association of Environmental and Engineering Geologists (AEG) in September 2018 in San Francisco, CA. Thirty-three oral presentations, three posters, an international panel discussion, and a field course were presented over the 3 days of the symposium. At the conclusion of the symposium, it was proposed to memorialize the findings and state-of-the-art practices described by the presenters by compiling many of the presentations into papers to be published in a special journal issue; this is the result. Additionally, several NOA Symposium presenters, along with the cochairs of AEG’s NOA Technical Working Group, submitted a proposal to IAEG’s Executive Committee to form an IAEG Commission on NOA, which was approved on September 22, 2019. This special edition of E&EG is the first product of IAEG Commission #39 on NOA with all commission members writing articles for and/or reviewing the included papers. Naturally occurring asbestos (NOA) is found in a wide variety of geologic rock types and environments, almost always as the result of metamorphism of pre-existing rocks. When mobilized into the air during ground-disturbing activities, NOA fibers may cause lung cancer, mesothelioma (cancer of the tissue that surrounds the lungs), and other types of lung disease, representing a potentially serious hazard to workers and the general public. The presence and mitigation of NOA can also have a significant impact upon infrastructure projects such as dams, highways, and tunnels, where heavy construction inevitably generates large quantities of dust, requiring extra measures to control. NOA-containing rocks are found around the world, though most often concentrated in areas that have experienced some degree of moun-

tain building where plate-tectonic collisions have occurred. Unlike industrial and building materials, where asbestos from a limited group of minerals often called “regulated asbestos” (serpentine chrysotile and the asbestiform amphibole varieties of riebeckite, grunerite, anthophyllite, tremolite, and actinolite with narrow chemical compositional ranges) was intentionally added, the range of asbestiform mineral-bearing rocks is far larger and includes the “non-regulated” varieties of the fibrous amphiboles winchite, richterite, glaucophane, ferro-glaucophane, magnesio-riebeckite, arfvedsonite, fluoro-edenite, and others that are still being discovered. Notably, erionite, a fibrous zeolite not considered to be structurally similar to chrysotile or amphiboles, has been implicated in causing mesothelioma in certain populations, and is included in discussions of NOA. This wide range of non-standard occurrences of asbestiform minerals often makes assessment, control, and abatement exceptionally difficult and expensive. The relatively new field of concern regarding NOA has experienced rapid changes in understanding. Just 8 years ago, at the Calaveras Dam Replacement Project (CDRP) in Fremont, CA, a new type of asbestos mineral, fibrous glaucophane (a close relative of riebeckite and sometimes considered to be a form of crocidolite) was identified for the first time. Fibrous glaucophane was present as a large-scale occurrence and had a major impact on the construction of the new dam. At the IAEG Congress/AEG Annual Meeting in 2018, AEG honored Kleinfelder and the San Francisco Public Utilities Commission (SFPUC) with the Outstanding Environmental and Engineering Geology Award for their work at the CDRP site, which acknowledged how the project dealt with the (at times overwhelming) issue of identification and mitigation of NOA in rock, soil, and dust. Since the identification of fibrous glaucophane at CDRP, several other similar occurrences have been identified in California, and it is reasonable to expect additional occurrences around the world will be identified in time.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 1–2

Bailey and Kalika

While this publication documents NOA occurrences around the world (in this journal represented by Italy, Germany, France, South Korea, New Caledonia, Australia, and United States), some with similar and others with very diﬀerent, geologic rock types and metamorphic histories, we have also included papers that discuss regulations concerning the protection of workers and the public from exposure to NOA in the various countries that are home to the authors, and information on innovative techniques to control NOA emissions. It also should be noted that two additional papers on NOA from Argentina were published separately in the XIII IAEG Congress Proceedings. Taking on the role of guest editors for this special edition of EG&G was a challenging learning process, and new to us, but our goal of combining the broad range of information presented at the IAEG/AEG

NOA Symposium into a special publication focused on the relatively new field of NOA, in a manner that had not been done before, drove us forward. We were not alone in this endeavor. We especially appreciate the guidance of EG&G co-editor Abdul Shakoor, assistance by editorial office manager Karen Smith, help from members of the IAEG Commission on NOA, and all the authors who worked tirelessly to create this special edition. Furthermore, it should be noted that peer reviewing is an arduous and often thankless task that is not officially credited with any type of direct citation, but it is critical to the creation of robust, highquality articles. The efforts of the peer reviewers for this special edition are greatly appreciated. It is hoped that this special edition of E&EG will stand as an important milestone in the development of thought and practice concerning the field of NOA.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 1–2

Naturally Occurring Asbestos: A Global Health Concern? State of the Art and Open Issues ALESSANDRO F. GUALTIERI* Department of Chemical and Geological Sciences, University of Modena and Reggio Emilia, Via G. Campi 103, 41125, Modena, Italy

Key Terms: Asbestos, Environmental Geology, Hazardous Waste, Mining, NOA, Tunnels ABSTRACT Naturally occurring asbestos (NOA) is a global public health issue because minerals that may be classified as asbestos are a common constituent of certain types of rock and soil, found in many regions on every continent. Disturbance of these rocks and soils, especially through construction activities, can result in airborne particles, leading to inhalation and risk of disease from these known human carcinogens. The presence of NOA in the environment affects all the human activities aimed at its modification, and all engineering/geological actions in the natural environment should take it into account. In the presence of NOA, specific procedures for sampling, evaluation of environmental risk, and monitoring should be applied to minimize the risk of exposure for the workers and the general public. Unfortunately, detailed procedures have been lacking to date, and consensus is difficult to achieve because basic issues, such as the definition of asbestos itself, are still open and being debated by scientists and regulators. While the term “asbestos” has been used in older geological publications, it is not currently defined by geologists. For the past century, “asbestos” was a commercial term used to describe minerals mined for specific purposes, and the term then entered the legal lexicon for purposes of control and compensation. All these basic matters are critically illustrated in the article. Finding clear and universally accepted definitions is mandatory; otherwise, there will continue to be controversial positions that can cause regulatory and legal issues and the outcome of lawsuits to be very subjective.

INTRODUCTION Naturally occurring asbestos (NOA) is a general term describing the six commercially classified forms of mineral fibers that are classified as asbestos when *Corresponding author email: alessandro.gualtieri@unimore.it

found in their natural state (Lee et al., 2008). The definition of NOA pertains to the asbestos fibers that occur in rock and soil as result of natural geological processes (Harper, 2008) and not the asbestoscontaining manufactured products that can be eventually found in the environment. The term commonly applies to areas where asbestos minerals are found in such low quantities that mining and commercial exploitation are not viable. While large commercial deposits of asbestos minerals are relatively rare, small non-economic occurrences of asbestiform (defined as having at least one characteristic of asbestos, including being composed of bundles with frayed or splayed ends, being flexible, and having great tensile strength) minerals are more common (Lee et al., 2008). Strictly, the term NOA applies to the natural geologic occurrence of any of the six regulated types of asbestos minerals, so it should not include other mineral fibers, but it has been used in this more inclusive way. Natural weathering and/or human activities may disturb NOA-bearing rocks or soils and release asbestos into the air, inducing potential human exposure by inhalation. Figure 1 shows a typical NOA site represented by an abandoned serpentine quarry in Romanoro, Northern Appennines (province of Modena, Italy). Although this quarry was not operated for asbestos production, asbestiform minerals have been identified in the rock. The rock piles and quarry faces from the former exploitation activity are persistently weathered by the action of atmospheric agents and percolating waters with a potential for the release of fibers to the air, although the extent of such release has not been quantified by air monitoring. Anthropogenic activities, such as driving, cycling, hiking, and geological investigations in areas where asbestos minerals occur, can result in airborne fibers, although the exposure is likely to be very low unless the area is heavily contaminated as, for example, in the areas immediately adjacent to the old asbestos mines and tailings dumps in the Clear Creek Management Area of California. NOA has raised concern worldwide with the appearance of scientific evidence of increased risk of malignant mesothelioma (MM) in the population

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 3–8

Gualtieri

exposed to airborne asbestos released from NOA sites. An example is a California study where residential proximity to naturally occurring asbestos was associated with increased risk of MM (Pan et al., 2005), although the methodology in this study is easily criticized, for example, by noting that it took no account of the heavy rates of immigration and emigration in California. The risk of MM decreased approximately 6.3% for every 10-km increase in residential distance from the nearest asbestos source (LaDou et al., 2010). However, studies of residents of Whatcom County, Washington, where gravels containing chrysotile have been extensively used by the community and where chrysotile is routinely deposited on residential property through river flooding, have shown no increase in morbidity despite the chrysotile presenting a much higher level of hazard than other types of asbestos in toxicological studies (Vander Kelen and Patrick, 2013). The greatest risk of exposure to asbestos-like minerals is in construction, and in some areas where these minerals are common, such as Fairfax County, Virginia, and in the state of California, specific procedures and regulations have been developed for construction activities (Lee et al., 2008).The presence of NOA in the environment needs to be considered in all the human activities aimed at its modification (e.g., engineering and geological activities, such as mining; tunnel, bridge, and dam construction; and road and highway pavement). In the presence of NOA, dedicated procedures for sampling, evaluation of environmental risk, and monitoring should be applied to reduce the risk of exposure for the workers and the general public. Regrettably,

Figure 1. A typical example of NOA site represented by an abandoned serpentine quarry, speciﬁcally the serpentine rock quarry of Romanoro in the northern Appennines (province of Modena, Italy). Rock piles and quarry faces from the former exploitation activity are persistently leached by the action of atmospheric agents (namely, wind and rain) and percolating waters, releasing chrysotile- and tremolite-rich airborne particulate. Anthropogenic actions may also cause disturbance (original picture A. F. Gualtieri).

shared unambiguous procedures are lacking to date because there is no international platform for addressing this issue. Limits of the existing definitions include (1) the mineral fibers to be classified and regulated as asbestos, (2) correct measurement of mineral fibers to be properly classified as asbestos, (3) the determination of the concentration of asbestos and mineral fibers in massive materials, (4) if elongate mineral particles other than those already classified as asbestos represent a hazard, and (5) the toxicity/pathogenicity potential of mineral fibers other than those classified as asbestos. As discussed in the core of this article, if we do not attempt to solve these basic issues, there will always be controversial positions, legal issues, and subjective outcomes of worldwide lawsuits, ultimately delaying the protections necessary to preserve human health. THE PROBLEM OF THE DEFINITION OF ASBESTOS As remarked by Harper (2008), the risk to health when living on soil and rock in the presence of NOA is not so obvious. The picture becomes even less clear when the minerals are subject to intensive investigation since our generally accepted definitions of asbestos are themselves put to the test. The discovery of asbestos or related minerals has consequences beyond any immediate risks to health, including profound effects on the value of and ability to use or enjoy property. To date, the literature reports at least 15 different definitions of asbestos (Gualtieri, 2017), and despite the fact that classification of asbestos minerals has been studied for about 40 years, there is still controversy as to which mineral particles should be classified as asbestos (Gualtieri et al., 2018b). Regrettably, the considerations made to date do not correspond to the common understanding of what asbestos is, and, to some degree, these two approaches sometimes contradict one another (Case et al., 2011). One of the early regulatory definitions of asbestos is the following: “Six fibrous silicates are classified as ‘asbestos’: the fibrous serpentine called chrysotile and the fibrous amphiboles actinolite, amosite, anthophyllite, crocidolite and tremolite” (U.S. Department of Labor, 1975). Since the 1970s, many other definitions have been proposed, but invariably these were ambiguous or had contradictory aspects. The National Institute for Occupational Safety and Health (NIOSH, 2011) has reviewed numerous resources and has not found any current reference for standard terminology and definitions that is complete and unambiguous. For example, in 1980, NIOSH and the Occupational Safety and Health Administration (OSHA) used the following definition for

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 3–8

Naturally Occurring Asbestos

workplace exposure to airborne asbestos in the United States (review and recommendations 81,103): “1. Asbestos is defined to be chrysotile, crocidolite, and fibrous cummingtonite-grunerite including amosite, fibrous tremolite, fibrous actinolite, and fibrous anthophyllite. 2. The ‘fibrosity’ of the above minerals is ascertained on a microscopic level with fibers defined to be particles with an aspect ratio ࣙ3:1.” While this definition may be appropriate for airborne asbestos particles, it is hardly applicable to bulk rocks and minerals since asbestos was prized and mined not for being composed of microscopic fibers but rather for fibers that were centimeters or even meters long that, for example, could be woven into cloth. The major stumbling block in the appreciation of asbestos has been the substantial disconnect between the macroscopic material that occurs in the earth, which was mined and used commercially, and the microscopic materials generated from it, which can lead to disease. This definition is a first attempt to give a definition of “microscopic fiber” but is too generic. There are no constraints on the length and width (or diameter) of the particles except that length and width are practically constrained by the methodology of sampling airborne particulates, which does not collect particles above a certain length or width. Particles with an aspect ratio (length/width, by geometric definition) >3 can display a non-fibrous habit (bladed, lamellar, columnar, or prismatic), but the definition is to be applied in asbestos mines and product factories, where the vast majority of particles with an aspect ratio >3 will be asbestos. Nevertheless, the OSHA analytical method includes procedures (albeit subjective) for the exclusion of non-fibrous particles, as these are not included in OSHA regulations. The definition of asbestos has been strongly linked to the concept of “fiber” on a microscopic level. Inadequate and incomplete definitions of asbestos have resulted, as noted by an International Agency for Research on Cancer (IARC) consensus panel, in “taxonomic confusion and lack of standardized operating definitions for fibers.” The term “asbestos” is often inappropriately used as a generic, homogeneous rubric, and even when an asbestos fiber type is specified, its source is rarely stated (Kane et al., 1996). Ambiguity in the definition of asbestos may lead to widespread confusion (Lee et al., 2008) and difficulties in properly handling activities in the presence of NOA. It is likely that no scientist would disagree that if we could turn back time and eliminate the word “asbestos,” we would. “Asbestos” has been used, without a good definition and not always in the same way, in commerce, regulation and law, and the health-related professions to such an extent that we are “stuck with it.” It is probably not now definable in any scientifically defensible way, and any further attempts

to refine such a definition would then be inevitably doomed. THE PROBLEM OF THE DEFINITION OF “ASBESTOS FIBER” Any definition of “asbestos” could not be separated from a definition of “fiber.” Being fibrous or not makes a difference (e.g., chrysotile, the fibrous variety of serpentine, is classified as asbestos, whereas lizardite, the lamellar variety of serpentine, is not classified as asbestos). Unfortunately, as with asbestos, there is no clear agreement on what we call a “fiber” (Belluso et al., 2017) other than it must be an object with elongation in one dimension compared to the other two. Although pictures of objects have been drawn in an attempt to guide others in the selection of terms such as “fibrous,” “acicular,” or “prismatic,” it remains that none of these terms has a good consensus definition. Because fiber is the issue of concern for the healthrelated professions (medical, epidemiological, and toxicological) and thus, in turn, for the regulatory and legal professions, a definition could be considered critical (Belluso et al., 2017). One may be tempted to use the World Health Organization (1997) guidelines for the determination of airborne fiber number concentrations: “Airborne fibers are objects with a length L >5µm, a width W <3µm, and L:W (aspect ratio) >3:1, using a phase-contrast optical microscope (PCOM).” However, this is not a definition but simply counting criteria, a protocol to count inhalable particles able to penetrate down to the alveolar space using PCOM. It is considered an index of exposure to airborne asbestos particles, and the value lies in its traceability to a body of measurements that could be compared to the prevalence of disease in workers whose exposures were measured in accordance with this procedure, thus allowing a dose–response relationship for use in risk assessment. Such parameters come from the assumption that the fiber length (L), width (W), and biopersistence (B) are the bearing walls of the fiber toxicity paradigm: long, thin and biopersistent fibers have the highest potential to induce adverse effects in vivo (Figure 2) and prompt carcinogenesis. However, others have proposed different dimensional definitions (e.g., >20:1; Chatfield, 2008). Even in regulations, we find differences. For example, >3:1 is considered the proper aspect ratio definition in some areas of jurisdiction but >5:1 in others, and while the majority count particles longer than 5 µm, others use 0.5 µm. Different definitions are sometimes for different purposes, for example, one definition to determine presence but another to determine risk. The issue is further confounded when attempting to apply definitions of fiber to bulk materials and their

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 3–8

Gualtieri

Figure 2. The three parameters defining the fiber toxicity paradigm: fiber length (L), fiber width (W), and biopersistence (B).

likely ability to release microscopic fibers. Microscopic fibers may be revealed by direct investigation of macroscopic materials, but far more microscopic fibers may be released by the handling of these materials, and the degree to which they will be released is a function of the type of handling and the energy involved. No proposals for a standardized procedure to determine the propensity of a bulk material to release microscopic fibers to the air have been agreed. Thus, it remains impossible to compare across studies and properly evaluate the relative risk of materials and processes. THE PROBLEM OF THE MEASUREMENT OF THE AIRBORNE CONCENTRATION OF ASBESTOS Because no unequivocal definition of asbestos exists, no unequivocal analytical procedures for identification and measurement of airborne concentration exist. All the available methods are the result of approximations and compromises and are invariably affected by some degree of bias and uncertainty. As noted above, the procedures using optical microscopy were developed for use in the asbestos industry (and were later expanded to include remediation of asbestos in buildings and polluted areas) but were never designed for the measurement of asbestos in ambient air, where there may be substantial presence of non-asbestos particles, both mineral and of biological origin, to interfere with the observation. Methods involving transmission electron microscopy and scanning electron microscopy have been developed to improve the situation, but these methods lack the underpinning of a dose–response relationship where dose has been assessed from either technique, and thus such measurements cannot be directly applied to risk assessment. Since such a relationship has been determined for the optical microscopy method, some measure of equiv-

alence to the optical microscopy measurements is assumed for the purpose. When these methods developed to the determination of airborne asbestos fibers are applied to the study of massive matrices, the situation is even more challenging. An open issue is the definition of the detection limits. For the Italian law (D.M. 471/1999), the asbestos limit concentration must not exceed 1 g of fibers for every kilogram of dry soil (0.1 wt%). However, the accurate and precise quantification of asbestos at or below 1 wt% in a real ground sample is an open analytical challenge (Foresti et al., 2003). Furthermore, the analysis of NOA in massive samples is complicated by the fact that the host matrix often shares the very same elemental composition and/or crystallographic features with asbestos. Electron microscopies, capable of discriminating the non-fibrous matrix from asbestos in natural samples, are preferred experimental methods with transmission electron microscopy that offers the highest resolution and the most accurate fiber identification but requires long analysis time for a statistically relevant quantitative analysis (Cossio et al., 2018). New analytical procedures appear in the literature to overcome the intrinsic limitations of these experimental methods. An example is described in Cossio et al. (2018), who developed a fully automatic quantification of asbestos fibers by scanning electron microscopy and energy dispersive X-ray spectroscopy to quantify asbestos in complex and confounding matrixes, including amphibole- and serpentine-rich ophiolites. THE PROBLEM OF THE ELONGATE MINERAL PARTICLE Because of the complicated issue of the definition and measurement of “asbestos,” “asbestiform,” “fiber,” and related terms, some years ago, a new term, elongate mineral particle (EMP), was coined for use in the NIOSH (2011) “roadmap” for the purpose of classifying all fiber-like mineral particles under a single term. Gunter (2009) noted issues around the use of the term “NOA,” where a constant component of the rocks and soils at NOA sites are elongate mineral particles unlikely to be classified as asbestos together with elongate mineral particles that would likely be classified as asbestos. By definition, any mineral particle with a minimum aspect ratio of 3:1 is considered an EMP (Belluso et al., 2017). A special case of EMP is the “elongate cleavage fragment” (ECF): “any particle with the same chemical formula of the asbestiform variety is considered an ECF if it is originated from a crystal that cleaves into fragments rather than separates longitudinally into fibrils” (Gunter et al., 2007). It should be remarked that there is widespread distribution of amphiboles in the

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 3–8

Naturally Occurring Asbestos

natural environment that can also occur with a 3:1 aspect ratio or greater when crushed. Thus, if the term EMP is to be used, it will require a more thorough description than a 3:1 aspect ratio, as this will make most natural mineral-containing materials (i.e., soils and rocks) “contaminated” with EMPs (Gunter, 2018). There are then two issues regarding ECFs. First, they must be broken from the original particle, and this is less likely to have occurred in situ, so that they will be more likely present in aerosol derived from disturbing the bulk than in the bulk. Second, to be separated for the purpose of risk assessment, they should present little or no hazard. While OSHA and some other national regulatory bodies do not regulate cleavage fragments, both NIOSH and the Environmental Protection Agency found insufficient evidence for a sufficiently low risk of disease, so that inclusion with asbestos is their current position of prudence. The discussion at the recent Monticello Conference on EMP, held in Charlottesville, Virginia, in 2017 (Weill, 2018), pointed out that lack of precision with terminology and clear definition of EMPs have impeded the scientific community’s ability to properly characterize health effects attributable to specific types of EMPs. There was broad agreement with the original usage by NIOSH that the term “EMP” does not speak specifically about the human health effects of the substances categorized under this umbrella term (Weill, 2018). Moreover, it was highlighted that the utilization of the aspect ratio alone should be avoided in EMP characterization and, again, that the term EMP does not alone imply elevated risk for an asbestos-related disease. A good example of this is the zeolite erionite. While exposure to airborne erionite is associated with MM, especially within a specific Turkish population, the same exposures, unlike with asbestos, have not also resulted in elevated rates of lung cancer. However, such differences also occur in minerals traditionally classified as asbestos; the MM rate for workers exposed only to chrysotile, a serpentine mineral, is undeniably much lower than for cohorts exposed to chrysotile and amphibole minerals or to amphiboles alone. The rates are in fact so low that some researchers were led to declare that MM resulted from only amphibole exposure (the “amphibole hypothesis”). Hence, while there may be a belief in some quarters that EMP is an overly inclusive term, the same is also true for asbestos. THE PROBLEM OF THE NATURALLY OCCURRING MINERAL FIBERS (NOMF) OTHER THAN ASBESTOS There are at least 400 known mineral species that display a fibrous (or fibrous-asbestiform) crystal habit. Besides the six asbestos minerals, there are few reg-

ulated fibers classified as carcinogenic by the IARC, namely, erionite and fluoro-edenite. In those cases, the general term NOA is no longer correct and should be modified according to the specific case (e.g., NOE for naturally occurring erionite). The rest of the mineral fibers are not regulated or classified in terms of potential toxicity/pathogenicity (Gualtieri 2017, 2018). There are many examples of natural occurrence of unclassified mineral fibers that may represent a potential hazard in Europe and the United States. Potential public exposure to respirable mineral dust other than asbestos from natural deposits is an issue of great concern in the United States (Swayze et al., 2009; California Air Resource Board, 2017). In California, NOMF (mostly NOA) occurs in 90% of the 58 counties. A remarkable example of unregulated potentially toxic/pathogenic fiber found in the blueschists of the Franciscan Complex (California) is fibrous glaucophane (Erskine and Bailey, 2018). The Franciscan Complex rocks are commonly excavated locally for building/construction purposes in northern and central California (e.g., the Calaveras Dam Replacement Project as well as many developmental projects in the city and county of San Francisco), and the dust generated by the excavation activities may potentially expose construction workers and the nearby populations to adverse health risks. Another example is represented by the fibrous zeolite ferrierite that occurs in southern Nevada. A recent study showed that fibrous ferrierite has the same chemical-physical properties that are deemed to prompt adverse effects in vivo by erionite (Gualtieri et al., 2018a). However, no restriction has been considered against it because neither toxicity/carcinogenicity nor epidemiological studies have been conducted to date. CONCLUSIONS This article describes the major problems encountered by the engineering/geological community when facing activities in the presence of NOA. Major problems are related to the lack of unique definitions and procedures for the identification and analysis of asbestos-containing materials worldwide. When dealing with asbestos-related problems, ambiguity in the definitions, health effects of asbestos and non-asbestos species, and experimental methods may lead to widespread confusion and miscommunication, which in turn can lead to unnecessary media hysteria and public alarm (Lee et al., 2008). Ambiguity in the regulatory definitions may also prompt twisted science and legal disputes. In perspective, the global academic and engineering/geological community should work in synergy by

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 3–8

Gualtieri

sharing experiences and overcoming ambiguities to achieve common and safe procedures to work in the presence of NOA. Specific solutions to problems such as mitigation of NOA during excavation, tunneling, and mining activity are necessary. To this aim, the continued efforts of dedicated working groups and commissions is highly recommended. ACKNOWLEDGMENTS The author wishes to thank the three competent referees for the work and especially Dr. M. Harper, who was of great help in improving the quality of the final version of this article. REFERENCES Belluso, E.; Cavallo, A.; and Halterman, D., 2017, Crystal habit of mineral fibres. In Mineral Fibres: Crystal Chemistry, Chemical-Physical Properties, Biological Interaction and Toxicity European Mineralogical Union-EMU Notes in Mineralogy, Vol. 18, pp. 65–109. California Air Resource Board, 2017, Implementation Guidance Document, Air Resource Board Test Method. Determination of Asbestos Content of Serpentine Aggregate: Field and Laboratory Practices ARB Monitoring and Laboratory Division, Quality Management Branch, Quality Management Section. 435 p. Case, B. W.; Abraham, J. L.; Meeker, G. D. P. F.; Pooley, F. D.; and Pinkerton, K. E., 2011, Applying definitions of “asbestos” to environmental and “low-dose” exposure levels and health effects, particularly malignant mesothelioma: Journal of Toxicology and Environmental Health, Part B, Vol. 14, No. 1–4, pp. 3–39. Chatfield, E. J., 2008, A procedure for quantitative description of fibrosity in amphibole minerals. In 2008 Johnson Conference: Critical Issues in Monitoring Asbestos: ASTM International, Burlington, Vermont. pp. 53–54. Cossio, R.; Albonico, C.; Zanella, A.; Fraterrigo-Garofalo, S.; Avataneo, C.; Compagnoni, R.; and Turci, F., 2018, Innovative unattended SEM-EDS analysis for asbestos fiber quantification: Talanta, Vol. 190, pp. 158–166. Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in the Franciscan Complex blueschist, northern Diablo Range, California: Toxicology and Applied Pharmacology, Vol. 361, pp. 3–13. Foresti, E.; Gazzano, M.; Gualtieri, A. F.; Lesci, I. G.; Lunelli, B.; Pecchini, G.; Renna, E.; and Roveri, N., 2003, Determination of low levels of free fibres of chrysotile in contaminated soils by X-ray diffraction and FTIR spectroscopy: Analytical and Bioanalytical Chemistry, Vol. 376, No. 5, pp. 653–658. Gualtieri, A. F., 2017, Mineral Fibres: Crystal Chemistry, Chemical-Physical Properties, Biological Interaction and Toxicity. EMU Notes in Mineralogy 18: European Mineralogical Union, Jena, Germany. 536 p. Gualtieri, A. F., 2018, Towards a quantitative model to predict the toxicity/pathogenicity potential of mineral fibers: Toxicology and Applied Pharmacology, Vol. 361, pp. 89–98. Gualtieri, A. F.; Gandolfi, N. B.; Passaglia, E.; Pollastri, S.; Mattioli, M.; Giordani, M.; Ottaviani, M. F.; Cangiotti, M.; Bloise, A.; Barca, D.; Vigliaturo, R.; Viani, A.;

Pasquali, L.; and Lassinantti Gualtieri, M., 2018a, Is fibrous ferrierite a potential health hazard? Characterization and comparison with fibrous erionite: American Mineralogist, Vol. 103, No. 7, pp. 1044–1055. Gualtieri, A. F.; Gandolfi, N. B.; Pollastri, S.; Rinaldi, R.; Sala, O.; Martinelli, G.; Bacci, T.; Paoli, F.; Viani, A.; and Vigliaturo, R., 2018b, Assessment of the potential hazard represented by natural raw materials containing mineral fibres—The case of the feldspar from Orani, Sardinia (Italy): Journal of Hazardous Materials, Vol. 350, 76–87. Gunter, M. E.; Belluso, E.; and Mottana, A., 2007, Amphiboles: Environmental and health concerns: Reviews in Mineralogy and Geochemistry, Vol. 67, No. 1, pp. 453–516. Gunter, M., 2009, Asbestos sans mineralogy: Elements, Vol. 5, No. 3, p. 141. Gunter, M. E., 2018, Elongate mineral particles in the natural environment: Toxicology and Applied Pharmacology, Vol. 361, pp. 157–164. Harper, M., 2008, 10th anniversary critical review: Naturally occurring asbestos: Journal of Environmental Monitoring, Vol. 10, No. 12, pp. 1394–1408. Kane, A. B.; Boffetta, P.; Saracci, R.; and Wilbourn, J. D., 1996, Mechanisms of Fibre Carcinogenesis: International Agency for Research on Cancer Scientific Publication 140. LaDou, J.; Castleman, B.; Frank, A.; Gochfeld, M.; Greenberg, M.; Huff, T. K. J.; Landrigan, P. J.; Lemen, R.; Myers, J.; Soffritti, M.; Soskolne, C. L.; Takahashi, K.; Teitelbaum, D.; Terracini, B.; and Watterson, A., 2010, The case for a global ban on asbestos: Environmental Health Perspectives, Vol. 118, No. 7, pp. 897–901. Lee, R. J.; Strohmeier, B. R.; Bunker, K. L.; and Van Orden, D. R., 2008, Naturally occurring asbestos—A recurring public policy challenge: Journal of Hazardous Materials, Vol. 153, No. 1–2, pp. 1–21. National Institute for Occupational Safety and Health, 2011, Asbestos Fibers and Other Elongate Mineral Particles: State of the Science and Roadmap for Research: Department of Health and Human Services, Centers for Disease Control and Prevention National Institute for Occupational Safety and Health. 147 p. Pan, X. L.; Day, H. W.; Wang, W.; Beckett, L. A.; and Schenker, M. B., 2005, Residential proximity to naturally occurring asbestos and mesothelioma risk in California: American Journal of Respiratory and Critical Care Medicine, Vol. 172, No. 8, pp. 1019–1025. Swayze, G. A.; Kokaly, R. F.; Higgins, C. T.; Clinkenbeard, J. P.; Clark, R. N.; Lowers, H. A.; and Sutley, S. J., 2009, Mapping potentially asbestos-bearing rocks using imaging spectroscopy: Geology, Vol. 37, No. 8, pp. 763– 766. U.S. Department of Labor, 1975, Occupational exposure to asbestos: Federal Regulation 40, pp. 47652–47665. Vander Kelen, P. and Patrick, G., 2013, Sumas Mountain/Swift Creek Asbestos Cluster Investigation: Electronic document, available at http://www.doh.wa.gov/Portals/1/ Documents/Pubs/334-331.pdf. Weill, D., 2018, Proceedings of the Monticello Conference on Elongate Mineral Particles (EMP): Toxicology and Applied Pharmacology, Vol. 361, pp. 1–2. World Health Organization, 1997, Determination of Airborne Fibre Number Concentrations: World Health Organization, Geneva, Switzerland.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 3–8

Overview of Naturally Occurring Asbestos in California and Southwestern Nevada R. MARK BAILEY* Asbestos TEM Laboratories, Inc., 600 Bancroft Way, Suite A, Berkeley, CA 94710

ABSTRACT Naturally occurring asbestos (NOA) is being discovered in a widening array of geologic environments. The complex geology of the state of California is an excellent example of the variety of geologic environments and rock types that contain NOA. Notably, the majority of California rocks were emplaced during a continental collision of eastward-subducting oceanic and island arc terranes (Pacific and Farallon plates) with the westward continental margin of the North American plate between 65 and 150 MY BP. This collision and accompanying accretion of oceanic and island arc material from the Pacific plate onto the North American plate, as well as the thermal events caused by emplacement of the large volcanic belt that became today’s Sierra Nevada mountain range, are the principal processes that produced the rocks where the majority of NOA-bearing units have been identified. INTRODUCTION California is among the most important states in terms of the extent and diversity of rock units bearing naturally occurring asbestos (NOA). Particularly noteworthy are two serpentine and serpentinized peridotite belts that cover approximately 3,200 km2 (Frazell et al., 2009), including in 42 of 58 California counties. Additionally, areas of serpentine-bearing gravels and soil shed from outcrop exposures are found in large alluvial fans and other drainage systems. Numerous other rock types described in the following paragraph also contain asbestos. An overview of the wide extent of asbestos occurrences is provided in a map compiled by the U.S. Geological Survey (USGS) and the California Geological Survey, as shown in Figure 1 (Van Gosen and Clinkenbeard, 2011). Ultramaﬁc rocks crop out in several distinct geomorphic provinces: (1) the Coast Range, (2) the western Sierra Nevada (referred to as the Sierra Nevada Foothills), and (3) the Klamath Mountains. Additionally, a considerable number of asbestos occurrences

*Corresponding author email: mark@asbestostemlabs.com

not related to ultramaﬁc rocks occur in the Basin and Range/Mojave Desert provinces. A general summary of California rock types known to contain asbestiform minerals and references to documents describing these types of occurrences include the following: Serpentinite (Van Gosen and Clinkenbeard, 2011): chrysotile, tremolite/actinolite Sedimentary serpentine deposits (Wakabayashi, 2011): chrysotile Metabasalt/metagabbro (blueschist) (Erskine and Bailey, 2018): glaucophane, winchite, riebeckite Metabasalt/metagabbro (greenstone) (Sacramento Metropolitan Air Quality Management District, 2004): tremolite/actinolite Meta-granitoids (Buck et al., 2013; Metcalf and Buck, 2015): actinolite, Mg-riebeckite, winchite/richterite Other meta-volcanic (Van Gosen et al., 2004; Wakabayashi, 2011): Na amphibole Fe-rich meta-chert (Van Gosen and Clinkenbeard, 2011): riebeckite, grunerite Iron-ore deposits replacing dolomite (Van Gosen and Clinkenbeard, 2011): tremolite/anthophyllite Meta-carbonates (Van Gosen et al., 2004; Van Gosen and Clinkenbeard, 2011): serpentine, tremolite/actinolite, winchite/richterite Phyllites (sodium-metasomatized) (Albino, 1995): riebeckite Schists (Higgins and Clinkenbeard, 2006): tremolite/actinolite, anthophyllite Shonkinite carbonatite and syenite (Olson et al., 1954): riebeckite Talc deposits (Van Gosen et al., 2004): tremolite, winchite/richterite Colluvium/alluvium and other rock/soil shed oﬀ hills containing the above rock types

CALIFORNIA COAST RANGE The Coast Range province is a belt of rocks ∼500 miles long by as much as 100 miles wide trending in a northwest-to-southeast direction and dominated by rocks of the Franciscan Complex (hereafter

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 9–14

Bailey

Figure 1. California maps. (A) Ultramaﬁc rocks (Harrison et al., 2004). (B) Asbestos mines and known occurrences of NOA (Van Gosen and Clinkenbeard, 2011). (C) Geomorphic provinces (California Geological Survey, 2016).

Franciscan) containing abundant outcrops of ultramafic rocks emplaced onto the west coast margin of the North American plate ∼65-150MY ago during subduction of the East Pacific and Farallon plates (Wakabayashi and Unruh, 1995). The type section of the Franciscan runs directly through the major population center of San Francisco (see Figure 2) as well as the Greater San Francisco Bay Area, and the ultramafic rocks it contains are responsible for a major effort and expense by local environmental, engineering, and construction firms as well as government regulatory bodies to control dust emissions during construction activities. NOA within ultramafic bodies in the Coast Range tends to be dominated by a short-fiber variety of

Figure 2. Geologic map of San Francisco. Note purple belt of ultramaﬁc rock extending from Hunters Point (southeast) to the Golden Gate Bridge (north) along the Hunters Point Shear Zone.

chrysotile (length <5 μm) while also containing occasional tremolite, actinolite, glaucophane, winchite, and hornblende fibrous amphiboles. In some ultramafic bodies, chrysotile is visible in hand samples and present in concentrations of 10% or more, commonly as a stockwork of veins. Curiously, although the veins are often several millimeters or more in thickness, they do not typically yield an abundance of long fibers. A possible explanation is given by Andreani et al. (2004) with veins growing through a small-scale (<5 μm) repetitive crack-seal cycle. The highest concentrations of asbestos in serpentine generally occur where it is invisible to the unaided eye. The defining example of this type of occurrence is in the New Idria serpentinite asbestos mining district (see Figure 3), which has been designated by the USGS as having the largest reserves of asbestos in the country (Van Gosen and Clinkenbeard, 2011). The ore material (Figure 4A; Rice, 1963) is rather non-descript with little obvious fibrosity. Rather, it is a pale waxy light greenish color. While the New Idria serpentinite had been known since the mid-1800s, it was not recognized until the mid-1950s when geologists analyzed the material by X-ray diffraction and discovered an exceptionally high purity level of sub-microscopic chrysotile (>50%). Notably, when fiber bundles are observed in the ore, they have been found by transmission electron microscopy to consist of collections of very short, generally <5-μm fibers that form interwoven aggregates that appear longer (Ilgren, 2008). Besides containing serpentine, the Franciscan in the Coast Range has numerous other types of NOAcontaining rocks with various fibrous amphiboles, including actinolite, tremolite, crocidolite, and winchite.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 9–14

NOA in California

Figure 3. Atlas mine, Coalinga, California. The largest of numerous asbestos mines in the area (∼450 acres) is now an EPA Superfund site.

A recent unusual occurrence of ﬁbrous amphiboles of a previously unrecognized form of asbestos discovered by the author is within the Calaveras Dam Replacement Project site (Erskine and Bailey, 2018) located in the hills east of Fremont, California. Here, both as-

bestiform glaucophane (a sodic amphibole closely related to regulated crocidolite) and winchite (a sodiccalcic amphibole similar to that found at Libby, MT) co-occur within altered block-in-matrix rocks composed of metabasalt mix greenstone and blueschist in

Figure 4. Asbestos ore. (A) Coalinga chrysotile, not obviously ﬁbrous, from the Coast Range. (B) Calaveras chrysotile in stockwork veins from the Sierra Nevada Foothills.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 9–14

Bailey

Figure 5. Right abutment of the Calaveras Dam in Fremont, California, in Franciscan mélange. Note the arrows pointing to an assortment of block-in-matrix rocks, three of which (underlined) were found to contain diﬀerent varieties of NOA.

a tectonic mélange (Figure 5). To complicate matters at the site, chrysotile-bearing serpentine and fibrous actinolite in an actinolitic schist are also present. The project site may well be the largest privately operated NOA management project in history with tens of thousands of transmission electron microscopic air samples collected and analyzed. SIERRA NEVADA FOOTHILL BELT Extending over 500 km in length, the Sierra Nevada Foothill serpentinite belt runs in a northwest-tosoutheast line along the western edge parallel to the Sierra Nevada mountain range, much of it contained within a the Foothill metamorphic belt group. It is generally rather narrow, varying from ∼10 km to as much as 50 km in width with numerous gaps. Although more remote than the San Francisco Bay Area, it has seen its population rise significantly due to its proximity to the growing Sacramento Valley area. Concern over NOA exposure first arose due to dust generated by vehicles traversing over NOA-bearing serpentinite gravels with a number of studies undertaken (Berman, 1994; Volpe Center, 2004; and Department of Toxic Substances Control, 2005) to assess the problem. The concern was well founded, resulting in considerable efforts to control hazardous dust, including paving existing roads and ceasing the use of offending material; implementation of the latter led to the shutdown of numerous serpentine rock quarries. Many of the occurrences in the Sierra Foothills lie on fault zones, particularly the Melones fault system. Within this belt lies the Calaveras mine in Copperopolis, California (Rice, 1963), which in the 1960s and 1970s was the largest producer of asbestos in the United States and operated until 1987 (Figure 6). Here,

asbestos occurs as veins within an ore body of serpentine (Figure 4B). Currently, the mine is operated as a hazardous waste dump for asbestos products. After the initial concern over the presence of asbestos in roadways abated, another type of more widespread NOA occurrence in the region was identified as asbestiform tremolite/actinolite or hornblende. Unlike the Coast Range, which experienced low- to high-pressure regional metamorphism at low temperature, the Sierra Foothills has experienced greenschist facies metamorphism, leading to the formation of amphiboles in many rocks, most notably in the El Dorado Hills area, where the U.S. Environmental Protection Agency and the USGS conducted detailed sampling and analysis as well as a full-scale risk assessment (Ladd, 2005; Meeker et al., 2006). During a series of studies of the area, fibrous tremolite was identified primarily in metamorphosed ultramafic rocks and actinolite-magnesiohornblende in mafic meta-volcanic rocks and meta-sedimentary rocks. The results of the studies led to the local high school athletic fields and playgrounds being buried with several feet of non– NOA-containing soil, new protocols being developed for dealing with NOA by the local air pollution control district, and warnings to residents on how to minimize exposure to NOA. An additional major NOA monitoring effort related to amphiboles in the area occurred during the emergency rebuilding effort at the Oroville Dam, which was severely damaged during the winter of 2016–2017, during which fibrous tremolite/actinolite was identified in metamorphosed mafic volcanic rocks that were being disturbed as repairs were made. KLAMATH MOUNTAINS The Klamath Mountains have the highest overall percentage of area of exposed serpentinite in the state, which occurs at the far northernmost intersection of the Coast Range and Sierra Nevada Foothills. However, the region is so remote and virtually nonpopulated that little knowledge exists of NOA in the area, and risk of exposure is concomitantly low. BASIN AND RANGE/MOJAVE DESERT (INCLUDING SOUTHERNMOST NEVADA AND SOUTHWESTERN ARIZONA) The Basin and Range/Mojave Desert area contains completely different rock types than the Coast Range, Sierra Foothills, and Klamath Mountains. The rocks are dominated by Precambrian though Paleozoic continental shelf and shallow sea sedimentary deposits that have been intruded and altered by various more recent volcanic and intrusive

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 9–14

NOA in California

Figure 6. The Calaveras mine in Copperopolis. The most productive asbestos mine in California is now an asbestos landﬁll operation.

rocks. Asbestos-bearing alterations most commonly occur at contacts of dolomitic rocks immediately adjacent to or nearby igneous intrusives ranging from basaltic/gabbroic (i.e., Death Valley; Van Gosen et al., 2004) to granitic/rhyolitic (i.e., Mazourka Canyon; Ross, 1966) and leading to a predominance of tremolite/winchite in the former and chrysotile/tremolite in the latter. Two particularly unusual occurrences of NOA have been identified within the last 5 years in silicic granitic rocks in southern Nevada/far western Arizona that have been subjected to thermal alteration, about as far chemically from the more common situation of ultramafic rock alteration described above. In the Boulder City area, the quartz monzonite Boulder City pluton (Miocene, ∼14 MY BP) was subjected to a recent thermal event that appears to have transformed through retrograde alteration pre-existing hornblende lathes into asbestiform actinolite (Buck et al., 2013; Metcalf and Buck, 2015). A few miles to the east lies the Wilson Ridge pluton (Miocene, ∼13 MY BP), which is hypothesized to have experienced an influx of sodium from hydrothermally mobilized saline evap-

orite deposits during cooling after intrusion, causing albitization of feldspars and deposition of ﬁbrous sodic amphiboles mg-riebeckite and sodic-calcic winchite/richterite (Buck et al., 2013; Metcalf and Buck, 2015).

CONCLUSION In summary, NOA in California and southwestern Nevada is found in an exceptionally wide range of rock types, with significant unexpected new deposits seeming to come every few years. Serpentine/ultramafic rocks are the primary source of NOA for chrysotile and/or tremolite/actinolite depending on the type and degree of metamorphism to which the rocks have been subjected. However, NOA has been found not only in ultramafic rocks but also in metamorphosed mafic and silicic igneous rocks as well as altered dolomites and a variety of others. This creates a serious challenge to geologists and others involved in assessing the potential for the presence of NOA at a given site where grounddisturbing activities will occur.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 9–14

Bailey

REFERENCES Albino, G. V., 1995, Sodium metasomatism along the Melones fault zone, Sierra Nevada foothills, California, USA: Mineralogical Magazine, Vol. 59, No. 3, pp. 383–399. Andreani, M.; Baronnet, A.; Boullier, A.-M.; and Gratier, J.P., 2004, A microstructural study of a “crack-seal” type serpentine vein using SEM and TEM techniques: European Journal of Mineralogy, Vol. 16, No. 4, pp. 585–595. Berman, W., 1994, Evaluation of Risks Posed to Residents and Visitors of Diamond XX Who Are Exposed to Airborne Asbestos Derived from Serpentine Covered Roadways: ICF Technology Prepared for U.S. EPA Region 9. Buck, B. J.; Goossens, D.; Metcalf, R. V.; McLaurin, B.; Ren, M.; and Freudenberger, F., 2013, Naturally occurring asbestos: Potential for human exposure, southern Nevada, USA: Soil Science Society of America Journal, Vol. 77, No. 6, pp. 2192–2204. California Geological Survey, 2016, California Geomorphic Provinces: Electronic document, available at https:// www.conservation.ca.gov/cgs/PublishingImages/Geotour/ GeolProvinces2.jpg. Department of Toxic Substances Control, 2005, Study of Airborne Asbestos from a Serpentine Road in Garden Valley, CA: California Environmental Protection Agency. Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in the Franciscan Complex blueschist, northern Diablo Range, California: Toxicology and Applied Pharmacology, Vol. 361, pp. 3–13. Frazell, J.; Elkins, R.; O’Geen, A. T.; Reynolds, R.; and Meyers, J., 2009, Facts about Serpentine Rock and Soil Containing Asbestos in California: University of California Division of Agriculture and Natural Resources Publication 8399. Harrison, S.; Safford, H.; and Wakabayashi, J., 2004, Does the age of exposure of serpentine explain variation in endemic plant diversity in California?: International Geology Review, Vol. 46, No. 3, pp. 235–242. Higgins, C. T. and Clinkenbeard, J. P., 2006, Relative Likelihood for the Presence of Naturally Occurring Asbestos in Placer County, California: California Department of Conservation, California Geological Survey Special Report 190. Ilgren, E., 2008, Review: The ﬁber length of Coalinga chrysotile: Enhanced clearance due to its short nature in aqueous solution with a brief critique on “Short Fiber Toxicity”: Indoor and Built Environment, Vol. 17, No. 1, pp. 5–26.

