F&S International Edition 2014 by VDL-Verlag

00_fs_international_2014_umschlag__ 21.05.14 20:15 Seite 1

D 11665F

International Edition

2014

International Edition

MFP 3000 Filter test system for flat filter media and small filter elements

New: DLB 2000 Condi oning of temperature and rela ve humidity for the compressed air used by the generator during powder dispersion

for Filtration and Separation Technologies

00_fs_international_2014_umschlag__ 21.05.14 20:16 Seite 2

02_fs_international_2014_seite_03-98__ 21.05.14 20:08 Seite 3

Highlights 2013

Dear Readers, In Germany and the neighbouring countries, the magazine F&S has been quite an institution for the past 27 years – after all, it is the only German-speaking trade journal exclusively dedicated to filtration and separation technology but also to the treatment of disperse substance systems. Our readers and advertisers especially value F&S because of the high quality of the published articles and essays that were often trendsetting and also describe today’s valid standards. For the fourteenth time, we have now had a small part of our broad editorial spectrum translated into English. These are contributions that were published in the year 2013. By doing so, we want to provide the contents of our magazine to process engineers in non-German-speaking countries as well. As we said, this is only a small selection of our articles. With a complete translation of all the articles that were published in the year 2013, you would now hold a thick book of about 450 pages in your hands. We would like to wish you a lot of reading pleasure and would be pleased to receive your feedback. If you would like to find out more about the German F&S, please do not hesitate to contact us at the address listed below (also see imprint: page 98).

With best regards

Eckhard von der Lühe Publisher

VDL-Verlag GmbH Heinrich-Heine-Straße 5 D - 63322 Rödermark Phone: + 49 (0) 60 74 / 92 08 80 Fax:

+ 49 (0) 60 74 / 9 33 34

E-mail: vdl-verlag@t-online.de Internet: www.fs-journal.de

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:08 Seite 4

MFP 3000 - Innovations in filter testing The MFP test rig is a modularly built filter test system for flat filter media and small filter elements. The following filter characteristics can be determined reliably and thereby economically within shortest time: • pressure drop development at the unloaded medium • initial fractional separation efficiency • fractional separation efficiency during loading • gravimetric separation efficiency and dust holding capacity. The MFP 3000 offers particular technical advantages in the choice of test conditions and easy operation: • Due to its modular set-up the MFP 3000 can be used either for loading or for fractional efficiency measurement with different types of challenge aerosols as e.g.: - ISO powders, e.g. ISO A2 Fine / ISO A4 Coarse etc. - Salt NaCl / KCl and others with maintenance free drying column - Droplets as e.g. DOP / DEHS - Soot • Widest range of face velocities from < 2 cm/s up to 1 m/s in suction mode. • Quasi simultaneous particle detection in upstream and downstream with welas® digital 3000 in one instrument, therefore easy correlation adjustment. • Wide measurement range from 200 nm up to 40 µm with high resolution (32 intervals per decade). The welas® digital 3000 aerosol spectrometer fulfills the requirements according to ASHRAE 52.2 and EN 779. • Wide range of aerosol concentrations from < 7 mg/m3 up to > 1 g/m3 (ISO A2 Fine). • Highest reproducibility in test results. • Integrated corona discharging system. • NEW: DLB 2000 rel. Humidity of dispersion air for powder generator RBG 1000 conditioned to 50 % rel. Humidity Thus, the MFP 3000 is the ideal system to test filter media in quality control and R&D for the following applications: • HVAC filter test in accordance with ASHRAE 52.2 and EN 779 • Cabin air filtration in accordance with ISO TS 11155-1 • Motor air inlet filtration in accordance with ISO 5011; ISO 19713 • Vacuum cleaners • HEPA/ULPA filters • and others Palas® has delivered almost 100 complete filter test systems to international customers and is therefore market leader worldwide. The success of Palas® filter test systems is based on their high reproducibility, flexibility and easy operation. Palas® – Your reliable partner

Contents ▼ Highlights 2013

Process intensification with membranes S. Ripperger

Ceramic membranes for the filtration of liquids: An actual overview St. Duscher 13 Separation technology in winemaking H. Lyko

The paper production and pulp industry on its way to sustainable production Report from the PTS Environmental Symposium 2013 H. Lyko 26 Centrifuge technology: Materials, machine design and applications Report from the 4th International Separation Technology Symposium of the Schmidt+Clemens Group H. Lyko 32 Systems and processes for drinking water and process water treatment H. Lyko

Nanotechnology in practical water application: Efficient solutions for the future Report of the event nano meets water IV at Fraunhofer UMSICHT H. Lyko

Status and perspectives of organophilic nanofiltration H. Lyko

Chemical-free treatment of cooling water in open and half-open circuits H. Lyko 52 Filters for clarification of auxiliary liquids R. Berndt

Universal similarity laws for the description of the initial pressure loss of metal woven wire cloths during the filtration of non-Newtonian fluids M. Müller, M. Piesche 60 Optimized filtration mesh for ballast water management systems Perfectly designed filtration media and filtration packages to meet increasing demands M. Knefel

Automatic backwash filter improves performance in pharmaceutical process St. Schöpf, St. Strasser, L. Ertl 70 Palas GmbH Greschbachstr. 3 b 76229 Karlsruhe, Germany Phone: +49 721 96213-0 Fax: +49 721 96213-33 E-Mail: mail@palas.de Internet: www.palas.de

Aerosol technology is flying high Report from the 27. Palas-Aerosol Technology Seminar H. Lyko

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:08 Seite 5

Contents Highlights 2013

Energy-efficient generation and treatment of compressed air Report from COMVAC 2013 H. Lyko Pressure loss of air filters for general ventilation at high relative humidity or exposed to water droplets F. Schmidt, A. Breidenbach, Suhartiningsih Semi-automated Gas Turbine Inlet Filter testing according ARAMCO standard or for performance optimization D. Renschen, J. Schamberg, N. Guttenbrunner, N. Schneider

Secure waste air filtration in laboratory autoclaves Integrated water intrusion test to fulfil the requirements in the area of biotechnology C. Grumbach, P. Czermak 89 Product development challenge S. Ripperger

Fine dust pollution decreases slightly A report by the Federal Environmental Agency

Prevention of the spreading of legionella H. Lyko

Eleven requirements for a future environmental policy

Imprint

02_fs_international_2014_seite_03-98__ 21.05.14 20:08 Seite 6

Highlights 2013

Process intensification with membranes S. Ripperger* In the field of process engineering, the term "process intensification" includes measures to improve the efficiency of a process drastically. The improvements may include, for example, an increase in the space-time yields of a system, a significantly higher energy and / or raw material utilization or relate to a new product or new product qualities. A process intensification may involve different concepts and measures, which are explained in the introductory part of the following article. Based on the specific properties of the membrane process is shown later that with the use of membranes a process intensification was associated mostly. Frequently thus could also be created new products and/or product qualities. 1. Introduction For several years, one trend within process engineering and chemical engineering has been described with the concept “Process intensification (PI)”. The concept was introduced during the seventies in connection