Ladd, K., 2005, El Dorado Hills Naturally Occurring Asbestos Multimedia Exposure Assessment El Dorado Hills, California: Preliminary Assessment and Site Inspection Report, Interim Final: Environmental Protection Agency Job No. 001275.0440, p. 01. Meeker, G.; Lowers, H.; Swayze, G.; Van Gosen, B.; Sutley, S.; and Brownfield, I., 2006, Mineralogy and Morphology of Amphiboles Observed in Soils and Rocks in El Dorado Hills, California: U.S. Geological Survey Open-File Report 20061362. Metcalf, R. V. and Buck, B. J., 2015, Genesis and health risk implications of an unusual occurrence of ﬁbrous NaFe3+amphibole: Geology, Vol. 43, No. 1, pp. 63–66. Olson, J. C.; Shawe, D.; Pray, L. C.; and Sharp, W. N., 1954, Rare-earth mineral deposits of the mountain pass district, San Bernardino County, California: Science, Vol. 119, No. 3088, pp. 325–326. Rice, S., 1963, California asbestos industry, Calif. Division of Mines and Geology, Mineral Information Service, Vol. 16, No. 9, p. 12. Ross, D. C., 1966, Stratigraphy of Some Paleozoic Formations in the Independence Quadrangle Inyo County, CA: U.S. Geological Survey Professional Paper 396. Sacramento Metropolitan Air Quality Management District, 2004, Map of NOA in Metro Sacramento: Electronic document, available at http://www.airquality.org/ StationarySources/Documents/NOA_Parcels_redux.pdf. Van Gosen, B. S. and Clinkenbeard, J. P., 2011, Reported Historic Asbestos Mines, Historic Asbestos Prospects, and Other Natural Occurrences of Asbestos in California: U.S. Geological Survey Open-File Report 2011-1188. Van Gosen, B. S.; Lowers, H. A.; and Sutley, S. J., 2004, A USGS Study of Talc Deposits and Associated Amphibole Asbestos within Mined Deposits of the Southern Death Valley Region, California: U.S. Geological Survey Open-File Report, 2004-1092. Volpe Center, 2004, Roadside Airborne Monitoring along an El Dorado County Serpentine Roadway: Volpe Center IAG VP262, Sponsor Document 01-T2226 Prepared for the California Department of Toxic Substances Control. Wakabayashi, J., 2011, Mélanges of the Franciscan Complex, California: Diverse structural setting, evidence for sedimentary mixing, and their connection to subduction processes: Special Paper of the Geological Society of America, Vol. 480, pp. 117– 141. Wakabayashi, J. and Unruh, J. R., 1995, Tectonic wedging, blueschist metamorphism, and exposure of blueschists: Are they compatible?: Geology, Vol. 23, No. 1, pp. 85–88.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 9–14

Clastic Sedimentary Rocks and Sedimentary Mélanges: Potential Naturally Occurring Asbestos Occurrences (Amphibole and Serpentine) JOHN WAKABAYASHI* California State University, Fresno, Department of Earth and Environmental Sciences, 2576 East San Ramon Avenue, Mailstop ST-24, Fresno, CA 93740-8039

Key Terms: NOA, Asbestos, Serpentinite, Glaucophane, Sandstones, Subduction Complexes ABSTRACT Petrography of mélange matrix and clastic sedimentary rocks in coastal California reveals the occurrence of detrital serpentine and detrital asbestiform sodic amphibole (glaucophane). Many sandstones of the Franciscan Complex have small amounts of detrital serpentine, with amounts of up to several percent in some cases. Detrital amphibole, including asbestiform glaucophane, is also present in some sandstones. Whereas rare sandstones have so much detrital glaucophane that they appear blue in hand specimen (up to nearly half of the rock volume), most glaucophane-bearing sandstones lack blue color, and the detrital glaucophane is not apparent in hand specimen. Most of the occurrences of detrital glaucophane are in blueschist facies sandstones, some of which also contain neoblastic (grew in place) glaucophane, but a notable exception is a widespread prehnite-pumpellyite facies unit that crops out primarily in Sonoma and Marin Counties. The detrital mineralogy of sandstones mirrors the block and matrix compositions of Franciscan mélanges that can be thought of as scaled-up equivalents of these clastic sedimentary rocks (mega-conglomerates/sedimentary breccias). Franciscan mélanges range from having a detrital siliciclastic to a detrital serpentinite matrix, and interfingering and gradation of the two matrix types is common. These findings suggest that clastic sedimentary rocks associated with current or past active orogenic settings elsewhere in the world may contain naturally occurring asbestos (NOA) even if the NOA component minerals are not visible in hand specimen. INTRODUCTION Whereas naturally occurring asbestos (NOA) was initially largely associated with serpentinite bodies, a growing number of studies (e.g., Metcalf and Buck, *Corresponding author email: jwakabayashi@csufresno.edu

2015; Erskine and Bailey, 2018) have found occurrences of asbestiform minerals in a wide range of geologic settings, including various metamorphic rocks and hydrothermally altered plutonic rocks. Many rock bodies with asbestiform minerals are exposed in positions to potentially shed NOA detritus into various distributary systems that will ultimately lead to accumulation in various depocenters in terrestrial and marine environments. Quaternary alluvial deposits downstream of known NOA localities can be reasonably suspected of having the potential to contain NOA. In contrast, clastic sedimentary rocks that no longer have a direct spatial connection to their sources have thus far been overlooked as potentially containing NOA. This contribution reports reconnaissance-level (petrography only) study of detrital serpentinite and detrital fibrous amphibole in clastic sedimentary rocks, including mélange matrix from the Franciscan Complex of coastal California. These findings have been previously reported in articles that emphasized interpretation of mélange origins (e.g., Wakabayashi, 2015, 2017a, 2019) rather than NOA. GEOLOGIC SETTING The Franciscan Complex makes up much of the bedrock of the California Coast Ranges (Figure 1) and is the world’s type subduction complex, formed by the transfer of rocks from the subducting to the upper plate (subduction-accretion), as the subduction thrust sporadically cuts into the top of the downgoing plate (Wakabayashi, 2015, 2017a). The upper part of the subducting plate (termed “ocean plate stratigraphy,” or OPS; Isozaki et al., 1990) consists of the uppermost part of the igneous oceanic crust (commonly basalt) overlain by pelagic sedimentary rocks (commonly chert), which is in turn overlain by clastic sedimentary rocks (mud rocks, sandstone, conglomerate) that represent trench fill. The clastic part of this triad is the most voluminous and it includes block-in-matrix units (mélanges), some of which have a serpentinite matrix (Wakabayashi, 2015, 2017a, 2019). The

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 15–19

Wakabayashi

Figure 1. General geology of northern and central California showing the Franciscan Complex, larger serpentinite bodies as well as the general localities (red unfilled circles) where detrital fibrous glaucophane and/or serpentinite were identified in this study. Note that the locations shown comprise multiple outcrops and thin sections. Map figure revised from Wakabayashi (2017a).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 15–19

Clastic Sedimentary Rocks: Potential NOA

Franciscan Complex includes both serpentinite mélanges and intact slabs of serpentinite, but the largest bodies of serpentinite within the California Coast Ranges are associated with the Coast Range ophiolite or Great Valley Group (Wakabayashi, 2017b, 2017c).

DETRITAL SERPENTINITE AND AMPHIBOLE IN SANDSTONES AND MÉLANGE MATRIX Detrital glaucophane and/or detrital serpentinite was identified by petrographic examination in over 100 thin sections from localities spanning a wide spatial range of units within the Franciscan Complex (some general localities shown on Figure 1). It is expected that these detrital components are present in a much larger part of the Franciscan that was not directly examined in this study. Detrital serpentinite/serpentine has been found in sandstones of the Franciscan Complex in a wide range of units, with metamorphism of prehnite-pumpellyite to blueschist facies (Wakabayashi, 2015, 2017a) (Figure 2). A significant fraction of the detrital serpentinite in these sandstones appears to have been replaced by chlorite. Most Franciscan sandstones have small amounts (<1 percent) of detrital serpentinite, but less common examples (probably <5 percent of Franciscan sandstones) have amounts of up to about 5 percent. The higher concentrations of detrital serpentinite are associated with sandstones and conglomerates that are considered mélange matrix. Sandstones and conglomerates with large amounts of detrital serpentinite (>50 percent) are present, but these have the appearance of serpentinite in hand specimen and outcrop, so the presence of serpentinite as a potential NOA source will be more easily recognized. The rocks composed of >50 percent detrital serpentinite have been mapped as serpentinite in previous studies and the detrital origins have been proposed on the basis of macro and micro textures, as well as stratigraphic relationships (e.g., Lockwood, 1971; Phipps, 1984; and Wakabayashi, 2012, 2015, 2017b, 2017c, 2019). Siliciclastic and detrital serpentinite mélange matrix in the Franciscan Complex has been interpreted to be interbedded, and they grade into one another on the basis of field and petrographic relationships (Wakabayashi, 2015, 2017a, 2017b). The sandstones and conglomerates with a large proportion of detrital serpentinite form serpentinite mélange in both the Franciscan Complex and the Great Valley Group (Wakabayashi, 2015, 2017a, 2017b, 2017c). Note that some otherwise siliciclastic sandstones of the basal Great Valley Group also contain detrital serpentinite (Figure 2).

Some blueschist facies and prehnite-pumpellyite facies sandstones of the Franciscan contain detrital glaucophane that is commonly fibrous, as are neoblastic (metamorphic) overgrowths of glaucophane found only in the blueschist facies sandstones (Figure 2). Whereas detrital glaucophane is found mainly in blueschist facies sandstones, it is also common in a widely distributed unit in Sonoma County that comprises prehnite-pumpellyite facies conglomerates, sedimentary breccias, and sandstones (Wakabayashi, 2015). Whereas blueschist clasts in a conglomerate are easily recognized in outcrop, neoblastic (metamorphic grew-in-place) or detrital glaucophane (detrital blueschist clasts) is seldom recognizable in sandstone hand specimens owing to the relatively small grain size (commonly <0.5 mm) and the relatively small volumetric proportions of this material (generally <5 percent). An exception is a rare blue sandstone from El Cerrito composed of about 50 percent detrital and neoblastic glaucophane (Wakabayashi, 2017a). The fibrous glaucophane seen in detrital blueschist clasts in Franciscan sandstones is typical of the glaucophane seen in fine-grained blueschists of the Franciscan, present both as blocks-in-mélange and intact kilometer-scale bodies (Figure 2j–l) (compare to identical textures shown in Erskine and Bailey [2018]). DISCUSSION AND CONCLUSIONS The occurrence of detrital serpentinite and glaucophane in sandstones of the Franciscan Complex and Great Valley Group mirrors the matrix, clast, and block populations in associated mélanges (Wakabayashi, 2015, 2017a). Whereas serpentinite and blueschist are easily identified as pebble- or largersized clasts and blocks, these detrital components are not visually obvious in outcrops of associated sandstones. Although this study reports detrital serpentinite and glaucophane from the Franciscan Complex and Great Valley Group of coastal California, the general geologic relationships there suggest the likelihood of such detritus in younger sedimentary units that are derived from exposures of blueschist and serpentinite. The most likely setting for clastic sediment with a high potential for detrital glaucophane and serpentinite is Quaternary alluvium downstream of major exposures of serpentinite and blueschist. In such cases, however, the presence of detrital blueschist (and, thus, detrital glaucophane) and detrital serpentinite is easy to detect because of the abundance of gravel-sized and larger clasts in the alluvial deposits. Miocene clastic rocks in the California Coast Ranges are sourced to variable degrees from the older Coast Range bedrock units such as the Franciscan Complex, Coast Range ophiolite, and Great Valley

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 15–19

Wakabayashi

Figure 2. Photomicrographs of detrital serpentinite and detrital glaucophane. Mineral abbreviations (for j): gln indicates glaucophane; hb, hornblende; jd, jadeite; qtz, quartz. a to h and k and l are presented as plane polarized and cross-polarized pairs, whereas i and j show cross-polarized and plane-polarized views alone, respectively. (a, b) A tremolite-chlorite-antigorite (meta-ultramafic) schist clast in prehnite pumpellyite facies pebbly sandstone from the cliffs of Black Sand Beach of the Marin Headlands. (c, d) Serpentinite clasts, one with chrome spinel, from a blueschist facies sandstone from Sunol Regional Wilderness, northern Diablo Range. (e, f, and i) From prehnite-pumpellyite facies pebbly sandstone from El Cerrito quarry, eastern San Francisco Bay area. (i) What appears to be cross-fiber chrysotile veins in the large serpentinite clast, whereas e and f show variable alteration of a serpentinite clast. (g, h) From basal Great Valley Group sandstone (zeolite facies or unmetamorphosed) at Chabot Reservoir, eastern San Francisco Bay area. (j) Detrital blueschist clasts with fibrous glaucophane as well as typical acicular glaucophane metamorphic (i.e., neoblastic, post-depositional) overgrowths on one of the clasts, from blueschist facies sandstone of El Cerrito quarry. The fibrous nature of glaucophane in fine-grained blueschists is typical in the Franciscan, as seen in k and l from a kilometer-scale blueschist body, the Rattlesnake Schist of O’Day (1974), near Cummings, northern California Coast Ranges. These textures are identical to those shown from blueschist from the Calaveras Dam area of the northern Diablo Range by Erskine and Bailey (2018). Photos a through d and j were presented with slightly different annotation in Wakabayashi (2017a), whereas the other photos are presented here for the first time. Photos a through d and i and j were taken at California State University, Fresno, whereas photos e through h and k and l were taken at Asbestos TEM Laboratories in Berkeley, CA.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 15–19

Clastic Sedimentary Rocks: Potential NOA

Group (e.g., Graham et al., 1984) and may be expected to have detrital glaucophane and/or detrital serpentinite. For example, conglomerates of the Miocene Orinda Formation of the Contra Costa Group in the eastern San Francisco Bay region (Figure 1) contain abundant clasts of blueschist, although the source outcrops may be >40 km away (Wakabayashi, 1999). In conclusion, the field and petrographic analysis presented in this article shows that clastic sedimentary rocks of the Franciscan Complex of California contain detrital serpentinite and detrital fibrous glaucophane. Assessment of the geology of coastal California suggests that such occurrences are widespread and that similar occurrences of detrital serpentinite and glaucophane are present in younger sedimentary units sourced from current and past exposures of blueschist and serpentinite. These findings have potential global implications for distribution of NOA. Clastic sedimentary rocks associated with current or past active orogenic settings may contain NOA even if the NOA component minerals are not visible in hand specimen. REFERENCES Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in Franciscan Complex blueschist, northern Diablo Range, California: Toxicology and Applied Pharmacology, Vol. 361, pp. 3–13. Graham, S. A.; McCloy, C.; Hitzman, M.; Ward, R.; and Turner, R., 1984, Basin evolution during change from convergent to transform continental margin in central California: American Association Petroleum Geologists Bulletin, Vol. 68, no. 3, pp. 233–249. Isozaki, Y.; Maruyama, S.; and Furuoka, F., 1990, Accreted oceanic materials in Japan” Tectonophysics, Vol. 181, nos. 1–4, pp. 179–205.

Lockwood, J. P., 1971, Sedimentary and gravity-slide emplacement of serpentinite: Geological Society America Bulletin, Vol. 82, no. 4, pp. 919–936. Metcalf, R. V. and Buck, B. J., 2015, Genesis and health risk implications of an unusual occurrence of ﬁbrous NaFe3+ amphibole: Geology, Vol. 43, no. 1, pp. 63–66. O’Day, M. S., 1974, Structure and Petrology of the Mesozoic and Cenozoic Rocks of the Franciscan Complex, Leggett-Piercy Area, Northern California: Ph.D. Dissertation, University of California, Davis. Phipps, S. P., 1984, Ophiolitic olistostromes in the basal Great Valley sequence, Napa County, northern California Coast Ranges. In Raymond, L. A. (Editor), Mélanges: Their Nature, Origin, and Significance: Special Paper 198, Geological Society of America, Boulder, CO, pp. 103–126. Wakabayashi, J., 1999, Distribution of displacement on and evolution of, a young transform fault system: The northern San Andreas fault system, California: Tectonics, Vol. 18, no. 6, pp. 1245–1274. Wakabayashi, J., 2012, Subducted sedimentary serpentinite mélanges: Record of multiple burial-exhumation cycles and subduction erosion: Tectonophysics, v. 568–569, p. 230–247. Wakabayashi, J., 2015, Anatomy of a subduction complex: Architecture of the Franciscan Complex, California, at multiple length and time scales: International Geology Review, Vol. 57, nos. 5–8 pp. 669–746. doi:10.1080/00206814.2014.998728. Wakabayashi, J., 2017a, Structural context and variation ocean plate stratigraphy, Franciscan Complex of California: Insight into mélange origins and subduction-accretion processes: Progress Earth Planetary Science, Vol. 4, No. 18, 23 p. doi:10.1186/s40645-017-0132-y. Wakabayashi, J., 2017b, Serpentinites and serpentinites: Variety of origins and emplacement mechanisms of serpentinite bodies in the California Cordillera: Island Arc. doi:10.1111/iar.12205. Wakabayashi, J., 2017c, Sedimentary serpentinite and chaotic units of the lower Great Valley Group forearc basin deposits, California: Updates on distribution and characteristics: International Geology Review, Vol. 59, nos. 5–6, pp. 599–620. doi:10.1080/00206814.2016.1219679. Wakabayashi, J., 2019, Sedimentary compared to tectonicallydeformed serpentinites and tectonic serpentinite mélanges: Unambiguous and disputed examples from California: Gondwana Research. v. 74, p. 51–67. doi:10.1016/j.gr.2019.04.005.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 15–19

Asbestiform Minerals of the Franciscan Assemblage in California with a Focus on the Calaveras Dam Replacement Project R. MARK BAILEY* Asbestos TEM Laboratories, Inc., 600 Bancroft Way, Suite A, Berkeley, CA 94710

Key Terms: Asbestos, NOA, Glaucophane, Blueschist, Amphibole, Franciscan ABSTRACT The San Francisco Bay Area is underlain by bedrock of the Franciscan Assemblage, which outcrops in numerous places. A significant portion of these outcrops consists of rock types that contain both regulated and unregulated asbestiform minerals, including ultra-mafic serpentinites, various greenstones, amphibolites, blueschist, and other schists (talc-tremolite, actinolite, etc.). These rocks are a legacy of tectonic activity that occurred on the west coast margin of the North American plate ∼65–150 MY ago during subduction of the East Pacific and Farallon plates. The Calaveras Dam Replacement Project (CDRP), located in Fremont, California, is an example of an area within the Franciscan Assemblage that is substantially underlain by metamorphosed oceanic sedimentary, mafic, and ultra-mafic rocks in a tectonic subduction zone mélange with highly disrupted relationships between adjoining rock bodies with different pressure/temperature metamorphic histories. In order to protect the health of workers and residents in the surrounding area, an extensive effort was taken to identify, categorize, and monitor the types, locations, and concentrations of naturally occurring asbestos at the site. Using a combination of geologic field observations and transmission electron microscopy, energy dispersive X-ray, and selected area electron diffraction analysis of airborne particulate and rock/soil samples, the CDRP was discovered to contain chrysotile-bearing serpentine. It also had as a range of amphibole-containing rocks, including blueschist, amphibolite schist, and eclogite, with at least 19 different regulated and non-regulated fibrous amphibole minerals identified. The extensive solid solution behavior of the amphiboles makes definitive identification difficult, though a scheme was created that allowed asbestos mineral fingerprinting of various areas of the project site.

*Corresponding author email: mark@asbestostemlabs.com

INTRODUCTION The Calaveras Dam site lies at the south end of the Calaveras Valley (Figure 1) in Fremont, California, within the Diablo Range, one of the major California coastal ranges. Bedrock at the site consists of Franciscan Assemblage and Great Valley Sequence rocks. In 2001, the existing dam, which was not engineered for earthquake safety, was found by the California Department of Water Resources, Division of Safety of Dams, to have an unacceptably high potential to fail during a maximum seismic event from nearby faults to the west: the Calaveras (∼¼ mi), the Hayward (∼5 mi), or the San Andreas (∼20 mi). As a result, the water level at the dam was lowered to 40 percent of capacity, a considerable loss of water storage capacity for the San Francisco Public Utility Commission, particularly as the Calaveras Reservoir is the second-largest source of freshwater for the city of San Francisco. In 2011, the Calaveras Dam Replacement Project (CDRP) was undertaken to build a new dam downstream of the old dam and minimize the risk of potential dam failure. Prior to dam construction, a geotechnical drilling program was performed to determine the rock types and competency of the proposed dam bedrock, which generated a large amount of drill core for study. Because Franciscan Assemblage serpentine was observed on geologic maps of the nearby area (Figure 2), indicating the potential presence of chrysotile naturally occurring asbestos (NOA), approximately 50 sections of drill core were submitted to Asbestos TEM Labs in Berkeley, California. Testing was conducted by transmission electron microscopy (TEM) bulk sample analysis to determine the asbestos content if any. As expected, chrysotile asbestos was detected within several of the drill cores (Figure 3). However, an unexpected finding was that numerous samples were found to contain significant concentrations of a suite of highly fibrous amphiboles (Figure 4) with chemistry, determined by energy dispersive X-ray (EDX) analysis (Figure 5), to be similar to riebeckite but with an exceptionally high aluminum and sometimes calcium content. Fibrous amphiboles of these chemical compositions are not classified by the

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

Bailey

Figure 1. Geologic map of California showing distribution of Franciscan Assemblage (light blue) (Irwin, 1990), modiﬁed.

International Mineralogical Association (Leake et al., 1997) as minerals that are regulated as asbestos per both U.S. and California regulations, even though they are in a solid solution series with them and have similar morphology.

Following this discovery, a careful review of the scientiﬁc literature on the glaucophane-riebeckite solid solution series (Deer et al., 1997; Leake et al., 1997; and National Institute for Occupational Safety and Health [NIOSH], 2005), as well as the documented

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

IAEG CDRP Glaucophane Presentation

Figure 2. Geologic map of Calaveras Dam site (Dibblee and Minch, 2005). Jurassic/Cretaceous Franciscan rocks. sp = serpentinite; f = shale; fm = mélange; gl = blueschist mixed with greenstone; fs = greywacke sandstone; Kp/Kps = Cretaceous Panoche formation (marine shale and sandstone); Tt = Tertiary Temblor sandstone; Tso = Tertiary Sobrante sandstone; Tm = Tertiary Monterey/Claremont formation; Tbr = Tertiary Briones sandstone; Qls = Quaternary landslide; Qa = Quaternary alluvium; af = artiﬁcial ﬁll. Approximate new dam location in red.

widespread occurrence of glaucophane blueschists in the Franciscan ranges, led to the determination that the dominant material was either a highaluminum riebeckite or a mix of glaucophane and Fe-

glaucophane. These amphiboles had not been previously recognized as occurring in an asbestiform habit in the literature on NOA occurrences (CARB, 1991; Van Gosen, 2007; and Van Gosen and Clinkenbeard,

Figure 3. TEM photomicrograph of a bundle of chrysotile asbestos ﬁbers.

Figure 4. TEM photomicrograph of glaucophane asbestos ﬁbers. Note slender width (∼0.25 μm), curvature, and high aspect ratios >10:1.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

Bailey

Figure 5. EDX spectra collected from amphibole ﬁbers. Note the lack of Al in (a) NIST crocidolite standard but its presence in all CDRP spectra; apparent solid-solution substitution of Fe3+ and Al between (b) glaucophane and (c) Fe-glaucophane; presence of signiﬁcant Ca in (d) winchite.

2011). Later, with further analysis, aluminum-rich varieties of the sodic-calcic amphibole mineral species winchite and Fe-winchite were also identified though in significantly lower concentrations. Unlike glaucophane and Fe-glaucophane, winchite and Fe-winchite have been found to occur in an asbestiform habit at the EPA Superfund site at Libby, Montana, within the vermiculite ore body of the former W.R. Grace mine. Because EDX is not considered a fully quantitative technique for detailing amphibole mineral identification and the exact ratio of Fe3+ to Al was unknown, the choice was made by the lab and the consulting geologist on the project to initially designate the fibrous glaucophane/Fe-glaucophane material as

“high-aluminum crocidolite”. This decision was based on the International Mineralogical Association (IMA) (IMA 1978) definition of crocidolite as any asbestiform sodic amphibole (Leake, 1978). TEM quantitative bulk concentrations of these sodic amphiboles were found to range to over 10 wt.%. This was of considerable concern and led to a number of studies in the field and in the lab to determine its extent at the site. One particularly interesting finding was the presence of greenstone and blueschist intermixed on all scales, making it virtually impossible to separate the two (Figures 6 and 7). As excavation for the new dam proceeded, it became evident that the rocks of the right abutment and

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

IAEG CDRP Glaucophane Presentation

Figure 6. Blueschist and greenstone boulders used as waterside armoring of old Calaveras Dam.

base of the new dam consisted of Franciscan mélange block-in-matrix rocks of non–NOA-containing rocks (greywacke, siliceous schist, and eclogite) and potentially NOA-containing rocks (greenstone, blueschist, serpentinite, actinolite amphibolite, and actinolite schist) in a shale matrix (Figure 8). Within the literature on the Franciscan Assemblage, it has been observed that tectonic mélange zones frequently contain rocks such as those found at the CDRP dam site that have been subducted to considerable depth (Figure 9) under high-pressure/low-temperature conditions (Wakabayashi and Unruh, 1995). At the CDRP dam site, greenstone and blueschist are the hosts that commonly contain the fibrous minerals glaucophane, Fe-glaucophane, winchite, and Fewinchite. Notably, most of the glaucophane/Feglaucophane fibers would fall into the pre-1994 IMA amphibole mineral crossite solid-solution field designation, which included parts of the current glaucophane, Fe-glaucophane, riebeckite, and Mg-riebeckite chemical solid-solution fields (Leake, 1978). Numerous authors have studied the Franciscan blueschists, a few of whom are listed here (Smith, 1906; Ernst, 1984; Wakabayashi, 2011; Wassmann and Stöckhert, 2012; Kim et al., 2013; and Erskine and Bailey, 2018). While several have implied that the material in some cases might be fibrous, noting highaspect-ratio glaucophane particles (Kim et al., 2013) or making direct observations by scanning electron microscopy (SEM) of glaucophane needles (Wassmann and Stöckhert, 2012), the author is the first to have characterized it as asbestiform (Erskine and Bailey, 2018). One of the first clues as to full the extent of glaucophane/winchite fibrosity occurred in 2013 when field emission scanning electron microscopy (FESEM)

Figure 7. Blueschist-greenstone mylonite cut parallel to foliation viewed in plane polarized light. Glaucophane fibers (pleocroic blue γ) oriented approximately east to west intermixed with rounded gray-green greenstone fragments. Low magnification, fibers subvisible.

analysis was performed on fracture surfaces of several blueschist rock samples. The results were startling, showing virtually the entire rock surface covered with ﬁbers (Figure 10), and the observed ﬁbers showed all the classic characteristics of asbestos (Figure 11):

r r r r

High-aspect-ratio ﬁbers Bundles with split ends Curving ﬁbers Fibers/bundles with lengths >5 μm

In addition to fibrous glaucophane, low concentrations of other fibrous calcic amphibole species were identified, including asbestiform actinolite as well as elongate particles of hornblende that met the CARB 3:1 aspect ratio to be counted as asbestos. In particular, actinolite was found to occur as fibrous

Figure 8. Right dam abutment hillcut in riprap Franciscan mélange with three types of asbestos-bearing rocks underlined.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

Bailey

Figure 9. Cross section of hypothesized Paciﬁc–North American plate margin ∼100 MY BP in the area of central California, looking north. Oceanic Paciﬁc plate (left) is subducting under the continental North American plate (right) (National Park Service, 2015).

overgrowths on large columnar non-asbestiform actinolite within an amphibolite sample as determined by polarized light microscopy and FESEM analysis (Figure 12). Due to the presence of this wide range of fibrous amphiboles at the CDRP dam site, a comprehensive air monitoring program was developed to monitor the presence of fibrous minerals. At the site fibrous minerals were observed in extreme cases to reach concentrations as high as ∼2 fibers/cc PCME determined by

NIOSH 7402 TEM analysis, which is well over regulatory permissible exposure limits if considered as asbestos. Furthermore, it became apparent that similar rocks also occurred off-site that at times contributed independently to high airborne fiber concentrations. This led to an effort to individually identify and speciate all fibrous amphiboles, a laborious analytical task, to allow tracking of both on-site and off-site NOA emissions.

CONCLUSION

Figure 10. FESEM photomicrographs of asbestiform CDRP glaucophane ﬁbers from a blueschist/greenstone rock fracture surface.

A wide range of asbestiform minerals have been identified at the CDRP site (chrysotile, glaucophane, Fe-glaucophane, winchite, Fe-winchite, actinolite, CARB 435–countable hornblende, and others) with the highest concentrations being found in a previously unidentified source of asbestos: blueschist/ greenstone metabasalt. The presence of the widespread occurrence of these fibrous amphiboles led to an intensive effort to understand the nature and extent of their presence as well as an intensive dust control and air monitoring effort to protect human health, including that of workers at the site and the public nearby. Blueschists are widely observed throughout the Franciscan Assemblage and should be expected to be found at future construction sites. It is important that these sites be identified to minimize the NOA hazard potential they represent during ground-disturbing activities.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

IAEG CDRP Glaucophane Presentation

Figure 12. Fibrous actinolite overgrowths on pre-existing nonfibrous actinolite in an amphibolite with no obvious asbestiform nature in hand sample. (a) Thin section viewed under cross-polarized light. (b) Thin-section blank surface viewed by FESEM. Figure 11. Glaucophane-winchite fibers in blueschist viewed by FESEM. Sample cut parallel (a) and perpendicular (b) to foliation/lineation. Note the high concentration of sub-parallel fibers and fiber bundles with consistent fiber width.

REFERENCES CARB, 1991, Method 435—Determination of Asbestos Content of Serpentine Aggregate: Electronic document, available at https://ww3.arb.ca.gov/testmeth/vol3/m_435.pdf. Deer, W. A.; Howie, R. A.; and Zussman, J., 1997, Rock-Forming Minerals: Double-Chain Silicates: Geological Society, London, U.K. Dibblee, T. W. and Minch, J. A., 2005, Geologic Map of the Calaveras Reservoir Quadrangle, Alameda & Santa Clara Counties, California: Santa Barbara Museum of Natural History, Santa Barbara, CA.

Ernst, W., 1984, Californian blueschists, subduction, and the signiﬁcance of tectonostratigraphic terranes: Geology, Vol. 12, No. 7, pp. 436–440. Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in the Franciscan Complex blueschist, northern Diablo Range, California: Toxicology and Applied Pharmacology, Vol. 361, pp. 3–13. Irwin, W. P., 1990, Geology and plate-tectonic development. In Wallace, R. E. (Editor), The San Andreas Fault System, California: US Geological Survey Professional Paper 1515, pp. 61–82. Kim, D.; Katayama, I.; Michibayashi, K.; and Tsujimori, T., 2013, Rheological contrast between glaucophane and lawsonite in naturally deformed blueschist from Diablo Range, California: Island Arc, Vol. 22, No. 1, pp. 63–73. Leake, B. E., 1978, Nomenclature of amphiboles: Mineralogical Magazine, Vol. 42, No. 324, pp. 533–563. Leake, B. E.; Woolley, A. R.; Arps, C. E.; Birch, W. D.; Gilbert, M. C.; Grice, J. D.; Hawthorne, F. C.; Kato, A.; Kisch,

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

Bailey H. J.; Krivovichev, V. G.; Linthout, K.; Laird, J.; Mandarino, J. A.; Maresch, W. V.; Nickel, E. H.; Rock, N. M. S.; Schumacher, J. C.; Smith, D. C.; Stephenson, N. C. N.; Ungaretti, L.; Whittaker, E. J. W.; and Youzhi, G., 1997, Nomenclature of amphiboles; report of the Subcommittee on Amphiboles of the International Mineralogical Association Commission on new minerals and mineral names: Mineralogical Magazine, Vol. 61, No. 405, pp. 295–310. National Institute for Occupational Safety and Health, 2005, Pocket Guide to Chemical Hazards: Publication No. 2005-149, Department of Health and Human Services, National Institute for Occupational Safety and Health, Washington, D.C. National Park Service, 2015, Geologic Thrusts from the Past: Electronic document, available at https://www.nps.gov/ prsf/learn/nature/geologic-thrusts-from-the-past.htm Smith, J. P., 1906, The paragenesis of the minerals in the glaucophane-bearing rocks of California: Proceedings of the American Philosophical Society, Vol. 45, No. 184, pp. 183–242.

Van Gosen, B. S., 2007, The geology of asbestos in the United States and its practical applications: Environmental & Engineering Geoscience, Vol. 13, No. 1, pp. 55–68. Van Gosen, B. S. and Clinkenbeard, J. P., 2011, Reported Historic Asbestos Mines, Historic Asbestos Prospects, and Other Natural Occurrences of Asbestos in California: U.S. Geological Survey Open-File Report 2011-1188. Wakabayashi, J., 2011, Mélanges of the Franciscan Complex, California: Diverse structural setting, evidence for sedimentary mixing, and their connection to subduction processes in melanges: Processes of formation and societal signiﬁcance: Special Paper of the Geological Society of America, Vol. 480, pp. 117–141. Wakabayashi, J. and Unruh, J. R., 1995, Tectonic wedging, blueschist metamorphism, and exposure of blueschists: Are they compatible?: Geology, Vol. 23, No. 1, pp. 85–88. Wassmann, S. and Stöckhert, B., 2012, Matrix deformation mechanisms in HP-LT tectonic mélanges—Microstructural record of jadeite blueschist from the Franciscan Complex, California: Tectonophysics, Vol. 568, pp. 135–153.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 21–28

Does Exposure to Naturally Occurring Asbestos (NOA) During Dam Construction Increase Mesothelioma Risk? DANIEL W. HERNANDEZ* Dragados USA, Inc., Flatiron West, Inc., Sukut Construction, Inc., Joint Venture (DFSJV), Calaveras Dam Replacement Project, 8022 Pinot Noir Court, San Jose, CA 95125

Key Terms: NOA, Amphiboles, Mesothelioma, Blueschist, Glaucophane, Particle Dimensions ABSTRACT The Calaveras Dam Replacement Project, a major construction project completed in 2019, involved hundreds of workers using heavy earth-moving equipment and mining operations, including blasting, drilling, rock crushing, and other operations designed to move millions of cubic yards of earth. Much of the material was composed of serpentinite, blueschist, and other rocks that contain chrysotile and a variety of amphibole minerals, including glaucophane, winchite, actinolite, tremolite, and other asbestos-related amphiboles. This article explores the unique characteristics of the blueschist that required extensive protective measures to be undertaken by the contractor to protect workers and surrounding sensitive receptors. This article will provide an overall summary of the dimensional characteristics of the airborne blueschist elongate mineral particles encountered during construction activities to compare and contrast current understanding of cleavage fragments versus asbestiform mineral fibers.

INTRODUCTION Located to the east of the San Francisco Bay Area, the Calaveras Dam Replacement Project is substantially underlain by exhumed Franciscan mélange composed of a variety of rocks. Rocks encountered during construction included serpentinite, which contains chrysotile, blueschist, greenstone, and other rocks that contain various amphiboles. At least 20 species of amphiboles, including actinolite and glaucophane (considered naturally occurring asbestos [NOA] minerals [NIOSH, 2011]), were discovered within the project boundary. In addition, winchite (associated with the well-known Libby Amphibole [U.S. EPA, 2014]) was also determined to be present.

*Corresponding author email: risicaredan@gmail.com

The construction of the dam required the disturbance of over 13 million yards of these materials involving operations that included mining, drilling, blasting, sorting, sizing, dozing, excavating, loading, transporting, dumping, and compaction. Operations complementing the construction of the replacement dam included slope shaping, rock breaking, and rock crushing. Early in the project, the contractor noted that countable fibers (fibers with a 3:1 or greater length:width aspect ratio, widths greater than 0.25 μm, and length greater than 5 μm) were substantially higher in personal breathing zone exposure samples collected during the disturbance of the blueschist, as compared to during the disturbance of other NOA-containing media (serpentinites). This was somewhat surprising, since the “blueschist rock” did not appear to be asbestiform, as in the classical definition of asbestos as “visually fibrous,” with individual fibers being separable by finger pressure (Figure 1). Since the dam required the mining and processing of approximately one million cubic yards of blueschist for the upstream rock shell, the contractor conducted a thorough industrial hygiene investigation of blueschist operations. Based on thousands of air samples collected, it was determined that glaucophane (compositionally similar to crocidolite) and winchite were the dominant amphiboles emitted when disturbing blueschist materials. The blueschist amphibole mineral belongs in the solid solution series of gluacophaneriebekite, which is currently recognized by the National Institute for Occupational Safety and Health (NIOSH) as a non-asbestiform habit of an asbestos mineral and is considered “asbestos” if cleavage fragments derived from this material meet the definition of fiber, as viewed microscopically (NIOSH, 2011). Considerable issues are associated with the disturbance of amphibole mineral rocks since the asbestiform crystalline habit is relatively rare in nature (Harper et al., 2008; Wylie, 2017), as compared to that of the coarsely crystalline non-asbestiform analog of the same mineral.. Disturbance of these non-asbestiform mineral analogs may result in the release of “cleaved” elongate mineral particles and/or

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 29–33

Hernandez Table 1. Length characteristics of bulk blueschist samples.

Length

Mean (µm)

%>5 (µm)

% > 10 (µm)

VO cleavage fragments Cleavage fragments w/o riebeckite VO asbestiform Blueschist 1 Blueschist 2

3.36 3.2 11.99 4.2 5.4

10.9 7.9 55.4 23.8 42.4

1.2 0.75 30.3 1.6 7.1

VO = Van Orden.

Figure 1. Greenstone breccia with blueschist matrix.

individual non-asbestiform crystals that have countable morphologies under current counting rules (NIOSH, 2011). The primary controversy relates to the relative toxicity of these particles as compared to asbestos. Research (Aust, 2011) has generally demonstrated that the potency of asbestos may be related to its fiber dimensions, its persistence in the lung, and the surface characteristics of the fibers. In addition, researchers (Harper et al., 2008; Van Orden et al., 2016) have also suggested that dimensional differences exist between asbestos and non-asbestos amphibole samples and that these differences can be used to identify true “asbestiform” fibers. This article includes the examination of bulk material and emitted particle size distributions, with comparison of particle size characteristics to other research findings, as well as exposure data associated with the particle size distributions. BLUESCHIST BULK SAMPLE PARTICLE SIZES Various researchers have examined particle size distributions of non-asbestiform amphibole particles and asbestiform amphibole particles to determine differences and similarities in order to identify means which could be useful for differentiating each. Harper et al. (2008) suggested that a maximum width criterion of 1 µm for asbestos measurement purposes could adequately include asbestos while minimizing the counting of cleavage fragments. In an extensive study, Van Orden et al. (2016) compared the particle size distributions of five regulated asbestiform species to their non-asbestiform analogs. The findings of this study indicated that very clear dif-

ferences exist between length and width distributions of asbestiform versus non-asbestiform particles. However, overlapping dimensional similarities were also demonstrated. To compare to the ﬁndings of Van Orden et al. and Harper et al., blueschist bulk samples collected from the dam’s blueschist mining area were disaggregated using a jaw crusher and pulverizer (Asbestos TEM Labs, Berkeley, CA). Using Van Orden’s counting rules, particle sizing was then performed by ALS Labs (Cincinnati, OH). Tables 1 and 2 compare the length and width characteristics of the blueschist amphibole to the average length and width characteristics of cleavage fragments and asbestiform particles reported by Van Orden et al. (2016). As can be seen in Tables 1 and 2, blueschist bulk sample particle sizes are thinner and longer than the average cleavage fragment populations of the regulated non-asbestiform amphiboles investigated by Van Orden et al.. It is important to note from a health risk perspective that average cleavage fragment particle sizes reported in Tables 1 and 2 included non-asbestiform riebeckite (a sodic-amphibole), which biased average non-asbestiform population dimensions to longer lengths and thinner widths. Health risk implications are further discussed below. HIGH-RISK ELONGATE MINERAL PARTICLE (EMP) DIMENSIONS As previously mentioned, the potency of asbestos EMPs is generally related to particle dimensions, durability of the particle in the lung, mineral Table 2. Width characteristics of bulk blueschist samples.

Width

Mean (µm)

% < 0.5 (µm)

% < 1.0 (µm)

VO cleavage fragments Cleavage fragments w/o riebeckite VO asbestiform Blueschist 1 Blueschist 2

1.24 1.35 0.244 0.52 0.54

7.2 1.8 94.3 71.8 65.9

42.3 33.8 100 88.7 87.2

VO = Van Orden.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 29–33

Does NOA Exposure Increase Mesothelioma Risk?

Figure 2. Cumulative length and width distribution of airborne blueschist EMPs > 5 µm in length emitted during construction operations. Length Distribution includes % of airborne EMPs with lengths greater than 10 um.

composition, and surface characteristics of the EMP. With respect to the non-asbestiform analogs of regulated asbestos minerals, particle dimensions are generally different than those of asbestos, and human, animal, and other evidence indicate lower potency (Gamble and Gibbs, 2008; Mossman, 2008; and Wylie, 2016). However, the relative difference in potency (if any) between an amphibole asbestos fiber and an individual non-asbestiform crystal, or a cleaved EMP from the same non-asbestiform mineral given the same dimensions, has not been resolved. Further complicating these issues, researchers (Gamble and Gibbs, 2008) have indicated that complete particle size data (length versus diameter) on distributions of cleavage fragments and asbestos fibers are extremely limited in number, making it difficult to compare length and width distributions. Stanton et al. (1981) hypothesized that biological activity is related to the dimension and durability of the particle. He found that neoplastic responses correlated well with the dimensional distribution of fibers and that durable particles less than 0.25 µm in width and greater than 8 µm in length correlated most strongly. Other researchers (Wylie, 2016) have indicated that most mesothelioma-causing fibers are very thin and are greater than 10 µm in length and found that particles longer than 5 µm with widths of <0.33 µm and <0.4 µm have been shown to be retained by the lung and preferentially transported to the pleura. Berman et al. (1995) found that structures greater than 5 µm in length and less than 0.4 µm in width appear to contribute to lung tumor risk in laboratory rat inhalation studies. Wylie et al. (1993) suggested that all fiber populations of similar width, length, and crystal morphol-

ogy of asbestos should be viewed with caution and given deference with respect to biological testing. With respect to dimensional aspects, longer, thinner fibers are considered more potent than shorter, thicker fibers, and amphibole fibers are more bio-durable in lung tissues than are those of chrysotile. Other things to consider are that fiber width controls penetration into the lung and that fiber length may control residence time (Wylie, 2017). PARTICLE SIZE ANALYSIS OF BLUESCHIST AIRBORNE EMPS Perhaps the most important information gathered by the contractor at the project site is the enumeration and particle sizing of EMPs emitted into the breathing zone of workers disturbing the blueschist amphibole. To better understand potential health risks associated with disturbing the blueschist amphibole during construction operations, 17 breathing zone samples were collected from workers involved in the blueschist mining operations. Each sample was evaluated for total countable fibers in addition to particle size analysis. Structures greater than 1 µm in length and meeting a minimum length to width aspect ratio (AR) of 3:1 were counted and sized using transmission electron microscopy methods by ALS Labs. Analysis of each sample was terminated upon completion of the grid opening containing the 100th EMP. A total of 1,946 particles were sized. All workers involved in these operations were equipped with air-purifying respirators and Tyvek protective coveralls. Water was sprayed continuously as a dust suppressant. Mining operations included sorting, sizing, dozing, and loadout of bulk blueschist rock

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 29–33

Hernandez

Figure 3. Relative frequency of the aspect ratios of airborne blueschist EMPs > 5 µm in length.

to be transported to the upstream rock shell of the dam. Phase Contrast Microscopy (PCM) results for these breathing zone samples ranged from 0.078 f/cc to 1.2 f/cc, with a geometric mean of 0.512 f/cc. The CAL/OSHA permissible exposure limit (PEL) for regulated asbestos is 0.1 f/cc (as an 8-hour time-weighted average). Figure 2 depicts the length and width distributions of airborne EMPs from 17 breathing zone samples, and Figure 3 shows the relative frequency of aspect ratios of the airborne blueschist EMPs of >5 µm in length. As can be seen in Figure 3, 35 percent of airborne blueschist EMPs greater than 5 µm in length have aspect ratios greater than 20:1. Figure 4 depicts an electron micrograph of a single blueschist EMP with a width of 0.204 µm, length of 13.8 µm, and aspect ratio of 68:1.

CONCLUSIONS Bulk and air samples collected during mining of bulk blueschist rock show that blueschist EMPs at the dam construction site were found in the “Asbestiform Size Range”; thus, blueschist EMPs at this locality should be considered ﬁbrous rather than cleavage fragments. The toxicity of the blueschist amphibole is unknown, but based on particle sizes found in exposure samples, at least 30 percent of the airborne EMPs should be considered very high risk, and perhaps up to 40 percent should be considered high risk. As indicated by personal air monitoring results (reported greater than the CAL/OSHA PEL of 0.1 f/cc), disturbing this material required dust controls, personal protective equipment, and exposure monitoring

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 29–33

Does NOA Exposure Increase Mesothelioma Risk?

Figure 4. Electron micrograph of a single blueschist EMP.

to ensure that protective equipment ensembles were adequate for the tasks at hand. REFERENCES Aust, A. E., Cook, P. M., Dodson, R. F., 2011, Morphological and chemical mechanisms of elongated mineral particle toxicities: Journal Toxicology Environmental Health, Part B Critical Reviews, Vol. 14, No. (1–4), pp 40–75. Berman, D. W., Krump, K. S., Chatfield, E. J., Davis, J. M., Jones, A. D., 1995, The sizes, shapes, and minerology of asbestos structures that induce lung tumors in AF/Han rats following inhalation: Risk Analysis, Vol. 15, No. 2, pp. 181–195. Gamble, J. F. and Gibbs, G. W., 2008, An evaluation of the risks of lung cancer and mesothelioma from exposure to amphibole

cleavage fragments: Regulatory Toxicology Pharmacology, Vol. 52, No. 1, pp. S155–S186. Harper, M., Lee, E. G., Doorn, S. S., Hammond, O., 2008, Differentiating non-asbestiform amphibole and amphibole asbestos by size characteristics: Journal Occupational Environmental Hygiene, Vol. 5, No. 12, pp. 761–770. Mossman, B. T., 2008, Assessment of the pathogenic potential of asbestiform vs. non-asbestiform particulates (cleavage fragments) in in vitro (cell or organ culture) models and bioassays: Regulatory Toxicology Pharmacology, Vol. 52, No. 1, pp. S200–S203. National Institute for Occupational Safety and Health (NIOSH), 2011, NIOSH Bulletin 62, April 2011: Asbestos Fibers and Other Elongate Mineral Particles: State of the Science and Roadmap for Research: Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health. Cincinnatti, Ohio. Revised edition, pp. 33–34. Stanton, M. F., Layard, M., Tergeris, A., Miller, E., May, M., Morgan, E., Smith, A., 1981, Relation of particle dimensions to carcinogenicity in amphibole asbestos and other fibrous minerals: Journal National Cancer Institute, Vol. 67, No. 5, pp. 965–976. U.S. Environmental Protection Agency (U.S. EPA), 2014, Toxicological Review of Libby Amphibole Asbestos: Integrated Risk Information, National Center for Environmental Assessment, Office of Research and Development, U.S. EPA, Washington, DC. Van Orden, D. R., Lee, R. J., Hefferan, C. M., Schlaegle, S., Sanchez, M., 2016, Determination of the size distribution of amphibole asbestos and amphibole non-asbestos mineral particles: Microscope, Vol. 64, No. 1, pp. 3–15. Wylie, A. G., 2017, Chapter 2 Asbestos and fibrous erionite. In Testa, J. R. (Editor), Asbestos and Mesothelioma, Current Cancer Research: Springer International Publishing. doi:10.1007/978-3-319-53560-92. Gewerbestrasse 11, 6330 Cham, Switzerland. Wylie, A. G., 2016, Amphibole dusts: Fibers, fragments, and mesothelioma: Canadian Mineralogist, Vol. 54, No. 6, pp. 1403–1435. Wylie, A. G., Bailey, K. F., Kelse, J. W., Lee, R. J., 1993, The importance of width in asbestos fiber carcinogenicity and its implications for public policy: American Industrial Hygiene Association Journal, Vol. 54, No. 5, pp. 239–252.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 29–33

NOA Air-Quality Lessons Learned during Calaveras Dam Replacement Project BART EKLUND* AECOM Corporation, 9400 Amberglen Boulevard, Austin, TX 78729

JOHN ROADIFER NOEL WONG MICHAEL FORREST AECOM Corporation, Kaiser Center, 300 Lakeside Drive, Oakland, CA 94612

Key Terms: NOA, Asbestos, Air Quality, Emission Controls ABSTRACT The Calaveras Dam Replacement Project (CDRP) pioneered technical approaches for addressing community exposure to naturally occurring asbestos (NOA) via the inhalation pathway. Over the course of the CDRP, approaches were developed for key issues, including determining the NOA particles of interest, defining the toxicity limits to apply to various types of NOA particles, establishing dust control, and creating appropriate feedback loops for using laboratory data. Specific issues of interest included whether to count only structures above a certain length and the inhalation unit risk value to use for amphiboles. The knowledge gained on the CDRP can and is being used to optimize NOA evaluation and control at other, similar projects. INTRODUCTION The original Calaveras Dam was constructed in 1925 in Alameda County, CA. At the time, it was the largest earth-ﬁll dam in the world. The resulting reservoir was situated primarily in Santa Clara County and had a capacity of 100,000 acre-ft (120 million m3 ) and a surface area of approximately 1,450 acres (590 ha). It was determined that the hydraulic ﬁll dam was subject to liquefaction failure during seismic activity, and the capacity of the reservoir was restricted to approximately 40 percent capacity by California regulators in 2001. To restore lost capacity, the San Francisco Public Utilities Commission (SFPUC) constructed a replacement dam located approximately 1,000 ft (300 m) downstream of the original dam. This was called the Calaveras Dam Replacement Project (CDRP). * Corresponding author email: bart.eklund@aecom.com

Construction on the new earth- and rock-ﬁll dam began in August 2011, and construction activities were completed in late 2018. The new dam is 220 ft (67 m) high and 1,180 ft (360 m) wide at the base. The volume of the dam is 3.5 million yd3 (2.7 million m3 ). About 11 million yd3 (8.4 million m3 ) of material was excavated during construction, some of which contained naturally occurring asbestos (NOA). A broad deﬁnition of NOA was applied (i.e., California Air Resources Board Asbestos Hazard Emergency Response Act [CARB-AHERA] structures), as discussed later in this paper.

Air-Quality Program A comprehensive air-quality program was undertaken to address species of interest. Pre-construction air monitoring was performed to develop a baseline of local air quality. Twenty-four-hour, time-integrated samples generally were collected on a monthly basis. In total, 22 monthly monitoring events were performed from August 2008 through December 2009 and from May 2010 through November 2010. Routine air monitoring began in January 2012 and continued through November 2018. The work was done in compliance with a Comprehensive Air Monitoring Program (CAMP) document that served as the test plan (URS, 2012). The CAMP document was amended about a dozen times between October 2013 and December 2016 as new information became available and the results-to-date were reviewed and evaluated. The dam construction took place in a valley that primarily ran north-south. There were relatively few residents located within 5 mi (8 km) of the construction activities in either direction, but the largest number of potential receptors was to the north. The monitoring network included six community monitoring locations located to the north over an extended distance, and

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 35–38

Eklund, Roadifer, Wong, and Forrest

about 12 perimeter monitoring locations located closer in and surrounding the construction activities. One key element of the air-quality program was the development of target concentrations for NOA at the community and perimeter monitoring locations. Target concentrations were developed for both total asbestos and for amphiboles. These target concentrations were revised several times over the course of the program to account for changes in site activities (e.g., the addition of nightshift work) or changes in assumed exposures (e.g., whether or not receptors aged 0–5 years were present). The target concentrations for the community monitoring stations were based on total risk from all NOA sources, with the relative contribution from CDRP being up to 10 percent of the total risk. The target concentrations for the perimeter monitoring stations were based on results of atmospheric dispersion modeling and were set to ensure that NOA concentrations further downwind would not exceed acceptable levels. An attenuation factor approach was used to set perimeter trigger levels, which resulted in the location of the perimeter stations being an important factor in how conservative the trigger levels were for evaluating potential exposure of receptors further downwind. The modeling addressed multiple emission areas, various types of construction equipment, fugitive dust from vehicle traﬃc, and emission sources that changed in elevation over time. The modeling involved complex terrain and “ﬂagpole” receptors (i.e., receptors located above the elevation where air emissions were released). Key Air-Quality Issues There were six air-quality issues that posed technical or logistical challenges: 1. 2. 3. 4. 5.

define what to measure; select analytical approach; define the acceptable ambient air concentration; determine where to measure; determine the feasible emission-control options; and 6. establish a process for tracking and evaluating airquality data. Each of these issues is discussed, in turn, below. Define What to Measure The original scope of work addressed NOA, inhalable dust (PM10 ), and elements associated with particulate matter (arsenic, chromium, cobalt, copper, and nickel). It became apparent during the early stages of the regular air-monitoring effort that particulate

matter and the various metals and arsenic were not an issue. Monitoring and reporting of these species eventually was discontinued, and over the later stages of the project, NOA was the only air-quality species of concern. Select Analytical Approach From the beginning of the project, it was deemed that Occupational Safety and Health Administration (OSHA) methods would not be sufficiently protective. Under OSHA, an asbestos fiber is defined as a particle that is 5 μm or longer, with a length-to-width ratio of 3:1 or longer (OSHA, 1988, revised 1997). Furthermore, the fibrous minerals defined as asbestos are those listed as commercially exploitable mineral fibers (i.e., chrysotile, crocidolite, amosite, tremolite, actinolite, and anthophyllite). Analysis is typically by phasecontrast microscopy (PCM), and results are typically reported in units of fibers per cubic centimeter (f/cc). For this project, CARB-AHERA structures were measured. The CARB-AHERA approach counts both asbestos structures visible by PCM and those up to an order of magnitude smaller. CARB-AHERA structures are defined as being 0.5 μm or longer, with an aspect ratio of 3:1 or greater, and largely parallel sides. Data were reported as structures per cubic centimeter (s/cc). The analysis included characterization of mineral type (chrysotile or amphibole). Two additional analyses were performed. In one analysis, fibers that met the OSHA definition, with fiber lengths 5 μm, were counted separately. In the second analysis, detailed fiber size ratio analysis and/or identification of specific minerals present were done for up to 5 percent of the samples. Samples were analyzed using transmission electron microscopy (TEM). Approximately 40,000 samples were analyzed over the course of the project. The detection limit achieved was dependent on the air volume sampled and the NOA loading. The number of grid openings that were counted had an effect on the analysis cost and the turnaround time. Define the Acceptable Ambient Air Concentration The usual “gold standard” for risk-based inhalation concentrations is the U.S. Environmental Protection Agency (USEPA) IRIS database. The USEPA has published an inhalation unit risk of 0.23 (f/cc)−1 for asbestos in IRIS (USEPA, 1988). Stated another way, the one-in-a-million risk level is 4E−06 f/cc. This value applies to both chrysotile and amphiboles. For the CDRP, a more conservative approach was employed, where the EPA value for chrysotile was used but amphiboles were assigned 10 times greater risk

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 35–38

Air Quality during Calaveras Dam Project

Figure 2. Air emissions of particulate matter near perimeter monitoring station P11 (note water spray used for dust control).

Figure 1. Perimeter monitoring locations in relation to work areas, where perimeter monitoring locations are shown as red dots, work areas are shown as shaded areas, and residences are shown as yellow dots.

than chrysotile. This entailed some risk of project shutdown, as it was not apparent early on whether the stricter target concentrations could be met. Some additional uncertainty was present because the target concentrations were based on atmospheric dispersion modeling of particulate matter. Asbestos could not be modeled directly because there were no emission factors for asbestos (e.g., emission rate per hour of dozer operation). Deﬁne Where to Measure The perimeter monitoring network is shown in Figure 1. There were relatively few receptors near the work site, as previously discussed, so the community monitoring sites extended over several miles. One site eventually was moved to avoid a localized emission source (i.e., gravel parking lot) that contributed to anomalous results. One of the perimeter monitoring locations also was moved to avoid a localized emission source (i.e., re-entrainment from a vehicle

washing station). Some of the work activities took place quite close to the perimeter monitoring locations, and it is likely that the monitoring results were biased high on those days (relative to downwind concentrations). Figure 2 shows an example of work being performed adjacent to a perimeter monitoring station. The community monitors proved to be less useful than originally anticipated. The baseline air monitoring took place during an economic recession when home building and construction activities were at a low point. As the economy recovered, increased activity resulted in higher levels of NOA being generated from non-CDRP sources than were encountered during the baseline period. This became apparent, in part, through evaluation of air-monitoring data collected on days when there was no construction activity at the CDRP site (e.g., Sundays) and comparison of the specific minerals present for monitors near the site versus monitors farther away. Therefore, it was difficult to readily distinguish the CDRP contribution to the total NOA loading at distant locations. Determine the Feasible Emission-Control Options At the start of the project, there were no data available on the control options that would best address NOA and the level of effectiveness to expect. The available information was for the effectiveness of water sprays with and without additives for dust control (USEPA, 1992). Water sprays were employed extensively at the site for dust control and for controlling serpentinites, but they were found to be of limited effectiveness for amphiboles. Increasing the water usage did not lead to a commensurate reduction in amphibole concentrations. Various water-delivery systems were tried (e.g., DustBoss, Kuma) in an effort to achieve better NOA control. It is known, however, that the water droplets need to be about the same size as the particles to be

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 35–38

Eklund, Roadifer, Wong, and Forrest

controlled for best effectiveness. If the droplets are too big, the dust particles will tend to move around the droplet and not collide with the droplet. Given the very small size of NOA structures, we believe the size difference between the structures and the smallest mist or water droplets that could be generated was the reason why the effectiveness did not meet expectations. Air emissions could be controlled or limited using standard options for construction sites, such as minimizing work under high wind conditions and adjusting the rate of production. These options were employed sparingly, given logistical and schedule constraints. Establish a Process for Tracking and Evaluating AirQuality Data The original plan was to use the off-site analytical results to guide on-site activities. This did not prove to be practical for several reasons. First, the time lag between sample collection and receipt of laboratory results meant that different activities might be under way at the site when results were received compared to when the data were generated. Second, the options for controlling NOA were limited, as discussed above. Knowing that NOA emissions were higher for a given activity was of limited help for addressing off-site, non-occupational exposures if there were not adequate means to mitigate those emissions. At many sites, it has been found that visual observations and real-time measurements of dust proved to be the best tools for on-site management of particulate matter air emissions (Eklund et al., 2014). The off-site analytical results still serve the important purpose of documenting the air quality over time. For the Calaveras project, however, visual observations and real-time measurements proved to be less useful than at other sites, due to the small size of the NOA particles versus particle size fractions such as PM2.5 or PM10 , which can be visible and are more amenable to real-time measurement using opacity-based or beta attenuation monitors. There was an intensive, on-going effort to evaluate the data as they were collected, and this proved to be crucial for understanding the fate and transport of NOA. This work demonstrated that activities such as blasting, which were expected to be major sources of NOA, were of too-limited duration to be significant in the overall accounting of air emissions. Although higher concentrations were measured downwind when blasting occurred, the blasting represented 1 percent of the overall time.

Relatively early in the project, elevated NOA concentrations were measured during a brief period when site activities were at a high level and the meteorological conditions were very unfavorable for atmospheric mixing and dispersion. The on-going data evaluation helped to show that this time period was an outlier and unlikely to be repeated. The ability to predict future NOA concentrations became more refined over time and provided some assurance over the final years of the project that air quality was under control and would continue to be within acceptable limits. The emission predictions took into account the number of hours worked per week as well as the type of activity, the level of activity, and the location of work. CONCLUSIONS NOA can pose large challenges for maintaining acceptable air quality at major construction projects. CDRP expended significant resources to better understand the fate and transport of NOA. The state of knowledge was much greater at the end of the project than at the start because of these efforts. It is not possible to achieve complete control of amphibole emissions for these types of large-scale construction projects. This needs to be taken into account during planning of future, similar work. A multi-disciplinary team proved to be an effective approach for dealing with the multi-media, complex issues posed by this work. Expertise in asbestos, construction practices, dust control, air-quality monitoring, etc., was utilized from an early stage in the process. REFERENCES Eklund, B.; Fitzgerald, C.; Wade, M.; Wilder, R.; and LaMond, D., 2014, Controlling air toxics emissions from remediation by monitoring of surrogate parameters: Remediation, Vol. 24, No. 4, pp. 127–138. Occupational Safety and Health Administration (OSHA), 2019, Method No. ID-160. Asbestos in Air, July 1988 (Revised July 1997): Electronic document available at: https://www. osha.gov/dts/sltc/methods/inorganic/id160/id160.pdf. URS Corporation, 2012, Comprehensive Air Monitoring Program (CAMP), Calaveras Dam Replacement Project: Report to the San Francisco Public Utilities Commission, San Francisco, CA. U.S. Environmental Protection Agency (USEPA), 1988, Integrated Risk Information System (IRIS), Chemical Assessment Summary, Asbestos: CASRN 1332-21-4: USEPA. Washington, DC. U.S. Environmental Protection Agency (USEPA), 1992, Control of Air Emissions from Superfund Sites: USEPA, Center for Environmental Research Information Report EPA/625/R92/012.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 35–38

Naturally Occurring Asbestiform Minerals in Italian Western Alps and in Other Italian Sites ELENA BELLUSO* Department of Earth Sciences, University of Torino, Via Valperga Caluso no. 35, 10125, Torino, Italy Interdepartmental Centre “G. Scansetti” for Studies on Asbestos and Other Toxic Particulates, University of Torino, Torino, Italy Institute of Geosciences and Earth Resources–IGG, National Research Council–CNR of Italy

ALAIN BARONNET Centre Interdisciplinaire de Nanosciences de Marseille, Université Aix-Marseille, France; Campus de Luminy, Case 913, 13288, Marseille, France

SILVANA CAPELLA Department of Earth Sciences, University of Torino, Via Valperga Caluso no. 35, 10125, Torino, Italy Interdepartmental Centre “G. Scansetti” for Studies on Asbestos and Other Toxic Particulates, University of Torino, Torino, Italy

Key Terms: NOA, Naturally Occurring Non-Asbestos Classified Asbestiform Minerals (NONA), TEM-EDS Identifications, Fiber Intergrowth ABSTRACT The natural occurrence of asbestos (NOA) in rocks and soil has been known for many years in several areas of the world, differently from the natural presence of asbestiform minerals. In Italy, the mapping of NOA is mandatory according to the 2001 and 2003 regulations. An investigation, not yet concluded, has revealed that in Italy, NOA is represented by chrysotile and tremolite asbestos with minor amounts of actinolite asbestos and anthophyllite asbestos. A field survey conducted in the Italian Western Alps (IWA), dealing with the natural occurrence of asbestiform minerals non-asbestos classified and not regulated, started many years ago and is still ongoing. It revealed that the following kinds of asbestiform silicates are present (in decreasing order of frequency): asbestiform polygonal serpentine and asbestiform antigorite, asbestiform diopside, asbestiform carlosturanite, asbestiform forsterite, asbestiform sepiolite, asbestiform balangeroite, and asbestiform talc. The asbestiform non-silicates brugnatellite and brucite have been rarely detected. Outside the IWA, asbestiform zeolite (erionite and offretite), asbestiform sodium amphibole (fluoro-edenite), and a few other asbestiform silicates have been also detected. For some asbestiform *Corresponding author email: elena.belluso@unito.it

minerals, the identification is problematic and needs the use of transmission electron microscopy combining imaging at high magnification and electron diffraction and chemical data. This investigation is particularly important to distinguish four kinds of asbestiform minerals (antigorite, polygonal serpentine, carlosturanite, and balangeroite) from chrysotile since only the last one is regulated. The issue is much more complicated by the intergrowth of different fibrous species on the submicrometer scale. INTRODUCTION In recent years and in several countries, health investigations dealing with asbestos have moved from occupational to environmental exposure (e.g., Baumann et al., 2015; Abakay et al., 2016; and Noonan, 2017). According to some authors, low but continuous exposure, as in the case of inhabitants of houses next to asbestos-bearing rock outcrops (i.e., naturally occurring asbestos [NOA]), could cause health problems (e.g., Luce et al., 2000; Bernardini et al., 2003). The same problem could concern the exposure to naturally occurring non-asbestos classified asbestiform minerals (NONA). For some of them, the carcinogenicity to high-dose exposure is known, as in the case of asbestiform fluoeredenite, asbestiform erionite, asbestiform winchite, and asbestiform richterite (e.g., Burragato et al., 2005; International Agency for Research on Cancer, 2012). In addition to these last minerals, many others with a similar asbestiform morphology have

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Belluso, Baronnet, and Capella

been discovered over the years in diﬀerent parts of the world. The most striking current example is represented by asbestiform antigorite. For many years, this antigorite asbestiform variety has not been recognized as having a full identity, and little information has appeared in publications (e.g., Keeling et al., 2006). But, exactly as it happens for NOA, NONA can be dispersed in air by natural causes (e.g., weathering, natural atmospheric agents, landslides) and by anthropogenic causes (e.g., excavation works). If these asbestiform minerals are noxious following continuous low dose and/or sporadic high doses, it will be known only in many years, similarly to asbestos given the long latency time (10–40 years) of related asbestos pathologies. For instance, it would be good practice to map the presence of these minerals in every country. As it concerns Italy, the mapping of NOA is mandatory according to the March 23, 2001, law, no. 93, and the related March 18, 2003, Environment Ministry Decree, no. 101. The mapping of NOA must be carried out by each region through its own environmental protection agency. Information including the following must be reviewed by each region:

r r r r

r r r r r

literature data; geological maps; historical research permits and mining concessions; reports and environmental monitoring (carried out as part of environmental and strategic impact assessment procedures for the construction of infrastructural works); analytical certiﬁcates; mining activities of lithotypes suspected for the presence of asbestos; surveys for the Geological Cartography project at a 1:50,000 scale; speciﬁc investigations and surveys; possible inspections and sometimes collection of samples and laboratory analyses.

Each year, the individual environmental protection agencies send any possible additional data to the Ministry of Health, which integrates those data into the evaluation of necessary remediation works. Currently, the map of Italian NOA is published by the Ministry of the Environment and Protection of the Territory and the Sea (2018); it was last updated in 2018 and reports NOA sites from three regions, but to the current state of scientiﬁc knowledge, NOA is present in at least seven other regions. This article describes the diﬀerent NOA and NONA detected thus far in Italy.

DETECTED NOA AND NONA Definition and Analytical Techniques According to the dimensional definition of the World Health Organization (1997), many regulatory agencies, and the literature (e.g., Belluso et al., 2017), in this note we use the definitions listed below.

r Fiber: inorganic particle with length ࣙ 5 μm, width

ࣘ 3 μm, length/width (aspect ratio) ࣙ 3:1, parallel sides when seen in two dimensions, perpendicularly to fiber axis. r Fibril: a single mineral fiber that cannot be further separated longitudinally into smaller components (without losing the fibrous properties or appearances). r Asbestiform: adjective for fibrous non-asbestos classified having the “fiber” dimensions and at least one of the asbestos properties, such as flexibility, splitting, and so on. r Fiber bundle: parallel aggregate of mineral fibers. In the scientific literature, much research exists on the topic of NOA. Some research has been carried out specifically for mineralogical interests and others for petrological, structural engineering, or geological studies, and sometimes the scientific investigations arise from mineral collectors. In any case, their identification is conducted by using different techniques, in some cases by using two or more in a complementary way. The most used identification techniques are X-ray powder diffractometry (XRPD), infrared spectroscopy (with Fourier transform), optical microscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM), both coupled with energy-dispersive spectrometry (EDS). Occurrences of NOA and NONA Based on scientific literature data, NOA is widespread in many areas of Italy and represented by chrysotile and tremolite asbestos (both very widespread), the less widespread actinolite asbestos, and, finally, anthophyllite asbestos, which is the much less diffused than others (e.g., Cavallo and Rimoldi, 2013; Gaggero et al., 2013, 2017; and Vignaroli et al., 2013). Figure 1 shows the 10 regions where NOA is present in rocks at the current state of knowledge. Crocidolite and amosite have not been found in Italy. Unlike NOA, the mapping of NONA is not mandatory in Italy. Several spot investigations, completed during asbestos investigations and specific mineralogical research, show that NONA is present in many Italian areas (eight Italian regions) and abundant in some places.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Naturally Occurring Asbestiform Minerals

Figure 1. Italian regions and mineral species of naturally occurring asbestos (NOA) in rocks: chrysotile, tremolite asbestos, less diﬀused actinolite asbestos, and rare anthophyllite asbestos. In the lower left corner, the world map with Italy in red is shown.

Italian Western Alps Regarding the Italian Western Alps (IWA) regions, that is, the Piedmont and the Aosta Valley, an ongoing ﬁeld survey of NOA has been conducted since 1980 (Belluso et al., 1995; Baronnet and Belluso, 2002; Leone, 2018; and Paccagnella, 2018). More than 300 samples have been collected and analyzed by us-

ing XRPD coupled with SEM-EDS and/or TEMEDS. NOA forms of chrysotile, asbestos tremolite, and, in lesser amounts, asbestos actinolite have been detected. The investigation in this area revealed not only that NOA is present but also that there are many kinds of asbestiform silicates (sometimes in very high amounts) of NONA not regulated in Italy. Ten are the identiﬁed

Table 1. List of naturally occurring non-asbestos classified asbestiform minerals (NONA) in decreasing order of finding frequency, detected in Piedmont and Aosta Valley regions, northwestern Alps, Italy, and the ideal chemical formula. NONA

Ideal Chemical Formula

Asbestiform polygonal serpentine, asbestiform antigorite Asbestiform diopside Asbestiform carlosturanite Asbestiform forsterite Asbestiform balangeroite Asbestiform sepiolite Asbestiform talc Asbestiform brugnatellite Asbestiform brucite

Mg3 Si2 O5 (OH)4 CaMgSi2 O6 (Mg,Fe,Ti)21 [Si12 O28 (OH)4 ](OH)30 •H2 O Mg2 SiO4 (Mg,Fe,Mn)21 O3 (OH)20 (Si4 O12 )2 Mg4 Si6 O15 (OH)2 •4H2 O Mg3 Si4 O10 (OH)2 Mg6 Fe(CO3 )(OH)13 •4H2 O Mg(OH)2

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Belluso, Baronnet, and Capella Table 2. Parallel intergrowths of asbestos and asbestiform minerals after investigations by transmission electron microscopy with–energy dispersive spectrometry. The most abundant asbestiform minerals (main) and the intergrown minerals (subordinate) in the same bundle are listed in the left column and in the right column, respectively (asbestos is indicated in italic, Italian legislation 277/91). Main Mineral Chrysotile + asbestiform polygonal serpentine Asbestiform antigorite Asbestos tremolite (asbestos actinolite) Asbestiform diopside

Asbestiform carlosturanite Asbestiform forsterite Asbestiform balangeroite Asbestiform brugnatellite Asbestiform brucite Asbestiform sepiolite

Subordinate Minerals Asbestiform antigorite, asbestiform diopside, asbestiform carlosturanite, asbestiform forsterite, asbestiform balangeroite Chrysotile + asbestiform polygonal serpentine, asbestos tremolite, asbestiform carlosturanite Chrysotile + asbestiform polygonal serpentine, asbestiform antigorite, asbestiform talc Chrysotile + asbestiform polygonal serpentine, asbestiform antigorite, asbestiform carlosturanite, asbestiform balangeroite, asbestiform brugnatellite Chrysotile + asbestiform polygonal serpentine, asbestiform diopside, asbestiform antigorite, asbestiform forsterite, asbestiform brucite Chrysotile + asbestiform polygonal serpentine, asbestiform carlosturanite Chrysotile + asbestiform polygonal serpentine, asbestiform diopside Asbestiform diopside Asbestiform diopside “Organic matter”

NONA, and eight of these are magnesium-containing silicates. The complete list, in decreasing order of frequency, is shown in Table 1. Notice that asbestiform antigorite is detected with very high frequency and that asbestiform brugnatellite and asbestiform brucite, the only two non-silicate asbestiform minerals, are rare.

As far as quantities are concerned, asbestiform carlosturanite and asbestiform balangeroite are very abundant. Asbestiform antigorite and asbestiform polygonal serpentine are not possible to quantify because these two minerals are always intergrown with other ﬁbrous asbestos and non-asbestos minerals (Table 2).

Figure 2. Italian regions and mineral species of naturally occurring non-asbestos classiﬁed asbestiform minerals (NONA) in rocks: asbestiform antigorite, asbestiform Ca-erionite with Ca-levyne, asbestiform F-edenite, asbestiform gedrite, asbestiform Mg-horneblende, asbestiform oﬀretite, asbestiform polygonal serpentine, and asbestiform sepiolite. For the Aosta Valley and Piedmont regions, details are in the text.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Naturally Occurring Asbestiform Minerals

Figure 3. Macroscopic similarity of mineral ﬁber bundle. (A) Vein ﬁlled with asbestiform antigorite (indicated by the yellow arrow). (B) Fibrous bundles of asbestiform antigorite. (C) Fibrous bundles of chrysotile.

Other Regions The NONA identiﬁed in other Italian regions beyond IWA are the following (in alphabetical order): asbestiform antigorite, asbestiform Ca-erionite (with

Ca-levyne), asbestiform F-edenite, asbestiform gedrite, asbestiform Mg-hornblende, asbestiform oﬀretite, asbestiform polygonal serpentine, and asbestiform sepiolite (e.g., Cattaneo et al., 2011; Bloise et al., 2014, 2016, 2017; Giordani et al., 2017; Lucci

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Belluso, Baronnet, and Capella

(EDS). If HTEM images are not obtainable, to distinguish serpentine minerals, the use of quantitative analysis of SAED and of chemical data is needed. This investigation is particularly important to distinguish the four kinds of NONA, specifically asbestiform antigorite (Figure 4), asbestiform polygonal serpentine (Figure 4), asbestiform carlosturanite, and asbestiform balangeroite, from chrysotile (Figure 4). Only the last one, chrysotile, is regulated as asbestos. The issue is very complicated because usually, on a scale of a few thousandths or even hundredths of micrometers, the fibers of two to four mineral species are intergrown (because of solid-state transformations), as can be seen in Figure 4. Based on the detailed investigation by TEM-EDS, practically all bundles of fibers turn out to be made up of at least two fibrous mineral species, except in the case of asbestiform sepiolite, which appears to have only intergrown with an organic phase (aliphatic hydrocarbons; Giustetto et al., 2014) (Table 2). CONCLUSIONS

Figure 4. Transmission electron microscopic image of cross section of an asbestiform composite made of axially textured intergrown ﬁbrils of asbestiform antigorite (atg) with minor asbestiform polygonal serpentine (PS) and chrysotile (chrys).

et al., 2018; and Mattioli et al., 2018). The Italian regions where these asbestiform minerals have been detected are shown in Figure 2. Difficulties and Necessity to Identify the Minerals For some asbestiform minerals, certain identification is not easy because they show similar (at times even equal) characteristics on a macroscopic scale. For NOA and NONA in IWA regions, in most cases it is not possible distinguish among chrysotile, asbestiform antigorite, asbestiform polygonal serpentine, asbestiform carlosturanite, and asbestiform balangeroite on the microscopic scale. Figure 3 shows an example of this similarity on the eye scale, both in outcropping rocks and after picking fibers from the rock, between asbestiform antigorite (Figure 3a and b) and chrysotile (Figure 3c). The bundles of fibers are flexible, and their characteristics are very similar to each other. Also, the color is similar because chrysotile, which contains a little amount of iron, is not white but cream to light brown colored. TEM is often needed for identification, combining imaging at high magnification (HTEM), selected area electron diffraction (SAED), and chemical data

In Italy, NOA and NONA are widespread in several areas. NOA in Italy is represented by only four mineral species; among these, only chrysotile and tremolite asbestos are abundant and diffused, whereas actinolite asbestos is less diffused, and anthophyllite asbestos is rare. NONA in Italy includes 15 mineral species, many more than NOA, although they do not have the same diffusion and abundance at the current state of knowledge. Detailed investigations on NOA and NONA, although not yet completed, have been conducted in the Piedmont and Aosta Valley regions. These investigations revealed many different NONA in rocks and, in some cases, in the same bundle of fibers. Because of the intergrowth on the sub-micrometric scale of different fibrous species, many times only a TEM-EDS investigation allows us to detect all the mineral species. For certain phases, only this kind of technique allows us to obtain a certain identification, for example, to distinguish among asbestiform antigorite, asbestiform polygonal serpentine, and chrysotile and therefore non-asbestos and asbestos classified minerals. Finally, many issues have arisen from the data identifying the relevance of correct and representative sampling, the importance of the use of the technique most suitable to identify the content of these composite materials (i.e., the bundle of fibers) and the need for extended surveys and extensive investigations where there are rocks that may contain mineral fibers. The investigation using TEM-EDS on fiber bundles collected around the world may provide surprises such as those from the IWA rocks.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Naturally Occurring Asbestiform Minerals

ACKNOWLEDGMENTS Skillful technical assistance by S. Nitsche and F. Quintric of the “Centre Interdisciplinaire de Nanoscience de Marseille–FR” is gratefully acknowledged. This article benefited from constructive improvements by three reviewers. REFERENCES Abakay, A.; Tanrikulu, A. C.; Ayhan, M.; Imamoglu, M. S.; Taylan, M.; Kaplan, M. A.; and Abakay, O., 2016, High-risk mesothelioma relation to meteorological and geological condition and distance from naturally occurring asbestos: Environmental Health and Preventive Medicine, Vol. 21, pp. 82–90. Baronnet, A. and Belluso, E., 2002, Microstructures of the silicates: Key information about mineral reactions and a link with the Earth and materials sciences: Mineralogical Magazine, Vol. 66, pp. 709–735. Baumann, F.; Buck, B. J.; Metcalf, R. V.; McLaurin, B. T.; Merkler, D. J.; and Carbone, M., 2015, The presence of asbestos in the natural environment is likely related to mesothelioma in young individuals and women from Southern Nevada: Journal of Thoracic Oncology, Vol. 10, pp. 731–737. Belluso, E.; Cavallo, A.; and Halterman, D., 2017, Crystal habit of mineral fibres: European Mineralogical Union Notes in Mineralogy, Vol. 18, pp. 65–109. Belluso, E.; Compagnoni, R.; and Ferraris, G., 1995, Occurrence of asbestiform minerals in the serpentinites of the Piemonte Zone, Western Alps. In Politecnico di Torino (Editor), Giornata di Studio in ricordo del Prof. Stefano Zucchetti: Politecnico di Torino Editor, Torino, Italy, pp. 57–64. Bernardini, P.; Schettino, B.; Sperduto, B.; Giannandrea, F.; Burragato, F.; and Castellino, N., 2003, Three cases of pleural mesothelioma and environmental pollution with tremolite outcrops in Lucania: Giornale Italiano Medicina Lavoro Ergonomia, Vol. 25, pp. 408–411. Bloise, A.; Catalano, M.; Critelli, T.; Apollaro, C.; and Miriello, D., 2017, Naturally occurring asbestos: Potential for human exposure, San Severino Lucano (Basilicata, Southern Italy): Environmental Earth Sciences, Vol. 76, pp. 648–660. Bloise, A.; Critelli, T.; Catalano, M.; Apollaro, C.; Miriello, D.; Croce, A.; Barrese, E.; Liberi, F.; Piluso, E.; Rinaudo, C.; and Belluso, E., 2014, Asbestos and other fibrous minerals contained in the serpentinites of the Gimigliano-Mount Reventino Unit (Calabria, S-Italy): Environmental Earth Sciences, Vol. 71, pp. 3773–3786. Bloise, A.; Punturo, R.; Catalano, M.; Miriello, D.; and Cirrincione, R., 2016, Naturally occurring asbestos (NOA) in rock and soil and relation with human activities: The monitoring example: Italian Journal of Geosciences, Vol. 135, pp. 268–279. Burragato, F.; Comba, P.; Baiocchi, V.; Palladino, D. M.; Simei, S.; Gianfagna, A.; Paoletti, L.; and Pasero, R., 2005, Geo-volcanological, mineralogical and environmental aspects of quarry materials related to pleural neoplasm in the area of Biancavilla, Mount Etna (Eastern Sicily, Italy): Environmental Geology, Vol. 47, pp. 855–868. Cattaneo, A.; Rossotti, A.; Pasquaré, G.; Somigliana, A.; and Cavallo, D. M., 2011, Analysis of fibrous zeolites in the vol-

canic deposits of the Viterbo Province, Italy: Environmental Earth Sciences, Vol. 63, pp. 861–871. Cavallo, A. and Rimoldi, B., 2013, Chrysotile asbestos in serpentinite quarries: A case study in Valmalenco, Central Alps, Northern Italy: Environmental Sciences: Processes Impacts, Vol. 15, pp. 1341–1350. Gaggero, L.; Crispini, L.; Isola, E.; and Marescotti, P., 2013, Asbestos in natural and anthropic ophiolitic environments: A case study of geohazards related to the Northern Apennine ophiolites (Eastern Liguria, Italy): Ofioliti, Vol. 38, pp. 29–40. Gaggero, L.; Sanguineti, E.; Gonzalez, A. Y.; Militello, G. M.; Scuderi, A.; and Parisi, G., 2017, Airborne asbestos ﬁbres monitoring in tunnel excavation: Journal of Environmental Management, Vol. 196, pp. 583–593. Giordani, M.; Mattioli, M.; Ballirano, P.; Pacella, A.; Cenni, M.; Boscardin, M.; and Valentini, L., 2017, Geological occurrence, mineralogical characterization, and risk assessment of potentially carcinogenic erionite in Italy: Journal of Toxicology and Environmental Health, Part B, Critical Reviews, Vol. 20, pp. 81–103. Giustetto, R.; Seenivasan, K.; and Belluso, E., 2014, Asbestiform sepiolite coated by aliphatic hydrocarbons from Perletoa, Aosta Valley Region (Western Alps, Italy): Characterization, genesis and possible hazards: Mineralogical Magazine, Vol. 78, pp. 919–940. International Agency for Research on Cancer, 2012, Asbestos (chrysotile, amosite, crocidolite, tremolite, actinolite, and anthophyllite). In Evaluation of Carcinogenic Risks to Humans. A Review of Human Carcinogens: Monographs, Vol. 100, Part C: Arsenic, Metals, Fibres, and Dusts, International Agency for Research on Cancer, Lyon, France, pp. 219–309. Keeling, J. L.; Raven, M. D.; and McClure, S. G., 2006, Identification of Fibrous Mineral from Rowland Flat Area, Barossa Valley, South Australia: Report Book 2006/00002, Primary Industries and Resources South Australia, Adelaide, Australia. 23 p. Leone, C., 2018, Mineralogical and petrographic characterization of fibrous minerals and of their hosting rocks exposed in the northwestern Voltri Group and Sestri-Voltaggio Zone, Alessandria province: Unpublished thesis, University of Torino, Torino, Italy, 180 p. Lucci, F.; Della Ventura, G.; Conte, A.; Nazzari, M.; and Scarlato, P., 2018, Naturally occurring asbestos (NOA) in granitoid rocks: A case study from Sardinia (Italy): Minerals, Vol. 8, pp. 442–464. Luce, D.; Bugel, I.; Goldberg, P.; Goldberg, M.; Salomon, C.; Billon-Galland, M. A.; Nicolau, J.; Quénel, P.; Fevotte, J.; and Brochard, P., 2000, Environmental exposure to tremolite and respiratory cancer in New Caledonia: A casecontrol study: American Journal of Epidemiology, Vol. 151, pp. 259–265. Mattioli, M.; Giordani, M.; Arcangeli, P.; Valentini, L.; Boscardin, M.; Pacella, A.; and Ballirano, P., 2018, Prismatic to asbestiform oﬀretite from Northern Italy: Occurrence, morphology and crystal-chemistry of a new potentially hazardous zeolite: Minerals, Vol. 8, pp. 69–84. Meeker, G. P.; Bern, A. M.; Brownfield, I. K.; Lowers, H. A.; Sutley, S. J.; Hoefen, T. M.; and Vance, J. S., 2003, The composition and morphology of amphiboles from the Rainy Creek Complex, near Libby, Montana: American Mineralogist, Vol. 88, pp. 1955–1969. Ministry of the Environment and Protection of the Territory and the Sea, 2018, Asbestos Mapping: Electronic

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Belluso, Baronnet, and Capella document, available at https://www.minambiente.it/sites/ default/ﬁles / boniﬁche / Mappatura_amianto / mappatura_ amianto_2018.pdf Noonan, C. W., 2017, Environmental asbestos exposure and risk of mesothelioma: Annals of Translational Medicine, Vol. 5, pp. 234–244. Paccagnella, G., 2018, Mineralogical and petrographical characterization of asbestiform phases and of their host rocks from the north-western Voltri Massif: Unpublished thesis, University of Torino, Torino, Italy, 182 p.

Vignaroli, G.; Ballirano, P.; Belardi, G.; and Rossetti, F., 2013, Asbestos ﬁbre identiﬁcation vs. evaluation of asbestos hazard in ophiolitic rock mélanges, a case study from the Ligurian Alps (Italy): Environmental Earth Sciences, Vol. 72, pp. 3679–3698. World Health Organization, 1997, Determination of Airborne Fibre Number concentrations; a Recommended Method, by Phase Contrast Optical Microscopy (Membrane Filter Method): World Health Organization, Geneva, Switzerland, 53 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 39–46

Naturally Occurring Asbestos in Valmalenco (Central Alps, Northern Italy): From Quarries and Mines to Stream Sediments ALESSANDRO CAVALLO* Department of Earth and Environmental Sciences, University of Milano-Bicocca, Milano, Italy

JASMINE RITA PETRIGLIERI Department of Chemistry, University of Torino, Torino, Italy

Key Terms: NOA, Asbestos, Serpentinite, Quarries, Soils, Sediments ABSTRACT The Valmalenco area (central Alps, northern Italy) is an excellent case study for naturally occurring asbestos (NOA) because of the huge outcrops of serpentinites and widespread quarrying and mining activities. Extensive sampling of rocks, soils, stream sediments, and airborne asbestos has been in progress since 2004. The combined use of scanning electron microscopy and transmission electron microscopy has proven to be effective for the correct discrimination between asbestiform and non-asbestiform mineralogical varieties (but falling into the World Health Organization fiber definition), whereas phase contrast microscopy has not proven suitable because of the very small size of fibrils after strong mechanical fragmentation. The quantitative analysis of “massive” samples (rocks, soils, and sediments) requires accurate and representative sampling as well as specific counting and discrimination criteria to determine NOA. Over a decade of experience has allowed us to identify critical issues and adopt effective preventive measures. INTRODUCTION The Valmalenco area (central Alps, northern Italy) is characterized by huge outcrops of serpentinites of the Malenco nappe (lower crust–mantle complex) at the Penninic to Austroalpine boundary zone. The Malenco nappe covers an area of about 130 km2 (Figure 1) and consists mainly of ultramafic rocks, especially schistose serpentinites, showing various degrees of deformation and serpentinization, ranging from massive, layered lherzolites to schistose, completely serpentinized rocks. The predominant

*Corresponding author email: alessandro.cavallo@unimib.it

rock type is a schistose serpentinite with nonpseudomorphic texture, and the rock-forming minerals are antigorite, olivine, diopside and minor magnetite, chlorite, and chrysotile (only in veins). The area is characterized by four main metamorphic events, and minerals such as antigorite, olivine, clinopyroxene, amphibole, and carbonates occur in several generations (Münterer & Hermann, 1996). Field data demonstrate that the Malenco ultramafics formed the lithospheric subcontinental mantle below the Margna basement during pre-Alpine, postVariscan times (Münterer et al., 2000). Many different mining and quarrying activities have been active in the past or still are being undertaken (Figure 1). Long-fiber chrysotile asbestos, which occurs in discrete cross-fiber and slip-fiber veins (Figure 2), gave rise to widespread asbestos mining, particularly between the end of the 19th century and 1975, and was used mainly for weaving tablecloths or for candle wicks. A big boost to mining activity occurred during World War II and immediately thereafter, with annual production up to 670 tons and more than 400 workers employed, until mining operations ended completely in 1975, leaving huge amounts of mining waste and tailings. The serpentinite is a well-known ornamental and building stone, a green “marble,” extracted at least since the 11th century, exported to Switzerland before the 18th century, and abundantly used in Sondrio and Valtellina since the 14th century. Today the Malenco serpentinite is appreciated and exported all over the world for its excellent chromatic and technical features. There are more than 20 quarrying enterprises in the Malenco valley, processing approximately 195 kt/yr of serpentinite, with more than 150 workers involved. The quarried serpentinites show various textures (schistose to massive) and color shades. The schistose variety is prevalently split in thin slabs for roof covering, whereas the “massive” serpentinite is processed in many ways: dressed, polished, bush hammered, and sandblasted. A

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 47–52

Cavallo and Petriglieri

the Italian company IMI Fabi is the third-largest producer of talc in the world. The huge outcrops of mafic and ultramafic rocks, as well as the important extractive activities, make this area an excellent case study of naturally occurring asbestos (NOA). Airborne fibers have been monitored systematically since 2004, especially in quarries and processing facilities (Cattaneo et al., 2012; Cavallo and Rimoldi, 2013), but there are few data about the serpentinite (dimension stone), the rock mass, soils, and stream sediments. The aim of this work is to increase knowledge of the distribution and abundance of asbestos in the environment. The Italian threshold for asbestos in rocks and soils is 1,000 mg/kg (or ppm; Italian D.Lgs. 152/06), whereas the occupational exposure limit for airborne asbestos in workplaces is 100 f/l of air (8-hour time-weighted average) and the environmental exposure limit is 2 f/l. MATERIALS AND METHODS Figure 1. Simplified sketch map of the Valmalenco area, showing the extension of the Malenco nappe, the location of the serpentinite quarries, the abandoned chrysotile, and the active talc mine.

huge amount of waste products (irregular blocks, rock chips, and cutting sludge) derive from the quarrying and processing activities (Cavallo, 2018). Steatite is also actively mined in the area: it is an impure talc, used mainly as a ﬁller for rubber and plastics, that occurs in metric, sub-vertical lodes within the serpentinites and/or at the contact with dolomitic marbles. The extraction occurs in underground mines, and

Extensive sampling of rocks, soils, and stream sediments was conducted from 2004 to 2018, and airborne asbestos concentrations have been monitored at least once a year since 2004. Airborne asbestos was analyzed mainly in the summer season in more than 250 personal air ﬁlters (41 phase contrast microscopy [PCM], 215 scanning electron microscopy/energy dispersive spectroscopy [SEM-EDS], and 11 transmission electron microscopy [TEM]) and 43 environmental air samples (SEM-EDS), collected from diﬀerent quarrying and processing locations, at quarry property borders, and at the closest villages. The World Health Organization (WHO, 1997) method was the reference method

Figure 2. Cross-ﬁber (left) and slip-ﬁber (right) chrysotile vein with hydrothermal alteration selvedge.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 47–52

NOA in Valmalenco Table 1. Airborne asbestos exposure values (f/l). Limit of detection (LOD) is approximately 0.1 f/l. Sampling Position

Chrysotile (f/l, Mean and Range)

Tremolite (f/l, Mean and Range)

22.3 (<0.5–72) 28.6 (0.5–160) 21.2 (<0.5–>200) 0.6 (<0.3–5.3) 0.2 (<LOD–0.5) 51.3 (0.5–>>200) 3.3 (0.5 – 10) 24.3 (1.9 – 92)

<LOD <0.1 (<LOD–0.5) <0.1 (<LOD–3) <LOD <LOD <LOD <LOD <LOD

Quarry, moving and handling Quarry, drilling and blasting Quarry, diamond wire cutting Quarry border Nearest town (1–5 km) Processing: gang saw cutting Processing: hand splitting Processing: dry ﬁnishing

for sampling and phase-contrast microscopy analysis (PCM), whereas SEM-EDS analysis was performed according to national legislation (Ministerial Decree 257, 1994) and Method 14966 of the International Organization for Standardization (2002). TEM analysis was carried out according to Method 7402 of the National Institute for Occupational Safety and Health (1994). The fiber definition criteria are those cited in the references above (length >5 µm, diameter <3 µm, and aspect ratio >3:1). Cellulose nitrate filters were used for PCM, whereas polycarbonate filters were used for SEM-EDS and TEM analyses. A total of 140 rock, 38 soil, and 27 stream sediment samples (each 1.5–2 kg) were collected and characterized for NOA concentration. A preliminary evaluation of rocks and mineralized vein samples was carried out with a handheld Enspectr RaPort Raman instrument, equipped with a 532-nm laser at maximum output power of 30 mW, with a real spatial resolution in the range of a few cubic millimeter. Raman spectra were collected in a single recording up to 60 seconds, in the extended wavenumber range 120–4,000 cm−1 , with a spectral resolution of 8 cm−1 . The bulk mineralogy (limit of detection between 0.5 and 2 wt%, depending on crystallinity) was determined by X-ray powder diffraction using a Bragg–Brentano θ–θ PANalytical X’Pert PRO PW3040/60 diffractometer, and quantitative phase analysis was carried out running FULLPAT software (Chipera and Bish, 2002). Quantitative asbestos determination was carried out by SEM-EDS, following specific technical criteria for

NOA (Italian D.Lgs. 152/06; Gualtieri et al., 2014, 2018; Belluso et al., 2017). Airborne Asbestos The analysis of airborne asbestos showed a complex environment, with significant analytical issues. The PCM data show quite low concentrations and poor agreement with SEM and TEM due to the abundance of very thin fibrils and/or fiber bundles, deriving from strong mechanical fragmentation and comminution (e.g., drilling, blasting, and diamond wire cutting), invisible under PCM (resolution 0.3 µm). Due to the abundance of lamellar antigorite with a high aspect ratio (especially schistose varieties), filters were heavily loaded with pseudo-fibrous antigorite splinters, falling into the WHO fiber definition. The distinction between chrysotile and pseudo-fibrous lamellar serpentine was based on both morphological and dimensional criteria (e.g., a cutoff diameter of 0.2 µm; Cattaneo et al., 2012; Belluso et al., 2017). TEM data showed good agreement with SEM, highlighting in some cases the abundance of chrysotile micro-fibrils (length <5 µm), not countable according to WHO criteria. Chrysotile and pseudo-fibrous lamellar serpentine were detected in almost all samples, whereas asbestiform tremolite was observed in very few cases. The assessed occupational exposure levels in quarries (Table 1) were mainly below 100 f/l (mean values 15–30 f/l), with the exception of drilling, diamond wire cutting, and block handling. Occupational exposure levels appeared to

Table 2. Asbestos concentrations (mean and range) in “massive” samples by SEM-EDS. Limit of detection (LOD) is approximately 50 ppm. Material Serpentinite, ﬁnished product Quarry waste Processing waste Soils Stream sediments Chrysotile mine tailings*

Chrysotile (ppm)

Tremolite (ppm)

80 (<LOD–250) 460 (120–1,700) 270 (60–2,300) 310 (150–2,800) 220 (60–560) 2.3 wt% (0.2–5.6 wt%)

<LOD 60 (<LOD–150) <LOD <60 (<LOD–180) <60 (<LOD–480) 230 (80–750)

*For chrysotile in mine tailings, the value is given in wt%.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 47–52

Cavallo and Petriglieri

Figure 3. SEM back-scattered electron images. (A) Asbestiform tremolite. (B) Chrysotile fiber bundles. (C) Abundant lamellar, pseudo-fibrous antigorite fragments. (D) Organic fiber.

be linked with the interception of chrysotile veins, which in some cases coincide with quarry block external faces. Airborne monitoring in processing facilities identified some critical fiber exposure in the process of dry finishing and multi-blade gang saw cutting of

blocks (values >>100 f/l). This can be explained by the chrysotile contamination of some external quarry block faces, emphasizing the importance of performing block squaring oﬀ directly in the quarry, removing all visible asbestos mineralization and accumulating

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 47–52

NOA in Valmalenco

asbestos minerals in a safe area of the quarry for subsequent disposal. The preventive measures taken (e.g., quarry floor wetting, wet drilling, and block squaring) have proven to be effective because the airborne asbestos values have shown a statistically significant decrease over the years. Values measured at quarry property borders and at the nearest town (1–5 km from the quarry) were mostly below the environmental exposure limit (2 f/l), confirming scarce dispersion of the fibers from the extraction sites. Rocks, Soils, and Stream Sediments The evaluation of NOA contamination of rock masses requires detailed structural and petrographic surveys, supported by mineralogical analyses. The portable Raman device was a user-friendly efficient analytical strategy. The Raman device is able to discriminate the potential asbestiform fibers from non-harmful fragments (Petriglieri et al., 2015), especially for preliminary evaluation of rock debris, veins, and visible fibrous material. The rock mass quality in serpentinite quarries was always observed to be good, with chrysotile veins occurring systematically along specific discontinuities (fractures and faults). Hydrothermal alteration selvedges of chrysotile veins pose a problem, where contamination occurs up to 20 cm from the edge of the veins (Figure 2). The quantitative analysis of “massive” samples (SEM) requires special care, especially for sample selection and preparation (ensuring collection of a representative sample), and discrimination between asbestiform minerals and nonasbestiform polymorphs (e.g., pseudo-fibrous antigorite splinters). The commercial finished rock products (both “massive” and schistose serpentinite; Table 2) can be considered virtually asbestos free (in almost all cases no detectable asbestiform fibers), whereas special care must be taken with raw quarry blocks, sometimes with visible slip-fiber chrysotile mineralization. In some cases, the contamination of quarry floors and tracks is a concern due to the continuous accumulation of chrysotile debris and the fragmentation linked to the passage of trucks and excavators. Small amounts of chrysotile (mean <400 ppm, range 150–2,800 ppm) were detected in soils and stream sediments, as were traces of asbestiform tremolite (mean <300 ppm, range <100–480 ppm). The highest concentrations of chrysotile were detected close to the abandoned mines, with peaks (up to 5 wt%; Table 2) in tailings, debris, and reclaimed mining landfills. While the presence of chrysotile was expected, asbestiform tremolite was a disturbing and widespread surprise (Figure 3): it is not reported in literature and is probably linked to talc veins and/or ophicarbonates occurring in both serpentinites and

dolomitic marbles. Therefore, in-depth studies are needed to ascertain the extent of tremolite in the area. CONCLUSIONS The over 10-year experience in quarries and mines of the NOA environment in Valmalenco has highlighted the importance of the following:

r geological and structural ﬁeld surveys, representative sampling, and accurate mineralogical analysis

r user-friendly analytical techniques (i.e., portable Raman) for easy and accurate preliminary evaluation of rock masses and fibrous material r the definition of specific criteria and the combination of analytical techniques (e.g., SEM and TEM) for the correct identification, discrimination, and quantification of asbestiform minerals and nonasbestiform polymorphs or varieties in both airborne and “massive” samples r the adoption of procedural and organizational solutions both in the quarries and in the processing facilities to easily recognize and reduce contamination r the need of appropriate training of employers and workers under coordinated supervision of the local authorities (e.g., workers’ compensation authority) and the need for using the correct personal protective equipment REFERENCES Belluso, E.; Cavallo, A.; and Haltermann, D., 2017, Crystal habit of mineral fibers. In A. F. Gualtieri (Editor), Mineral Fibres: Crystal Chemistry, Chemical—Physical properties, Biological Interaction and Toxicity: EMU Notes in Mineralogy, 18. European Mineralogical Union and Mineralogical Society of Great Britain and Ireland, London, pp. 65–110. Cattaneo, A.; Somigliana, A.; Gemmi, M.; Bernabeo, F.; Savoca, D.; Cavallo, D. M.; and Bertazzi, P. A., 2012, Airborne concentrations of chrysotile asbestos in serpentine quarries and stone processing facilities in Valmalenco, Italy: Annals of Occupational Hygiene, Vol. 56, No. 6, pp. 671–683. Cavallo, A., 2018, Serpentinitic waste materials from the dimension stone industry: Characterization, possible reuses and critical issues: Resources Policy, Vol. 59, pp. 17–23. Cavallo, A. and Rimoldi, B., 2013, Chrysotile asbestos in serpentinite quarries: A case study in Valmalenco, Central Alps, Northern Italy: Environmental Science: Processes & Impacts, Vol. 15, pp. 1341–1350. Chipera, S. J. and Bish, D. L., 2002, FULLPAT: A full-pattern quantitative analysis program for X-ray powder diffraction using measured and calculated patterns: Journal of Applied Crystallography, Vol. 35, No. 6, pp. 744–749. Gualtieri, A. F.; Bursi Gandolfi, N.; Pollastri, S.; Rinaldi, R.; Sala, O.; Martinelli, G.; Bacci, T.; Paoli, F.; Viani, A.; and Vigliaturo, R., 2018, Assessment of the potential hazard represented by natural raw materials containing mineral fibres—The case of the feldspar from Orani, Sardinia (Italy): Journal of Hazardous Materials, Vol. 350, pp. 76–87.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 47–52

Cavallo and Petriglieri Gualtieri, A. F.; Pollastri, S.; Bursi Gandolfi, N.; Ronchetti, F.; Albonico, C.; Cavallo, A.; Zanetti, G.; Marini, P.; and Sala, O., 2014, Determination of the concentration of asbestos minerals in highly contaminated mine tailings: An example from abandoned mine waste of Crètaz and Èmarese (Valle d’Aosta, Italy): American Mineralogist, Vol. 99, pp. 1233–1247. International Organization for Standardization, 2002, Ambient Air—Determination of Numerical Concentration of Inorganic Fibrous Particles—Scanning Electron Microscopy Method 14966: International Organization for Standardization, Geneva, Switzerland. Münterer, O. and Hermann, J., 1996, The Val Malenco lower crust-mantle complex and its ﬁeld relations (Italian Alps): Schweizerische Mineralogische und Petrographische Mitteilungen, Vol. 76, pp. 475–500.

Münterer, O.; Hermann, J.; and Trommsdorff, V., 2000, Cooling history and exhumation of lower-crustal granulite and upper mantle (Malenco, eastern Central Alps): Journal of Petrology, Vol. 41, pp. 175–200. National Institute for Occupational Safety and Health, 1994, Asbestos by TEM: Method 7402: Issue 2, August 15, 1994. Petriglieri, J. R.; Salvioli-Mariani, E.; Mantovani, L.; Tribaudino, M.; Lottici, P. P.; Laporte-Magoni, C.; and Bersani, D., 2015, Micro-Raman mapping of the polymorphs of serpentine: Journal of Raman Spectroscopy, Vol. 46, pp. 953–958. World Health Organization, 1997, Determination of Airborne Fiber Number Concentrations: A Recommended Method, by Phase Contrast Microscopy (Membrane Filter Method): World Health Organization, Geneva, Switzerland.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 47–52

Naturally Occurring Asbestos in France: Geological Mapping, Mineral Characterization, and Technical Developments FLORENCE CAGNARD* DIDIER LAHONDÈRE BRGM, French Geological Survey, Orléans, France

Key Terms: NOA, Asbestos, Geological Mapping, Regulations, France ABSTRACT In France, asbestos was banned by national decree (no. 96-1133) in 1996. The regulatory texts and standards adopted to implement this ban are concerned primarily with asbestos-containing manufactured products and are difficult to apply to asbestos-bearing natural materials (i.e., rocks and soils). Considering problems related to asbestos-bearing natural materials, the French Ministry of Ecology, Sustainable Development, and Energy has mandated the French Geological Survey to map locations where asbestos-bearing rocks are found. Mapping was prioritized to geological domains where naturally occurring asbestos (NOA) was predictable (e.g., the Western Alps and Corsica). These studies integrated field expertise, sampling, and laboratory analysis data to characterize the potential of geological units to contain NOA. Additionally, studies were conducted on geological formations exploited to produce aggregates. These studies were focused on quarries excavating massive, basic or ultrabasic rocks likely to contain NOA and quarries mining alluvium likely to contain asbestos-bearing rock pebbles. These studies highlight the difficulty of establishing robust analytical procedures for natural materials. The distinction between cleavage fragments (resulting from the fragmentation of non-asbestos particles) and proper asbestos fibers is particularly problematic for laboratories. Thus, a recent study by the National Agency for Health Safety, Food, Environment, and Work recommends applying the asbestos regulation to elongated mineral particles (length/depth > 3:1, length > 5 μm, depth < 3 μm) with chemical composition corresponding to one of the five regulated amphibole species regardless of their mode of crystallization (asbestiform or non-asbestiform). The upcoming regulatory changes are part of a decree published in 2017, including the prior identification of asbestos in natural soils or rocks likely to be impacted by *Corresponding author email: f.cagnard@brgm.fr

ground-disturbing construction activities. Specific protocols will be defined for sampling, analysis, and characterization of natural materials that may contain asbestos. INTRODUCTION In France, the use of asbestos was extensive through most of the 20th century. Asbestos was mined in France from 1920 to 1965, especially in two main quarries localized in the Alps (the Val de Peas mine extracting tremolite-asbestos) and in Corsica (the Canari mine excavating chrysotile). After 1965, the domestic demand for asbestos was satisfied by importation, especially from Canada. In 1977, the first regulation by the international agency for research on cancers classified all the forms of asbestos as carcinogenic. Even though asbestos was known to be harmful since the beginning of the 20th century, protection and prevention measures were adopted later in France. In fact, a complete ban on the use of asbestos in France was declared in 1996 and confirmed at the European level by Directive 99/77/EC, which prohibited any extraction, manufacture, or processing of asbestos fibers. The issue of naturally occurring asbestos (NOA) has emerged in only the past decade in France, with the first identification of 20 sites containing NOA in 2005 (Dessandier and Spencer, 2005). There is currently no law concerning the specific problem of NOA in France, and the existing regulations concern only “industrial” asbestos. However, a group of researchers working with two French national agencies recently proposed recommended guidelines (ANSES, 2010, 2015, 2017; INRS, 2013). Based on these studies, national regulations concerning NOA are currently in progress. France is made up of various geological domains consisting mainly of Neoproterozoic to Paleozoic basements partly affected by the Variscan orogeny, juxtaposed with Meso-Cenozoic sedimentary basins, which are locally reworked during the Alpine orogeny (Figure 1). Except the sedimentary basins, many basement lithologies, localized within the Armorican Massif, the Vosges Massif, the Massif Central, the

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Cagnard and Lahondère

Figure 1. Simpliﬁed lithological map of metropolitan France with the localization on studied areas concerned with projects mapping NOA occurrence susceptibility at a regional scale (1/50,000) (modiﬁed from Lahondère et al. 2018 and references therein).

Western Alps, the Pyrenees, and Corsica, are likely to contain NOA. The hazardous lithologies in basement areas concern, in particular, the basic and ultrabasic rocks, the meta-carbonates, and different types of rocks affected by greenschist facies metamorphism (e.g., dolerites and acid magmatic rocks). Another French territory, the island of the New Caledonia, is strongly affected by the NOA issue. Ultrabasic rocks containing significant volumes of NOA cover onethird of the island’s surface. Moreover, New Caledonia is one of the world’s largest producers of nickel ore, concentrated mainly within alteration profiles of ultrabasic rocks. Aware of the NOA issue, the French Ministry of Ecology, Sustainable Development, and Energy mandated the French Geological Survey (BRGM) to establish a systematic review and mapping of the geologi-

cal units that may contain NOA. In parallel with these projects, the BRGM was also mandated to establish a list of quarries with potential to contain NOA-bearing rocks. The main results obtained during these projects and associated studies undertaken by the BRGM are presented in this article. TYPOLOGY OF THE MAIN NOA OCCURRENCES IN FRANCE NOA in France occurs mostly within basic and ultrabasic rocks. Chrysotile constitutes the most typical asbestiform occurrence, as veins and shear planes localized within serpentinites. Cross veins of chrysotile are extensively developed in the center of serpentinite massifs. Long ﬁbers of chrysotile appear also as slip veins and are crystallized along shear planes within

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Naturally Occurring Asbestos in France

highly sheared serpentinites, close to tectonic contacts (Lahondère et al., 2019) (Figure 2a-b). Anthophylliteasbestos is also commonly observed, especially in New Caledonia, filling cross veins cutting across serpentinites (Figure 2c-d). In addition, tremolite-asbestos is frequently developed as long asbestos fibers within veins cross-cutting serpentinites or within shear planes (Figure 2e-f). Such tremolite-asbestos occurrences are found extensively within serpentinite massifs, especially in tectonic contacts and/or along the boundaries of these massifs, in close proximity to calcic-rich surrounding rocks, such as meta-gabbros, meta-basalts, or meta-carbonate rocks (Lahondère et al., 2019). Alteration and hydrothermal weathering of tremolite-rich veins and planes result in the formation of white bundles of tremolite fibers, easily dispersible into the air. Meta-basalts and meta-gabbros may be rich in fibrous and asbestos tremolite and actinolite, which may sometimes be located within synfolial planes and more frequently within open veins at high angles to the foliation, in association with greenschist facies metamorphic minerals such as albite ± calcite ± epidote ± chlorite (Lahondère et al., 2019). Hydrothermally altered dolerites are also frequently rich in asbestos-actinolite fibers, which are located mainly within shear planes and veins cross-cutting the dolerites (Lahondère et al., 2018) (Figure 2g-h). Other unusual occurrences and non-regulated fibrous minerals can locally be observed and concern, particularly (1) tremolite-asbestos within marbles, (2) fibrous Mgriebeckite within orthogneisses, and (3) fibrous actinolite within hydrothermally altered granitoids. MAPPING OF POTENTIAL NOA OCCURRENCES AT THE REGIONAL SCALE (1/50,000) AND THE COMMUNE SCALE (1/5,000) Because of the abundance of rocks having a potential to contain NOA in French territory, the BRGM was mandated to determine the localities of hazardous geological units. A project of mapping (at a regional scale: 1/50,000) geological units that are likely to contain NOA is funded by the French Ministry of Ecology, Sustainable Development, and Energy. This project began in 2009 and is still in progress. Mapping is being conducted in different areas of France (i.e., the Armorican Massif, the Vosges Massif, the Massif Central, the Western Alps, the Pyrenees, and the Corsica) (Figure 1). For each area, a threefold approach is used. First, a determination of lithologies and target zones with potential NOA is made from bibliographic and database compilations. Second, geological mapping is done in the field, followed by the characterization of sampled fibrous minerals by different analyt-

ical techniques in the laboratory (polarized light microscopy, transmission electron microcopy, scanning electron microscopy, electron probe microanalysis, and Raman spectroscopy). The final step consists of the drawing of an NOA occurrence susceptibility map at a regional scale (1/50,000 scale) with three NOA occurrence susceptibly classes (1 = none to weak, 2 = medium, 3 = high to very high) (Figure 3). At the end of the project, a list of townships that are likely to contain lithologies characterized by a medium to high susceptibility to contain NOA is established for each studied area. Following the establishment of the list of these townships, new projects are initiated and funded by local French agencies, such as the Environmental Office of Corsica and the Regional Directorate for Environment, Development, and Housing. The main goal is to establish detailed maps of the NOA occurrences at precise scale. These documents are then included in the territorial development plans. In such projects, the detailed geological mapping does not concern the entire “commune” areas but is focused mainly on key zones (i.e., urbanized zones, potentially constructible areas, towns, hiking paths, tourist areas, and so on) and key roads. A selection of such key zones and roads is realized by GIS treatments of different sources of data, such as planning documents, geological maps, and land cover. Detailed geological mapping is performed in the field and is followed by the analysis of fibrous samples at the laboratories leading to the production of a precise final map of NOA occurrence susceptibility at 1/5,000 scale (Figure 4). Finally, the main emissive outcrops and areas that are likely to release NOA in the air are identified and remediation proposals to fix the problems made. IDENTIFYING QUARRIES THAT ARE LIKELY TO EXPLOIT NOA-BEARING ROCKS In parallel with the mapping projects, the BRGM was mandated by the Ministry of Ecology, Sustainable Development, and Energy to identify the quarries that are likely to exploit NOA-hosting rocks. A list of 75 priority quarries exploiting solid rocks has been established and visited. A geological inspection of the sites with sampling was followed by the characterization of fibrous materials in the laboratory. This work led to the establishment of a list of quarries ranked according to their abundance of fibrous minerals. During this project, the difficulty of establishing reliable diagnoses of fibrous minerals, especially the fibrous amphiboles, was underlined. From this, new sampling and analytical protocols are currently being tested and discussed to improve the identification of fibrous versus asbestiform amphiboles in rock materials. The

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Cagnard and Lahondère

Figure 2. Macroscopic aspects and scanning electron microscopic (SEM) photographs of typical fibrous occurrences in France. (a) Fibers of chrysotile on a shear plane within highly sheared serpentinites (Canari mine, Corsica). (b) SEM photograph of chrysotile fibers within a serpentinite (Corsica). (c) Bundles of antigorite fibers in a serpentinite (New Caledonia). (d) SEM photograph of fibrous antigorite in serpentinite (New Caledonia). (e) Tremolite-asbestos in a serpentinite (Val de Peas mine, Western Alps). (f) SEM photograph of tremoliteasbestos from a serpentinite (Val de Peas mine, Western Alps). (g) Fibrous-actinolite in a shear plane within a hydrothermally altered dolerite. (h) SEM photograph of bundles of fibrous actinolite from a hydrothermally altered dolerite (Massif Central).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Naturally Occurring Asbestos in France

Figure 3. Map of NOA occurrence susceptibility of Northern Corsica (France) with three classes (1 = none to weak, 2 = medium, 3 = high to very high) at a regional scale (1/50,000 scale) (modiﬁed after Lahondère and Zammit, 2012).

ministry recommended that the problematic quarries conduct supplementary detailed geological mapping and airborne surveys. Results of the supplementary studies allowed the ministry to make ﬁnal decisions concerning the continuation or cessation of activity at the quarries. Similar studies currently concern alluvial quarries and are especially challenging because several hundred quarries are potentially concerned, involving a wide range of lithologies. This raises new methodological problems concerning primarily the representativeness of the sampling protocol. CURRENT ACTIVITIES The expertise gained from all these projects was beneﬁcial for discussions in national working groups on

the health effects from NOA hazards, leading to the production of technical recommendations for workers within NOA-bearing lithologies (ANSES, 2010) and the identification of cleavage fragments of amphiboles from quarried minerals (ANSES, 2015). Current activities concern the publication of new regulations and standard definitions. In parallel to the mapping projects, several research projects pertaining to the NOA issue are under way in the BRGM. The first project concerns the development of the use of hyperspectral imaging combined with geometric data processing approaches to identify, map, and monitor NOA. Such non-destructive approaches provide the potential to detect and map NOA minerals during continuous sampling instead of discrete sampling. In addition, innovative studies are consid-

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Cagnard and Lahondère

Figure 4. Examples of (a) a regional scale (1/50,000) map of NOA occurrence susceptibility in an area from Northern Corsica with three classes (1 = none to weak, 2 = medium, 3 = high to very high), (b) a detailed (1/5,000 scale) map of NOA occurrence susceptibility in an area from Northern Corsica with three classes (1 = none to weak, 2 = medium, 3 = high to very high) (a–b, modiﬁed after Gutierrez et al., 2016), and (c) a photograph of an identiﬁed emissive outcrop (ancient pit of serpentinites).

ering the mineral and morphological characterization of non-regulated mineral fibers and atypical fibrous occurrences, such as characterization of fibrous sodic amphiboles and occurrence of actinoliteasbestos within various lithologies, such as granitoids, dolerites, and orthogneisses, as already described in the literature (e.g., Metcalf and Buck, 2015; Lahondère et al., 2018). Finally, studies developed in the BRGM concern the problematic characterization of fibrous amphiboles with the challenge to obtain a robust and reliable methodology to converge similar definitions of fibrous and/or asbestiform minerals. New method-

ological developments concerning an experimental protocol allowing the release of fibers from a solid rock are under way and benefit from an extensive sampling of different types of amphiboles in the field. CONCLUSIONS The NOA issue has been an emerging subject of interest in France for the past decade. Currently, no regulations or national laws concerning the NOA problem exist in France. Different projects have been carried out by the BRGM, leading to the production

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Naturally Occurring Asbestos in France

of NOA potential occurrences maps at regional and commune scales and to the establishment of a list of quarries potentially exploiting NOA-bearing rocks. Results of such projects allow BRGM and the French government to inform the public and industry about the location of potential NOA-bearing lithologies with the goal to reduce the potential for exposure during human activities such as has occurred, for example, in the United States (Van Gosen, 2007), Australia (Hendricks, 2009), Italy, and California (Churchill and Hill, 2000). Different research projects are in progress and concern (1) the use of spectral-geometric approaches to the mapping and prediction of NOA occurrences, (2) the characterization of non-regulated and atypical species of fibrous minerals, (3) the development of experimental protocols and methodologies to distinguish cleavage fragments of asbestos, and (4) the development of a database as well as a process for the management and the diffusion of data concerning NOA in France. ACKNOWLEDGMENTS The authors gratefully acknowledge the DirectorateGeneral for Risk Prevention of the French Ministry of Ecological and Solidarity Transition, the BRGM and the Corsican Environmental Office for discussions and for their financial support. We thank the reviewers for their constructive comments. REFERENCES ANSES, 2010, Affleurements naturels d’amiante. Etat des connaissances sur les expositions, les risques sanitaires et pratiques de gestion en France et à l’étranger: Rapport d’étude, 248 p.

ANSES, 2015, Effets sanitaires et identification des fragments de clivage d’amphiboles issus des matériaux de carrière: Rapport d’expertise collective, 218 p. ANSES, 2017, Particules minérales allongées. Identification des sources d’émission et proposition de protocoles de caractérisation et de mesures: Rapport d’expertise collective, 184 p. Churchill, R. K. and Hill, R. L., 2000. A General Location Guide for Ultramafic Rocks in California—Areas More Likely to Contain Naturally Occurring Asbestos: California Department of Conservation, Division of Mines and Energy, Open File Report 2000, 19. Dessandier, D. and Spencer, C., 2005, Recensement et classement des sites naturels amiantifères et des formations géologiques potentiellement amiantifères en France: BRGM/RP-53599, 59 p. Gutierrez, T.; Lahondère, D.; and Cagnard, F., 2016, Reconnaissance des zones naturelles amiantifères sur neuf communes de la région du Nebbio (Haute-Corse): BRGM/RP-66345-FR, 185 p. Hendricks, M., 2009, Naturally occurring asbestos in eastern Australia: A review of geological occurrence, disturbance and mesothelioma risk: Environmental Geology, Vol. 57, pp. 909– 926, doi:10.1007/s00254-008-1370-5. Lahondère, D.; Cagnard, F.; Wille, G.; and Duron J., 2019. Naturally occurring asbestos in an Alpine ophiolitic complex (Northern Corsica, France): Environmental Earth Sciences, Vol. 78, doi:10.1007/s12665-019-8548-x. Lahondère, D.; Cagnard, F.; Wille, G.; Duron, J.; and Misseri, M., 2018, TEM and FESEM characterization of asbestiform and non-asbestiform actinolite fibers in hydrothermally altered dolerites (France): Environmental Earth Sciences, Vol. 77, doi:10.1007/s12665-018-7549-5. Lahondère, D. and Zammit, C., 2012, Déclinaison en trois classes de l’aléa “amiante environnemental” dans le département de la Haute-Corse: BRGM/RP-61734-FR, 21 p. Metcalf, R. V. and Buck, B. J., 2015, Genesis and health risk implications of an unusual occurrence of fibrous NaFe3+ amphibole: Geology, Vol. 43, pp. 63–66. Van Gosen, B. S., 2007. Reported Historic Asbestos Mines, Historic Asbestos Aspects and Natural Asbestos Occurrences in the Rocky Mountain States of the United States (Colorado, Idaho, Montana, New Mexico and Wyoming): U.S. Geological Survey Open File Report 2007-1182.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 53–59

Naturally Occurring Asbestos in France: a Technical and Regulatory Review ERELL LÉOCAT* Iﬀendic, Bretagne 35750, France

Key Terms: Naturally Occurring Asbestos, Elongated Mineral Particle, Transmission Electron Microscopy, Regulation, France ABSTRACT Naturally occurring asbestos (NOA) has been a wellknown issue within rock quarries for a long time. In France, the subject has recently become more controversial, particularly since 2013. In fact, some mineral fibers with the chemical composition of regulated asbestos (i.e., actinolite) have been discovered in roadbase aggregates and associated air filter samples. The main problem concerns the determination of the asbestiform versus non-asbestiform character of such mineral particles. The in-force standard based on the morphological identification of a fiber does not allow one to make this distinction. Presently, in France, the asbestos analysis of building material is based on a “yes” or “no” result. This method has limitations for analyzing NOA, as NOA may be present in lower concentrations in natural materials, especially in road-base aggregates. The health effects of the non-asbestiform particles, also called “cleavage fragments,” with fiber morphology are not well established. The French government mandated the National Agency for Food, Environmental and Occupational Health and Safety to conduct a review on the “state of the art” concerning the cleavage fragment issue. The conclusions of the report highlight the fact that elongate mineral particles (EMPs) are up for debate and address remaining questions concerning this subject. The next fundamental step is to secure agreement on the terminology of EMPs with the aim of comparing the studies in different disciplines. INTRODUCTION There are many issues in France related to the identification of asbestos, especially naturally occurring asbestos (NOA). Some of the issues can be explained by the non-concordance of terms used by the different stakeholders (Table 1). The legislation refers to “asbestos,” that is, a commercial term referring to *Corresponding author email: leocat.erell@gmail.com

six mineral species extracted from specific rocks and deposits because of their asbestiform qualities. It is well known that these deposits may contain in trace amounts non-asbestiform varieties of these same mineral species (Langer, 1975; Van Orden et al., 2008). Identifying commercial asbestos in building material is easier, as this material is mainly composed of asbestiform fibers with typical shape and morphology, in particular curved and thin fibers with a length over width ratio of 20:1 or more (U.S. EPA, 1993), and interfering mineral fibers are easily distinguishable. Most of the standards dedicated to the analysis of building material and identification of commercial asbestos refer to the “fiber” term, especially to inhaled fibers, defined by a diameter of <3 µm. However, the dimension and morphology criteria of a fiber might be differently defined in these documents. These criteria lead to include mineral particles with various origins such as asbestiform minerals, other fibrous minerals, and cleavage fragments with fiber dimension. Under physical strengths, some single minerals can break along the weakness plane, called the “cleavage plane,” leading to particles called “cleavage fragments” that may have various shapes such as irregular particles with non-parallel sides, particles with fiber dimension, and particles with dimension of asbestiform fiber. In this context, the identification of asbestos in natural material is not obvious, and the determination of the nature of fibrous material remains challenging. The term “elongate mineral particle” (EMP) is a recently used term for particles with an aspect ratio of >3:1 and with approximately parallel sides, without a distinction in terms of the particle’s origin (ANSES, 2017). Within this dimensional and morphological definition are numerous particles with various origins, including fibrils originated from asbestiform bundles, single minerals with a fiber habit, cleavage fragments with asbestiform dimensions, cleavage fragments with fiber morphology, and cleavage fragments with nonparallel sides. REGULATORY REVIEW Asbestos has been forbidden in France since 1997 (Décret, 1996). Only the six following mineral fibers,

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 61–65

Léocat Table 1. The terminologies in the domain of asbestos. For more explanation, see the text.

classified as “asbestos,” are regulated in France: chrysotile, amosite (asbestos grunerite), crocidolite (asbestos riebeckite), asbestos tremolite, asbestos actinolite, and asbestos anthophyllite. The following fibers are not regulated: the non-asbestiform homologues of the six regulated asbestos and the four mineral fibers recognized as carcinogens for the human by the U.S. EPA (2014) for the winchite and the richterite and by the International Agency for Research on Cancer (IARC; 2014) for the erionite and the fluoro-edenite.

For Asbestos Survey in Bulk Material The decree (Décret, 2012) requires completion of an asbestos survey before work, demolition, and sale related to any construction older than 1997. If asbestos is present in a material, some specific processes must be followed depending of the nature of the material and the work equipment used to remove it, but not on the content of the asbestos. Consolidated by a regulatory document (UFETAM, 2013), this survey must be done on bituminous pavement, as chrysotile has been added in some bituminous coating materials. Actinolite has been identified in some cases, but the distinction between asbestiform fibers and cleavage fragments was not obviously performed. This mineral occurs naturally in most of the rocks used for road coated aggregates. NOA became a significant issue within the construction business (Léocat et al., 2018). Consequently, the Work Ministry published a directive on the regulated asbestos and cleavage fragments (Note DGT, 2014). This document gives information on the management of

this issue in transportation and building domains. Under a request of the French government, the National Agency for Food, Environmental and Occupational Health and Safety led to the development of a report on the health effects of cleavage fragments (ANSES, 2015). The conclusion of this report is that the toxicity of these particles cannot be excluded, as any scientific studies could argue that there is no hazard. Consequently, the agency recommends following the precaution principle and using the same dispositions for the cleavage fragments as for asbestos. It also recommends extending the regulation to the non-asbestiform homologues of the six asbestos minerals and to the four human carcinogenic minerals. The French Work Ministry published a second directive (Note DGT, 2018) on the dust-creating work activities during disturbance of building materials containing aggregates. In the case of presence of non-asbestiform fibers, the worker protective dispositions to be applied are those related to dust hazards and not those related to asbestos. A recent decree (Décret, 2017) outlines the obligation to perform an asbestos survey before work in seven domains, according to the respective standards: 1. Building (NF X46-020, 2017); 2. Other buildings (NF X46-102, in progress) engineering structures, transportation infrastructures, and other networks; 3. In-place rocks and soils (NF P94-001, in progress). This domain is concerned with NOA and refers to quarries, earthworks, etc.; 4. Railway equipment and other transportation equipment (NF F01-020, in progress); 5. Boat, ship, and other sea equipment (NF X46-101, 2019); and

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 61–65

Naturally Occurring Asbestos in France Table 2. The six cases of the COFRAC Lab INF 44 (2018) document. * indicates a non-ﬁbrous elongate mineral particle that has an aspect ratio of >3 and sides that are not parallel and/or stepped. Case

Material

1 2 3 4

Building material Contaminated soil (with debris of building material) Rocks (e.g., soils, aggregates)

5 6

Building materials containing natural material (roads, cement, concrete, coating)

6. Aircraft (NF L80-001, in progress). 7. Facilities, structures, and equipment of the industrial domains (NF X46-100, 2019). For Asbestos Testing Laboratories The in-force standard (NF X43-050, 1996) is mandated for transmission electron microscopy (TEM) asbestos analysis of air filter sampling. This standard gives the definition of a fiber as a mineral particle with parallel or stepped sides. Moreover, a fiber has a length to width ratio of over 3 μm and a minimal length of 0.5 µm; that is the technical limit. It defines “asbestos” as silicate minerals belonging to the amphibole and serpentine groups that have crystallized with asbestiform facies. This term “asbestiform” refers to a specific habit of a mineral for which the fibers and fibrils present a high tensile strength and a high flexibility. However, this document does not explain what an asbestiform mineral is from an analytical point of view. In fact, no dimensional and morphological criteria are attributed to the asbestiform habit. This standard also says that fibers analyzed in air samples need to have a diameter of under 3 µm. The asbestos regulatory rulings for the ambient air in buildings (Arrêté, 2011) and for the worker exposure limit (Arrêté, 2018) specify that the analyzed fibers in air samples are the fibers defined by the World Health Organization (WHO, 1998). This fiber called “WHO fiber” has a length of >5 µm. As no specific standard refers to bulk materials, the NF X43-050 (1996) standard is used for the analysis of this material and mainly for the identification of fibers. Qualitative analysis of short and long fibers should be performed to detect asbestos and should conclude by a “yes” or “no” result. The detected fibers must have a width under 3 μm, a length over 0.5 μm, and an aspect ratio of ࣙ3:1. The French Association for Accreditation, the government body overseeing asbestos testing laboratories, published a document for laboratories that were applying to be accredited for asbestos analysis in bulk ma-

Type of Analyzed Mineral Particles Intentionally added asbestos Intentionally added asbestos Naturally occuring asbestos Non-fibrous elongate mineral particles,* fibers from cleavage fragments, and fibers from asbestiform habit Naturally occuring asbestos Non-fibrous elongate mineral particles,* fibers from cleavage fragments, and fibers from asbestiform habit

terial (COFRAC, Lab INF 44, 2018) (Table 2). This document deﬁnes six areas of accreditation depending on the material type and the nature of the ﬁbers (industrial asbestos, NOA, cleavage fragments) Three methods can be used, as follows: 1. PLM (polarized light microscopy); 2. SEM (scanning electron microscopy); 3. TEM. The following six cases are distinguished:

r In cases 1 and 2, the industrial asbestos is identified, respectively, in building material and in soils contaminated with asbestos-containing material. The term “intentionally added asbestos” refers to the asbestos added to building material. In case 2, the intentionally added asbestos refers to the one incorporated in building material that may contaminate the ground. r In cases 3 and 5, the laboratory can identify the six regulated asbestos types in rocks and industrial materials containing natural material, but it is not allowed to report on non-asbestiform fibers. r In cases 4 and 6, respectively, in rocks and in building material containing rocks fragments, the laboratory has the ability to identify three different particles: ◦ cleavage fragments; ◦ asbestiform fibers; and ◦ non-fibrous EMPs. In this document, the term “non-fibrous” refers to particles having non-parallel and/or non-stepped sides and an aspect length:diameter ratio of >3. With an absence of standard testing materials, a guide (COFRAC, Lab GTA 44, 2018) has been published for testing laboratories that want to be accredited for asbestos identification in bulk materials. It explains, among other things, what the requirements are for validating the detection limit of 0.1 percent asbestos in bulk materials. The detection limit is based on the fact

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 61–65

Léocat

that asbestos additions in building materials were performed only in amounts greater than 0.1 percent (Directive européenne, 1999). The laboratory should have a referent person who has knowledge in mineralogy to support NOA identification. This guide also says that TEM analysis allows for identification of asbestos using the following three criteria: 1. Morphology to distinguish the fibers from the whole particles; 2. Chemistry to get the chemical composition of the fiber; and 3. Diffraction to identify the crystalline structure and distinguish the mineral fibers from the other fibers. This guide says that SEM analysis allows only classification of asbestos because identification of crystalline structure by diffraction analysis is not possible. However, the SEM three-dimensional images allow for distinguishing cleavage fragments from fibers from a morphological point of view. This document also requires using the Locock (2014) spreadsheet based on the International Mineralogical Association (IMA) classification to identify the five amphibole mineral species belonging to asbestos, which are actinolite, tremolite, anthophyllite, grunerite, and riebeckite species. TECHNICAL REVIEW After the first report, the French government mandated that the ANSES agency conduct a review on the emission source of EMPs of interest (EMPi; ANSES, 2017). These particles correspond to the asbestiform and non-asbestiform particles of the six regulated minerals and the four human carcinogen minerals. In this report, the agency proposes a protocol for identifying minerals in materials that are susceptible to liberate EMPi and a protocol to measure EMPi in the air. The aims are to explore the potential exposure of the professional population and the public to EMPi in some construction work activities (for example, the gravel quarries) and to study the emissivity of materials containing EMPi. Following these recommendations, BRGM (French Geological Survey), the Institut National de la Recherche Scientifique (INRS), and the Particle, Fiber, Asbestos Laboratory (LAFP) worked on the asbestos emissivity of aggregates submitted to attrition tests during the PIMAC project (BRGM, 2018). The French government asked the Professional Organization for Accident Prevention in the Building and Public Works Sector to coordinate a project on the EMPi with health, work, and environment ministries and the

French scientific organizations. Based on the recommendations of the ANSES report (2017), the project yielded the EMPi measurement campaign in construction activities. The different French scientific organizations have carried on various works on the subject for the last few decades. Among other tasks, BRGM works on mapping of rocks that contain or may contain NOA in natural outcrops and in quarries (BRGM, 2005, 2013). Based on the potential of containing NOA, the rocks are classified into five groups. Recent works weigh in on the metrological issue of NOA identification (Lahondère et al., 2018; Misseri and Lahondère, 2018). The INRS published a guide on NOA management during work on soils and rocks for engineering construction and transportation infrastructure terrain (INRS, 2013). DISCUSSION AND CONCLUSIONS The NOA issue has been apparent for a long time, and some scientific organizations, such as BRGM, are currently working on the subject. The presence of NOA became a big issue when it affected the domain of public works and buildings. Research projects and work are mainly concentrated on the emissivity of natural materials or building materials containing natural materials and to a lesser degree on the toxicity of these EMPs. Further essential work will concern a clear and univocal definition of EMPs and a nomenclature of the amphibole mineral species with asbestiform habit to support international agreement. In fact, an integrated NOA knowledge database may lead to improved understanding of the issues. ACKNOWLEDGMENTS I would like to thank all the individuals who encouraged me in my work in the field of asbestos and NOA. My thanks go to the reviewers Francesco Turci, Mark Bailey, and Florence Cagnard, who patiently read a first version of this article to greatly improve it. REFERENCES ANSES (Agence Nationale de Sécurité Sanitaire de l’Alimentation, de l’Environnement et du Travail [French National Agency for Food, Environmental and Occupational Health and Safety]), 2015, Effets sanitaires et à l’identification des fragments de clivage d’amphiboles issus des matériaux de carrière: Avis de l’Anses, Saisine 2014-SA-0196, 13 p. ANSES, 2017, Particules minérales allongées. Identification des sources d’émission et proposition de protocoles de caractérisation et de mesures: Avis de l’Anses, Saisine 2016-SA-0034, 13 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 61–65

Naturally Occurring Asbestos in France Arrêté du 19 août 2011 relatif aux modalités de réalisation des mesures d’empoussièrement dans l’air des immeubles bâtis: Journal Officiel de la République Française (JORF [Gazette of the French Republic]) 0202. Arrêté du 30 mai 2018 modifiant l’arrêté du 14 août 2012 relatif aux conditions de mesurage des niveaux d’empoussièrement, aux conditions de contrôle du respect de la valeur limite d’exposition professionnelle aux fibres d’amiante et aux conditions d’accréditation des organismes procédant à ces mesurages: JORF 0148. BRGM (Bureau de Recherches Géologiques et Minières [French Geological Survey]), 2005, Recensement et classement des sites naturels amiantifères et des formations géologiques potentiellement amiantifères en France: BRGM/RP-53599-FR, 59 p. BRGM, 2013, Exposition aux fibres asbestiformes dans les industries extractives: Identification des sites potentiellement concernés en France métropolitaine: BRGM/RP-61977-FR, 163 p. BRGM, 2018, PIMAC: capacité de libération et d’émission de particules amiantifères des granulats de carrières: Electronic document, available at https://www.brgm.fr/projet/pimaccapacite-liberation-emission-particulesamiantiferesgranulats-carrieres COFRAC (Comité Français d’Accréditation [French Association for Accreditation]), 2018, Guide technique d’accréditation: Recherche d’amiante dans les échantillons massifs: Lab GTA 44, 17 p. COFRAC, 2018, Nomenclature et expression des lignes de portée d’accréditation pour la recherche d’amiante dans les échantillons massifs: Lab INF 44, 6 p. Décret 96-1133 du 24 décembre 1996, relatif à l’interdiction de l’amiante, pris en application du code du travail et du code de la consummation: Premier ministre. Décret 2012-639 du 4 mai 2012, relatif aux risques d’exposition à l’amiante: JORF 0106. Décret 2017-899 du 9 Mai 2017, relatif au repérage de l’amiante avant certaines opérations: JORF 0109. Directive européenne, 1999, Directive 1999/45/CE du Parlement européen et du Conseil du 31 mai 1999 concernant le rapprochement des dispositions législatives, réglementaires et administratives des États membres relatives à la classification, à l’emballage et à l’étiquetage des préparations dangereuses: Journal Officiel de l’Union européenne L 200 du 30/07/1999, pp. 0001–0068. IARC (International Agency for Research on Cancer), 2014, Carcinogenicity of fluoro-edenite, silicon carbide fibres and whiskers, and carbon nanotubes: Lancet, Vol. 15, pp. 1427– 1428. INRS (Institut National de la Recherche Scientifique [National Center for Scientific Research]), 2013, Travaux en terrain amiantifère. Opérations de génie civil de bâtiment et de travaux publics: INRS ED 6142, 121 p. Lahondère, D.; Cagnard, F.; Wille, G.; and Duron, J., 2018, TEM and FESEM characterization of asbestiform and non-asbestiform actinolite fibers in hydrothermally altered dolerites (France): Environmental Earth Sciences, Vol. 77, p. 385. Langer, A. M.; Mackler, A.D.; and Pooley, F. D., 1974, Distinguishing between Amphibole Asbestos Fibers and Elongate Cleavage Fragments of their Non-Asbestos Analogues: Environmental Health Perspectives, Vol. 9, p. 63–80. Léocat, E.; Rielland, C.; and Letessier, P., 2018, Analysis and identification of elongate mineral particles in road coated aggregates: Toxicology Applied Pharmacology, Vol. 361, pp. 149– 154.

Locock, 2014, An Excel spreadsheet to classify chemical analyses of amphiboles following the IMA 2012 recommendations: Computer Geosciences, Vol. 62, pp. 1–11. Note DGT (Direction générale du travail), 2014, Cadre juridique applicable aux travaux réalisés sur des matériaux de BTP contenant des fibres d’amiante et/ou des fragments de clivage issus de matériaux naturels: Ministère du Travail, de l’Emploi, de la Formation professionnelle et du Dialogue social, Paris, France. Note DGT, 2018, Amiante – Cadre juridique applicable aux travaux réalisés sur des matériaux de BTP contenant des fibres d’amiante et/ou des fragments de clivage issus de matériaux naturels: Ministère du Travail, de l’Emploi, de la Formation professionnelle et du Dialogue social, Paris, France. Misseri, M. and Lahondère, M., 2018, Characterisation of chemically related asbestos amphiboles of actinolite: Proposal for a specific differentiation in the diagram (Si apfu versus Mg/Mg+Fe2+ ): International Journal Metrology Quality Engineering, Vol. 9. NF X43-050, 1996, Qualité de l’air – Détermination de la concentration en fibres d’amiante par microscopie électronique à transmission – Méthode indirecte: Association Française de Normalisation (AFNOR), Paris, France. NF X46-020, 2017, Repérage amiante – Repérage des matériaux et produits contenant de l’amiante dans les immeubles bâtis: AFNOR, Paris, France. NF X46-100, 2019, Repérage amiante – Repérage des matériaux et produits contenant de l’amiante dans les installations, structures ou équipements concourant à la réalisation ou à la mise en œuvre d’une activité: AFNOR, Paris, France. NF F01-020, 2019, Repérages avant travaux de l’amiante dans le matériel roulant ferroviaire: AFNOR, Paris, France. NF L80-001, in progress, Repérage avant travaux de l’amiante dans les aéronefs: AFNOR, Paris, France. NF P94-001, in progress, Repérage avant travaux – repérage amiante environnemental dans les sols et roches en place: AFNOR, Paris, France. NF X46-101, in progress, Repérage amiante – Repérage des matériaux et produits contenant de l’amiante dans les navires, bateaux et autres constructions flottantes: AFNOR, Paris, France. NF X46-102, in progress, Repérage des matériaux et produits contenant de l’amiante dans les ouvrages de génie civil, infrastructures de transport et réseaux divers: AFNOR, Paris, France. UFETAM (Union Fédérale de l’Environnement, des Territoires, des Autoroutes et de la Mer [Federal Union of the Environment, Territories, Highways and the Sea]), 2013, Gestion des risques sanitaires liés à l’amiante dans le cas de travaux sur les enrobés amiantés du réseau routier national non concédé: Circulaire du 15 Mai 2013, 8 p. U.S. EPA (U.S. Environmental Protection Agency), 1993, Method for the Determination of Asbestos in Bulk Building Materials: EPA/600/R-93/116, 98 p. U.S. EPA, 2014, Toxicological Review of Libby Amphibole Asbestos. In Support of Summary Information on the Integrated Risk Information (IRIS): EPA/635/R-11/002F, 685 p. Van Orden, D. R.; Allison, K. A.; and Lee, R. J., 2018, Differentiating amphibole asbestos from non-asbestos in a complex mineral environment: Indoor Built Environment, Vol. 17, No. 1, pp. 58–68. World Health Organization, 1998, Détermination de la concentration des fibres en suspension dans l’air. Méthode recommandée: la microscopie optique en contraste de phase (comptage sur membrane filtrante), Geneva, Switzerland, 64 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 61–65

Regulations Concerning Naturally Occurring Asbestos (NOA) in Germany—Testing Procedures for Asbestos STEFAN PIERDZIG* CRB Analyse Service GmbH, Bahnhofstrasse 14, D-37181 Hardegsen, Germany

Key Terms: NOA, Asbestos, Analysis, Regulations, Germany, TRGS 517 ABSTRACT In Germany, potential asbestos-containing rocks are used as raw materials for a number of engineering applications. These rocks are ultrabasites (dunite, harzburgite), igneous rocks (basalt, gabbro, norite), and metasomatic or metamorphic rocks like talcum, greenschist and amphibolite. Based on the German Gefahrstoffverordung (Hazardous Substances Ordinance), regulatory statutes exist for operations using these rocks and resultant composites and products. The authorities state that in Germany no natural rocks exist with more than 0.1 mass-% of one of the six regulated asbestos minerals. But it is well known that there are rocks with a high modal concentration of these minerals with a nonasbestiform, columnar to prismatic habitus. Under mechanical stress during handling, they can lead to fibrous cleavage fragments, which conform to the World Health Organization (WHO) “respirable asbestos fiber” definition. In view of this fact, the regulations changed in 2009, with revision of the Technical Rules for Hazardous Substances (TRGS) 517: any fibrous asbestos particles, regardless of whether or not they represent naturally occurring asbestos or are of cleavage origin, are evaluated for potential hazards associated with handling of these rocks. If the WHO fiber concentration is <0.1 mass-%, rocks and products can be used and re-used under protective measures. At concentrations >0.1 mass-%, the material is considered hazardous waste. These regulations apply to many industrial sectors that exploit and process rocks, using them in road building and track construction and when they are recycled. Analysis (by scanning electron microscopy, SEM/energy dispersive x-ray spectroscopy, EDS) to determine the asbestos concentration of rocks, gravels, or dusts is carried out in the <100-µm, grain-size fraction produced by sieving or grinding. The results provide a representation of a worst-case examination of the air quality during mechanical treatment of these materials. Workplace monitoring is done by air sampling to survey an exposure limit of 10,000 fibers/m3 of air (0.01 f/cc). *Corresponding author email: pierdzig@crb-gmbh.de

INTRODUCTION Compiled by Germany’s Committee on Hazardous Substances (AGS), Technical Rules for Hazardous Substances (TRGS) reflect the state of the art and the state of occupational health and occupational hygiene, as well as other sound scientific knowledge relating to activities involving hazardous substances, including their classification and labeling. In compliance with the requirements of the Gefahrstoffverordnung (Hazardous Substances Ordinance; GefStoffV, 2017), TRGS 517 (2013) applies to activities with potentially asbestos-containing mineral raw materials and mixtures and products produced from them and describes the protective measures to be applied to these activities. It applies especially to the following:

r the extraction and purification of naturally occurring mineral raw materials containing asbestos in quarries (e.g., gravel, grit, crushed sand, filler); r the further processing of asbestos-containing mineral raw materials and mixtures and products manufactured from them in construction and civil engineering (e.g., road and rail construction, concrete, asphalt); r the re-processing (recycling) and re-use in road construction (e.g., the treatment and reincorporation of recycled materials, the manufacture of asphalt); and r the processing of natural stone (e.g., soapstone in construction furnaces) and cold milling machines in traffic areas. Following the GefStoffV, the extraction, preparation, further processing, and re-use of mineral raw materials that occur naturally and preparations and articles manufactured therefrom that contain asbestos with a mass content of more than 0.1 percent are prohibited. INFORMATION GATHERING AND RISK ASSESSMENT To assess working conditions according to GefStoffV section 6, the employer must, before commencing activities with potentially asbestos-containing materials, competently determine by means of

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 67–71

Pierdzig

Figure 1. Diabase quarry in the Harz Mountains, Hercynian Forest.

appropriate measures whether asbestos exposure for employees should be expected and to what extent. Potentially Asbestos-Containing Rocks In the mineral deposits found when mining in the Federal Republic of Germany, for particular types of rock, the occurrence of asbestos minerals chrysotile, tremolite, actinolite, amosite, and, to a lesser extent, anthophyllite needs to be taken into account. Above all, basic magmatites are aﬀected. Figure 1 shows a Diabase quarry in the Harz Mountains, where one million tons of gravel for road and railroad construction are produced per year. The following rock types are particularly considered to contain asbestos (TRGS, 2013):

r Ultrabasite/peridotite (e.g., dunite, lherzolite, harzburgite); r Basic eﬀusives (e.g., basalt, diabase, spilite, basanite, tephrite, phonolite); r Basic intrusives (e.g., gabbro, norite); and r Metamorphic and metasomatically inﬂuenced rocks (e.g., metasomatic talc occurrences, green schist, chlorite and amphibole schist/bedrock [e.g., nephrite], serpentinite, amphibolite).

ratio of <5:1 and a larger diameter, as displayed on Figure 2. Asbestos is defined as six regulated minerals (chrysotile, amosite, tremolite, actinolite, crocidolite, anthophyllite, compare TRGS 517) that exhibit dimensions determined by the World Health Organization (WHO, 1986): length > 5 µm, diameter < 3 µm, with an aspect ratio >3:1). The regulation applies regardless of whether an asbestos fiber has been released from a primary fibrous deposit or from a non-fibrous, rockforming mineral, which belongs to the category of regulated asbestos minerals. Analysis and quantification of asbestos content is carried out according to BIA/IFA Code 7487 (BIA/IFA, 1989) by suspension in water and filtration through a 0.2-µm membrane filter. Quantification of asbestos with a limit of detection (LOD) of 0.008 percent is performed to estimate the potential of exposure to asbestos. In addition to the determination of the mass percentage of asbestos, the number of asbestos fibers per milligram of material examined should also be determined. This information is gathered when determining the mass percentage. If asbestos is detected in the material, asbestos exposure must be determined (TRGS, 2010) by a scanning electron microscopy (SEM)/energy dispersive x-ray spectroscopy (EDS) method designed for the monitoring of workplaces. Risk Assessment The risk assessment calculation for workplace and work activities is required to be performed by a competent person (DGUV Information 213–546, 2014). When performing the risk assessment, the following aspects should be taken into account:

r The extent and duration of the inhalation exposure; r Working conditions and work practices, including work equipment; and

r Required protective measures, based on exposure potential. When a change in the operating conditions occurs, which can lead to a major change to the hazard, the risk assessment must be performed again and the results of the risk assessment documented.

Procedures for Determining the Mass Content of Asbestos

Event-Related Advice

When processing asbestos-containing technical products, respirable asbestos ﬁbers are released, which predominantly display an aspect ratio of >10:1 and mostly exhibit a small diameter of <1 μm. Elongated particles from an asbestos mineral, released in the processing of mineral raw materials, diﬀer morphologically. The majority of these particles have an aspect

The employer is required to obtain event-related advice on protective measures to be taken to minimize the risks related to exposure to asbestos. If the employer’s own expertise is non-existent, advice may be provided through a qualiﬁed employee or subcontracted supervisor. If custom recommendations are not obtained, the procedures within “Dust-free ablation of

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 67–71

NOA Germany

Figure 2. Asbestos cleavage fragment of actinolite, released from a primary non-ﬁbrous deposit of an asbestos mineral under physical stress, such as that involved in crushing and milling.

asphalt pavements with cold milling machines” (BG Bau, 2011) can be followed. When a change to the operating conditions occurs, which can lead to a major change to the hazard situation, the event-related risk assessment should be carried out again.

sures and then at regular intervals, at least once per year. PROTECTIVE MEASURES General Protective Measures

Notiﬁcation to Local Mining Authorities If the investigation has revealed that workers are or may be exposed to asbestos during their activities, the employer must notify the local mining authorities of these activities. The notiﬁcation to authority must be carried out by the employer before the start of the activity and must contain the following information:

r r r r r

The location of the work site; The activities and procedures; The number of workers concerned; The beginning and duration of the activities; and Measures to limit asbestos exposure for workers.

For activities and procedures of a similar nature, a single enterprise-related notification is acceptable. The notification must be repeated when there is a significant change in the working conditions. After notification has been completed the protective measures should be selected and documented according to the result of the risk assessment. The effectiveness of the protective measures should be reviewed by subsequent exposure measurements after implementing the mea-

For activities involving asbestos-containing materials, the breathing air at the workplace must, as much as possible using state-of-the-art controls, be free from asbestos fibers. Where the asbestos fiber concentration falls below 10,000 F/m³, minimum basic measures to protect workers must be carried out (TRGS, 2008). Where the asbestos fiber concentration exceeds 10,000 F/m³, the following ranking of protective measures must be complied with to minimize risk: 1. Use of low-emission work procedures and work equipment; 2. Implementation of collective protection measures at the hazard source, such as extraction, aeration and ventilation, and appropriate organizational measures; and 3. Use of personal protective equipment if a risk cannot be prevented by the measures referred to in points 1 and 2 above. The general protective measures describe, among others, the following: the use of machinery and equipment, the design of work spaces, ventilation measures,

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 67–71

Pierdzig

air return, hygienic facilities, material storage and handling, cleaning procedures, and handling of wastes and residues. Protective measures also include deﬁnitions of responsibilities and supervision, limiting of the number of persons exposed, procedures for minimizing exposure, and information and training of workers. Complementary Protective Measures In addition to the general protection measures, complementary protective measures exist for the following work areas and activities:

r r r r r r r

Extraction and processing in quarries; Reprocessing and recycling; Processing of natural stone; Release agents and lubricants; Fillers and aggregates; Tunneling; and Cold milling of road surfaces.

OCCUPATIONAL HEALTH PREVENTION AND CARE In the case of activities with potentially asbestoscontaining raw materials and mixtures, and products derived therefrom, occupational medicine prevention generally comprises the participation of the occupational physician in the risk assessment, general medical advice, and occupational health care. The focus here is on imparting knowledge about carcinogenic and other chronically damaging properties of asbestos, as well as information regarding the necessity of wearing personal protective equipment. The workload must be included in the assessment of the inhalation burden. Occupational medicine toxicological advice is used to inform endangered employees, for example, within the framework of a brieﬁng. Instruction is to be provided if possible, with the participation of the company’s physician, and is also intended to provide information on the uses and scope of occupational health care examinations. Information is presented in a manner to motivate workers to participate. Occupational health care includes assessing the individual interrelation between work, physical and mental health, and early detection of work-related health problems. An additional purpose is to assess whether there is an increased health risk associated with a particular work activity. If this is this case, advice for employees on exposure and resulting hazards to their health is the focus. If physical or clinical investigation are not required by the physician’s assessment or are rejected by the worker, occupational health care may be limited to a consultation.

Figure 3. Flowchart showing the sequence of the survey of potentially asbestos-containing mineral raw materials and their products.

Occupational health care must be arranged for the workers by the employer prior to the activity and then at regular intervals when repeated exposure to asbestos in the workplace cannot be ruled out. The whole process of surveying potentially asbestos-containing raw materials is shown in Figure 3. SUMMARY In Germany, the Hazardous Substances Ordinance (GefStoffV, 2017) and the Technical Rule for Hazardous Substances, TRGS 517 (2013), provide extensive regulatory statutes for the examination of potentially asbestos-containing materials. The use and re-use of asbestos-containing materials with an asbestos content of <0.1 percent is allowed. For risk assessment, both naturally occurring asbestos and cleavage (elongated) fragments of asbestos minerals are taken into account. The determination of asbestos concentration follows BIA/IFA Code 7487 (1989) and is conducted using SEM/energy dispersive x-ray spectroscopy (EDS) after grinding of material, suspension, and filtration at the grain-fraction of <100 µm and reflects a worst-case-examination of the health hazard. Extensive general and complementary protective measures are required for activities with asbestoscontaining materials. REFERENCES DGUV Information 213-546 (previously ZH 1/120.46, BGI 505-46), 2004, Verfahren zur getrennten Bestimmung der Konzentrationen von lungengängigen anorganischen Fasern in

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 67–71

NOA Germany Arbeitsbereichen—Rasterelektronenmikroskopisches Verfahren [Procedure for the Separate Determination of the Concentration of Inorganic Fibers in Work Areas—Scanning Electron Microscopy Procedures]. Analytical Methods Recognized by the Professional Associations for Determining the Concentration of Carcinogenic Substances in the Air at the Workplace]: Deutsche Gesetzliche Unfallversicherung (DGUV), Federation of Professional and Trade Associations in Industry, Sankt Augustin/Carl Heymanns Verlag, Cologne. BIA/IFA Code 7487, 1989, Verfahren zur Bestimmung geringer Massengehalte von Asbestfasern in Pulvern, Pudern und Stäuben mit REM/EDX (Kennzahl 7487) [Procedure for Analytically Determining Low Mass Concentrations of Asbestos Fibers in Flours, Powders and Dusts with SEM/EDX (Code 7487)]. In: IFA-Arbeitsmappe Messung von Gefahrstoffen [IFA Work Folder on the Measurement of Hazardous Substances], 18, Supplement IV/97: Professional Association for Industrial Safety (BIA), Sankt Augustin/Erich Schmidt, Bielefeld–Losebl.-Ausg. Bundesgesetzblatt (BGBI) IS 626, 2017, Verordnung zum Schutz vor Gefahrstoffen, [Regulation on the Protection Against Hazardous Substances] Hazardous Substances Ordinance— GefStoffV”.

BG BAU, 2011, Branchenlösung “Asphaltbeläge staubarm abtragen mit Kaltfräsen” [branch solution Dust-free ablation of asphalt pavements with cold milling machines]: http://www.bgbau.de, Webcode 3096458. TRGS 402, 2010, Ermitteln und Beurteilen der Gefährdungen bei Tätigkeiten mit Gefahrstoffen: Inhalative Exposition [Identifying and Assessing the Risks in Activities Involving Hazardous Substances: Inhalation Exposure]: Bundesanstalt für Arbeitsschutz und Arbeitsmedizin (BAuA), Berlin. TRGS 500, 2008, Schutzmaßnahmen [Protective Measures]: Bundesanstalt für Arbeitsschutz und Arbeitsmedizin (BAuA), Berlin. TRGS 517, 2013, Tätigkeiten mit potenziell asbesthaltigen mineralischen Rohstoffen und daraus hergestellten Gemischen und Erzeugnissen [Activities with Potentially Asbestos Containing Minerals and Mixtures and Products Manufactured from Same]: Bundesanstalt für Arbeitsschutz und Arbeitsmedizin (BAuA), Berlin. World Health Orhanization (WHO), International Programme on Chemical Safety & WHO Task Group on Asbestos and other Natural Mineral Fibres, 1986, Asbestos and other natural mineral fibres: Published under the joint sponsorship of the United Nations Environment Programme, the International Labour Organisation, and the World Health Organization.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 67–71

Fibrous Tremolite in Central New South Wales, Australia MARC HENDRICKX* Marc Hendrickx and Associates, Australia, P.O. Box 61, Berowra Heights, New South Wales, Australia, 2082

Key Terms: NOA, Asbestos, Tremolite, Elongated Mineral Particles, Australia, Toxicology ABSTRACT Tremolite schists in Ordovician meta-volcanic units in central New South Wales (NSW) consist of fine fibrous tremolite-actinolite. They host tremolite asbestos occurrences, and small quantities of asbestos were mined from narrow vein deposits in central NSW during the last century. When pulverized, the tremolite schist releases mineral fragments that fall into the classification range for countable mineral fibers and may be classed as asbestos despite not having an asbestiform habit. The ambiguity in classification of this type of natural material raises significant health and safety, legal, and environmental issues that require clarification. While the health effects of amphibole asbestos fibers are well known, the consequences of exposure to non-asbestiform, fibrous varieties is not well studied. This group of elongated mineral particles deserves more attention due to their widespread occurrence in metamorphic rocks in Australia. Toxicological studies are needed to assess the health risks associated with disturbance of these minerals during mining, civil construction, forestry, and farming practices. INTRODUCTION

This article documents preliminary results of a geological assessment for natural occurrences of asbestos (NOA) undertaken in Rockley and Byng Volcanics between Orange and Dog Rock State Forest in central NSW (Figure 2). The assessment included review of available geological information, such as new NOA potential maps (HACA, 2015a, 2015b), and included new geological mapping and assessment of regional airborne geophysics (aeromagnetic and radiometric datasets) to improve base geology maps along with sampling and mineral analysis. The work was undertaken following discovery of tremolite asbestos veins during minor roadworks (Figure 3). GEOLOGICAL SETTING The area occurs in the eastern Lachlan Fold Belt, a major geological province in eastern Australia that was active from the Cambrian to the Carboniferous. The geological history of the area includes marine sedimentation in troughs and basins adjacent to a long volcanic arc that was active through the Ordovician to early Silurian. Ordovician volcanism and sedimentation across the central Lachlan Fold Belt was terminated in the early Silurian by a regional deformation event termed the Benambran Orogeny. This caused extensive shortening across the region, folding and

Lower Ordovician meta-volcanics in central New South Wales (NSW) are host to minor asbestos occurrences. Some of these were mined in the first half of the 1900s, yielding small quantities of poorquality tremolite asbestos used mainly to line boilers (Hendrickx, 2009). The host rock to asbestos occurrences are tremolite-actinolite-chlorite schists in the Rockley, Byng, and Sofala Volcanics that form distinctive outcrops throughout the region (Figure 1). When crushed for industrial use as a road construction material, the schists release crystal fragments with size and aspect ratios that fall into the classification for countable fibers and may be classed as asbestos. These elongated mineral particles (EMPs) raise health and safety, legal, and environmental issues that require clarification. *Corresponding author email: marchgeo@gmail.com

Figure 1. Outcrop of tremolite-actinolite-chlorite schist from Byng Volcanics, near Orange, central NSW (−33.27537, 149.19834).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 73–77

Hendrickx

Figure 2. Geological map highlighting location of outcropping Byng and Rockley Volcanics and other locations mentioned in the text. From Pogson and Watkins (1998).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 73â&#x20AC;&#x201C;77

Fibrous Tremolite in NSW, Australia

below. In thin section, the tremolite schist comprises altered clinopyroxene phenocrysts in a strongly foliated matrix of laths of tremolite, chlorite and very fine, fibrous and asbestiform tremolite groundmass. Relict pyroxene crystals in the tremolite schist suggest a mafic volcanic origin, possibly a basaltic lava. ASBESTOS OCCURRENCES

Figure 3. Tremolite-actinolite-chlorite schist in Rockley Volcanics, Dog Rocks Road. Inset shows narrow slip ﬁber vein with tremolite asbestos (−33.73434, 149.58693).

faulting the rocks, and was accompanied by granite intrusion and regional metamorphism to actinolitebiotite grade (greenschist facies). Ultramafic and mafic volcanic and intrusive rocks were metamorphosed during this event, and primary igneous mineral assemblages were replaced by metamorphic minerals, including amphiboles (Pogson and Watkins, 1998). Dog Rocks Area Geological assessments for NOA were undertaken in a number of areas across the region underlain by Rockley and Byng Volcanics. This report focuses on results from the Dog Rocks State Forest area, approximately 35 km south of Bathurst in central NSW (see Figure 2) 150 km west of Sydney. Late Ordovician Rockley and Byng Volcanics form a mixed unit comprising metamorphosed siltstone, shale, chert, and sandstone along with metamorphosed mafic and ultramafic volcanic rocks (meta-basalt, meta-gabbro, amphibolite, pyroxenite, and serpentinite with tremolite-chlorite schist). Byng Volcanics include a few larger metamorphosed ultra-mafic and serpentinite bodies that contain both tremolite and chrysotile asbestos (Hendrickx, 2009). The Rockley Volcanics occur in the Dog Rocks area (Figure 2) and are divided into four unnamed units— Ocr, Ocrc, Ocru, and Ocrs—based on the dominant lithology (Pogson and Watkins, 1998). Asbestos mineralization occurs mainly in Ocru, which comprises tremolite-actinolite-chlorite schist (Figure 3) and minor meta-basalt. Approximately 30 narrow veins (<2 cm across) of asbestiform tremolite (slip fiber and cross fiber) were located in tremolite-chlorite schist and meta-basalt along the road across a 1-km section (Figure 3), and these are described in more detail

Asbestos was first reported in the Rockley area at Briar Park in 1877, just west of the Dog Rocks State Forest. Asbestos was also worked at Sewells Creek, and 50 tons was reportedly produced in 1942 but not sold. The location of this mined material is not known, but it may be the source of dumped mine waste material found in the area during this work. The main type of asbestos is tremolite-forming discrete narrow (up to 2 cm wide) slip fiber veins in metamorphosed mafic volcanics (tremolite-actinolitechlorite schist). Fibrous tremolite is also present within the fine-grained matrix of tremolite-chlorite schist host rock. This is different in style to serpentinite-hosted asbestos deposits found elsewhere in NSW. Serpentinitehosted deposits, such as those from the Coolac-Tumut area, consist of very fine closely spaced veins of cross fiber and slip fiber chrysotile disseminated through the host rock (Hendrickx, 2009). The asbestos vein system at Dog Rocks is inferred to form a semi-continuous elongate zone parallel to regional structures. It is inferred that this vein system has been truncated by a prominent transverse fault structure where tremolite asbestos was mined at Sewell’s Creek. The fault structure appears to be unconformably overlain by Late Silurian meta-sedimentary rocks to the east of Dog Rocks State Forest. This suggest the asbestos mineralization predates faulting and is likely associated with the early Silurian Benambran Orogeny; hence, younger rocks are unlikely to have this style of asbestos mineralization. Scanning electron microscopy (SEM) with energy dispersive X-ray spectroscopy (EDS) assessment on tremolite schist and asbestos veins from the Dog Rocks Road section was undertaken. Ground samples of tremolite schist broke down to a mix of fine tremolite cleavage fragments and non-fibrous crystal fragments (Figure 4a, b, and c). The proportion of fibers classed as respirable ranged up to about 1%–5% for any given field of view on the SEM. Generally, the length of these fibers was between 7 and 20 μm, with typical aspect ratios of about 10–20. Further SEM analysis is required to determine the precise mineralogical characteristics of this material. SEM imagery of asbestiform fibers from the veins confirms the asbestiform nature of the vein material. Aspect ratios >100 are very common. Individual fiber

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 73–77

Hendrickx

Figure 4. SEM images of crushed schist (a, b, and c) note acicular and elongate crystal fragments and tremolite asbestos (d).

bundles are clearly observed breaking into ﬁner ﬁbrils approximately 0.1–0.2 μm wide (Figure 4d).

DISCUSSION The definition of a “countable fiber” includes elongate and fibrous mineral fragments that do not necessarily have an asbestiform habit. These fragments have been defined as EMPs (ANSES, 2014). Detailed testing using SEM or transmission electron microscopic methods help to differentiate EMPs from true asbestiform particles. However, ambiguity in what constitutes a “fiber” and lack of information and some uncertainty about the toxicological properties of these elongate amphibole mineral particles are hampering concise assessment of the health risks. There are mixed views about the health dangers posed by respirable non-asbestiform amphiboles.

Williams et al. (2013) concluded, “No evidence of demonstrable cancer effects from exposure to nonasbestos amphiboles that may be counted as fibers, under certain assessment protocols, was found.” However a review by the French Agency for Food, Environmental, and Occupational Health and Safety (ANSES, 2014) concluded, There is no reason to make a distinction between the cleavage fragments meeting the “WHO” dimensional criteria for fibers (L > 5 µm; D < 3 µm and L:D > 3:1) and asbestiform fibers of calcic and sodic-calcic EMPs (Elongated Mineral Particle), in particular due to the uncertainties and difficulties related to their characterisation and to their differentiation by routine analytical methods.

The consequences of confusion about “ﬁber” terminology and classiﬁcation were highlighted by Thompson et al. (2011), who showed for the United States

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 73–77

Fibrous Tremolite in NSW, Australia

that “based on the regulatory definition, 13% of soil pedons and 5% of soil horizons in the U.S.A. are ‘naturally contaminated.’ ” Similar issues arise with confusion over terminology and the widespread occurrence of amphiboles in the geological environment in eastern Australia. CONCLUSIONS NOA occurrences in Ordovician meta-volcanic rocks in central NSW comprise tremolite asbestos and EMPs, which pose a problem of classification when working on the assessment of NOA hazard. EMPs have not been fully considered in assessments of NOA exposure risk, and further work is required to better understand the distribution and health risks associated with these non-asbestiform amphiboles, including toxicological studies. REFERENCES ANSES, 2014, Health Effects and the Identification of Cleavage Fragments of Amphiboles from Quarried Minerals: French Agency

for Food, Environmental, and Occupational Health and Safety Opinion Request No. 2014_SA_0196 MaisonsAlfort. https://www.anses.fr/en/system/ﬁles/AIR2014sa0196 RaEN.pdf HACA, 2015a, Mapping of Naturally Occurring Asbestos in NSW– Known and Potential for Occurrence: Heads of Asbestos Coordination Authority, NSW Trade and Investment, Division of Resources and Energy. HACA, 2015b, Naturally Occurring Asbestos in NSW NOA Potential Maps: Heads of Asbestos Coordination Authority, NSW Trade and Investment, Division of Resources and Energy. https://trade.maps.arcgis.com/apps/PublicInformation/index. html?appid=87434b6ec7dd4aba8cb664d8e646fb06 Hendrickx, M., 2009, Naturally occurring asbestos in eastern Australia: A review of geological occurrence, disturbance and mesothelioma risk: Environmental Geology, Vol. 57, No. 4, pp. 909–926. Pogson, D. J. and Watkins, J. J., 1998, Bathurst 1:250000 Geological Sheet SI/55-8: Explanatory Notes: Geological Survey of New South Wales, Sydney. Thompson, B. D.; Gunter, M. E.; and Wilson M. A., 2011, Amphibole asbestos soil contamination in the U.S.A.: A matter of deﬁnition: American Mineralogist, Vol. 96, pp. 690–693. Williams, C.; Dell, L.; Adams, R.; Rose, T.; and Van Orden D., 2013, State-of-the-science assessment of non-asbestos amphibole exposure: Is there a cancer risk?: Environmental Geochemistry and Health, Vol. 35, No. 3, pp. 357–377.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 73–77

Management of Naturally Occurring Asbestos Area in Republic of Korea SUNGJUN YOON Chonnam National University, 77 Yongbong-ro, Buk-gu, Gwangju 61186, Republic of Korea

KYUBONG YEOM 11 Doum6-ro, Sejong Special Self-Governing City, Republic of Korea

YONGUN KIM BYUNGNO PARK JAEBONG PARK HYESU KIM HYEONYI JEONG YUL ROH* Chonnam National University, 77 Yongbong-ro, Buk-gu, Gwangju 61186, Republic of Korea

Key Terms: Naturally Occurring Asbestos, Construction, Management, Site Investigations ABSTRACT The Republic of Korea Government has adopted a whole-of-government approach in the management of naturally occurring asbestos (NOA) through a nationwide asbestos management plan. Regional and detailed mapping, and examination of NOA effects are still ongoing for NOA management by indoor air, noise and asbestos management division, Ministry of Environment. Plans by the Korea Rail Network Authority are under way to rebuild the Janghang double-track railway. The proposed Jannghang double-track railway route is through an area of high NOA probability that has serpentine and ultramafic rock. Chrysotile, tremolite, and actinolite asbestos were among the rocks identified within the project site (initial planning line and the operational design line). The level of asbestos in most soils was low (ࣘ0.25 percent), while some soils contained 0.75 percent asbestos. Monitoring and analyses of air quality revealed below 0.01 fibers per cm3 (f/cc). However, there were no traces of asbestos detected in the groundwater and stream water. Despite the low asbestos content of the soil and rock, the disturbance of NOA-containing soils and rocks during railway construction could trigger the release of asbestos fibers into the air. NOA mitigation plans and measures are necessary for workers and residents during the construction of the railway.

*Corresponding author email: rohy@jnu.ac.kr

INTRODUCTION The Korean Peninsula consists of three Precambrian massifs, from north to south, the Nangrim, Gyeonggi, and Yeongnam Massifs, which form the basement rocks. All three Precambrian massifs consist of unclassified high-grade gneiss complexes and overlying supracrustal sequences. The Precambrian massifs are joined by two intervening belts, the Ogcheon and Imjingang belts. The Gyeongsang Supergroup comprises the Gyeongsang basin and other small basins. In the Korean Peninsula, the most recent volcanic activity occurred between the Cretaceous and Early Tertiary periods and was closely related to the contemporaneous plutonism (Figure 1; Kim et al., 1999. Due to these geological characteristics, there are two main ultramafic rocks known to have naturally occurring asbestos (NOA) in the Republic of Korea, serpentinite and metamorphosed carbonate rocks. NOA is formed through the reaction of olivine in ultramafic rocks with hydrothermal fluids or as a result of hydrothermal alteration of dolomite (Lee and Park, 1995; Woo and Kang, 1999; Woo and Suh, 2000; Woo and Kim, 2003; Song et al., 2004; Kim and Woo, 2005; Koh et al., 2006; Park et al., 2012; and Jung et al., 2014, with references). 3Mg2 SiO4 (Olivine) + SiO2 + H2 O → 2Mg3 Si2 O5 (OH)4 (Serpentine) .

(1)

5CaMg(CO3 )2 (Dolomite) + 8SiO2 + H2 O → Ca2 Mg5 Si8 O22 (OH)2 (Tremolite) (2) + 3CaCO3 + 7CO2 .

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

Yoon, Yeom, Kim, Park, Park, Kim, Jeong, and Roh

Figure 1. Geologic map of Korean Peninsula published by KIGAM (1995).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79â&#x20AC;&#x201C;87

NOA in Republic of Korea

Figure 2. A sample of regional NOA maps (Hongsung-gun, scale = 1:50,000).

Development in the NOA areas, where rock and soil contain NOA at construction sites, has resulted to asbestos hazards. Disturbance of the materials within NOA areas increases the possibility of asbestos ﬁbers being released into the air. Diseases that are associated with asbestos occur as a result of cumulative and long-term respiration exposure to airborne asbestos. To address this challenge, the Republic of Korean Government adopts a whole-of-government

approach to managing NOA through a nationwide asbestos management plan. Currently, there are ongoing surveys, including regional and detailed NOA mapping, and examination of NOA eﬀects to aid NOA management. The Korean Ministry of Environment carried out a regional NOA mapping project from the year 2010 to 2017. A detailed NOA mapping project by the Ministry of Environment commenced in 2013 and is still ongoing. According to

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

Yoon, Yeom, Kim, Park, Park, Kim, Jeong, and Roh

Figure 3. A sample of detailed NOA maps (Susan-myeon, scale = 1:25,000).

the regional NOA survey, NOA areas in the Republic of Korea cover about 5.48 percent (2,154 mi2 ) of the land area. Ultramaﬁc rocks, which are the most likely to contain NOA, are found in parts of Chungnam, Kyonggi, and Gyeongbuk provinces. Metamorphosed basic rock, which are scattered locally, are moderately likely to contain NOA. The least likely rocks to contain NOA are metamorphosed carbonate rocks and metamorphic rocks, which are found in parts of Chungbuk, Chungnam, Jeonbuk,

and Gangwon provinces. However, ﬁeld investigation has shown that some of metamorphosed carbonate rocks, which are scattered locally, are likely to contain NOA. The objective of this study is to present the status of regional and detailed mapping of NOA and to examine the eﬀects of NOA in the Republic of Korea and then introduce the Janghang Railway Project as an example of an investigation into development in NOA areas.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

NOA in Republic of Korea

Figure 4. Examination procedures of NOA eﬀects in Republic of Korea.

REGIONAL AND DETAILED MAPPING OF NOA AND EXAMINATION OF NOA EFFECTS The Asbestos Safety Management Act (2012), which contains the general regulations on asbestos, including asbestos-containing materials, asbestoscontaining building materials, and NOA, was enacted in 2012 in the Republic of Korea. It includes a method to conduct both regional and detailed mapping and for examination of NOA effects. The brief field survey and mapping used to generate the regional NOA map (scale = 1:50,000) were based on Ministry of Environment, Korea, Notification 2018-23 (2018a), Mapping Technique of Naturally Occurring Asbestos, Asbestos Safety Management Act. Geologic maps published by the Korea Institute of Geoscience and Mineral Resources (KIGAM, 1995) and the relative likelihood for the presence of NOA in the rocks formed the basis for NOA mapping. The Republic of Korean classification of relative likelihood for the presence of NOA is very similar to that of Placer County, CA, USA (Higgins and Clinkenbeard, 2006). There are four categories of areas of relative likelihood based on rocks as listed below: (1) area most likely to contain NOA, which includes ultramafic rock and serpentine rocks (serpentinite);

(2) area moderately likely to contain NOA, which includes metamorphosed basic rocks; (3) area least likely to contain NOA, which includes metamorphosed carbonate rocks and metamorphic rocks (e.g., schist and gneiss); and (4) fault or shear zones, which may locally increase the likelihood for the presence of NOA where they exist in or adjacent to areas that are most likely to contain NOA (buﬀer zone = 500 m on each side of the fault line). The regional NOA map consists of information such as the relative likelihood for the presence of NOA, defunct asbestos factories, deposits, abandoned asbestos mines, occurrences, and faults (Figure 2). Counties where geologic and soil mapping for detailed NOA mapping (scale = 1:25,000) was done were selected based on the regional NOA map, development, population, etc. Mapping of the detailed NOA map (scale = 1:25,000) was based on Ministry of Environment, Korea, Notiﬁcation 2018-23 (2018a), Mapping Technique of Naturally Occurring Asbestos, Asbestos Safety Management Act. Additionally, it was based on geologic maps published by KIGAM, the regional NOA Map, and geologic and soil surveys. The project is still ongoing. The detailed NOA map includes information such as the area of NOA presence, defunct asbestos factories, deposit locations, abandoned

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

Yoon, Yeom, Kim, Park, Park, Kim, Jeong, and Roh

Figure 5. Geologic map of the Janghang Double-Track Railway Project area.

asbestos mines, occurrences, and faults (Figure 3). Based on the detailed NOA map, analysis of NOA effects is conducted to establish the health risks of residents as a result of exposure to NOA-containing rocks and soils. Examination of NOA effects is based on Ministry of Environment, Korea, Notification 2018-23 (2018b), Method of Examination of NOA Effects, Asbestos Safety Management Act. The examination of NOA effects is divided into two categories: preliminary survey and main survey. The objective of the prelimi-

nary surveys is to determine the need for the main survey by collecting data through analysis of local conditions such as land use, population distribution, literature review on NOA, and ﬁeld surveys. The main survey is dependent on the results of the preliminary survey. The main survey entails investigation of asbestos in rocks, soils, water, and air at the selected area (near or around residential areas) for purposes of assessing the health risks to the local population posed by NOA. The NOA Veriﬁcation Committee should verify the

Table 1. Number of soil, air monitoring, and water samples for NOA survey. Classiﬁcation

Surface Soil (0–0.1 m)

Subsurface Soil (0.1–0.4 m)

Deep Soil (0.4–1 m)

Total

40 68 108

22 19 41

6 9 15

68 96 164

First Sampling

Second Sampling

Third Sampling

Total

14 15 29

42 45 87

Soil Original planning line Working design line Total

Air Original planning line Working design line Total

Water Groundwater Stream water Total

First Sampling

Second Sampling

Total

8 3 11

16 6 22

NOA = naturally occurring asbestos.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

NOA in Republic of Korea Table 2. Number of core and outcrop samples for NOA survey. Classiﬁcation

Borehole Number of Samples

Core Original planning line Tunnel section (L = 310 m) Cutting section (L = 2,684 m) Bridge section (L = 6,490 m) Working design line Tunnel section (L = 1,145 m) Cutting section (L = 4,869 m) Bridge section (L = 4,520 m) Total Outcrop Original planning line Working design line Total

2 18 32

10 67 38

30 38 87 207

153 255 87 610 16 25 41

NOA = naturally occurring asbestos.

results of the preliminary survey. The committee then decides whether or not to conduct the main survey. The main survey should be conducted urgently if deemed necessary (Figure 4). JANGHANG DOUBLE-TRACK RAILWAY PROJECT The Korea Rail Network Authority is planning to rebuild the Janghang double-track railway. The proposed Janghang double-track railway will traverse an area of high NOA probability with rocks such as serpentinite or ultramaﬁc rock. Based on the geological maps of Hongsong and Daecheon (KIGAM), serpentinite or ultramaﬁc rocks occur as exposed, isolated bodies in the Precambrian Kyeonggi gneiss complex located within the railway project area (Figure 5). Fur-

thermore, there are abandoned asbestos mines such as the Singokri mine and the Daeheung mine adjacent to the proposed railway track. The local media and environmental lobby groups have raised concerns regarding the generation of dust potentially containing asbestos during implementation of the project. Soil, air, and water samples were collected and analyzed in compliance with the Korean Ministry of Environment survey guidelines to evaluate the general risk as a result of the NOA (Table 1). Surface soil sampling, ambient air monitoring, and water sampling were conducted, and the asbestos was subsequently analyzed to determine the concentration of NOA within the project area. Rock core drilling and field surveying for the rock samples were performed and analyzed to determine qualitative amounts of asbestos (Table 2). Polarized light microscope (PLM), X-ray diffraction (XRD), phase contrast microscope (PCM), scanning electron microscope and energy dispersive Xray spectroscope (SEM-EDS), and transmission electron microscope and energy dispersive X-ray spectroscope (TEM-EDS) analyses were utilized to characterize the types and contents of the asbestos. PLM and XRD analyses of the rocks and soils revealed that the rocks and soils contained chrysotile, tremolite, and actinolite asbestos (Tables 3 and 4). The asbestos concentrations of soil in the project area varied depending on the type of geologic unit and ranged from below detection limit to about 1percent as analyzed by the 400 point-counting method of CARB 435 using PLM (Table 4). The majority of the soils’ asbestos content was low (ࣘ 0.25 percent). SEM-EDS analyses of the rocks and soils revealed that tremolite asbestos was more fibrous as compared with chrysotile and actinolite asbestos, which can grade into asbestiform morphology (Figure 6). PCM and TEM analyses

Table 3. Asbestos analyses of core and outcrop samples.

Classiﬁcation

Borehole

Core Original planning line Tunnel section Cutting section Bridge section Working design line Tunnel section Cutting section Bridge section Total Outcrop Original planning line Working design line Total

Boreholes with Asbestos Detected

Number of Samples

Samples of Asbestos Detected

Detected Asbestos Type Chysotile, tremolite, actinolite

2 18 32

1 6 13

10 67 38

1 13 14

30 38 87 207

19 27 12 78

153 255 87 610

49 72 12 161

16 25 41

1 4 5

Chrysotile, tremolite, actinolite

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

Yoon, Yeom, Kim, Park, Park, Kim, Jeong, and Roh Table 4. Quantitative analyses of asbestos in soil by 400 point-counting method using polarized light microscope. Classiﬁcation Original planning line Working design line Total

Number of Samples

Not Detected

<0.25%

0.25%–1.0%

>1.0%

Detected Asbestos Type

68 (40)* 96 (68) 164 (108)

42 (21) 78 (52) 120 (73)

21 (14) 17 (15) 38 (29)

5 (4) 1 (1) 6 (5)

— — —

Chysotile, tremolite, actinolite

* Sampling station number indicated by ( ).

of ambient air samples recorded lower than 0.01 ﬁbers per cm3 (f/cc), which is below the legal threshold (0.01 f/cc), and asbestos was not detected in any of the water samples (Table 5). Although the asbestos content of the soil was below the legal threshold (1 percent), the dust mitigation measures and work management practices adopted for the project plans should keep generation of any such dust to a minimum. Additionally, the air-quality monitoring program, which is integrated with dust-control measures, should guarantee that asbestos-containing dust does not leave the site at concentrations suﬃcient to pose health risks to workers or the general public.

CONCLUSIONS The NOA survey project, which entails regional/detailed NOA mapping and analysis of NOA eﬀects, is currently ongoing for purposes of NOA management in the Republic of Korea. A survey for detailed NOA mapping (scale = 1:25,000) is currently ongoing based on the Ministry of Environment, Korea, Notiﬁcation 2018-23 (2018a), Mapping Technique of NOA, Asbestos Safety Management Act. The Korea Rail Network Authority is planning to rebuild the Janghang double-track railway. The proposed Janghang double-track railway is expected to traverse

Figure 6. SEM-EDS analyses of chrysotile, actinolite, and tremolite asbestos.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

NOA in Republic of Korea Table 5. Asbestos analyses of air and water samples using phase contrast microscope and transmission electron microscope (TEM). PCM (f/cm3 ) Classiﬁcation Air Number of ﬁrst sampling Number of second sampling Number of third sampling Total Water Groundwater Stream water Total

TEM (f/cm3 )

>0.01

0.01–0.0022

< Detection limit

>0.01

0.01–0.001

0.001–0.0001

< Detection limit

0 0 0 0

9 14 4 27

20 15 25 60

0 0 0 0

4 4 1 9*

Number of ﬁrst sampling 8 3 11

TEM (>7 MFL) 0 0 0

Number of second sampling 8 3 11

TEM (>7 MFL**) 0 0 0

*9 air samples were randomly selected from 87 air samples for TEM analysis. **MFL: million ﬁber per liter.

an area of high NOA probability, which has serpentine or ultramaﬁc rock. Although the asbestos content of the soil and rock is low, the dust mitigation measures and work management practices adopted for project plans should maintain generation of any such dust to a minimum. In addition, the air-quality monitoring program, which is integrated with dust-control measures, should guarantee that asbestos-containing dust does not leave the site at concentrations high enough to pose health risks to workers and the general public. ACKNOWLEDGMENT We are grateful to Mr. Jeong in CCRF for XRD and Dr. Moon in KBSI-Gwangju for SEM-EDS analyses. REFERENCES Asbestos Safety Management Act, 2012, Asbestos Safety Management Act: Electronic document, available online at http://www.law.go.kr/lsInfoP.do?lsiSeq = 199137&efYd = 20180529#0000 (In Korean.) Higgins, C. T. and Clinkenbeard, J. P., 2006, Relative Likelihood for the Presence of Naturally Occurring Asbestos in Placer County, California: California Geological Survey Public Information Oﬃces Special Report 190. Jung, H. M.; Shin, J. D.; Kim, Y. M.; Park, J. B.; and Roh, Y., 2014, Mineralogical characteristics of naturally occurring asbestos (NOA) at Daero-ri, Seosan, Chungnam, Korea: Economic and Environmental Geology, Vol. 47, No. 5, pp. 467–477. (In Korean.) Kim, J. H.; Na, G. C.; Yang, S. Y.; Won, J. K.; Lee, D. Y.; Lee, J. H.; Lee, C. J.; Choi, D. K.; Choi, H. I.; Yoon, S.; Park, C. E.; Jeon, M. S.; Kim, H. S.; Oh, M. S.; Park, Y. A.; and Jo, S. K., 1999, Geology of Korea: Sigma Press Book Company, Seoul, Korea, 802 p. Kim, Y. T. and Woo, Y. K., 2005, Serpentinization of olivine and pyroxene in Chungnam serpentinites, Korea: Journal of Korean Earth Science Society, Vol. 26, No. 3, pp. 297–304. (In Korean.)

Koh, S. M.; Park, C. K.; and So, W. J., 2006, Preliminary study on the formation environment of serpentinite occurring in Ulsan area: Journal of the Mineralogical Society of Korea, Vol. 19, No. 4, pp. 325–336. (In Korean.) Korea Institute of Geoscience and Mineral Resources (KIGAM), 1995, Geological Maps: Electronic documents, available online at https://mgeo.kigam.re.kr/map/main. do?process=geology_50k (In Korean.) Lee, J. J. and Park, B. J., 1995, A geochemical study on genesis of Ulsan serpentine deposits: Journal of the Korean Earth Science Society, Vol. 16, No. 1, pp. 9–19. (In Korean.) Ministry of Environment, Korea, Notification 2018-23, 2018a, Mapping Technique of Naturally Occurring Asbestos: Electronic document, available online at http:// www.law.go.kr/admRulInfoP.do?admRulSeq=2100000116349 (In Korean.) Ministry of Environment, Korea, Notification 201823, 2018b, Method of Examination of NOA Effects: Electronic document, available online at http://www. law.go.kr/admRulInfoP.do?admRulSeq= 2100000135990 (In Korean.) Park, G. N.; Hwang, J. H.; Oh, J. H.; and Lee, H. M., 2012, Occurrence and mineralogy of serpentinite from Bibong Mine in Chungyang area, Korea: Journal of Mineralogical Society of Korea, Vol. 25, No. 1, pp. 9–21. (In Korean.) Song, S. H.; Choi, S. K.; Oh, C. H.; Seo, J. E.; and Choi, S. H., 2004, Petrography and geochemistry of the ultramafic rocks from the Hongseong and Kwangcheon areas, Chungcheongnam-Do: The Korean Society of Economic and Environmental Geology, Vol. 37, No. 5, pp. 477–497. (In Korean). Woo, Y. K. and Kang, H. J., 1999, Alteration of serpentinites on Weolhyeon serpentine ore deposits in Hongseong County, Choongnam, Korea: Journal Korean Earth Science Society, Vol. 20, No. 2, pp. 189–198. (In Korean.) Woo, Y. K. and Kim, S. H., 2003, Original rock and serpentinization of serpentinites on serpentine ore deposits in Cheongyang, Choongnam, Korea: The Korean Association for Science Education, Vol. 34, pp. 181–196. (In Korean.) Woo, Y. K. and Suh, M. C., 2000, Petrological study on the ultramafic rocks in Choongnam area: Journal Korean Earth Science Society, Vol. 21, No. 3, pp. 323–336. (In Korean.)

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 79–87

Identiﬁcation and Preliminary Toxicological Assessment of a Non-Regulated Mineral Fiber: Fibrous Antigorite from New Caledonia JASMINE RITA PETRIGLIERI* Department of Earth and Environmental Sciences, University of Milano-Bicocca, Piazza della Scienza 4, 20126 Milano (Italy)

CHRISTINE LAPORTE-MAGONI Institute of Exact and Applied Sciences, University of New Caledonia, Campus de Nouville, BP R4, 98851, Nouméa Cedex, New Caledonia (France)

EMMA SALVIOLI-MARIANI Department of Chemistry, Life Sciences and Environmental Sustainability, University of Parma, Parco Area delle Scienze 157/A, 43124 Parma (Italy)

MAURA TOMATIS Department of Chemistry and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and other Toxic Particulates, University of Torino, Via P. Giuria 7, 10126 Torino (Italy)

ELENA GAZZANO Department of Oncology and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and other Toxic Particulates, University of Torino, Via Santena 5, 10126 Torino (Italy)

FRANCESCO TURCI Department of Chemistry and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and other Toxic Particulates, University of Torino, Via P. Giuria 7, 10126 Torino (Italy)

ALESSANDRO CAVALLO Department of Earth and Environmental Sciences, University of Milano-Bicocca, Piazza della Scienza 4, 20126 Milano (Italy)

BICE FUBINI Department of Chemistry and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and other Toxic Particulates, University of Torino, Via P. Giuria 7, 10126 Torino (Italy)

Key Terms: Fibrous Antigorite, NOA, Weathering, Toxicity, New Caledonia ABSTRACT The rising awareness about the risk due to asbestos environmental exposure has led to a new interest in the

*Corresponding author email: jasmine.petriglieri@gmail.com

investigation of non-regulated mineral fibers. Evidence of chronic diseases has been described in individuals exposed to naturally occurring asbestiform (NOA) minerals in Turkey (erionite), Italy (fluoro-edenite), and the United States (winchite/richterite). In New Caledonia, an increased incidence of asbestos-related diseases was correlated with the natural occurrence of fibrous serpentines chrysotile and fibro-lamellar antigorite in outcrops, roadways, and soils. A minor amount of tremolite asbestos was also observed, increasing the health

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Petriglieri, Laporte-Magoni, Salvioli-Mariani, Tomatis, Gazzano, Turci, Cavallo, and Fubini

hazard. By adopting a precautionary principle, New Caledonia legislation classified antigorite as regulated asbestos, even if limited toxicity assessment is available. Caledonian antigorite exhibits a wide range of natural shapes, morphologies, and degrees of alteration as a result of pedogenic alteration induced by subtropical conditions. As the alteration increases, lamellar antigorite gradually cleaves into fibrous-like particles, assuming a fibro-lamellar habit. An increase in the emission of inhalable (potentially asbestiform) fibers in air was observed. To understand this mechanism, a multidisciplinary mineralogical and geochemical investigation was carried out. Additionally, several in vitro tests have been performed on three antigorite samples, subjected to different levels of alteration, to collect preliminary information on antigorite toxicity. Alteration modifies the surface reactivity of antigorite. The circulation of fluids induces a mechanical stress and an elemental exchange at mineral/water interface, promoting the loss of cohesion of the mineral structure and affecting the surface chemistry and toxicity of fibrous (asbestiform) antigorite. INTRODUCTION Inhalation is the primary route of exposure of mineral fibers that becomes a cause of concern in the case of exposure from natural deposits of asbestos (IARC, 2012). Asbestos fibers may be released from asbestosbearing deposits and, without appropriate dust management, may pose a potential health hazard when rocks are crushed or exposed to natural weathering and erosion or to human activities. Natural contexts are therefore unconfined sites of study with a great intrinsic diversity, not only related to the activities that can cause the suspension of mineral fibers but also to environmental sources. Lee et al. (2008) emphasize how difficult it is to reliably correlate the presence of mineral outcrops belonging to the carcinogenic mineral fibers (Group I = Carcinogenic to humans; IARC, 2012) and the impact on health. This depends upon the different physico-chemical properties, the amount of the fibers emitted by each source, and the local environmental conditions (IARC, 2012; Turci et al., 2016; and Erskine and Bailey, 2018). In the past few decades, epidemiological in vitro and in vivo studies have linked chronic diseases to the presence of nonasbestos fibrous minerals. A high-profile case is the example of the mesothelioma epidemic in Cappadocia (Turkey), where the impact on the health of exposed people was observed before the fibrous minerals responsible for the epidemic could be determined; these minerals were finally discovered to be fibrous erionite (Carbone et al., 2011), a zeolite more carcinogenic than the six regulated asbestos minerals. This led to a

new interest in the scientific community in investigating potentially hazardous non-asbestos fibrous minerals (e.g., balangeroite; Gazzano et al., 2005; Turci et al., 2005). The lack of a comprehensive scientific knowledge on the toxicology of non-regulated fibrous minerals makes it difficult to assess the potential risk due to environmental exposure. New Caledonia provides a good example with which to assess the toxicity of fibrous antigorite, considering the impact of pedogenesis on the formation and release of these fibers into the environment. ASBESTOS HEALTH HAZARDS IN NEW CALEDONIA Located in the southwest Pacific Ocean, in a complex set of marginal basins and continental or volcanic ridges along the Circum-Pacific Belt, the island of New Caledonia is one of the world’s largest producers of nickel ores formed by the alteration of ultramafic rocks. The New Caledonia ophiolite complex is one of the largest (300 km long, 50 km wide, and 2 km thick) and best-exposed continuous peridotite complexes in the world, covering more than a third of the land area (Ulrich et al., 2010). The investigation of naturally occurring asbestiform (NOA) in New Caledonia started after the diagnosis of asbestos-related pathologies in human populations that had been non-occupationally exposed to asbestos (Luce et al., 2004). An excess of malignant mesothelioma, observed in the 1980s in the northern Kanak communities, was associated with the use of “Pö,” a tremolite-containing whitewash (Goldberg et al., 1991). Other cases of mesothelioma and pleural cancer were noted through the year 2008, affecting people associated with mining sites and municipalities. Recently, Baumann et al. (2011) linked these cases of lung malignancies to the presence of serpentinite outcrops, rich in chrysotile and fibrous antigorite. Caledonian populations, living and/or working in proximity to natural outcrops, are therefore subjected to a double environmental and domestic exposure. In this scenario, mining companies need to implement NOA risk management in order to protect workers, sites, and residents. In the assessment of risk of exposure, an extensive geological survey of the different (fibrous) varieties of amphibole and serpentine present in the outcrops was performed (Lahondère, 2012, and references therein). The natural occurrences of asbestos and related fibrous minerals were overlain onto a detailed geological map (Figure 1; DIMENC-SGNC, 2010). As a result, most outcrops of nickel (Ni)–laterite deposits are found to contain serpentine and amphibole, not infrequently occurring as fibrous (asbestiform) varieties.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Fibrous Antigorite from New Caledonia

Figure 1. Geological sketch map of natural occurrences of fibrous-asbestiform minerals in New Caledonia. The three major sites of nickel-mining activity are magnified (modified after DIMENC-SGNC, 2010). Sampling sites are indicated with a square.

While tremolite-amphibole is mainly present in central and northern New Caledonia terranes, serpentine chrysotile and fibro-lamellar antigorite occur in peridotites (Lahondère, 2012). The large distribution of fibrous antigorite over a large part of the island makes its environmental exposure a potential public health issue for New Caledonia (Laporte-Magoni et al., 2018). To deal with this occupational and environmental issue, the Government of New Caledonia legislated and promulgated its first regulation on asbestos materials (Délibération N°82 du 25 aout 2010). In contrast to European and worldwide asbestos regulations, the New Caledonia decree classifies serpentine antigorite as asbestos, on a precautionary basis. It is worth noting that the regulation makes no distinction between antigorite and fibrous (asbestiform) antigorite. Moreover, this legislative text does not specify an analytical method for the identification and quantification of fibers emitted, relying in this respect on French regulations (NF X43 269). It should be noted

that no standard samples exist for measurement of airborne antigorite ﬁber concentration, which has led to some diﬃculties in asbestos risk prevention and management. Finally, the New Caledonian decree, similar to the vast majority of asbestos regulations currently enforced, does not provide a guideline about the management of the NOA risk.

NOA OCCURRENCES IN LATERITIC UNITS In Caledonian ultrabasic units the serpentinized peridotite assemblages exhibit the widespread presence of serpentine minerals, combined with minor amounts of tremolite-actinolite amphibole. Owing to its ability to better withstand the oxidation processes, ﬁbrous serpentine is commonly found in the saprolitic zones currently mined for Ni (Trotet, 2012). Serpentine occurs along tectonic structural discontinuities as fractures, faults, and shear zones, likely due to

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Petriglieri, Laporte-Magoni, Salvioli-Mariani, Tomatis, Gazzano, Turci, Cavallo, and Fubini

different thermodynamic conditions according to the geodynamic context. When exposed to natural weathering, NOA-bearing rocks are subjected to a humid tropical to subtropical climate, influenced by trade winds, and an alternating hot-dry and rainy-cool season. Under these climate conditions, natural deposits of asbestos are subject to a secondary process of alteration. As a result, mineral fibers occur with different morphologies, likely connected to different degrees of alteration. In this context, the term “alteration” refers to a physicomechanical modification in the shape and cohesion of rock fabric. With an increase in the degree of alteration, massive assemblages gradually cleave into lamina or needle-like acicular crystals. This progressive loss of cohesion leads to the disappearance of the original structure and, conversely, increases the appearance of individual asbestos-like fibers. Minerals that have been subjected to alteration may vary from prismaticplaty to asbestiform through acicular-lamellar. A complete mineralogical-petrological (optical microscopy, scanning electron microscopy [SEM], transmission electron microscopy, micro-Raman) and geochemical (major and trace elements) approach was applied to the structural, chemical, and morphological characterization of fibro-lamellar antigorite (Petriglieri et al., 2019). Thirty-five rock fragments collected at mining sites (outcrops, quarries, tracks, pits) of different ultrabasic units were analyzed (Figure 1). FIBRO-LAMELLAR ANTIGORITE Serpentinized peridotites show a large network of fault planes and veins containing lamellar crystals of antigorite, measuring centimeters to decimeters in length, cross-cut by more or less continuous veinlets of chrysotile. In the less altered areas, generally at the base of the saprolite horizon, antigorite blades show a compacted, moderately hardened appearance, dominated by a pale-green to green color. Platyshaped lamellae are welded and parallel to each other (Figure 2A). Moving up in the regolith profile, antigorite occurs in the form of stacks of laminas exhibiting a progressively more friable aspect. Blades appear fragmented and are associated with fibers. These fibers may originate from the extreme cleavage (fraying) of the same lath-shaped crystals (Figure 2B and C). Antigorite assumes a fibro-lamellar habit in highly altered horizons. As a result of strong mechanical separation and cleavage, antigorite has a completely transformed morphological appearance and is associated with a porous low-density material. The friable nature of these specimens is evident (Figure 2D). Therefore, antigorite, non-fibrous when fresh, gradually cleaves with pedogenic alteration, presenting fibrous-like par-

ticles, which are not strictly asbestos fibers, according to legal and commercial definitions, but their fibrousasbestiform nature may have a potential impact on human health. In the evaluation of morphological and textural features, optical and SEM images of Caledonian antigorite show an intimate intergrowth of lamellar and fibro-lamellar shapes (Figure 3). Several key distinctions relating to the mineralogy, texture, and alteration state of antigorite were obtained through the examination of petrographic thin sections. According to Wicks and Whittaker (1977), antigorite is typically recognized for its non-pseudomorphic “interpenetrating” or “interlocking” texture (Figure 3B). Actually, Caledonian samples consist of randomly oriented aggregates of fibro-lamellae and show a wider variety of shapes and intergrowths. They appear as star- and fan-formed aggregates (Figure 3A), lathshaped lamella (Figure 3C), and fibro-lamellar blades (Figure 3D). Even a single sample can display the coexistence of several different textures. Although the two-dimensional nature of petrographic thin sections makes it difficult to distinguish the crystal habit (e.g., fibrous, acicular, and lamellar), polarized light microscopy observations allow one to evaluate the intergrowth of different fibrous or non-fibrous phases in their textural context. Samples that appear massive, lamellar, and unaltered in hand sample commonly display their fibrous shape at the optical microscopy scale. Increasing the magnification, SEM images display the huge morphological variability exhibited by Caledonian antigorite, which has the form of fibrolamellar crystals, characterized by the co-existence of both lamellar and fibro-lamellar shapes (Figure 4). Bundles of parallel, elongated, lath-shaped crystals exhibit the typical habit of phyllosilicate minerals, characterized by the overlapping of platy sheets. However, aggregates of randomly oriented non-elongated blades may also occur. Most particles maintain their lamellar habit, displaying a gradually bent, slinky to curvilinear ending. IMPACT OF PEDOGENIC ALTERATION ON FIBER RELEASE Physical and mechanical stress appears to be one of the main reasons for the various degrees of alteration displayed by the mineral fibers of New Caledonia. As the alteration increases, a gradual increase in distance between closely overlapped fibers and/or fibrolamellae occurs, resulting in a greater macro-porosity. This is likely related to the circulation of surface water, which percolates down and penetrates rocks, permeating cracks, fractures, and shear zones. It is proposed that the penetration of fluids within fibrils is thus

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Fibrous Antigorite from New Caledonia

Figure 2. Macroscopic features of hand-scale antigorite samples. (A) Massive-lamellar fragment; (B–C); stacks of fragmented blades associated with ﬁbers; and (D) friable-dusty aggregate of ﬁbro-lamellae. An evident lack of cohesion and a very altered appearance characterizes the New Caledonia rock fragments.

favored and causes a chemical elemental exchange at the mineral/water interface, creating a severe mechanical stress that results in the complete loss of cohesion of the original structure. To evaluate the role of chemical element exchange with regard to its capacity to break apart and disperse antigorite fibers, a preliminary geochemical investigation was conducted. In this context, the main chemical reactions involved at the crystal/water interface are dissolution, redox reactions, hydration, decarbonation, and the most common reaction, hydrolysis. Thus, the most soluble elements may be leached by water (e.g., magnesium), leading to the dissociation of fibrous minerals and consequently favoring the emission of fibers. It should be remembered that the variation in element solubility is strictly related to the type of element and silicate mineral involved in the mineral/water reactions. The study of major and trace element concentrations represents a first tracer of the impact of weathering on altered rocks. Analyses of major and trace elements were conducted using optical and mass spectrometry (ICP-OES and ICP-MS). Chemical signatures of (fibrous) antigorite

reveal a systematically lower MgO and higher FeOtot content compared to what is typically reported in the scientific literature (from 35 to 45, and 2 to 5 wt.%, respectively; Deschamps et al., 2013; Cannaò et al., 2016). Additionally, a higher concentration of chromium, manganese, cobalt, vanadium, scandium, copper, and Ni, was observed. An advanced stage of alteration is observed for all antigorite specimens, as well as for samples macroscopically observed to be unaltered. These results are consistent with the laterization process involved in Ni-ore deposit formation (Butt and Cluzel, 2013). POTENTIAL TOXICITY OF FIBROUS ANTIGORITE To date, only preliminary data on the potential toxicity of fibrous antigorite are available (ANSES, 2014, and references therein). To better assess its pathogenicity, a set of in vitro cell-free and cellular tests were performed. To this purpose, three antigorite samples presenting different levels of cohesion (from little to highly altered) and containing about 50 percent fibrous

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Petriglieri, Laporte-Magoni, Salvioli-Mariani, Tomatis, Gazzano, Turci, Cavallo, and Fubini

Figure 3. Textures of ﬁbrous antigorite observed by polarized light microscopy (cross-polarizing images). (A) Star- and fan-formed aggregates; (B) interpenetrating texture; (C) lath-shaped lamella; and (D) ﬁbro-lamellar blade.

particles were compared with chrysotile (UICC Chrysotile A, Rhodesian) in terms of physico-chemical properties known to modulate asbestos toxicity and cellular responses. Asbestos toxicity is based on fibrous habit, surface reactivity, and high biopersistence, which altogether yield persistent inflammation and DNA damage. For this reason, 1) size and morphology, 2) surface reactivity toward free radical release and iron bioavailability, and 3) dissolution in simulated body fluids were investigated. Data acquired were also compared to those obtained from a lamellar antigorite from the western Alps, Italy (Groppo and Compagnoni, 2007). Size and morphology, including aspect ratio, of the four antigorite samples were determined by an automated image analysis system (FPIA 3000, Malvern Instruments, Malvern, Worcestershire, UK). Morphometrical analysis was performed to discriminate between respirable fibers (Length/Diameter (L/D) > 3, Diameter (D) < 3 μm), non-respirable fibers (L/D > 3, D > 3 μm), and non-fibrous particles (L/D < 3), according to regulated critical dimensions (IARC, 2012). Caledonian samples are all in the form of fibro-

lamellar crystals. After a gentle mechanical stress they fracture easily, releasing elongated fibrous particles, most of which have the dimensional characteristics of respirable fibers (L/D > 3, D < 3 μm). Antigorite samples were ground in a ball mixer mill (Retsch MM200, Haan, Germany) for 2 to 5 minutes (27 Hz) to obtain a similar size distribution. Agate jars were used to avoid metal contamination. After the grinding procedure was complete, particles appear fractured, mainly in the form of acicular or isometric crystals. Caledonian antigorite contains about 40 to 55 percent of respirable fibers, compared to 12 to 15 percent of the Italian sample. The lamellar Italian antigorite is made up of mostly prismatic fragments. In all Caledonian samples the amount of respirable fibers was not correlated to the alteration status. Surface reactivity was evaluated by measuring the ability of antigorite to catalyze generation of hydroxyl and carbon-centered radicals in cellfree tests and to release iron into solution (bioavailable iron). Mid- to highly altered antigorite showed the same reactivity in hydroxyl radical release as did the UICC chrysotile, but, opposite to chrysotile, it did not

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Fibrous Antigorite from New Caledonia

Figure 4. SEM images of diﬀerent morphologies of antigorite. Most particles maintain their lamellar habit, displaying a gradually bent, slinky to curvilinear ending.

catalyze carbon-centered radical generation and contained smaller amounts of bioavailable iron. Dissolution was investigated in Gamble’s solution, which mimics interstitial fluid within the deep lung, and phagolysosomial simulant fluid. All samples dissolved slower than chrysotile. Finally, cellular effects were investigated in human epithelial cells (A549) and in murine macrophages (MH-S). Figure 5 shows the release of LDH (lactate dehydrogenase), an intracellular cytosolic enzyme that is released in the culture medium when cell membranes are damaged (cytotoxicity). Highly altered antigorite showed a similar, dose-dependent cytotoxic effect. On the other hand, less-altered lamellar antigorite, as well as the nonfibrous Italian sample, were not toxic, even at the highest doses. The increasing higher activity of LDH is associated with a higher degree of alteration. Moreover, at high dose (four times higher than chrysotile), highly

altered samples induced oxidative stress and production of nitric oxide, a cytotoxic and pro-inﬂammatory intracellular messenger. They also damaged DNA in alveolar cells. The unaltered antigorite showed very weak surface reactivity and did not trigger any cellular eﬀect.

CONCLUSIONS The comprehensive approach involved in the study of Caledonian ﬁbro-lamellar antigorite delivered three main results:

r Caledonian antigorite exhibits a ﬁbro-lamellar habit, resulting in a greater variability in texture and morphology than was associated with unaltered antigorite.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Petriglieri, Laporte-Magoni, Salvioli-Mariani, Tomatis, Gazzano, Turci, Cavallo, and Fubini

Figure 5. Cytotoxicity LDH released by alveolar macrophages after a 24-hour incubation with increasing doses of antigorite samples (Italian non-ﬁbrous Atg and Caledonian less- to highly altered), 30 µg/ml of chrysotile (Ctl) or 120 µg/ml of synthetic vitreous ﬁbers (MMVF-CTRL).

r Pedogenic alteration affects the surface reactions and increases the genesis and release of fibers. The penetration of fluids within fibrils, associated with a chemical elemental exchange at the mineral/water interface, causes progressive internal mechanical stress; ultimately, there is a complete loss of cohesion of the original structure. r The different reactivity of three antigorite samples in cell-free and cellular tests suggests a role of pedogenic alteration, which modifies surface chemistry, in the potential pathogenicity of fibrous antigorite. Cell-free and cellular tests revealed a lower reactivity of antigorite samples compared to chrysotile. This reactivity is fully absent in the less-altered specimen, suggesting a lower hazard associated with fibrous antigorite. The slow dissolution in simulated bodily fluids, however, indicates that antigorite biopersistence could be higher than that of chrysotile. Further research is needed to confirm the lower toxicity of antigorite with respect to chrysotile. ACKNOWLEDGMENTS This study is a part of a multidisciplinary project “Amiantes et Bonnes Pratiques (ABP),” supported by the Centre National de Recherche Technologique (CNRT–Nouvelle Calédonie). JRP’s post doc position is partially funded by the University of Torino (CHI.2019.08/XXI “Development of innovative tools

for the risk assessment of elongated mineral particles (EMP) in natural environment”). REFERENCES ANSES, 2014, Évaluation de la Toxicité de l’Antigorite: ANSES Rapport d’expertise collective, Maisons Alfort, France. 116 p. Baumann, F.; Maurizot, P.; Mangeas, M.; Ambrosi, J.-P.; Douwes, J.; and Robineau, B. P., 2011, Pleural mesothelioma in New Caledonia: Associations with environmental risk factors: Environmental Health Perspectives, Vol. 119, No. 5, pp. 695–700. Butt, C. R. M. and Cluzel, D., 2013, Nickel laterite ore deposits: Weathered serpentinites: Elements, Vol. 9, No. 2, pp. 123–128. Cannaò, E.; Scambelluri, M.; Agostini, S.; Tonarini, S.; and Godard, M., 2016, Linking serpentinite geochemistry with tectonic evolution at the subduction plate-interface: The Voltri Massif case study (Ligurian Western Alps, Italy): Geochimica Cosmochimica Acta, Vol. 190, pp. 115–133. Carbone, M.; Baris, Y. I.; Bertino, P.; Brass, B.; Comertpay, S.; Dogan, A. U.; Gaudino, G.; Jube, S.; Kanodia, S.; Partridge, C. R.; Pass, H. I.; Rivera, Z. S.; Steele, I.; Tuncer, M.; Way, S.; Yang, H.; and Miller, A., 2011, Erionite exposure in North Dakota and Turkish villages with mesothelioma: Proceedings National Academy Sciences, Vol. 108, No. 33, pp. 13618–13623. Délibération n° 82 du 25 août 2010 – Relative à la protection des travailleurs contre les poussières issues de terrains amiantifères dans les activités extractives, de bâtiment et de travaux publics. JONC du 9 Septembre 2010. Deschamps, F.; Godard, M.; Guillot, S.; and Hattori, K., 2013, Geochemistry of subduction zone serpentinites: A review: Lithos, Vol. 178, pp. 96–127.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Fibrous Antigorite from New Caledonia DIMENC-SGNC, 2010, Cartographie des terrains potentiellement amiantifères en Nouvelle-Calédonie–Etat des connaissances, mars 2010. Direction de l’Industrie, des Mines et de l’Energie - Service Géologique de Nouvelle Calédonie (DIMENC-SGNC), Bureau de Recherches Géologiques et Minières (BRGM), Nouméa, Nouvelle Calédonie. Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in the Franciscan Complex blueschist, northern Diablo Range, California: Toxicology Applied Pharmacology, Vol. 361, pp. 3–13. Gazzano, E.; Riganti, C.; Tomatis, M.; Turci, F.; Bosia, A.; Fubini, B.; and Ghigo, D., 2005, Potential toxicity of nonregulated asbestiform minerals: Balangeroite from the Western Alps. Part 3: Depletion of antioxidant defenses: Journal Toxicology Environmental Health, Part A, Vol. 68, No. 1, pp. 41–49. Goldberg, P.; Goldberg, M.; Marne, M. J.; Hirsch, A.; and Tredaniel, J., 1991, Incidence of pleural mesothelioma in New Caledonia: A 10-year survey (1978–1987): Archives Environmental Health, Vol. 46, No. 5, pp. 306–309. Groppo, C. and Compagnoni, R., 2007, Ubiquitous fibrous antigorite veins from the Lanzo Ultramafic Massif, Internal Western Alps (Italy): Characterisation and genetic conditions: Periodico Mineralogia, Vol. 76, pp. 169–181. IARC, 2012, Arsenic, metals, fibres, and dusts. A review of human carcinogens. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 100C. IARC: Lyon, France. 501 p. Lahondère, D., 2012, Serpentinisation et Fibrogenèse dans les Massifs de Péridotite de Nouvelle-Calédonie. Atlas des Occurrences et des Types de Fibres d’Amiante sur Mine: Nouméa, Nouvelle Calédonie. 128 p. Laporte-Magoni, C.; Tribaudino, M.; Meyer, M.; Fubini, B.; Tomatis, M.; Juillot, F.; Petriglieri, J. R.; GunkelGrillon, P.; and Selmaoui-Folcher, N., 2018, Amiante et Bonnes Pratiques. Rapport Final: Nouvelle Calédonie, Nouméa. 214 p.

Lee, R. J.; Strohmeier, B. R.; Bunker, K. L.; and Van Orden, D. R., 2008, Naturally occurring asbestos—A recurring public policy challenge: Journal Hazardous Materials, Vol. 153, No. 1–2, pp. 1–21. Luce, D.; Billon-Galland, M.-A.; Bugel, I.; Goldberg, P.; Salomon C.; Févotte, J.; and Goldberg, M., 2004, Assessment of environmental and domestic exposure to tremolite in New Caledonia: Archives Environmental Health, Vol. 59, No. 2, pp. 91–100. Petriglieri, J. R.; Laporte-Magoni, C.; Gunkel-Grillon, P.; Tribaudino, M.; Bersani, D.; Sala, O.; Le Mestre, M.; Vigliaturo, R.; Bursi Gandolfi, N.; and SalvioliMariani, E., 2019, Mineral fibres and environmental monitoring: A comparison of different analytical strategies in New Caledonia: Geoscience Frontiers, Ophiolites. In press. Trotet, F., 2012, Fibrous serpentinites in oxyded nickel ores from New Caledonia: Risk management in a modern mining company—Societal implications. In 3rd Serpentine Days, Porquerolles Island, France, pp. 87. Turci, F.; Favero-Longo, S. E.; Gazzano, C.; Tomatis, M.; Gentile-Garofalo, L.; and Bergamini, M., 2016, Assessment of asbestos exposure during a simulated agricultural activity in the proximity of the former asbestos mine of Balangero, Italy: Journal Hazardous Materials, Vol. 308, pp. 321–327. Turci, F.; Tomatis, M.; Gazzano, E.; Riganti, C.; Martra, G.; Bosia, A.; Ghigo, D.; and Fubini, B., 2005, Potential toxicity of nonregulated asbestiform minerals: Balangeroite from the Western Alps. Part 2: Oxidant activity of the fibres: Journal Toxicology Environmental Health, Part A, Vol. 68, No. 1, pp. 21–39. Ulrich, M.; Picard, C.; Guillot, S.; Chauvel, C.; Cluzel, D.; and Meffre, S., 2010, Multiple melting stages and refertilization as indicators for ridge to subduction formation: The New Caledonia ophiolite: Lithos, Vol. 115, No. 1–4, pp. 223–236. Wicks, F. J. and Whittaker, E. J. W., 1977, Serpentine texture and serpentinization: Canadian Mineralogist, Vol. 15, No. 4, pp. 459–488.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 89–97

Geologic Investigations for Compliance with the CARB Asbestos ATCM BRADLEY G. ERSKINE* Erskine Environmental Consulting, Inc., 401 Marina Place, Benicia, CA 94510

Key Terms: ATCM, NOA, Ultramafic Rocks, CARB, Franciscan Complex ABSTRACT The California Air Resources Board Airborne Toxic Control Measure for Construction, Grading, Quarrying, and Surface Mining Operations (ATCM) provides requirements for the evaluation for naturally occurring asbestos (NOA) on a construction site. There are two compliance triggers: (1) a determination that the site is located within a geographic ultramafic rock unit, defined as a geographic area designated as an ultramafic rock on referenced maps, and (2) the presence of NOA, serpentinite, or ultramafic rock. The California Geological Survey requires that NOA evaluations be conducted by a licensed professional geologist. However, under the ATCM, a professional geologist is required only when a property owner wishes to demonstrate that a geographic ultramafic rock unit is not actually represented by ultramafic rocks. The professional geologist who must advise whether the ATCM applies at a construction site is therefore placed in a precarious position. Does a limited desktop review of geologic maps meet any standard of practice? If the ATCM is triggered by the presence of asbestos, is the geologist negligent if no evaluation is recommended or conducted? Could geologic units be pre-screened for asbestos potential? Using case studies and geologic data in the city of San Francisco and East Bay, this presentation reviews these issues and provides a context for the geologist to conduct the appropriate level of investigation for compliance with the ATCM. INTRODUCTION California geology is exceedingly complex and among the most diverse in the world. Igneous crystallization and metamorphic mineralization processes, combined with past and present tectonic activity, have produced a diverse range of rocks and minerals, many of which are or were economically signiﬁcant. Gold in the Sierra Nevada foothills, rare earth elements near Mountain Pass, and tourmaline in the Pala District east of San Diego are but a few examples. *Corresponding author email: Erskine.geo@gmail.com

Asbestos, primarily chrysotile, was one economically important mineral that was mined for its use in building materials. The KCAC mine in the Coalinga District was once the largest asbestos mine in the world and was the last asbestos mining operation in the United States when it closed in 2002. To illustrate how common asbestos is in California, consider that chrysotile is a ubiquitous component of serpentinite, the California state rock. Serpentinite is exposed in the Sierra Nevada foothills and in a western belt that extends from the Klamath Mountains to as far south as Santa Barbara. Based on several studies by the U.S. Environmental Protection Agency (EPA) and the U.S. Agency for Toxic Substances and Disease Registry (ATSDR), notably the Clear Creek Management Area near Coalinga (U.S. EPA, 2008) and the El Dorado Hills exposure studies (U.S. ATSDR, 2011), California implemented the nation’s most developed regulations and guidance documents to mitigate the potential exposure to asbestos on construction sites. An example (and the subject of this article) is the California Air Resources Board (CARB) asbestos Airborne Toxic Control Measure for Construction, Grading, Quarrying, and Surface Mining Operations (ATCM), which provides requirements for the evaluation of naturally occurring asbestos (NOA) on earth-disturbing construction sites (CARB, 2002). When the ATCM is triggered via criterion listed in Section (b)—Applicability, mandatory asbestos-specific dust control measures are required by CARB, and perimeter air monitoring may be required at the discretion of the regional Air Pollution Control District. The two relevant compliance criteria specified in the Applicability section are the following: Subsection (b)(1): “Any portion of the area to be disturbed is located in a Geographic Ultramafic Rock Unit”; or Subsection (b)(2): “Any portion of the area to be disturbed has naturally-occurring asbestos, serpentine, or ultramafic rock.”

The criterion in subsection (b)(1) refers to a “Geographic Ultramaﬁc Rock Unit,” deﬁned in the ATCM as follows:

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

Erskine Geographic Ultramaﬁc Rock Unit means a geographic area that is designated as an ultramaﬁc rock unit or ultrabasic rock unit, including the unit boundary line, on any of the maps referenced in Appendix A.

Ultramafic and ultrabasic rocks are equated within the ATCM, and defined as follows: Ultramafic Rock means an igneous rock composed of 90 percent or greater of one or a combination of the following iron/magnesium-rich, dark-colored silicate minerals: olivine, pyroxene, or more rarely amphibole. For the purposes of this section, “ultramafic rock” includes the following rock types: dunite, pyroxenite, and peridotite; and their metamorphic derivatives.

Thus, a geographic ultramafic rock unit (GURU) is not a physical lithologic unit; rather, it is an ultramafic unit where represented as such on a map. However, most of the maps referenced in Appendix A of the ATCM are large-scale compilations of maps often from the late 1950s and 1960s of varying detail and quality, and a desktop review of these maps is the specified procedure to comply with subsection (b)(1) of the ATCM. For compliance with subsection (b)(1), there is no requirement to do the following:

r Map review by a licensed professional geologist r Verify in advance of a compliance determination that the referenced maps correctly represent the distribution of geologic units r Conduct a review of other geologic maps that may be available r Review the associated reports that provide the basis for the mapping r Field check the maps, which is standard professional geology practice In addition, it is incorrectly assumed that the maps accurately depict all occurrences of ultramafic rocks and are absent where not depicted. The criterion in subsection (b)(2) refers to a physical presence of ultramafic rocks or asbestos but is silent on the methodology that should be employed to verify that NOA, serpentinite, or ultramafic rocks are actually present on a site. The only specified requirement for a geologic assessment is provided within the Exemption section of the ATCM as follows: Geologic Evaluation: The Air Pollution Control Officer may provide an exemption from this section for any property that meets the criterion in subsection (b)(1) if a registered geologist has conducted a geologic evaluation of the property and determined that no serpentine or ultramafic rock is likely to be found in the area to be disturbed. Before an exemption can be granted, the owner/operator must provide a copy of a report detailing the geologic evaluation to the Air Pollution Control Officer for his or her consideration.

100

Therefore, the ATCM requires a geologic assessment only when the owner of a site wishes to prove that the geologic map depicting the presence of a GURU is incorrect and underlying rocks at a site are in fact not ultramafic in composition. It does not require that the geologist verify that asbestos is not present. For compliance with subsection (b)(2), there is no requirement to verify or provide a geologic opinion as to whether NOA, serpentine minerals, or ultramafic rock are present or not likely to be present. Subsection (b)(2) cannot be adequately addressed without a geologic assessment that meets the standard of practice for professional geologists. The ATCM in Practice Subsections (b)(1) and (b)(2) of the ATCM indicate that notification to air districts and asbestosspecific dust control measures are triggered when asbestos is present or “likely to be present” on a site, including chrysotile and the five amphiboles compositions named in the ATCM’s definition of asbestos. Some California agencies have expanded the definition of NOA to include amphiboles and other minerals that were not mined commercially and incorporated into building materials. In practice, however, the standard has become largely serpentinite-centric with the assumption of chrysotile being the primary target mineral. Amphibole-bearing rocks, generally igneous and metamorphic rocks that are not ultramafic, are overlooked in the process. For example, the San Luis Obispo Air Pollution Control District publishes an online map, “Areas of Concern Near Known Serpentinite Rock Formations,” that is used to determine whether the ATCM is triggered. Three air districts publish maps that are drawn from asbestos-potential maps compiled by the California Department of Conservation, and these derivatives are used to determine whether the ATCM is triggered: the El Dorado County Air Quality Management District following Churchill et al. (2000), the Placer County Air Pollution Control District following Higgins and Clinkenbeard (2006a), and the Sacramento Metropolitan Air Quality Management District following Higgins and Clinkenbeard (2006b). Of these, only the map of the Sacramento Metropolitan Air Quality Management District includes within potentially NOA-bearing areas, a unit that is not defined as a GURU in asbestos-potential maps, whereas in other air district maps, only GURUs are mapped. The State of California requires NOA evaluations, including review and interpretation of geologic maps, to be conducted by a professional geologist (Clinkenbeard et al., 2002). However, under the ATCM, a geologist conducting a geologic assessment is required only

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

Professional Geologists/CARB ATCM

when a property owner wishes to demonstrate that a site within a GURU is not actually underlain by serpentinite or other ultramafic rocks (refer to the previous section that describes a geologic exemption within the ATCM). The professional geologist who must advise whether the ATCM will be triggered at a construction site is therefore placed in a precarious position. Does the restricted desktop review of geologic maps, with varying and uncertain quality and accuracy, meet any standard of practice? Because the ATCM is triggered by the presence of asbestos, can the geologist be found negligent if no additional evaluation is recommended or conducted, particularly if relying on the air district potential maps? What is the appropriate level of inquiry? Is it possible that an unlicensed geologist who often makes this determination can provide an informed opinion? These questions are probed below using theoretical and actual case studies. Compliance Using Review of Geologic Maps as Specified in the ATCM Assume that a professional geologist is tasked to determine whether an Asbestos Dust Mitigation Plan (ADMP) is required by a local air district at two development sites located near each other in the Oakland hills of California. Assume also that the only criterion of interest to the air district is subsection (b)(1), which is generally the case. In this example, the two sites selected reside in the Oakland hills near the Oakland Zoo (Figure 1a and b). Following the protocol of the ATCM, the geologist is referred to the geologic map of the San Francisco–San Jose quadrangle (Wagner et al., 1990) to determine whether the sites are located within a GURU. If this large-scale (1:250,000) compilation map is used as specified, the sites fall in unit “KJF,” which is denoted as the Franciscan Complex (Figure 2). The legend of the map denotes Franciscan mélange by horizontal lines, which are not indicated at the location of the sites. Ultrabasic rocks are indicated to the east, colored dark purple on the map. Thus, the requirement under subsection (b)(1) has been satisfied, and without further assessment, these sites would not require notification to the air district, nor would an ADMP or air monitoring be required. However, the professional geologist would undoubtedly know that a review of only the specified map is not sufficient because it is a compilation of many sources of varying quality and accuracy. Other map resources are available and would be consulted. One source is the San Francisco Bay Area geologic map by the U.S. Geological Survey (Graymer et al., 2006). When plotted, the sites fall within unit “fsr,” which denotes Franciscan Complex mélange (Figure 3). A professional geologist may conclude that the ATCM is not triggered

Figure 1. (a) Google Earth oblique photograph showing the location (triangle) of the two sites in the Oakland hills east of San Francisco, California. (b) Google Earth oblique photograph showing the locations (squares) of the northern and southern sites in the Oakland hills, California. Highway 24 and the Oakland Zoo are shown for locational reference.

because the site does not fall in a GURU as deﬁned, even though the reference to mélange suggests the potential for serpentinite to be present. Another important resource is the national geologic map database, which is easily accessed on the Web (https://ngmdb.usgs.gov/ngmdb/ngmdb_home.html). The locations of the two sites are plotted in Figures 4 and 5. Note that the database is a compilation of

Figure 2. Location of the two sites (square) on the geologic map prescribed in Appendix A of the ATCM (Wagner et al., 1990).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

101

Erskine

Figure 3. Site location (square) on the U.S. Geological Survey map of the San Francisco Bay region (Graymer et al., 2006). The unit “fsr” denotes Franciscan mélange.

many maps, and in this case, the sites fall near the intersecting corners of four map quadrangles. The southern site, shown on a map by Graymer (2000), falls within unit “KJfm”: Franciscan mélange. Again, the unit is not a GURU as defined, and an ADMP is not required. The map for the northern site, however, has been replaced by a map by Dibblee (2005a), placing the northern site within a very different rock unit, “Ob”: Coast Range Ophiolite. This unit is denoted in the legend as “ultramafic rocks, mostly gabbro and diorite.” In this case, the determination of whether the site lies in a GURU will differ depending on which map is used. Further Investigation through Assessment of Asbestos Potential Several air districts have developed guidance for how to determine a project’s applicability to the ATCM, one example being the Placer County As-

Figure 4. Sites (square) plotted on the compilation map as depicted on the U.S. Geological Survey National geologic map database (Graymer, 2000).

102

Figure 5. The two sites plotted on the image from U.S. Geological Survey geologic map database as it appeared at the time of this publication. The northern and eastern quadrants are mapped by Dibblee and Minch (2005a, 2005b, 2005c). The southwestern quadrangle is mapped by Graymer (2000).

bestos Dust Mitigation Plan Guidance for NOA (Placer County, 2014). The guidance document states that the ATCM is applicable when the site is located in a GURU or “has naturally-occurring asbestos, serpentine or ultramaﬁc rock as determined by owner/operator, registered geologist or the District APCO.” Professional geologists are aware that a geologic map review is the ﬁrst step of a geologic investigation and should not be the sole source used to assess whether asbestos may be present on a site (note the reference to the presence of “asbestos” as a trigger in subsection (b)(2) of the ATCM). When assessing a site for ATCM applicability, the geologist should conduct an appropriate inquiry following the standard of practice for geologists in California. This standard begins with protocols within the Guidelines for Geologic Investigations of Naturally Occurring Asbestos in California (Clinkenbeard et al., 2002). The assessment could take an investigatory path similar to “Phase I Environmental Assessments” designed to assess whether chemical spills, releases, or uncontrolled use of hazardous chemicals may have occurred at the site and whether further action in the form of a “Phase II soil and groundwater investigation” is warranted. For NOA, the site could be evaluated on the basis of NOA potential: sites with underlying rock and soil with a high NOA potential (e.g., serpentinite and former asbestos mines and prospects) could be assumed to have an asbestos concentration greater than 1 percent, and dust control and Occupational Safety and Health Administration (OSHA) compliance measures be implemented without further investigation. If the determination of concentrations relative to the ATCM 0.25-percent surfacing and cover threshold or the Cal/OSHA 1 percent threshold for “Class II Work” is important, then a soil and rock

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

Professional Geologists/CARB ATCM

investigation is needed. Rocks or soil with moderate NOA potential require further investigation, beginning with a literature and map review, followed by a site verification. The professional geologist would then develop an opinion as to whether a “Phase II” rock and soil investigation is warranted. The concept of asbestos potential is well documented in several publications by the California Geological Survey (Churchill et al., 2000; Higgins and Clinkenbeard, 2006 a, 2006b). Of these, Higgins and Clinkenbeard (2006a) is particularly important because it assigns high, medium, and low NOA potential to the various geologic units in that region, provides the basis for each designation, and presents an NOA potential map. However, the professional geologist should not rely solely on the map and should use it and the concepts provided within the publication to arrive at his or her own conclusions. For example, alluvium is listed as low potential, but further evaluation may show that a particular alluvial unit may have been sourced from rocks with high NOA potential and therefore is also a high potential unit. A rock unit that has been assigned a moderate NOA potential may be heterogeneous, and individual sub-units underlying a particular site may be of high or low NOA potential. Only an appropriate site-specific investigation by a professional geologist that includes a field assessment can differentiate which of the two is most likely.

Programmatic Implementation of NOA Potential Mapping: An Example The concepts regarding asbestos potential provided in Higgins and Clinkenbeard (2006a) are gaining acceptance by geologists, and techniques that utilize these concepts have been developed to prevent exposure to workers and the public. One excellent example is the program implemented by the San Francisco Public Utilities Commission (SFPUC) Department of Health and Safety for their internal use. The SFPUC owns large tracts of land across northern California associated with the extensive network of facilities, reservoirs, and pipelines of San Francisco’s Hetch Hetchy water delivery system. A database of NOA potential that was produced by a professional geologist, (Figure 6) covers the counties where SFPUC facilities reside. The local detailed segment of this map is reviewed when soil disturbance that may produce exposure to dust is expected at a work site, which ranges in scope from large construction projects to small weedwhacking tasks for ﬁre suppression purposes. Based on the review, a follow-up assessment that may include a site visit and testing for NOA may be initiated.

Figure 6. NOA potential map of northern California produced from SFPUC’s NOA potential GIS database (not published; for internal use only and not available to the public). Green = high NOA potential, predominantly serpentinite and other ultramaﬁc rocks; orange = moderate NOA potential, predominantly metamorphic rocks where regulated amphiboles are likely to be present in certain rock units.

Further Evaluation Following Geologic Map and NOA Potential Map Review A professional geologist tasked with an NOA evaluation at a site should, at a minimum, provide data to determine whether a regulatory threshold may be exceeded. For earthmoving projects, the primary regulatory agencies are CARB for the protection of off-site receptors and Cal/OSHA for the protection of on-site workers. In the case of the CARB ATCM, the threshold is “No Asbestos Detected” (note that the 0.25 percent threshold specified within the ATCM refers to post-construction covering and surfacing applications, not excavation). There is no threshold specified for excavation activities during construction. Cal/OSHA regulations are triggered when asbestos is present in any amount (the 1 percent threshold in the regulation triggers additional engineering controls and personal monitoring). Thus, the threshold that triggers compliance response actions during a typical construction project is “No Asbestos Detected” as measured with an appropriate lower limit of detection or analytical sensitivity. It follows that the threshold that geologists should use to assess whether a geologic assessment is appropriate is that “asbestos is not likely to be present” rather than “asbestos is likely to be present.” Both the California Geological Survey and SFPUC produced NOA potential maps that depict the relative likelihood for the presence of NOA. In both cases, the high-likelihood rocks are serpentinite and other ultramafic rocks. It is generally assumed that asbestos is likely to occur in these rocks, and Cal/OSHA and CARB regulations are automatically triggered. Both of the NOA potential maps tend to designate nonmetamorphic and sedimentary rocks as low potential. Both maps include a third category: rocks with a moderate potential for asbestos to be present. This

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

103

Erskine

group includes metamorphic rocks where amphiboles are likely to be present. Should these rocks be further investigated considering that the threshold by the two agencies cited above is “No Asbestos Present”? The answer is yes, as will be illustrated in the following example. Consider the Franciscan Complex in northern California and compare the non-serpentinite component in two locations. The first is the city and county of San Francisco, where tectonic terranes of relatively unmetamorphosed Franciscan Complex rocks are separated by two serpentinite mélange belts. The second is the Franciscan Complex in the Diablo Range southeast of San Francisco, where the Franciscan Complex is present as a blueschist-grade mélange (for details regarding the tectonic histories of various Franciscan Complex terranes, see Wakabayashi, 2015). Using the criteria presented by Higgins and Clinkenbeard (2006a), SFPUC designated the greenstone metabasalt as moderate potential. However, further evaluation of the greenstone metabasalt in both locations indicate that the greenstone should be redesignated, one to high potential and the other to low potential, as described below. Franciscan Greenstone in the Diablo Range The Diablo Range Franciscan rocks were studied during SFPUC’s Calaveras Dam Replacement Project (CDRP) as part of an extensive NOA control and monitoring program. The rocks outcropping at this location were subducted to relatively deep levels and subjected to low-temperature/high-pressure blueschist facies metamorphism. Through the metamorphic process, blueschist rocks originated in which fibrous glaucophane amphibole crystallized, as shown in Figure 7a and b. Erskine and Bailey (2018) showed through a fiber dimensional analysis that the glaucophane met the characteristics of the asbestiform habit. Testing of the blueschist by transmission electron microscopy revealed asbestiform glaucophane concentrations that exceed 20 percent by weight and more than 200,000 million fibers per gram (MFG). The U.S. Geological Survey concurred and recently designated glaucophane at the CDRP as a known occurrence of asbestos. Thus, the greenstone at the CDRP site, as well as the other metamorphic rocks of the Franciscan in the Diablo Range, has a high likelihood for asbestos to be present.

Figure 7. (a) Hand specimen of ﬁbrous blue glaucophane crystallizing within greenstone at the CDRP construction site, northern Diablo Range, California. (b) Thin-section photomicrograph of greenstone in the northern Diablo Range showing ﬁbrous glaucophane (blue mineral) crystallizing within a matrix hosting relic greenstone porphyroclasts. Photomicrograph taken under plane polarized light.

reach blueschist facies metamorphism, as evidenced by well-preserved primary structures and textures and no glaucophane (Figure 8a). In thin section, primary intersertal structure is recognizable, with skeletal plagioclase phenocrysts set in a ﬁne-grained matrix (Figure 8b). No indication of glaucophane or other low-temperature/high-pressure mineralization was present in these rocks, and asbestos was not detected by transmission electron microscopy in the samples tested. Therefore, with further veriﬁcation, the NOA potential for these rocks may be downgraded from moderate to low likelihood for asbestos to be present.

Franciscan Greenstone in San Francisco

SUMMARY AND CONCLUSIONS

Greenstone of the Marin Headlands terrane collected near Lands End in north San Francisco and at the Marin Headlands show that these rocks did not

The CARB ATCM provides criteria to evaluate regarding whether a site is subject to mandatory asbestos dust control measures and potential air monitoring.

104

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

Professional Geologists/CARB ATCM

Figure 8. (a) Outcrop of greenstone (metabasalt) at Lands End in San Francisco. No evidence of blueschist metamorphic index minerals (glaucophane or lawsonite) is present. (b) Thin-section photomicrograph of greenstone showing primary plagioclase within a ﬁne-grained matrix. Photomicrograph taken under crossed polarizing light.

The evaluation process, however, falls far short of the standard of practice for professional geologists. The following conclusions and recommendations are oﬀered based on the information and data presented in this article:

r All assessments, including the preliminary review of geologic maps, should be conducted by a professional geologist as deﬁned by the state licensing board. The ATCM requires that a professional geologist conduct an assessment for the purposes of an exemption from the GURU rule and envisions that a geologist who could detect unforeseen asbestoscontaining rock and soil will be on-site. It does not, however, require that a professional geologist conduct the initial assessment through the review and interpretation of geologic maps. The initial assessment is the most important phase of the ATCM process, and the review and interpretation of geo-

logic maps by unqualified persons for the purpose of assessing NOA potential presents a high risk of misinterpretation. The procedure outlined in the ATCM where a determination is made solely through a desktop review of maps is inadequate to assess the likelihood of asbestos to be present at a site. Geologic maps provide a representation of the distribution of rocks and structure as observed by the mapper and do not include information regarding asbestos. Rock units are identified differently depending on the context of the map; for example, serpentinite may also be mapped as ultramafic rocks, mélange, or incorporated without identification into a larger unit. Smaller bodies or those with poor exposure may be missed altogether. A review of geologic maps should be the first step in an investigation, as is standard practice, to be followed by further characterization that includes a site visit and perhaps soil and rock testing. More consideration should be given to subsection (b)(2) in the Applicability section, which refers to the presence of asbestos as a trigger for compliance with the ATCM. The determination of whether asbestos is or is not present requires far more inquiry than a map review. The professional geologist should conduct a reasonable level of inquiry before he or she may conclude that “asbestos is not likely to be present.” While NOA potential maps are powerful tools, further assessment, including a field reconnaissance at a minimum, is needed to adequately assess the likelihood for asbestos to be present. The determination that the ATCM is triggered should be tied to the permitting process in the same manner as asbestos in buildings materials. Building demolition permitting requires an asbestos inspection conducted by a California Department of Occupational Safety and Health–certified asbestos consultant. Similarly, a grading permit should require an NOA assessment report by a qualified professional geologist. At present, some California air districts have such a system, but the majority do not.

REFERENCES California Air Resources Board (CARB), 2002, Asbestos Airborne Toxic Control Measure for Construction, Grading, Quarrying and Surface Mining Operations: Publication No. 2001-07-29, 23 p. Churchill, R. K.; Higgins, C. T.; and Hill, B., 2000, Areas More Likely to Contain Natural Occurrences of Asbestos in Western El Dorado County, California: California Department of Conservation Open-File Report 2000-2002, 66 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

105

Erskine Clinkenbeard, J. P.; Churchill, R. K.; and Lee, K., 2002, Guidelines for Geologic Investigations of Naturally Occurring Asbestos in California: California Geological Survey Special Report 124, 85 p. Dibblee, Thomas W. Jr. and Minch, J. A., 2005a, Geologic Map of the Oakland East Quadrangle, Contra Costa & Alameda Counties California. Dibblee Geological Foundation Map DF-160. Dibblee, Thomas W. Jr. and Minch, J. A., 2005b. Geologic Map of the Las Trampas Ridge Quadrangle, Contra Costa & Alameda Counties California. Dibblee Geological Foundation Map DF-161. Dibblee, Thomas W. Jr. and Minch, J. A., 2005c. Geologic Map of the Hayward Quadrangle, Contra Costa & Alameda Counties, California. Dibblee Geological Foundation Map, DF-163. Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in Franciscan Complex blueschist, northern Diablo Range, California: Journal of Toxicology and Applied Pharmacology, Vol. 361, pp. 3–13. Graymer, R. W., 2000, Geologic Map and Map Database of the Oakland Metropolitan Area, Contra Costa and San Francisco Counties, California. U.S. Geological Survey Report MF-2342. Graymer, R. W.; Moring, B. C.; Saucedo, G. J.; Wentworth, C. M.; Brabb, E. E.; and Knudsen, K. L., 2006. Geologic Map of the San Francisco Bay Region, United States Geological Survey: Scientiﬁc Investigations Map 2918.

106

Higgins, C. T. and Clinkenbeard, J. P., 2006a, Relative Likelihood for the Presence of Naturally Occurring Asbestos in Placer County, California: California Geological Survey Special Report 190, 54 p. Higgins, C. T. and Clinkenbeard, J. P., 2006b, Relative Likelihood for the Presence of Naturally Occurring Asbestos in Eastern Sacramento County, California: Geological Survey Special Report 192, 43 p. Placer County, 2014, Asbestos Dust Mitigation Plan (ADMP) Guidance for Naturally Occurring Asbestos: Revision 14, May 21, 2014, 11 p. U.S. Agency for Toxic Substances and Disease Registry (U.S. ATSDR), 2011, Evaluation of Community-Wide Asbestos Exposures: El Dorado Hills Naturally Occurring Asbestos Site, El Dorado Hills Boulevard, El Dorado Hills, California: EPA Facility ID Can000906083, 194 p. U.S. Environmental Protection Agency (U.S. EPA), 2008, Clear Creek Management Area Asbestos Exposure and Human Health Risk Assessment: 75 p. Wagner, D. L.; Bortugno, E. J.; and McJunkin, R. D., 1990, Geologic Map of the San Francisco-San Jose Quadrangle (set of five sheets). California Division of Mines and Geology Regional Map 5A. Wakabayashi, J., 2015, Anatomy of a subduction complex: Architecture of the Franciscan Complex, California, at multiple length and time scales: International Geology Review, Vol. 5, Nos. 5–8, pp. 669–746.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 99–106

Geological Model for Naturally Occurring Asbestos Content Prediction in the Rock Excavation of a Long Tunnel (Gronda di Genova Project, NW Italy) LUCA BARALE FABRIZIO PIANA* SERGIO TALLONE Institute of Geosciences and Earth Resources, National Research Council of Italy, Via Valperga Caluso 35, 10125 Torino, Italy, and “G. Scansetti” Interdepartmental Centre for Studies on Asbestos and Other Toxic Particulates, University of Torino, Via Pietro Giuria 7, 10125 Torino, Italy

ROBERTO COMPAGNONI Department of Earth Sciences, University of Torino, and “G. Scansetti” Interdepartmental Centre for Studies on Asbestos and Other Toxic Particulates, University of Torino, Via Valperga Caluso 35, 10125 Torino, Italy

CHIARA AVATANEO SERENA BOTTA IGOR MARCELLI Gi-RES srl, Via Gottardo 223, 10154 Torino, Italy

ANDREA IRACE Institute of Geosciences and Earth Resources, National Research Council of Italy, University of Torino, Via Valperga Caluso 35, 10125 Torino, Italy

PIETRO MOSCA Institute of Geosciences and Earth Resources, National Research Council of Italy, Via Valperga Caluso 35, 10125 Torino, Italy

ROBERTO COSSIO Department of Earth Sciences, University of Torino, Via Valperga Caluso 35, 10125 Torino, Italy

FRANCESCO TURCI Department of Chemistry, University of Torino, and “G. Scansetti” Interdepartmental Centre for Studies on Asbestos and Other Toxic Particulates, University of Torino, Via Pietro Giuria 7, 10125 Torino, Italy

Key Terms: NOA, Asbestos, Rock Tunneling, Geological Model, Meta-Ophiolites ABSTRACT For a reliable evaluation of the geo-environmental risk due to naturally occurring asbestos (NOA) during rock excavation of large infrastructure projects, a proper procedure is needed. First, it is necessary to *Corresponding author email: fabrizio.piana@cnr.it

provide a detailed geological model tailored to the NOArelated issues that should drive the rock sampling procedures in order to obtain a representative sampling. The sampling procedures should take into account lithological variability and relative spatial distributions of the rock units. The geological model for NOA should be thus constrained by the main NOA petrofacies occurring in a given geotectonic context, which take into consideration both the mineralogical and structural features, and the identification of NOA homogeneous zones in which the NOA petrofacies are distributed. In this paper, some geo-environmental problems faced during the

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 107–112

107

Barale, Piana, Tallone, Compagnoni, Avataneo, Botta, Marcelli, Irace, Mosca, Cossio, and Turci

excavation, in meta-ophiolites, of a long highway tunnel are described. The geological model of the complex setting of the tunnel area (northern Italy, Alps-Apennines junction) is described, focusing on how the NOA-related problems were addressed to allow reliable and detailed estimations of NOA contents for each NOA homogeneous zone and the relevant tunnel layout segment. INTRODUCTION The Gronda di Genova project concerns the realization of a highway bypass around the city of Genoa in NW Italy, aiming at relieving the traﬃc from the existing highway system, which connects the port and the city of Genoa with the rest of Italy and adjoining countries (Figure 1). The project, presently in the working plan phase, proposes 72 km of new highway, including 54 km of tunnels; of these, about half will be excavated in potentially asbestos-bearing rocks (meta-ophiolitic units). The construction of large infrastructure in rocks with naturally occurring asbestos (NOA) poses several geo-environmental issues, such as adequate preparation of the construction site, workers’ safety, and spoil management. The estimated total amount of spoils that will derive from the Gronda di Genova tunnel excavation is about 11,000,000 m3 (ASPI, 2011). Some of these spoils are expected to contain NOA above the concentration threshold ﬁxed by the Italian regulation (NOA limit, in the following); this limit corresponds to a concentration of asbestos in the rock of 1,000 mg/kg (or ppm; Italian regulation Decreto Legislativo 152/06 (Repubblica Italiana, 2006).

Spoils below the NOA limit or without NOA will be mainly re-used to fill the tunnel invert and for filling work in the stilling channel of the Genoa airport to enlarge the runway strip. The latter use will allow disposal of about 8,000,000 m3 of spoils (ASPI, 2011). Conversely, spoils above the NOA limit represent, by law, a hazardous waste and have to be disposed of in special dumps, thus resulting in high disposal costs. For this reason, a preliminary evaluation of the expected amount of asbestiferous spoils was deemed necessary in the first phases of project development (results of early studies were reported in Giacomini et al., 2010; Turci et al., 2015). This task required a multi-disciplinary team of researchers and professionals with varied expertise including structural geology, petrography, mineralogy, and analytical chemistry. The first step was the creation of an ad hoc geological model tailored to NOA issues (cf. Labagnara et al., 2013; Vignaroli et al., 2014) to guide representative sampling of NOA-bearing rocks. The whole set of information derived from field observations, drill core analysis, and petrographic studies, together with the data on NOA content provided by scanning electron microscopy–energy dispersive X-ray spectroscopy (SEM-EDS) quantitative analyses, allowed definition of several NOA-targeted petrostructural facies (Botta et al., 2019), categorized into four NOA-content classes (very high, high, low, nil). The distribution of NOA petrofacies was then presented in detailed geological sections along the project layout, thus allowing prediction of the total volume of NOA-bearing rocks along the tunnels.

Figure 1. (a) Simpliﬁed geological scheme of the Genoa area; SVZ = Sestri-Voltaggio Zone. (b) Excerpt of the geological map produced for the present study, with indication of the main geological units crossed by the Gronda di Genova project layout. In the inset, the position of the Genoa area within the highway system of NW Italy is shown (existing highways are indicated by green lines).

108

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 107–112

Geological Model to Predict NOA Content

Geological Outline of the Project Area The Genoa area is located in the Alps-Apennines “interference zone” (sensu Piana et al., 2017), made up of several geological units that shared, at least in part, the evolution of both orogens. From west to east, the geological units crossed by the Gronda di Genova layout are (Capponi and Crispini, 2008; Capponi et al., 2016, with references therein; Figure 1):

r the Voltri “massif,” made up of meta-ophiolites and relevant meta-sedimentary cover; in the study area, the Voltri “massif” is composed of two tectono-metamorphic units, the Voltri and PalmaroCaﬀarella units, which experienced, respectively, eclogitic- and blueschist-facies peak metamorphism during the Alpine orogeny; r the Sestri-Voltaggio Zone, a several-kilometer-large tectonic slice zone containing both meta-ophiolitic and non-ophiolitic rock sequences, which recorded blueschist- to prehnite-pumpellyite–facies Alpine metamorphism; and r the Ligurian units, corresponding to a very lowgrade to non-metamorphic hemipelagic sedimentary succession of Cretaceous age. These units underwent a polyphase tectonometamorphic evolution, including early phases of isoclinal to tight folding (D1 and D2 phases), which generated composite, transposed regional foliation, and later phases (D3 and D4) responsible for the generation of meter- to kilometer-scale gentle to open folds, associated with reverse faults and shear zones (Capponi and Crispini, 2008, with references). The Voltri massif and the Sestri-Voltaggio Zone units are presently juxtaposed along a N-S–striking deformation zone, known in the literature as “Sestri-Voltaggio line, which acted as a zone of preferential shear concentration at diﬀerent crustal levels during the polyphase evolution of the AlpsApennine system (Crispini et al., 2009; Piana et al., 2017, with references). NOA minerals (chrysotile and actinolite-tremolite amphiboles) are widespread in the meta-ophiolitic sequences of the Voltri massif and of the SestriVoltaggio Zone, whereas they are absent in the (meta-)sedimentary Ligurian units. METHODS NOA-Oriented Geological Model and Sampling Strategy The construction of the NOA-oriented geological model primarily relied on detailed geological mapping at 1:5,000 scale along a 3- to 5-km-large belt around

the tunnel layout (Figure 1b). This was accompanied by detailed structural analysis to achieve a thorough view of mesoscale structural features of NOA-bearing lithotypes and subdivision of the study area into homogeneous structural domains. Field surveys were integrated using the analysis of a large amount of available drill cores and logs (about 300 drillings were produced for the Gronda di Genova project, and data from hundreds others drilled in the area were obtained from public and private agencies). Integration of field and drilling data allowed the production of 1:5,000 scale geological sections along both tubes of the main tunnels and underground access ramps. In parallel, a detail petrographic analysis was conducted, focusing on potentially NOA-bearing lithotypes (Botta et al., 2019). Particular attention was dedicated to the description of veins, with respect to their composition, thickness, and presence of fibrous minerals (either regulated or non-regulated) and elongated/acicular minerals. All of this information provided an important support to subsequent SEM observations made for quantitative analysis. Optical microscopy has always been coupled with microRaman spectroscopy, which has been useful both in the univocal distinction of serpentine polymorphs (e.g., Rinaudo et al., 2003; Groppo et al., 2006), and in the determination of fine-grained minerals or mineral aggregates. At this point, after considering all the acquired geological and petrographic information, a proper strategy for representative sampling could be planned. Sampling was guided by two criteria: the lithological and the statistical representativeness. The first criterion guaranteed that all the lithotypes and lithological varieties identified in the study area were represented by at least one sample (logically, more abundant lithotypes were granted a higher number of samples than the rarest ones). The statistical representativeness was intended to have samples covering all the structural domains, independently of their lithology, evenly distributed along the project layout. Each sample was divided into two sub-samples, with one utilized for thin section preparation and the other (around 1 kg) prepared for quantitative analysis. Quantitative Evaluation of NOA Content SEM-EDS quantitative analyses represented the central part of the work and were made on 190 selected rock samples (106 from outcrops and 84 from drill cores). Twenty-one soil samples were also analyzed to evaluate the asbestos content in the surficial cover at the main open-air construction sites (e.g., tunnel entrances). The broad range of analytical cases encountered during this project, together with the previous

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 107–112

109

Barale, Piana, Tallone, Compagnoni, Avataneo, Botta, Marcelli, Irace, Mosca, Cossio, and Turci

Figure 2. NOA petrofacies deﬁned for the Gronda di Genova project, and their subdivision in very high, high, low, and nil content classes.

knowledge acquired in analogous studies, allowed our group to address important issues regarding the analytical procedure, including the definition of guidelines and best practices for sample preparation and analysis (see Turci et al., in press). The results of quantitative analyses were integrated with the data derived from petrography and structural analysis to define several NOA petro-structural facies (NOA petrofacies). NOA petrofacies are classes of rocks that share common lithological-structural features and NOA content. NOA petrofacies were subdivided into four NOA content classes (Figure 2) according to their asbestos content ratio (ACR), i.e., the ratio between the number of samples in which NOA content is higher than 1,000 mg/kg and the total number of analyzed samples. Very-high-content petrofacies and high-content petrofacies contain asbestos commonly above 1,000 mg/kg (ACR > 30 percent); these include most types of serpentinites and serpentinite breccias (petrofacies 2b, 2c, 2d), as well as actinolite/tremolitechlorite schists (SAC, petrofacies 2a) and fault rocks developed at the expense of serpentinites (petrofacies 1). Low-content petrofacies are those that do contain NOA, but commonly below 1,000 mg/kg (ACR < 30 percent); they comprise partly serpentinized peridotites (petrofacies 3a), and metabasites (petrofacies 3b). The nil content group corresponds to NOA petrofacies 4; this petrofacies groups all lithotypes that do not contain NOA, for which further distinction was useless for the aims of the present work (Botta et al., 2019). Following the identification of the NOA petrofacies present in the area, the tunnels were subdivided along their entire length into NOA homogeneous zones, i.e., tunnel segments characterized by the presence of a distinct NOA petrofacies or an association of a few NOA petrofacies. For each NOA homogeneous zone, the ACRz was defined, representing the ratio between 110

Figure 3. Excerpt of the geological section along the main tunnel of the Gronda di Genova project. The geological section is accompanied by a table identifying the main geological information, including NOA information in the last four rows.

the rock volume above the NOA limit and the total rock volume: ACRz =

rock volume above NOA limit . total rock volume

Each NOA zone had its own ACRz based on the type and relative proportions of NOA petrofacies. The ACRz was calculated as the sum of the ACR of all the petrofacies within a given NOA zone, each one multiplied by a correction factor (ranging between 0 and 1), which represented the relative volume of each NOA petrofacies in that NOA zone. The obtained ACRz values were reported on the complete geological sections of the tunnels (Figure 3), together with general geological data (e.g., structural domain, lithology, presence and type of faults and fault rocks) and other NOA information (homogeneous zone, petrofacies, NOA content class, and number of analyzed samples). The ACRz was then utilized to calculate, for each homogeneous zone, the volume of rocks above 1,000 mg/kg, by multiplying the zone length by the tunnel section area and by the ACRz itself: Volume above 1, 000 mg/kg = homogeneous zone length × tunnel section area × ACRz . By adding the volumes obtained for each NOA homogeneous zone, the total volume of rocks above NOA limit (1,000 mg/kg) for the Gronda di Genova project was estimated as 978,000 m3 of in situ rock. This

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 107–112

Geological Model to Predict NOA Content

corresponds to about 1,270,000 m3 of spoils, considering a bulking factor of 1.3. The “Frac_Asbestos” Test In parallel with the above described work, a test on semi-quantitative prediction of asbestos content in a given rock volume was conducted based on mathematical modeling of fracture systems (“Frac_Asbestos”; Piana et al., in press). The test was made on an outcrop of intensely fractured lizardite serpentinites with lizardite and chrysotile veins. On this outcrop, the main fracture systems were identiﬁed and characterized both geometrically (orientation, spacing, persistence) and petrographically (occurrence and type of vein minerals). This characterization was followed by a semi-automatic extraction of fracture traces on a three-dimensional laser scanner image; discrete fracture network modeling was then completed to obtain the “fracture intensity” (i.e., the total amount of fracture surface per unit volume). The obtained fracture intensity was utilized to calculate the NOA weight per unit volume, by multiplying by a series of factors obtained from outcrop observation and thin section study. To validate the test, the calculated NOA concentration was compared with the results of quantitative analyses made on samples from three cores drilled on the same outcrop in roughly perpendicular directions and intended to be representative of the whole rock mass. CONCLUSIONS The evaluation of NOA content in the Gronda di Genova highway project highlighted the importance of:

r a geological model tailored to NOA issues, to define structural domains and to constrain representative sampling; r definition of NOA petro-structural facies by comparing data from structural analysis, petrography, and quantitative NOA analyses, beginning with first stages of sampling; and r different approaches, including NOA petrostructural facies or fracture network modeling to calculate NOA-bearing rock volumes. ACKNOWLEDGMENTS This work was developed in the frame of the geoenvironmental model for the “Gronda di Genova” working plan (highway bypass of Genoa city, Italy,

designed by Autostrade per l’Italia [ASPI], Rome), commissioned by Spea Engineering spa, Milan (Italy), to a consortium comprising the “G. Scansetti” Interdepartmental Center for Studies on Asbestos and Other Toxic Particulates of the University of Torino (Italy), the Institute of Geosciences and Earth Resources of the National Research Council of Italy, Torino, and Gi-RES srl (a Consiglio Nazionale delle Ricerche [CNR] spin-oﬀ company), Torino, Italy. The authors are indebted to Girolamo Belardi, Daniele Passeri, and Francesca Trapasso (Institute of Environmental Geology and Geoengineering [CNRIGAG], Rome) for designing and performing comminution of the samples. A research paper on this last complex and unexplored subject is currently being prepared. Francesco Cipolli (Spea Engineering, Genoa), and Vittorio Boerio and Simona Polattini (Spea Engineering, Milan) are kindly acknowledged for their continuous support and feedback throughout all working stages. Mark Bailey and two anonymous reviewers are kindly acknowledged for their comments on an earlier version of the manuscript.

REFERENCES ASPI (Autostrade per l’Italia), 2011, Opera a Mare nel Canale di Calma. Parte Descrittiva Generale. APG9030: Electronic document (in Italian), available at http://www.va.minambiente.it/ File/Documento/25374. Botta, S.; Avataneo, C.; Barale, L.; Compagnoni, R.; Cossio, R.; Marcelli, I; Piana, F.; Tallone, S.; and Turci, F., 2019, Petrofacies for the prediction of NOA content in rocks: Application to the ‘Gronda di Genova’ tunnelling project: Bulletin Engineering Geology Environment, https://doi.org/10.1007/s10064-019-01539-6. Capponi, G. and Crispini, L., 2008, Carta Geologica d’Italia alla scala 1.50.000 e Note Illustrative. Foglio 213-230 Genova: APAT, Rome, Italy. Capponi, G.; Crispini, L.; Federico, L.; and Malatesta, C., 2016, Geology of the Eastern Ligurian Alps: A review of the tectonic units: Italian Journal Geosciences, Vol. 135, No. 1, pp. 157–169. Crispini, L.; Federico, L.; Capponi, G.; and Spagnolo, C., 2009, Late orogenic transpressional tectonics in the “Ligurian Knot”: Italian Journal Geosciences, Vol. 128, pp. 433–441. Giacomini, F.; Boerio, V.; Polattini, S.; Tiepolo, M.; Tribuzio, R.; and Zanetti, A., 2010, Evaluating asbestos fibre concentration in metaophiolites: A case study from the Voltri Massif and Sestri-Voltaggio Zone (Liguria, NW Italy): Environmental Earth Sciences, Vol. 61, pp. 1621–1639. Groppo, C.; Rinaudo, S.; Cairo, D.; Gastaldi, R.; and Compagnoni, R., 2006, Micro-Raman spectroscopy for a quick and reliable identification of serpentine minerals from ultramafics: European Journal Mineralogy, Vol. 18, No. 3, pp. 319–329. Labagnara, D.; Patrucco, M.; Rossetti, P.; and Pellegrino, V., 2013, Predictive assessment of asbestos content in the Western Italian Alps: An essential tool for an effective approach to risk analysis in tunneling operations and muck reuse: Environmental Earth Sciences, Vol. 70, pp. 857–868.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 107–112

111

Barale, Piana, Tallone, Compagnoni, Avataneo, Botta, Marcelli, Irace, Mosca, Cossio, and Turci Piana, F.; Barale, L.; Botta, S.; Compagnoni, R.; Fidelibus, C.; Tallone S.; Avataneo C.; Cossio, R.; and Turci, F., in press, Direct and indirect assessment of the amount of naturally occurring asbestos in fractured rocks: Boletín Geológico Minero. Piana, F.; Fioraso, G.; Irace, A.; Mosca, P.; d’Atri, A.; Barale, L.; Falletti, P.; Monegato, G.; Morelli, M.; Tallone, S.; and Vigna, G.B., 2017, Geology of Piemonte region (NW Italy, Alps-Apennines interference zone): Journal Maps, Vol. 13, No. 2, pp. 395–405. Repubblica Italiana, 2006, Decreto Legislativo 3 aprile 2006, n. 152. Gazzetta Uﬃciale, Serie Generale n.88 del 14-04-2006 Suppl. Ordinario n. 96: Electronic document (in Italian), available at http://www.gazzettauﬃciale.it/eli/gu/2006/04/14/ 88/so/96/sg/pdf Rinaudo, S.; Gastaldi, D.; and Belluso, E., 2003, Characterization of chrysotile, antigorite and lizardite by FT-Raman spectroscopy: Canadian Mineralogist, Vol. 41, pp. 883–890.

112

Turci, F.; Avataneo, C.; Botta, S.; Marcelli, I.; Barale, L.; Tomatis, M.; Cossio, R.; Tallone, S.; Piana, F.; and Compagnoni, R., in press, New tools for the evaluation of asbestosrelated risks during excavation in NOA-rich geological setting: Environmental and Engineering Geoscience (in press). Turci, F.; Compagnoni, R.; Piana, F.; Delle Piane, L.; Tomatis, M.; Fubini, B.; Tallone, S.; Fuoco, S.; and Bergamini, M., 2015, Geological and analytical procedures for the evaluation of asbestos-related risk in underground and surface rock excavation. In Lollino, G., et al. (Editors), Engineering Geology for Society and Territory, Vol. 5: Springer, Berlin, Germany, pp. 619–622. Vignaroli, G.; Ballirano, P.; Belardi, G.; and Rossetti, F., 2014, Asbestos ﬁbre identiﬁcation vs. evaluation of asbestos hazard in ophiolitic rock mélanges, a case study from the Ligurian Alps (Italy): Environmental Earth Sciences, Vol. 72, pp. 3679–3698.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 107–112

New Tools for the Evaluation of Asbestos-Related Risk during Excavation in an NOA-Rich Geological Setting FRANCESCO TURCI* Department of Chemistry and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and Other Toxic Particulates, University of Torino, Via Pietro Giuria, 7, 10125 Torino, Italy

CHIARA AVATANEO SERENA BOTTA IGOR MARCELLI Geological Risk Analysis Gi-RES Srl, Via Gottardo 223, 10154 Torino, Italy

LUCA BARALE National Research Council of Italy, Institute of Geosciences and Earth Resources, Via Valperga Caluso, 35, 10125 Torino, Italy

MAURA TOMATIS Department of Chemistry and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and other Toxic Particulates, University of Torino, Via Pietro Giuria, 7, 10125 Torino, Italy

ROBERTO COSSIO Department of Earth Sciences and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and Other Toxic Particulates, University of Torino, Via Valperga Caluso, 35, 10125 Torino, Italy

SERGIO TALLONE FABRIZIO PIANA National Research Council of Italy, Institute of Geosciences and Earth Resources, Via Valperga Caluso, 35, 10125 Torino, Italy

ROBERTO COMPAGNONI Department of Earth Sciences and “G. Scansetti” Interdepartmental Center for Studies on Asbestos and Other Toxic Particulates, University of Torino, Via Valperga Caluso, 35, 10125 Torino, Italy

Key Terms: Naturally Occurring Asbestos, Asbestos Analysis, Rock Tunneling, Sampling, SEM-EDS, Ligurian Alps, Asbestos ABSTRACT The presence of naturally occurring asbestos (NOA) in many areas worldwide requires an enhanced geological risk evaluation to ensure workplace safety from asbestos during large construction projects. Due to the complexity of the geological risk definition, health and safety regulations for working with asbestos-bearing

*Corresponding author email: francesco.turci@unito.it

materials are often not enforceable in NOA settings. Therefore, to correctly estimate the risk of NOA in these scenarios, new procedures are urgently needed to provide (1) a detailed geological model representative of the possible presence of the asbestos, (2) representative sampling, and (3) a reliable quantitative determination of asbestos content in rocks. This work aims to discuss the improvements on the two latter points specifically developed during the design of the “Gronda di Genova” project, a 50-km-long tunnel bypass partially designed in the NOA-bearing meta-ophiolites of the Ligurian Alps and ophiolites of the northern Apennines in Italy. Implementation of Gy’s theory on sampling was used to maintain statistical validity during sample processing from the primary sample to the analytical

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

113

Turci, Avataneo, Botta, Marcelli, Barale, Tomatis, Cossio, Tallone, Piana, and Compagnoni

Figure 1. Worldwide distribution of active and inactive asbestos mining sites. Some NOA occurrences cited in the text are also indicated: 1, Biancavilla, Italy; 2, Libby, Montana; 3, Franciscan Complex, California; 4, 5, 6, erionite sites in Turkey (Cappadocia), the United States (Dunn County, North Dakota), and Mexico (Tierra Blanca de Abajo), respectively. Data from “Asbestos. Overview and Handling Recommendations” (Deutsche Gesellschaft für Technische Zusammenarbeit, 1996) and Virta (2002).

sample and is here described. The scanning electron microscopy/energy dispersive spectroscopy procedure for the quantification of NOA was improved with an error analysis delivering the minimum number of fibers to be measured to achieve the best analytical results. INTRODUCTION Asbestos is a human carcinogen currently causing the majority of occupational lung cancers and several other malignant and non-neoplastic pathologies (International Agency for Research on Cancer, 2012). The occupational exposure to asbestos, a set of six minerals defined by the World Health Organization/International Agency for Research on Cancer, the National Institute for Occupational Safety and Health, the European Union, and many other international agencies worldwide, is unambiguously correlated to an increase of mesothelioma, a fatal malignancy currently causing more than 100,000 deaths per year worldwide. Asbestos are not the only minerals to have a fibrous habit. More than 300 minerals are known to occur in fibrous form (Skinner et al., 1988), and many of them share toxic effects with asbestos (Baumann et al., 2013). A more inclusive definition of these fibers that might have an impact on human health is urgently needed, and the definitions

114

of elongated mineral particles and naturally occurring asbestos (NOA) are currently debated (Gunter, 2018). The term NOA is used here to define asbestos and other fibrous minerals (mainly fibrous antigorite) that occur in the investigated area. In many industrialized countries, the ban on or the strong regulatory limitations to asbestos use have drastically reduced the exposure to asbestos fibers, though many low- and middle-income countries are currently mining and using asbestos. In the western Alps, NOA occur as rockforming minerals in many rocks of the ophiolitic suite, including green-colored serpentinites derived from the hydration of the peridotites. Excavation, tunneling, construction, and generally all anthropic activities carried out in ophiolitic rocks may occasionally release NOA. During unintended or designed disturbance of natural asbestos-rich areas, such as agricultural activities (Turci et al., 2016), road traffic (Baumann et al., 2011; Petriglieri et al., 2019), and tunnel construction (Turci et al., 2015; Gaggero et al., 2017; and Barale et al., this issue), the workers and the population at large might incur occupational or environmental exposure to NOA (Hendrickx, 2009). To gain insight into the possibility of the above scenario occurring worldwide, we could look at geographic areas where asbestos minerals were abundant enough to be commercially exploited (Figure 1). Figure 1 also highlights the

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

Evaluate Asbestos Risk During Excavation

Figure 2. Schematic representation of the possible dispersion routes and human exposures from NOA sources during large construction projects.

occurrence of fibrous minerals that were never exploited commercially but where health-related issues were recorded by mesothelioma epidemiology, including the occurrence of fluoroedenite, in Biancavilla, Sicily, Italy; fibrous amphibole in Libby, Montana; and erionite in Turkey, the United States, and Mexico (Case and Marinaccio, 2017, and references therein). Fibrous glaucophane and winchite, found in the blueschist Franciscan Formation in California, are also marked after the work of Erskine and Bailey (2018). Although not all fibrous minerals may constitute a risk to human health, NOA risk should properly be assessed (Gualtieri, 2018) during the design and the realization of construction works that may liberate fibers in both the workplace and the environment. Appropriate risk assessment has to be carried out prior to and during excavation/construction in NOA-rich sites. Specifically, NOA mobilization from rock/soil to waters and air and vice versa has to be taken into account (Figure 2), and site-specific health and environmental protection measures have to be enforced to limit airborne and waterborne dispersion of the fibers. The greatest challenge posed by NOA is indeed due to their spatial localization and quantification in rocks. The first challenge requires the adoption of an asbestos-oriented geological model, such as the approach used for the design of the “Gronda di Genova” highway bypass and described in this special issue (Barale et al., this issue). The NOA quantification in situ generates further challenges, including (1) a

harmonized definition of NOA, (2) a reliable sampling methodology, and (3) a proper analytical procedure. The first point is currently under consideration by two independent international panels within the International Mineralogical Association and the International Association of Engineering Geology and the Environment. This work deals with the two latter points in the specific setting of the Gronda di Genova project. A parallel approach can be found in the high-speed railway project designed in a similar geological context (Clerici, 2018).

RESULTS AND DISCUSSION Sampling Methodology To obtain a reliable measure, a proper sampling method is required, and although few analysts are aware of the availability of the theory of sampling, it is useful to remember that the analyst’s work is useless if a sample is biased. Specifically, the sample variance of a poorly designed sampling method might easily be up to 1,000%, while analytical methods are usually required to have a variance lower than 1%. All sampling errors derive from the heterogeneity of the material to sample, and the specific case of a NOA-bearing outcrop, the distributional heterogeneity (i.e., the spatial distribution of the constituent elements of the rock with respect to NOA) is very high. NOA minerals are indeed confined in

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

115

Turci, Avataneo, Botta, Marcelli, Barale, Tomatis, Cossio, Tallone, Piana, and Compagnoni

Figure 3. Sampling site (A) and primary ﬁeld samples (soil and hand sample from drill core, B and C, respectively), from the Gronda di Genova highway bypass.

micrometric veins, irregularly distributed along a network of fractures, within the rock unit. During the evaluation of the NOA occurrence for the Gronda di Genova project, two main scenarios were taken into consideration, and two different sampling approaches were designed: (1) a “preventive evaluation” was used to define, during the preliminary stages of the project, the volume of excavated rocks where NOA could occur above the regulatory limit (Botta et al., 2019), and (2) an “ongoing quantification” was designed to be implemented throughout the project lifetime and to deliver real-time quantitative assessment of NOA in the excavated rock and soil. The sampling methodology adopted during the preventive evaluation, thoroughly described in Barale et al. (this issue), included the definition of site-specific NOA petrofacies. All the lithotypes identified in the study area were represented by at least one sample (logically, more abundant lithotypes were granted a higher number of samples than the rarest ones). The statistical representativeness was achieved by collecting primary samples from the defined petrofacies for each of the structural domains distributed along the tunnel layout. That approach was also followed on portions of drill cores from the geognostic survey as well as on soil samples collected by geologists during the field survey (Figure 3). In all cases, a reasoned non-random sampling approach was chosen. Despite obvious limitations, a non-stochastic approach offers the advantage of identifying discontinuities between lithotypes and collecting representative portions of each fundamental asbestos-bearing rock type, consisting of recurrent lithological and structural features controlling asbestos occurrence: the NOA petrofacies (Piana et al., 2019). Only a non-stochastic sampling, carried out by a trained geologist, allows the composite sample to be representative of the entire spectrum of homogeneous

116

petrofacies occurring in the primary sample. To take into account such intrinsic heterogeneity, the size and mass reduction steps required to create the analytical sample were carried out with an approach derived from Pierre Gy’s theory (Gy, 1967, 1971, 1998). Gy’s theory and the theory of sampling in general usefully describe heterogeneity with two terms: the constitutional and the distributional heterogeneity. These terms have to be defined to quantify the minimum mass of the analytical sample that correctly represents the primary (or composite) sample collected in the field (Belardi et al., 2018). A detailed discussion of constitutional and distributional heterogeneity in the present case study is beyond the purpose of this work and will be described in a following paper. However, it is known that heterogeneity is related to some fundamental factors of the investigated material. In the case of solid fragments, the heterogeneity can be described by an invariant term H expressed as the product of five descriptive parameters H = cβ f gd 3 , where c is the constitution parameter (g/cm3 ) and depends on the amount of the analyte of interest and its relative density; f is the shape factor (dimensionless) and depends on the shape of grains; g is the size range factor (dimensionless) and describes the dimensional heterogeneity; and d is the top-particle size (cm), which is also used to calculate β, the liberation parameter (dimensionless), as the square root of the dlib /d ratio, where dlib is the liberation diameter (cm), the size of grains at which all the analyte is liberated. During the Gronda di Genova project, the descriptive factor values were used (Table 1). As the variance of the sampling error (σ2 ) is a function of heterogeneity (H), Gy’s approach conveniently correlates the minimum mass of the material to be

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

Evaluate Asbestos Risk During Excavation Table 1. Descriptive factors for the evaluation of the sample heterogeneity in NOA-rich ophiolitic rock of the Gronda di Genova highway bypass. Descriptive Factors

Factor Description and Range

Values Used in the Study

Constitution parameter, c

It depends on the mineralogical composition of the material. For a two-component system, the parameter is calculated using the following equation: [(1 − a)ρ1 + aρ2 ], c = (1−a) a where a is the expected value of the analyte and ρ1 and ρ2 are the analyte and matrix density, respectively.

Particle shape factor, f Size range factor, g

From round particles (0.4–0.5) to lamellar ones (0.1) From non-calibrated material (0.25) to material selected by means of sieves (0.55). For an analyte free from matrix, the factor is around 0.4–0.8; if the particle top size (d) is higher than the liberation size (dlib ), the factor is around 0.05–0.02.

c = 2,697 Setting a as the threshold limit for asbestos detection (1,000 ppm), ρ1 = 2.7 g/cm3 (average asbestos density, considering chrysotile more abundant than amphiboles) and ρ1,2 = 2.7 g/cm3 (matrix and NOA average density) f = 0.30 g = 0.25

Liberation factor, β

sampled (MS ) with the error introduced during the sampling: σ2 ∼

c β f g d3 . MS

As all the descriptive parameters are strictly positive, Gy’s theory allows us to set an acceptable sampling error and obtain the minimum sampling mass (MS ) for a given top-particle size (d). To correctly operate in real-case scenarios, d is often adjusted to meet the practical and theoretical constraints of the system investigated, and sample preparation procedures can be ad hoc designed as a series of alternating size (d) and mass (MS ) reductions (Figure 4A). Speciﬁcally, the primary sample collected in the ﬁeld (ca. 1 kg) was size reduced to d = 1 mm and mass reduced to 16 g; a further size reduction to d = 0.1 mm was completed with a mass reduction to ca. 5 mg (Figure 4B, curve a) obtained by an automatic rotational splitter.

β = 0.05, determined by posing the particle top size

The second sampling methodology was designed to collect samples for the quantitative analyses of NOA content throughout the project lifetime. That sampling again had to deal with several types of materials to be collected and required diﬀerent sampling strategies. Sampling strategies were designed to collect samples mainly from the output slurry of the tunnel boring machine, but ad hoc sampling was designed also for excavation faces, heaps, and drill cores. As per the previous case, the distributional heterogeneity has to be taken into account, and the primary samples have to be reduced by a series of alternating size (d) and mass (MS ) reduction steps (Figure 4B, curves b and c). The recursive steps allow production of a statistically representative analytical sample of few milligrams, suitable for the quantitative scanning electron microscopy/energy dispersive spectroscopy (SEM-EDS) assessment of NOA, from the original 200–300 kg of rocks.

Figure 4. General scheme (A) and project-speciﬁc size and mass reduction curves (B) adopted during the study. The analytical samples were obtained from primary samples following Gy’s theory to maintain representativeness in each mass reduction step.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

117

Turci, Avataneo, Botta, Marcelli, Barale, Tomatis, Cossio, Tallone, Piana, and Compagnoni

Figure 5. Evolution of the measured NOA concentrations (black squares, ppm) and the relative errors of the measures (red squares, %) plotted against the number of fibers detected on the membrane of two independent samples, A and B. In both cases, when the number of measured fibers (N) increases, the error progressively decreases with a plateau effect and stabilizes on a value that is dependent on the dimensional variability of the fibers on the membrane.

Quantitative Measurement of NOA The analytical samples obtained were quantified by adopting the only methodology, SEM-EDS, able to discriminate between a mineral with either fibrous or massive habit, to resolve structures of 100–200 nm, to gain insight into the elemental composition of the minerals detected, and to investigate a statistically relevant portion of the sample to achieve 100-ppm sensitivity. The method is described in the Italian legislation on asbestos (DM 6/9/94, Annex 1B), which is similar to ISO 14966 (2002), but it extends the quantification of the fiber concentration from a per-number to a per-mass basis. About 5 mg of the fine analytical sample are weighed and suspended in a water solution containing 0.1% surfactant. A known amount of the suspension is transferred to a polycarbonate membrane, dried, and prepared for electron microscopy. Quantitative NOA analyses are carried out manually by scanning a known area of the membrane and measuring the dimensions (length and diameter) of each NOA fiber detected on the membrane. The observed area of the membrane is clearly a key parameter for the statistical significance of the measure. For each fiber detected, length and diameter are used to approximate the fiber to a cylinder, derive its volume, and calculate its weight by multiplying the volume by the average density of the mineral (2.6 and 3.0 g/cm3 for serpentine and amphiboles, respectively). Despite SEM-EDS being among the most sensitive quantitative analyses of fibers, it often generates data with a high error due to some instrumental- and sample-related factors. Specifically, (1) due to the manual operational setting, only 0.4% to 0.6% of the membrane can be observed; (2) the discrimination between fibrous and non-fibrous

118

mineral is subjective, and a harmonized definition of fibers is urgently needed; and (3) a fiber or, even worse, an aggregate of fibers cannot always be properly approximated to a cylinder. While the two latter points may be solved by working on the analytical definition of NOA, the former issue can be tackled by analyzing the variation of the error during the measurement phase. The experimental error of quantitative SEMEDS analysis of NOA is essentially due to the fiber sampling statistics generated by reading of the filter, as the number of fibers sampled on a given surface is approximated by a Poisson distribution if the fibers are randomly distributed on the membrane. Furthermore, the error is dependent on the dispersion of the grainsize distribution of the fibers, generally well described by a lognormal distribution. The experimental error C on the fiber concentration C is ⎞ ⎛ 2 ¯ i ( f − fi ) ⎜ 1 N·(N−1) ⎟ ⎟, C ≈ C × ⎜ ⎠ ⎝ √N + f¯ where C is the concentration (ppm) of the fibers in the sample; N the number of identified fibers; f¯ is the average weight of an asbestos fiber determined as the average of the weights of the identified N asbestos fibers; and fi the weight of the ith counted fiber. It can be observed that the main contribution to the error is due to the number of fibers analyzed for low numbers of fibers (N < 30). As the number√of fibers increases, the contribution of the term 1/ N becomes negligible with respect to the standard deviation included in the second term of the equation. The standard deviation is still influenced by the

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

Evaluate Asbestos Risk During Excavation

Figure 6. SEM images of (A) a representative NOA complex aggregate of ﬁbers and (B) an asbestos bundle of uncertain diameter and apparent density evidenced during the quantitative analysis of NOA in ophiolitic rock from the Gronda di Genova project.

number of fibers detected, but it strongly depends on the dispersion of the weights of the individual fibers analyzed ( f¯ − fi )2 . The dispersion of the weights is intrinsically related to the dimensional variability of the sample on the membrane, and it can be reduced only by narrowing the particle size distribution during the preparation phases. To evaluate the evolution of the experimental error, the NOA concentrations (black squares) and the relative errors of the measures (red squares) are reported as a function of the number of the fibers detected (Figure 5). As the number of the fibers measured increases, the calculated NOA concentration varies quite unpredictably due to inclusion of outliers in the distribution. Conversely, the relative error of the measure progressively decreases with the increase in measured fibers, producing a plateau effect, and stabilizes to a value that is independent on the number of the fibers detected; that is, it cannot be further reduced by analyzing a larger portion of the membrane. For large numbers of fibers, the error is indeed related only to the dimensional variability of the fiber masses, hence their sizes, described by the second term of the error equation. By performing the error evaluation in parallel with SEM-EDS measurement, we estimated the minimum number of fibers to be analyzed to achieve the best analytical result in terms of experimental error of the measure. The error-minimizing approach can be particularly effective if integrated in unattended SEM-EDS analyses of asbestos concentrations, described in a previous work (Cossio et al., 2018). To further increase the reliability and interlaboratory reproducibility of SEM-EDS quantitative analysis of NOA, a few other issues have to be solved. Future research will include the non-trivial approxi-

mation to a geometrical shape of complex fiber aggregates, which often occur during microscopic analysis (Figure 6A), and the apparent density and the diameter of large bundles, in which inter-crystal spaces cannot be disregarded when evaluating the total bundle volume (Figure 6B). CONCLUSIONS To correctly predict the risk of naturally occurring asbestos in large construction projects, a methodology for reliable sampling and quantitative analysis has to be sought. In this article, we described the approaches used in the Gronda di Genova highway bypass to achieve a reliable sampling statistic and methodology to minimize the analytical error during SEM-EDS quantitative analysis of NOA. An implementation of Gy’s theory allowed us to define the minimum quantity of sample and the size and mass reduction curves to maintain statistical representativeness throughout the sample processing steps. An error-minimizing approach was adopted by investigating the variation of error during the NOA analysis and allowed us to optimize the analytical time without affecting accuracy. ACKNOWLEDGMENTS This research was developed in the frame of the Gronda di Genova research project between University of Torino, Institute of Environmental Geology and Geoengineering-National Research Council (IGGCNR), GiRES srl and SPEA Engineering S.p.A., Genoa, Italy, under contract no. 400006258/2018. The authors are indebted to Dr. G. Belardi,

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

119

Turci, Avataneo, Botta, Marcelli, Barale, Tomatis, Cossio, Tallone, Piana, and Compagnoni

Dr. F. Trapasso, and Dr. D. Passeri, IGG-CNR, Monterotondo, Rome, for carrying out comminution experiments and for the fruitful discussion on the theory of sampling. The authors kindly acknowledge Dr. J. R. Petriglieri for reviewing the manuscript. The second author is indebted to Agenzia Regionale per la Protezione Ambientale (ARPA) Valle d’Aosta for funding her scholarship. REFERENCES Barale, L.; Piana, F.; Compagnoni, R.; Tallone, S.; Avataneo, C.; Botta, S.; Irace, A.; Marcelli, I.; Cossio, R.; Mosca, P.; and Turci, F., 2019, Geological model for NOA content prediction in the rock excavation of a long tunnel (“Gronda di Genoa” project—NW Italy): Environmental and Engineering Geoscience, this issue. Baumann, F.; Ambrosi, J.-P.; and Carbone, M., 2013, Asbestos is not just asbestos: An unrecognised health hazard: Lancet Oncology, Vol. 14, No. 7, pp. 576–578. Baumann, F.; Maurizot, P.; Mangeas, M.; Ambrosi, J.-P.; Douwes, J.; and Robineau, B. P., 2011, Pleural mesothelioma in New Caledonia: Associations with environmental risk factors: Environmental. Health Perspectives, Vol. 119, No. 5, pp. 695–700. Belardi, G.; Vignaroli, G.; Trapasso, F.; Pacella, A.; and Passeri, D., 2018, Detecting asbestos fibres and cleavage fragments produced after mechanical tests on ophiolite rocks: Clues for the asbestos hazard evaluation: Journal of Mediterranean Earth Sciences, Vol. 10, pp. 63–78. Botta, S.; Avataneo, C.; Barale, L.; Compagnoni, R.; Cossio, R.; Marcelli, I.; Piana, F.; Tallone, S.; and Turci, F., 2019, Petrofacies for the prediction of NOA content in rocks: Application to the “Gronda di Genova” tunnelling project: Bulletin of Engineering Geology and the Environment, pp. 1–20. Case, B. W. and Marinaccio, A., 2017, Epidemiological approaches to health effects of mineral fibres: Development of knowledge and current practice. In A. F. Gualtieri (Editor), Mineral Fibres: Crystal Chemistry, Chemical-Physical Properties, Biological Interaction and Toxicity, Vol. 18: EMU Book, London, UK, pp. 367–416. Clerici, C., 2018, Procedure di campionamento del fronte nello scavo di gallerie con l’eventuale presenza di rocce potenzialmente contenenti amianto. In: Protocollo Gestione Amianto per Il Terzo Valico Ferroviario Dei Giovi: Osservatorio Ambientale per il Terzo Valico Ferroviario dei Giovi, Alessandria (I), pp. 1–32. (In Italian) Cossio, R.; Albonico, C.; Zanella, A.; Fraterrigo-Garofalo, S.; Avataneo, C.; Compagnoni, R.; and Turci, F., 2018, Innovative unattended SEM-EDS analysis for asbestos fiber quantification: Talanta, Vol. 190, pp. 158–166. Decreto Ministeriale 6/9/94, Ministero della Sanità (G.U. n.288). Annex 1: Normative e metodologie tecniche per la valutazione del rischio, la bonifica, il controllo e la manutenzione dei materiali contenenti amianto presenti negli edifici (in Italian) Deutsche Gesellschaft für Technische Zusammenarbeit, 1996, Asbestos: Overview and Handling Recommendations: Vieweg, Braunschweig (D), 203 p.

120

Hendrickx, M., 2009, Naturally occurring asbestos in eastern Australia: A review of geological occurrence, disturbance and mesothelioma risk: Environmental Geology, Vol. 57, pp. 909–926. Internationsl Standard Organization ISO 14966, 2002 Ambient air: determination of numerical concentration of inorganic fibrous particle – scanning electron microscopy method. Erskine, B. G. and Bailey, M., 2018, Characterization of asbestiform glaucophane-winchite in the Franciscan Complex blueschist, northern Diablo Range, California: Toxicology and Applied Pharmacology, Vol. 361, pp. 3–13. Gaggero, L.; Sanguineti, E.; Yus González, A.; Militello, G. M.; Scuderi, A.; and Parisi, G., 2017, Airborne asbestos fibres monitoring in tunnel excavation: Journal of Environmental Management, Vol. 196, pp. 583–593. Gualtieri, A. F., 2018, Towards a quantitative model to predict the toxicity/pathogenicity potential of mineral fibers: Toxicology and Applied Pharmacology, Vol. 361, pp. 89–98. Gunter, M. E., 2018, Elongate mineral particles in the natural environment. Toxicology and Applied Pharmacology, Vol. 361, pp. 157–164. Gy, P., 1967, L’échantillonnage des minerais en vrac, Tome 1: Societe de l’industrie minerale, Paris. Gy, P., 1971, L’échantillonnage des minerais en vrac, Tome 2: Societe de l’industrie minerale, Paris. Gy, P., 1998, Sampling for Analytical Purposes: John Wiley and Sons, New York, 172 p. International Agency for Research on Cancer, 2012, Asbestos (chrysotile, amosite, crocidolite, tremolite, actinolite, and anthophyllite): IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol. 100C, pp. 219–309. Petriglieri, J. R.; Laporte-Magoni, C.; Gunkel-Grillon, P.; Tribaudino, M.; Bersani, D.; Sala, O.; Le Mestre, M.; Vigliaturo, R.; Bursi Gandolfi, N.; and SalvioliMariani, E., 2019, Mineral fibres and environmental monitoring: a comparison of different analytical strategies in New Caledonia: Geoscience Frontiers, in press. Piana, F.; Barale, L.; Botta, S.; Compagnoni, R.; Fidelibus, C.; Tallone, S.; Avataneo, C.; Cossio, R.; and Turci, F., 2019, Direct and indirect assessment of the amount of naturally occurring asbestos in fractured rocks. Accepted for publication Bol. Geo Min. Skinner, H. C. W.; Ross, M.; and Frondel, C., 1988, Asbestos and Other Fibrous Materials: Mineralogy, Crystal Chemistry, and Health Effects: Oxford University Press, Oxford. Turci, F.; Compagnoni, R.; Piana, F.; Delle Piane, L.; Tomatis, M.; Fubini, B.; Tallone, S.; Fuoco, S.; and Bergamini, M., 2015, Geological and analytical procedures for the evaluation of asbestos-related risk in underground and surface rock excavation. In: Engineering Geology for Society and Territory, Vol. 5: Springer, Berlin, pp. 619–622. Turci, F.; Favero-longo, S. E.; Gazzano, C.; Tomatis, M.; Gentile-Garofalo, L.; and Bergamini, M., 2016, Assessment of asbestos exposure during a simulated agricultural activity in the proximity of the former asbestos mine of Balangero, Italy: Journal of Hazardous Materials, Vol. 308, pp. 321–327. Virta, R. L., 2002, Asbestos: Geology, Mineralogy, Mining, and Uses: U.S. Department of the Interior, U.S. Geological Survey, Washington, DC, 28 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 113–120

Sampling, Analysis, and Risk Assessment for Asbestos and Other Mineral Fibers in Soil ED CAHILL* EMSL Analytical, Inc., 200 Route 130 North, Cinnaminson, NJ 08077

Key Terms: Activity-Based Sampling, Fluidized Bed Asbestos Segregator, Incremental Sampling Methodology, Natural Occurrences of Asbestos, Risk Assessment, Soil ABSTRACT Asbestos may be present in soil as a natural occurrence or by contamination from asbestos-containing building materials, illegal dumping of asbestos, or other human activities. When trying to properly assess asbestos and other mineral fiber content in a sample by microscopy, soil is a problem matrix in all respects. Even defining the sample to be collected requires forethought and can greatly influence the final analytical result. Determining the sampling approach as well as the best sample preparation and analysis techniques are critical to obtaining accurate results in a metric that is useful to the end user. This article provides an overview of the various approaches that can be applied to assist those involved with asbestos in soil projects. There are many analytical techniques that can be applied for the determination of asbestos content in soil, including visual observation in the field, stereomicroscopy, polarized light microscopy, scanning electron microscopy, transmission electron microscopy, x-ray diffraction, and others. All of these techniques have their own inherent strengths and weaknesses. Fortunately all of the analysis options are complementary, and using multiple techniques can help to better characterize a sampling site and provide a more comprehensive assessment. Time and cost constraints will typically play a role in determining the final sampling and analysis plan. INTRODUCTION As high-profile examples of asbestos in soil projects (such as the naturally occurring asbestos [NOA] discovery and response in El Dorado County in California; Meeker et al., 2006) appear in the news media there is increasing awareness and attention paid to the potential for asbestos in soil on all types of project sites (ATSDR, 2011; Buck et al., 2013). NOA sites, brown*Corresponding author email: ed.cahill@outlook.com

fields, sites of previous building demolitions, sites of known or suspected illegal dumping, etc., are all being sampled and tested much more than in the past. Unfortunately, though there are well-established policies, guidance, and regulations covering asbestos in building materials and the indoor or built environment—such as those associated with the Environmental Protection Agency (EPA) Asbestos Hazard Emergency Response Act (US EPA, 1986), the EPA National Emissions Standard for Hazardous Air Pollutants (1973), Occupational Safety and Health Administration (OSHA, 1986), and state and local regulations—there is little guidance dealing with asbestos in soil and rock. Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) regulations, commonly known as the EPA’s Superfund Program (US EPA, 1980), provide some guidance, but their applicability is limited. CERCLA does recognize the fact that dealing with asbestos-containing building materials (typically covered under NESHAP) is much different from dealing with asbestos in soil scenarios. It cites that using the typical <1 percent action limit for asbestos in building materials is inappropriate for asbestos in soil, as concentrations below 1 percent may still pose unacceptable health risks. The EPA Framework for Investigating Asbestos Contaminated Superfund Sites (US EPA, 2008) does provide some relevant but limited guidance for asbestos in soil projects. The document does suggest various sampling and analytical approaches, but recognizing how disparate various project sites can be, it recommends developing your own site-specific procedures and risk-based action levels.

SAMPLING Unlike sampling building materials in the built environment, soil sampling often involves much larger and less well-defined sampling areas. If asbestos is present it is typically heterogeneously distributed in the soil, which makes it difficult to decide where to sample and how many samples to collect. The goal is to collect enough samples to reasonably reflect the asbestos content in the soil or rock over a defined area. The collection of discrete grab samples is a common sampling

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

121

Cahill

strategy and can be useful for finding hot spots of asbestos, especially when there is some evidence or indication of where asbestos might be located. Because of the heterogeneity of asbestos in soil, however, individual grab samples can be non-representative of larger areas of concern. This can result in false-negative assumptions, on the one hand, to overestimation of asbestos content on the other. It can take large numbers of grab samples to adequately represent asbestos content on even modest-sized sampling sites. For this reason, incremental sampling is increasingly being used as a means of obtaining more representative samples while controlling for analytical costs. The incremental sampling methodology (ISM) is designed as a means to collect composite samples that better represent a defined area. In 2009 the Interstate Technology Regulatory Council (ITRC) established a team of experts to investigate, evaluate, and refine the ISM for sampling soils. This technique is valuable for numerous target analytes, not just asbestos. ITRC’s comprehensive Web site (www.itrcweb.org) provides excellent online training and lays out detailed instructions for designing a quality sampling plan. Development of a quality sampling plan is best achieved with the involvement of all parties associated with the project, including property owners, field personnel, geologists, consultants, the lab technicians performing the preparation and analysis of samples collected, and the end users of the data. One of the key components of ISM is defining a decision unit, that area (or, more precisely, the volume, since a decision unit is three dimensional) from which a composite sample is to be collected. The size of the decision unit depends on a variety of factors, including what the response action will be if that area is determined to exceed a pre-determined threshold. The specific steps to be taken are detailed on the ITRC Web site and elsewhere. In addition to the field collection procedures, there are quite-specific ISM preparation steps to be performed in the laboratory as well (Figure 1). It is important to work with a laboratory that is versed in ISM and can be a strategic partner on the project team. ANALYSIS Microscopy is the primary analytical technique typically applied. X-ray diffraction (XRD) can be useful for mineral identification; however, it lacks the sensitivity of microscopy. XRD also lacks the critical ability to differentiate fibrous from non-fibrous particles. The “detector” for a microscopic analysis is the human eye. This makes it uniquely suited for distinguishing very thin asbestos fibers from a field of view filled with obscuring non-asbestos particulate as a background. One

122

downside of a microscopic analysis is that it has a subjective component, which can add to variability and bias. In addition, regardless of what method is used, there is currently disagreement among experts about what should be counted as a fiber per the method and per existing regulations, what fiber dimensions are of concern toxicologically, how to distinguish true asbestiform particles from cleavage fragments, and if that distinction is even important. More regulations and guidance are needed in these areas. Even with these issues, microscopy is generally considered the best available technology for detecting low levels of asbestos in soil and rock. This is especially true when both light and electron microscopy are used in conjunction. Polarized light microscopy (PLM) analyses combine visual, stereomicroscopic, and PLM observations for detecting, identifying, and quantifying asbestos. The magnification typically used for a PLM analysis is 400×. Because this type of analysis is performed at relatively low magnifications, the amount of subsample analyzed is much greater than is possible for analyses that occur at higher magnifications; however, small fibers below the limit of resolution for light microscopy go undetected. Electron microscopy offers much higher magnifications as well as the ability to perform elemental analysis. Scanning electron microscopy (SEM) can reach higher magnifications than can PLM, and being a surface analysis, SEM provides a different perspective than both PLM and transmission electron microscopy (TEM). Although SEM cannot match TEM in terms of magnification or resolution, its images are far superior in depth of field, which provides a more three-dimensional view of the surface of a sample and better morphological information of suspect fibers. SEM is arguably weaker at identifying fibers than are both PLM and TEM as a result of SEM’s inability to easily obtain diffraction pattern information from individual fibers; however, this technique bridges the magnification gap between the two and may be the preferred initial analytical technique in some cases. TEM operates at much higher magnifications than both PLM and SEM and can readily detect asbestos fibers that are well beyond the limit of resolution of light microscopy. TEM analyses are typically performed at magnifications of 20,000× and higher. The TEM’s extremely high magnification and its ability to collect diffraction patterns and to perform elemental analysis on a fiber-by-fiber basis via energy-dispersive spectroscopy make it well suited to asbestos analysis. Its extremely high magnification, however, limits the amount of subsample that can be analyzed. All of the analytical techniques mentioned are complementary, and multiple approaches can be applied to a single

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

Asbestos in Soil Table 1. Comparison of analytical techniques. Technique

Strengths

Limitations

X-ray diﬀraction (XRD)

Strong mineral identiﬁcation

Limited sensitivity No morphological data obtained; cannot distinguish ﬁbrous from non-ﬁbrous

Polarized light microscopy (PLM)

Relatively large subsample can be analyzed Good mineral identiﬁcation Morphological information can be obtained

Relatively low magniﬁcation and limited resolution, so small ﬁbers can go undetected

Scanning electron microscopy (SEM)

Higher magniﬁcation than PLM Better resolution than PLM Excellent morphological data obtained Elemental analysis capability for mineral identiﬁcation

Quantitation difficult Inability to easily obtain diffraction data makes mineral identification less definitive than TEM

Transmission electron microscopy (TEM)

Highest magniﬁcation and resolution Elemental analysis capability

Small subsample analyzed

sample. The strengths as well as the limitations of each approach are summarized in Table 1. Microscopic analysis works best on homogeneous fine powders. In this regard, soil and rock represent a problem matrix, as they are typically heterogeneous and have relatively large particle sizes. To address these issues, a well-outfitted soil lab has all manner of equipment (Figure 2), including jaw crushers, sieve shakers, riffle splitters, turbula mixers, muffle furnaces, mills, and grinders, to render the soil more amenable to a quality preparation and analysis. In fact, the main differences between the methods typically used for analysis of soil and rock comprise the preparation steps used to address the two issues of particle size and heterogeneity. All of the preparation methods described below have the potential to release and disperse asbestos fibers. The release of fibers is a concern not only for potential exposure to lab personnel but also for potential cross-contamination between samples. For this reason, all preparation steps are performed in negative pressure, high-efficiency particulate air (HEPA)– filtered enclosures. Specific procedures and choice of

equipment should be made with an eye toward minimizing the release of fibers. Sieving Approach With the ASTM method D7521 (2016) approach the fine fraction of the sample that is amenable to preparation and analysis is separated out through sieving. The sample is typically dried, weighed, and then passed through a sieve stack of 19 mm, 2 mm, and 106 µm. The fine (<106-µm) fraction is amenable to both PLM and TEM preparation and analysis. A benefit of this approach is that this fine fraction is analyzed essentially in the same state in which it was collected in the field, with minimal manipulation or alteration. The downside of the sieving approach is that unlike the fine fraction, the coarse and medium fractions are still not amenable to preparation and analysis by PLM or TEM. The particle size of these fractions precludes making proper slide preparations for PLM or grid preps for TEM. Instead, both fractions are examined by low-magnification stereomicroscopy for suspect asbestos materials that might then be isolated and ground fine enough to be mounted for analysis. This approach can potentially lead to false-negative results, and when asbestos is found it makes accurate quantitation difficult. If all three fractions are “non-detectable” for asbestos then the fine fraction is prepared and analyzed qualitatively by TEM. This helps to ensure that small fibers beyond the resolution of light microscopy are not being missed. The coarse and medium fraction are not prepped for or analyzed by TEM since their particle size is prohibitive. Milling Approaches

Figure 1. Two-dimensional slab cake, part of the lab portion of the incremental sampling methodology.

Rather than separating out the ﬁne fraction, the milling methods reduce the entire sample to a parti-

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

123

Cahill

Figure 2. Soil lab instrumentation (clockwise from top left): jaw crusher, riﬄe splitter, turbula mixer, and sieve shaker.

cle size amenable to preparation and analysis for microscopy (US EPA, 1993; California Air Resources Board, 1991). Once the rock and/or soil have been reduced to a ﬁne powder they are more readily homogenized and prepared for light or electron microscopy (Figure 3). The advantage of milling over sieving is that the entire sample is now represented by the subsample

124

being analyzed. This can reduce the chance for falsenegatives and helps to provide a more accurate overall quantitation of asbestos content. Just as PLM and TEM analyses are complementary, so too can the sieving and milling preparation approaches be used in conjunction. If, for example, the ﬁne fraction is sieved out of the sample and

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

Asbestos in Soil

Figure 3. Typical heterogeneous soil and rock sample, as submitted and after milling.

analyzed prior to milling, we get the advantage of looking at the fine fraction unaltered by milling. This can be important in NOA situations, since milling can alter fiber dimensions, break up large bundles of asbestos, and potentially create cleavage fragments from non-asbestiform minerals that can be counted as fibers by the method used. The subsequent milling of the coarser fraction(s) makes those fractions amenable to proper PLM and TEM analyses. DATA INTERPRETATION Unequivocal identification of individual mineral fibers found in soil can be difficult. Many of the amphiboles are part of solid solution series, with no clear boundaries between members. It can likewise be difficult to distinguish true asbestiform minerals, those

that grow in a mainly unidirectional manner to form extremely long and thin fibers, from those that have broken or cleaved into countable “fibers” by the analytical methods being applied. These are subjective decisions, and it is currently debated just how important those distinctions are. Most existing regulations focus on the indoor environment and six specific regulated asbestos types. But minerals other than the regulated six, such as those that are part of what is known as the “Libby Amphibole” suite of minerals (Meeker et al., 2003), and Erionite, a fibrous Zeolite, are also implicated in causing disease. The National Institute for Occupational Safety and Health (NIOSH) Current Intelligence Bulletin 62 (NIOSH, 2011) known as the “Asbestos Roadmap” advocates looking beyond just the regulated six asbestos minerals. More guidance, policy, and regulations are needed. Most methods that are applied to soil and rock for asbestos content provide results as a percentage of asbestos. While this is informative to a degree, it provides little information regarding risk. At any given percentage it is the number of respirable fibers that really inform as to the potential risk. For this reason many consultants will augment their sampling of the soil or rock with activity-based air sampling (ABS). ABS typically involves collecting air samples on personnel outfitted with personal protective equipment while mimicking an activity that is likely to occur on the site. The results (in fibers per cubic centimeter) can be compared with existing air action limits such as the OSHA Permissible Exposure Limits (OSHA, 1986). Values can also be compared with historic exposure levels and corresponding toxicity values via the EPA’s Integrated Risk Information System (US EPA, 1988) for risk characterization. ABS can be useful but is relatively expensive to implement and is hindered by the myriad variables associated with field sampling. Wind velocity and direction, moisture content of soil, height of the sampler (distance from the source), etc., can lead to variability between measurements. To avoid some of the pitfalls of ABS the EPA has begun using the Fluidized Bed Asbestos Segregator (Januch et al., 2013). With this instrument (Figure 4), a known quantity of soil is made to behave like a fluid in an enclosed glass vessel by flowing HEPA-filtered air in the opposite direction of sedimentation, thus liberating respirable dust while a TEM air sample is collected isokinetically from above. The results are expressed in releasable fibers per gram of soil. This instrument was developed for the EPA by the Idaho National Lab (Wright and O’Brien, 2007) and is based on work by Spurney et al. (1975) and others. The EPA made several improvements to the initial design and is using the technique on select projects. The fluidized bed procedure is currently cited

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

125

Cahill Table 2. Comparison of sampling techniques. Technique

Strengths

Limitations

Activity-based sampling (ABS)

Personal air sampling results in ﬁbers per cubic centimeter can be used directly for risk assessment

Relatively expensive Field sampling can result in wide variabilty due to changing environmental conditions

Fluidized bed asbestos segrator (FBAS)

Field sampling variables are eliminated for more consistent results Less expensive than ABS, since no personal air monitoring is required Can be used to determine if the more expensive ABS sampling should be performed

Results in ﬁbers per gram are only indirectly applicable for risk assesssment models

as “Other Test Method 42” (US EPA, 2018) as it awaits the Federal rulemaking process. The results from the ﬂuidized bed analysis can augment ABS data or be used to help determine if ABS is necessary. Both of these techniques are complementary and can be used in conjunction to take advantage of the strengths of each approach, as summarized in Table 2. CONCLUSIONS The determination of asbestos and other mineral ﬁber content in soil and rock is complex and fraught with variables. It is important to recognize these variables as well as the limitations of each approach to sampling and analysis in order to obtain quality data in the appropriate metric for the data end users. Fortunately, the various approaches and methodologies are

Figure 4. Fluidized bed asbestos segregator.

126

complementary, and each can contribute to the overall characterization of the soil and/or rock on a given project site. How much sampling and analytical effort are to be applied depend on many variables and can be determined on a site-by-site basis. REFERENCES ASTM D7521-16, 2016, Standard Test Method for Determination of Asbestos in Soil:ASTM International, West Conshohocken, PA. Buck, B.; Goosens, D.; Metcalf, R.; McLaurin, B.; Freudenberger, F.; and Ren, M., 2013, Naturally occurring asbestos: Potential for human exposure, southern Nevada, USA: Soil Science Society America Journal, Vol. 77, No. 6, pp. 2192–2204. California Air Resources Board, 1991, Determination of Asbestos Content of Serpentine Aggregate: Method 435, 19 p. Interstate Technology Regulatory Council (ITRC), 2009, Incremental Sampling Methodology Training Courses available at https://www.itrcweb.org/Guidance/ ListDocuments?topicID=11&subTopicID=16 Januch, J.; Brattin, W.; Woodbury, L.; and Berry, D., 2013, Evaluation of a fluidized bed asbestos segregator preparation method for the analysis of low levels of asbestos in soil and other solid media: Analytical Methods, Vol. 5, No. 7: 1658–1668. Meeker, G. P.; Bern, A. M.; Brownfield, I. K.; Lowers, H. A.; Sutley, S. J.; Hoefen, T. M.; and Vance, J. S., 2003, The composition and morphology of amphiboles from the Rainy Creek complex, near Libby, Montana: American Mineralogist, Vol. 88, No. 11-12, pp. 1955–1969. Meeker, G. P.; Lowers, H. A.; Swayze, G. A.; Van Gosen, B. S.; Sutley, S. J.; and Brownfield, I. K., 2006, Mineralogy and Morphology of Amphiboles Observed in Soils and Rocks in El Dorado Hills, California: U.S. Geological Survey Open-File Report 2006-1362, 47 p. National Institute of Occupational Safety and Health (NIOSH), 2011, Asbestos Fibers and Other Elongate Mineral Particles: State of the Science and Roadmap for Research: Current Intelligence Bulletin 62, Publication No. 2011-159, 174 p. Occupational Safety & Health Administration [OSHA]. (1986) 1910, Subpart Z. Regulations (Standards-29 CFR 1910.1001. Retrieved from https://www.osha.gov/lawsregs/regulations/standardnumber/1910/1910.1001 Spurney, K.; Boose, C.; Hochraine, D.; and Geselschaft, F., 1975, On the pulverization of asbestos fibers in a fluidized-bed aerosol generator: Staub Reinhaltung Luft [in English], Vol. 35, No.12, pp. 440–445.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

Asbestos in Soil U.S. Department of Health and Human Services Agency for Toxic Substances and Disease Registry (ATSDR), Division of Health Assessment and Consultation, 2011, Evaluation of Community-Wide Asbestos Exposures—El Dorado Hills Naturally Occurring Asbestos Site, El Dorado Hills Boulevard, El Dorado Hills, California: Health Consultation, EPA Facility ID: CAN000906083, 194 p. U.S. Environmental Protection Agency (US EPA), 1973, National Emission Standards for Hazardous Air Pollutants Compliance Monitoring (NESHAP), 40 CFR Part 61 Subpart M; amended 1990 and 1995. US EPA, 1980, Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA or Superfund); amended by the Superfund Amendments and Reauthorization Act, 1986.

US EPA, 1986, Asbestos Hazard Emergency Response Act, 40 CFR Part 763, Subpart E. US EPA, 1988, Integrated Risk Information System Assessments: Asbestos, CASRN 1332-21-4. US EPA, 1993, Method for the Determination of Asbestos in Bulk Building Materials: EPA600/R-93/116, 98 p. US EPA, 2008, Framework for Investigating AsbestosContaminated Superfund Sites: OSWER Directive 9200.0-68, 71 p. US EPA, 2018, Sampling, Sample Preparation and Operation of the Fluidized Bed Asbestos Segregator: Other Test Method (OTM) 42, 21 p. Wright, K. and O’Brien, B., 2007, Fluidized Bed Asbestos Sampler Design and Testing: INL/EXT-07-13122, 23 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 121–127

127

Reﬁnement of Sampling and Analysis Techniques for Asbestos in Soil JULIE WROBLE* U.S. Environmental Protection Agency, Region 10, 1200 6th Avenue, Seattle, WA 98101

TIM FREDERICK U.S. Environmental Protection Agency, Region 4, 61 Forsyth Street SW, Atlanta, GA 30303

DANIEL VALLERO U.S. Environmental Protection Agency, Oﬃce of Research and Development, National Exposure Research Laboratory, 109 T.W. Alexander Drive, Research Triangle Park, NC 27709

Key Terms: NOA, Asbestos, Soil Sampling, Incremental Sampling, Fluidized Bed ABSTRACT Measuring the concentrations of asbestos in contaminated soils is challenging. Data are often highly variable. Variability in soil measurements has led to limitations in comparing results from sites nationally and difficulties in reproducing results, even from the same sites over time. The difficulties in collecting reproducible soil data limit the ability to extrapolate from concentrations in soil and compare to concentrations in air. This extrapolation is necessary if soil data are to be used in human health risk assessments. To address this substantial limitation of asbestos soil data, researchers from Environmental Protection Agency (EPA) regions and the National Exposure Research Laboratory are conducting a series of efforts to advance the use of data that are collected, processed, and analyzed using the most reproducible methods. These soil data, collected from a variety of sites across the country, will be compared to air data from activity-based sampling in an attempt to establish a quantitative relationship between asbestos soil concentrations and airborne fiber concentrations. This research plan summary provides an update on the EPA efforts under way and the challenges that lie ahead. INTRODUCTION The Framework for Investigating AsbestosContaminated Superfund Sites of the Environmental Protection Agency (EPA) establishes activity-based sampling (ABS) as the preferred method for estimat*Corresponding author email: Wroble.Julie@epa.gov

Figure 1. Sampling team collecting incremental samples in area contaminated with asbestos-containing material (ACM). (Inset) ACM debris on the ground in sample area.

ing potential risks from asbestos in soil at Superfund sites (U.S. Environmental Protection Agency, 2008). ABS usually follows a site determination that asbestos is present or may be present based on soil sampling data and/or the visual identification of asbestos debris in soil. Using this approach, the air concentrations measured using ABS can be used to quantify the potential risks to current or future site users. Implementing ABS at potentially contaminated sites can be costly and time consuming. It can also be difficult to implement when members of the general public are located nearby. Devising methods that would allow for using soil samples in risk-based decision making at Superfund sites would lead to a more efficient process. In a previous study, EPA researchers compared three soil analytical methods and two soil sampling

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 129–131

129

Wroble, Frederick, and Vallero

methods to determine whether one method or a combination of methods would yield more reliable soil asbestos data than other methods (Wroble et al., 2017). Samples were collected using both traditional discrete (“grab”) sampling and incremental sampling methodology (ISM) (Interstate Technology and Regulatory Council, 2012). Analyses were conducted using methods established by the California Air Resources Board (California Environmental Protection Agency, 1991, 2017), the American Society of Testing and Materials (2013), and the Environmental Protection Agency (Januch et al., 2013; U.S. Environmental Protection Agency, 2011, 2018). Our data show that the fluidized bed asbestos segregator (FBAS), a process developed by EPA scientists, followed by analysis using transmission electron microscopy, was the most sensitive analytical method (Wroble et al. 2017). It was anticipated that ISM would provide less variable data than discrete sampling. While this was true for metals data collected to provide an additional measure of variability, the asbestos data remained highly variable. The next phase of this work is intended to address some of the issues identified in the earlier study. Soil will be collected from a variety of asbestoscontaminated sites (naturally occurring asbestos as well as soils contaminated with asbestos-containing materials) and contaminated with a variety of fiber types. Increasing the number of increments collected for each sample from 30 to 100 may help to overcome the problems of representativeness when asbestos in soil is not homogeneously distributed in soil. The number of replicate samples collected at each site will be increased from three to five to allow for better statistical comparisons. To be more rigorous, soil processing will be conducted in a laboratory rather than the field. Laboratory preparation of samples is expected to be more consistent across samples than field preparation because preparation conditions can be more controlled. Finally, activity-based air samples will be compared to FBAS data. The collected data may provide a scientific basis for using FBAS data directly for risk-based decision making at asbestos-contaminated Superfund sites. SOIL SAMPLING Soil samples from three asbestos-contaminated sites will be collected using ISM (Fig. 1). Four performance evaluation (PE) samples (two asbestos types at two different concentrations) will also be included in the study in attempt to provide a comparison of known concentrations of asbestos in a consistent medium to uncertain concentrations of asbestos in environmental samples. PE samples are prepared by adding a known

130

mass of asbestos to a known volume of soil, whereas soil measurements often rely on point counts or visual area estimation, which may not relate directly to mass. The PE samples will be prepared according to an approved work plan by a contracted laboratory. Analysis of PE samples will help to provide context to the environmental soil measurements for asbestos in this study. Both PE samples and site samples will be processed at a single location. Processing will include vigorous mixing of each sample using a Turbula mixer. Samples will then be systematically subsampled to provide the necessary soil volume for each analytical method: American Society of Testing and Materials, California Air Resources Board, and FBAS. The remaining soil from each sample will be reserved and combined with other soil from each site for ABS conducted in a laboratory setting. AIR SAMPLING A key part of the next phase of this project will be conducting activity-based sampling in a laboratory setting. One combined sample from each location will be used for the ABS activity. ABS will be performed inside of an exposure chamber at the EPA’s Oﬃce of Research and Development laboratories in Research Triangle Park, North Carolina or a similar facility. Investigators in personal protective equipment will mimic the activity of a child playing in the dirt while wearing sampling pumps with sample media attached. Investigators will also perform ABS activities with the PE soil samples. DATA COMPARISONS When all the data have been collected as described above, a variety of comparisons can be made. The team will compare the relative variability of each soil sample location and the PE samples for each soil analytical method. The purpose of these comparisons will be to determine which of the analytical methods, if any, provides reproducible soil data. The comparisons of site samples to PE samples will determine whether samples collected from contaminated sites are inherently more variable than laboratory-prepared PE samples. The investigators will also compare the ABS samples to the FBAS sample data. The data may provide a scientiﬁc basis for using FBAS data directly for risk-based decision making at asbestos-contaminated Superfund sites. CONCLUSIONS When completed, the researchers intend to have established FBAS as the preferred method for detecting

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 129–131

Soil Analysis for Asbestos

low levels of asbestos in soils based on ability to detect and reproducibility. The researchers also hope to demonstrate that by reﬁning ISM techniques, reproducibility of asbestos sampling/analytical methods for soil is improved. The inclusion of PE samples will provide context to relate real-world concentrations of asbestos in soil to standards.

REFERENCES ASTM D7521-13, 2013, Standard Test Method for Determination of Asbestos in Soil: ASTM International, West Conshohocken, PA. doi:10.1520/D7521-13. www.astm.org California Environmental Protection Agency, Air Resources Board, 1991, Determination of Asbestos Content of Serpentine Aggregate, Test Method 435: Electronic document, available at https://www.arb.ca.gov/testmeth/vol3/m_435.pdf California Environmental Protection Agency, Air Resources Board, 2017, Implementation Guidance Document, Field Sampling and Laboratory Practices, Test Method 435: Determination of Asbestos Content of Serpentine Aggregate: Electronic document, available at https://www.arb. ca.gov/toxics/asbestos/tm435/guidancedocument.pdf Interstate Technology and Regulatory Council, 2012, Incremental Sampling Methodology: Electronic document, available at http://www.itrcweb.org/ism-1

ISO 10312, 1991, Ambient Air—Determination of Asbestos Fibres— Direct-Transfer Transmission Electron Microscopy Method: International Organization of Standardization, Geneva, Switzerland. Januch, J.; Brattin, W.; Woodbury, L.; and Berry, D., 2013, Evaluation of a ﬂuidized bed asbestos segregator preparation method for the analysis of low-levels of asbestos in soil and other solid media: Analytical Methods, Vol. 5, No. 7, pp. 1658–1668. U.S. Environmental Protection Agency, 2008, Framework for Investigating Asbestos-Contaminated Superfund Sites: Asbestos Committee of the Technical Review Workgroup of the Oﬃce of Solid Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, DC. OSWER Directive No. 9200.0-68. U.S. Environmental Protection Agency, 2011, Field Standard Operating Procedure: Sampling, Sample Preparation and Operation of the Fluidized Bed Asbestos Segregator: U.S. Environmental Protection Agency, Seattle, WA. OEAFIELDSOP-102. U.S. Environmental Protection Agency, 2018. Other Test Method—42: Sampling, Sample Preparation and Operation of the Fluidized Bed Asbestos Segregator: Air Emission Measurement Center, U.S. Environmental Protection Agency, Washington, DC. Wroble, J.; Frederick, T.; Frame, A.; and Vallero, D., 2017, Comparison of soil sampling and analytical methods for asbestos at the Sumas Mountain Asbestos Site—Working towards a toolbox for better assessment: PLoS ONE, Vol. 12, No. 7, p. e0180210. https://doi.org/10.1371/journal. pone.0180210

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 129–131

131

Discerning Erionite from Other Zeolite Minerals during Analysis ROBYN RAY* EMSL Analytical, Inc., 200 Route 130 North, Cinnaminson, NJ 08070

Key Terms: Erionite, Transmission Electron Microscopy, Liquid Nitrogen Cryogenic Holder, Cold Stage, Offretite, Quantitative EDS, Asbestos ABSTRACT Erionite, a naturally occurring fibrous mineral that belongs to the zeolite group has been designated by the International Agency for Research on Cancer (IARC) as a Group 1 Carcinogen on the basis of mesothelioma, a disease also resulting from the inhalation of asbestos fibers. Significant outcrops of fibrous erionite have been reported in California, North Dakota, Nevada, Oregon, and other states. For geologists and industrial hygienists dealing with mining, construction, or various aspects of community protection, it is vital to understand the basics of detecting and handling erionite, since it is similar to asbestos and can cause similar disease. There are many fibrous zeolites, and discerning erionite from these other minerals requires modifications to current asbestos analysis methods. Without these modifications, identification and quantification are questionable and could increase the likelihood of both false negatives and false positives. There is currently no published method specific to erionite analysis; without guidance standards, each laboratory has approached erionite analysis independently. With a few small but significant changes to asbestos analysis methodologies, we developed a reproducible analytical procedure for rapid identification of erionite fibers in air, bulk, and soil samples by transmission electron microscopy (TEM). Using specialized preparation techniques, energy dispersive Spectrometry (EDS) calibrations, and a liquid nitrogen cryo-holder (cold stage), we were able to overcome the difficulties associated with erionite analysis. By incorporating these changes, commercial analytical laboratories can contribute reliable data to air-exposure studies and characterization guidelines, which may help in determining regulations and further understanding the health risks of erionite.

*Corresponding author email: rray@emsl.com

INTRODUCTION Erionite is a member of a large group of hydrated aluminosilicate minerals called zeolites. Zeolites are found in altered volcanic tuffs, ash, and the soils derived from them. In Turkey, these volcanic tuffs are used as building stone for houses (Carbone et al., 2011). Similar to asbestos mineral fibers, erionite mineral fibers have been shown to cause to cancer (International Agency for Research on Cancer [IARC], 2012; Saracci, 2015). Internationally, there are areas in Turkey (Dogan et al. (2008)) and Mexico (Ilgren et al., 2008; Kliment et al., 2009) where malignant mesothelioma has been attributed to erionite that was found in the local environment. There are several occurrences of erionite worldwide. In the United States, it is predominately found in the intermountain west: California, North Dakota, Nevada, Oregon, and other states (Van Gosen et al., 2013, Van Gosen et al. (1996), USGS (1996)). Erionite has not been mined for commercial use since the late 1980s (National Toxicology Program [NTP], 2004). Therefore, many commercial laboratories focusing on asbestos in bulk building material are inexperienced in handling erionite. In fact, in 2009, when erionite was provided as an unknown sample on the transmission electron microscopy (TEM) proficiency round for the National Voluntary Laboratory Accreditation Program (NVLAP), only 3 out of 76 accredited laboratories identified it as erionite (National Institute of Standards and Technology, 2009). Erionite is not regulated as an asbestos mineral in the United States, and no exposure limits have been published. The potency of erionite has been shown in animal studies, where erionite is 500–800 times more tumorigenic than chrysotile and 200 times more tumorigenic than crocidolite (Carbone et al., 2011). Exposure to erionite may occur during mining and production of other zeolites where erionite is found as a contaminant (Rom et al., 1983). Erionite-related diseases are being studied among road construction and maintenance workers who may have been exposed to erionite containing gravel used in road surfacing (National Institute of Occupational Safety and Health [NIOSH], 2014). In 2017, the Nevada Department of Transportation (NDOT) started conducting

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

133

Ray

surveys of their right of ways, pit material, and import material for natural occurrences of asbestos and erionite (NDOT, 2018). Evidence for non-cancerous outcomes due to erionite exposure has also been studied. Erionite has been shown to cause pleural fibrosis and produce the same auto-antibodies commonly seen in people with systemic auto-immune diseases (Zebedo et al., 2014). With this evidence of increased negative health effects, it is important to know who is most likely to be exposed to erionite, where it can be found today, and how to test it. Erionite vs. Asbestos Preparation With erionite’s similarities to asbestos, it is easy to expect that asbestos analysis methods would be sufficient. However, erionite and asbestos could not be more different. Erionite is very delicate and does not endure well when subjected to asbestos preparation and analysis techniques. The majority of the erionite samples that have come into our laboratory for testing have been rock and soil, or air samples collected when rock and soil has been disturbed. Rock and soil material must be reduced to a fine powder for analysis by optical microscopy. We performed milling tests using material collected by NDOT from Pine Valley, NV. The material was milled to two nominal sizes: 250 μm and 75 μm. These two milling targets were chosen because the 250 μm target was used for the soil processing at the Libby Environmental Protection Agency (EPA) Superfund Site, and the 75 μm target is referenced in the California Air Resource Board (CARB) Method 435 for Serpentine in Aggregates (CARB, 2017). The results showed that erionite is fragile and extremely susceptible to overgrinding (Figure 1). Once over milled, bundles and fibers are destroyed and broken into non-fibrous particles, and these structures would no longer be countable by an analyst during microscopic examination. An IARC study in 1989 showed this to be the case with erionite from Oregon as well, though no milling size targets were specified (IARC, 1989). In an effort to remove organics, soils can undergo the same gravimetric reduction as floor tiles and roofing shingles and assist the analyst in fiber identification (EPA, 1993). However, when dealing with erionite in soil, it is extremely vulnerable to this preparation procedure because it interferes with the chemistry that is being measured. When performing quantitative energy dispersive spectrometry (EDS) chemistry for identification, the elemental composition is less than it should be for identification of erionite. For a sample from Reese River, NV, the average charge balance, a measure of chemical reliability, prior to gravimet-

134

Figure 1. Erionite is susceptible to over-grinding and can be shattered into non-ﬁbrous particulate. (a) Erionite from Pine Valley, NV, milled to nominal 250 μm. (b) Erionite from Pine Valley, NV, milled to nominal 75 μm, the target set in the CARB 435 method (CARB, 2017).

ric reduction was 13.72 percent, whereas after gravimetric preparation, the average charge balance was 148.77 percent. ERIONITE ANALYSIS In the United States, TEM is the preferred electron microscopy technique for asbestos because one can quickly obtain chemistry, morphology, and diffraction of the fibers under analysis. Even so, that is only true if the sample can withstand the high energy that is being used during analysis. Asbestos, which is known for its thermal stability, can withstand this energy with limited damage, though even asbestos is not completely impervious to damage (Brindley and Zussman, 1957; Martin et al., 2016). Most zeolites are quite sensitive to the electron beam. Once the energy of the electron beam collides with these delicate fibers, they

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

Discerning Erionite from Other Zeolite Minerals

Figure 3. Erionite from Reese River, NV, showing particulate loading of a soil sample under TEM.

The majority of erionite research has focused on using chemistry alone for identification. It is typically measured by inductively coupled plasma mass spectrometry or scanning electron microscopy with an electron microprobe on large rock samples. These techniques are not suitable for individual fibers collected on air samples, so TEM equipped with an EDS system is employed. While TEM EDS is often used in asbestos analysis, it is a technique that can be heavily influenced by particulate and matrix that has also collected on the filter. Due to the hydrous nature of zeolites, when running quantitative elemental chemistry, it is important to verify the reliability of the analysis through the use of the equation for charge balance (E-balance) as described in Passaglia et al. (1998) and Dogan and Dogan (2008): Charge balance Al + Fe3+ − Na + K + 2 × (Ca + Mg + Sr + Ba) = Na + K + 2 × (Ca + Mg + Sr + Ba) ×100. (1) Figure 2. Erionite on a standard TEM holder. (a) The fiber prior to focused beam interaction. (b) Same fiber after focused beam interaction.

deform (Figure 2). The degradation caused by the electron beam also influences the chemistry and crystal structure. In standard TEM analyses of erionite, the selected area electron diffraction (SAED) pattern does not last long enough to be documented and measured for identification. Moreover, once these interactions start, they cannot be stopped. This problem affects quality control and re-analysis because the damaged fiber is no longer available for confirmation during the subsequent analysis.

This test assesses the Al³+ substitution in the T-site and compares it to Al³+ calculated from the measured ionic cations. Ideally, these numbers should balance each other to keep the crystal at a neutral charge. However, the weakness in this equation for TEM EDS is that the equation assumes all the Al measured in the spectrum is substituting into the T-site and is not a contribution from another source, such as a surface particulate. Indeed, Gualtieri et al. (2016) showed that Fe found in some erionite analyses was actually coming from ironbearing nano-particles on the surface of the erionite ﬁbers. On TEM samples, one cannot always overcome high particulate loadings, which can compromise EDS collection (Figure 3).

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

135

Ray

Figure 4. (Left) EDS spectra of oﬀretite. (Right) EDS spectra of erionite from the same sample (AZ).

Another difficulty with EDS arises in the fact that distinguishing between the different zeolites on a fiber by fiber basis is challenging. In Figure 4, we show the EDS spectrum of two different zeolites from the same sample: offretite and erionite. The mineral identification was determined by diffraction; however, one can see the similarity between the two spectra are—they are almost indistinguishable. During EDS analysis, the energy of the beam displaces elements from the structure. A study in 2011 (Carbone et al., 2011), showed just how the beam affected chemistry. In this study, they found that the K counts were reduced by one third during 15 seconds of low-energy beam interaction. Under the much higher energy of TEM, we are “losing” the chemical elements we are trying to measure quantitatively to confirm erionite. As described above, the problems related to TEM analysis of erionite stem from its interaction with the electron beam, and the key to overcoming these problems is stabilization. Cryogenic electron microscopy holders (referred to as a cold stage) have been shown to help stabilize zeolite fibers in the TEM (Gualtieri et al., 1998). The cryogenic holder is a simple addition to the laboratory protocol; cooling by liquid nitrogen

(LN2), it protects the sample from the energy of the electron beam. After employing the cold stage, and applying Cliff Lorimer quantitative calibrations, we saw improved chemistry as demonstrated in the charge balance test. Table 1 shows how, with this technique, more particles pass the chemical reliability (charge balance) test of 20 percent as used by the USGS (USGS, 2010). In a comparison using both a standard (non-cooled) TEM holder and the cold stage, we estimated that on an erionite sample from Rome, OR, we were able to reduce the rate of Na loss from 41 percent to 12 percent. The cold stage cannot overcome all of the limitations of TEM EDS, and the user should still expect that interference from particulate and some cation loss will still be present. Consequently, identification by chemistry alone will remain inconclusive on some fibers. For our samples, all TEM analysis was conducted on a JEOL 1200 EX II analytical TEM operating at 100 kV. This microscope was outfitted with an AMT Digital camera, IXRF Iridium Ultra Software, and Gatan single-tilt cryo-holder, cooled with liquid nitrogen. The samples were milled to ∼250 μm particles size in a puck mill. To calibrate the EDS system, we used ground Icelandic basaltic glass (BIR-1G) purchased

Table 1. Reese River quantitative chemistry and SAED with cryogenic holder. Criteria Si + Al TSi Mg/(Ca + Na) Mg (72 O) Si/Al Abundant cation Charge balance (%) SAED inter-row

136

Reference Value

36 0.68–0.79 <0.15 <0.8 2.85–3.6 Ca, Na, K

36.03 0.77 0.01 0.03 3.34 K 2.56 15.1

36.3 0.77 0.03 0.09 3.31 K 19.3 15.2

35.77 0.76 0.17 0.36 3.25 K 28.59 15.5

36.04 0.78 0.05 0.16 3.48 K 9.56 15.3

35.91 0.77 0.02 0.06 3.38 K 2.92 15.2

36.06 0.77 0.01 0.03 3.43 K 13.17 15.4

36.17 0.77 0.02 0.05 3.39 K 11.07 15.3

36.11 0.77 0.02 0.06 3.34 K 14.55 14.9

36.43 0.77 0.01 0.03 3.36 K 26.76 15.4

35.87 0.78 0.02 0.06 3.48 K 8.73 15.4

15.07

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

Discerning Erionite from Other Zeolite Minerals

Figure 5. SAED patterns of erionite from diﬀerent locations showing the ∼15.0 Å layer line spacing taken on a single-tilt cold stage. (a) Killdeer, ND. (b) Rome, OR. (c) Reese River, NV.

from the USGS and Microcline NMNH 143966 from the Smithsonian Institute, which were analyzed under the same conditions used for erionite analysis (tilt of 30 degrees and spot size of ∼160 nm on the Gatan cryoholder). The reference values obtained were used to create calibration files for the IXRF system. These calibration files were then applied to the unknown spectra and analyzed using the Foil Quantitative routine in the software. For erionite, all EDS results were based on 72 oxygen atoms. The camera constant used for indexing diffraction patterns was calibrated from the Au grid from Ted Pella. Utilizing the cryogenic holder, the biggest difference was seen in the successful recording of diffraction patterns, with most lasting longer than 5 minutes (Figure 5). Unlike analysis with the routine holder, even when chemistry and diffraction analysis had previously been performed, the diffraction results were repeatable during quality-control analysis. This is a worthwhile alternative to having to rely solely on EDS chemistry, since it is easily compromised by surrounding particulates. Electron diffraction (ED) can now also be used to differentiate erionite from other zeolites because the 15 Å layer line spacing of erionite is unique amongst the fibrous zeolites. Therefore, when one is trying to differentiate between offretite, mordenite, and erionite, it can be easily done with ED, because offretite has half of the layer line spacing that erionite does. This is not to say EDS chemistry is not valuable—it is, but having multiple ways to confirm a mineral is necessary for defensible erionite identification (Harper et al., 2017). We also tested other zeolites from various localities, including erionite from Arizona; and chabazite, offretite, and mordenite from Italy. Using the cold stage, we examined each mineral by chemistry and diffraction and compared them to published reference values. In compiling the data, trends were noted which showed

that when the chemistry did not balance, diffraction patterns could still confirm the analysis. Recommended Approach When Seeking Erionite Analysis With samples such as soil and stone, it is important to ensure that a representative subsample is collected. Incremental sampling may help in this regard (Interstate Technology and Regulatory Council [ITRC], 2012). When looking for a laboratory to analyze erionite, it is important to use a laboratory that has cryogenic TEM experience and that has demonstrated competency with quantitative EDS analysis. For analysis, request TEM EPA 600/R-93/116 with the following modifications:

r Do not gravimetrically reduce the sample or treat with acid.

r Samples are to be milled to a nominal 250 μm size. r Samples are to be analyzed on the TEM cryo-holder or similar cold stage.

r Fibers confirmed as erionite must meet the following criteria: 1) Diffraction is ∼15 Å, and qualitative chemistry is consistent with erionite. 2) If no diffraction pattern can be achieved, then quantitative chemistry can be used provided the following conditions are met: a. The charge balance formula, as described in Passaglia et al. (1998) is modified to ࣘ 20 percent as used in USGS (2010). b. Calibrations for Cliff Lorimer K-factors include K. c. Si + Al ∼ 36. d. Mg < 0.8.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

137

Ray

CONCLUSIONS An abundance of care and prior experience is needed when sampling, preparing, and analyzing erionite samples. During preparation, we demonstrated how over-milling an erionite sample could lead to false negatives, since erionite loses its elongated shape. Erionite is also sensitive to gravimetric reduction, which can affect quantitative chemistry. Since many zeolites look similar by qualitative chemistry, quantitative chemistry should also be performed. Laboratories will need to acquire additional standards to determine Cliff Lorimer K factors (k-factors) for erionite analysis. Regardless of the sample matrix, cryogenic TEM is the best technique to acquire reproducible and defensible data on erionite. REFERENCES Brindley, G. W. and Zussman, J., 1957, A structural study of the thermal transformation of serpentine minerals to forsterite: American Mineralogist, Vol. 42, pp. 461–474. California Air Resources Board (CARB), 2017, Implementation Guidance Document Air Resources Board Test Method 435 Determination of Asbestos Content of Serpentine Aggregate Field Sampling and Laboratory Practices: Electronic document available at https://www.arb.ca.gov/toxics/asbestos/ tm435/workshops/m435-asbestosguidance-2017.pdf Carbone, M.; Baris, Y. I.; Bertino, P.; Brass, B.; Comertpay, S.; Dogan, A. U.; Gaudino, G.; Jube, S.; Kanodia, S.; Partridge, C. R.; Pass, H. I.; Rivera, Z. S.; Steele, I.; Tuncer, M.; Way, S.; Yang, H.; and Miller, A., 2011. Erionite exposure in North Dakota and Turkish villages with mesothelioma: Proceedings of the National Academy of Sciences of the United States of America, Vol. 108, No. 33, pp. 13618–13623. Dogan, A. U. and Dogan, M., 2008, Re-evaluation and reclassification of erionite series minerals: Environmental Geochemistry and Health, Vol. 30, No. 4, pp. 355–366. Dogan, A. U.; Dogan, M.; and Hoskins, J., 2008, Erionite series minerals: Mineralogical and carcinogenic properties: Environmental Geochemistry and Health, Vol. 30, No. 4, pp. 367–381. Environmental Protection Agency (EPA), 1993, Test Method for the Determination of Asbestos in Bulk Building Materials: Environmental Protection Agency Report EPA/600/ R-93/116. Gualtieri, A.; Artioli, G.; Passaglia, E.; Bigi, S.; Viani, A.; and Hanson, J. C., 1998, Crystal structure–crystal chemistry relationships in the zeolites erionite and offretite: American Mineralogist, Vol. 83, No. 5–6, pp. 590–606. doi:10.2138/am-19985-619. Gualtieri, A. F.; Gandolfi, N. B.; Pollastri, S.; Pollok, K.; and Langenhorst, F., 2016, Where is iron in erionite? A multidisciplinary study on fibrous erionite-Na from Jersey (Nevada, USA): Scientific Reports, Vol. 6, pp. 37981. doi:10.1038/srep37981. Harper, M.; Dozier, A.; Chouinard, J.; and Ray, R., 2017, Analysis of erionites from volcaniclastic sedimentary rocks and possible implications for toxicological research: American Mineralogist, Vol. 102, pp. 1718–1726. doi:10.2138/am-20176069.

138

Ilgren, E. B.; Pooley, F. D.; Larragoitia, J. C.; Talamantes, M.; Navarrete, G. L.; Krauss, E.; and Breña, A. F., 2008, First confirmed erionite related mesothelioma in North America: Indoor and Built Environment, Vol. 17, pp. 567–568. International Agency for Research on Cancer (IARC), 1989, Modification of fibrous Oregon erionite and its effects on in vitro activity. In Bignon, J.; Peto, J.; and Saracci, R. (Editors), Non-Occupational Exposure to Mineral Fibers: Scientific Publication 90, IARC, Lyon, France, pp. 74–80. International Agency for Research on Cancer (IARC), 2012, IARC Monograph on the Evaluation of Carcinogenic Risks to Humans. Volume 100C: Arsenic, Metals, Fibres and Dusts: IARC, Lyon, France, 311 p. Interstate Technology & Regulatory Council (ITRC), 2012, Incremental Sampling Methodology: Electronic document, available at https://www.itrcweb.org/ism-1/pdfs/ISM1_021512_Final.pdf Kliment, C.; Clemens, K.; and Oury, T., 2009, North American erionite-associated mesothelioma with pleural plaques and pulmonary fibrosis: A case report: International Journal of Clinical and Experimental Pathology, Vol. 2, No. 4, pp. 407–410. Martin, J.; Beauparlant, M.; Sauvé, S.; and Espérance, G., 2016, On the threshold conditions for electron beam damage of asbestos amosite fibers in the transmission electron microscope (TEM): Journal of Occupational and Environmental Hygiene, Vol. 12, pp. 924–935. National Institute for Occupational Safety and Health (NIOSH), 2014, Health Hazard Evaluation Report: Evaluation of Erionite and Silica Exposure During Dirt Road Maintenance: NIOSH HHE Report No. 2012-0141-3220, U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, Cincinnati, OH: Electronic document, available at http://www.cdc.gov/niosh/hhe/reports/pdfs/2012-01413220.pdf National Institute of Standards and Technology, 2009, NVLAP TEM Proficiency Test 2009-1 Summary Report: National Institute of Standards and Technology, Boulder, CO. National Toxicology Program (NTP). 2004, Erionite. In Report on Carcinogens, 11th ed.: NTP, Durham, NC, pp. III114–III115. Nevada Department of Transportation (NDOT), 2018, Documenting Naturally Occurring Asbestos and Erionite in Import Material for Nevada Department of Transportation Projects: State of Nevada Department of Transportation Environmental Services Division, Carson City, NV. Passaglia, E.; Artioli, G.; and Gualtieri, A., 1998, Crystal chemistry of the zeolites erionite and offretite: American Mineralogist, Vol. 83, No. 5–6, pp. 577–589. Rom, W.; Casey, K.; Parry, W.; Mjaatvedt, C.; and Moatamed F., 1983, Health implications of natural fibrous zeolites in the Intermountain West: Environmental Research, Vol. 30, No. 1, pp. 1–8. Saracci, R., 2015, Erionite and cancer in a Mexican village: Occupational and Environmental Medicine, Vol. 72, pp. 163–164. U.S. Geological Survey (USGS), 1996, Occurrences of Erionite in Sedimentary Rocks of the Western United States: U.S. Department of the Interior, Denver, Colorado: Electronic document, available at http://pubs.usgs.gov/of/1996/ 0018/report.pdf U.S. Geological Survey (USGS), 2010, Chemical and Morphological Comparison of Erionite from Oregon, North Dakota, and Turkey: U.S. Geological Survey Open-File Report 2010-1286, 8 p.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

Discerning Erionite from Other Zeolite Minerals Van Gosen, B. S.; Blitz, T. A.; Plumlee, G. S.; Meeker, G. P.; and Pierson, M. P., 2013, Geologic occurrences of erionite in the United States: An emerging national public health concern for respiratory disease: Environmental Geochemistry and Health, Vol. 35, No. 4, pp. 419–430. doi:10.1007/s10653-012-9504-9.

Zebedeo, C. N.; Davis, C.; Peña, C.; Ng, K. W.; and Pfau, J. C., 2014,. Erionite induces production of autoantibodies and IL-17 in C57BL/6 mice: Toxicology and Applied Pharmacology, Vol. 275, No. 3, pp. 257–264.

Environmental & Engineering Geoscience, Vol. XXVI, No. 1, February 2020, pp. 133–139

139

EEG Journal - February 2020 Vol. XXVI, No. I (2)

Articles inside

Discerning Erionite from Other Zeolite Minerals during Analysis

New Tools for the Evaluation of Asbestos-Related Risk during Excavation in an NOA-Rich Geological Setting

Sampling, Analysis, and Risk Assessment for Asbestos and Other Mineral Fibers in Soil

Refinement of Sampling and Analysis Techniques for Asbestos in Soil

Geological Model for Naturally Occurring Asbestos Content Prediction in the Rock Excavation of a Long Tunnel (Gronda di Genova Project, NW Italy

Geologic Investigations for Compliance with the CARB Asbestos ATCM

Identification and Preliminary Toxicological Assessment of a Non-RegulatedMineral Fiber: Fibrous Antigorite from New Caledonia

Management of Naturally Occurring Asbestos Area in Republic of Korea

Fibrous Tremolite in Central New South Wales, Australia

Regulations Concerning Naturally Occurring Asbestos (NOA) in Germany—Testing Procedures for Asbestos

Naturally Occurring Asbestos in France: a Technical and Regulatory Review

Naturally Occurring Asbestos in France: Geological Mapping, Mineral Characterization, and Technical Developments

Naturally Occurring Asbestiform Minerals in Italian Western Alps and in Other Italian Sites

Asbestiform Minerals of the Franciscan Assemblage in California with a Focus on the Calaveras Dam Replacement Project

Naturally Occurring Asbestos in Valmalenco (Central Alps, Northern Italy): From Quarries and Mines to Stream Sediments

Does Exposure to Naturally Occurring Asbestos (NOA) During Dam Construction Increase Mesothelioma Risk?

NOA Air-Quality Lessons Learned during Calaveras Dam Replacement Project

Overview of Naturally Occurring Asbestos in California and Southwestern Nevada

Naturally Occurring Asbestos: A Global Health Concern? State of the Art and Open Issues

Clastic Sedimentary Rocks and Sedimentary Melanges: Potential Naturally Occurring Asbestos Occurrences (Amphibole and Serpentine

Foreword to the Environmental & Engineering Geoscience Special Edition on Naturally Occurring Asbestos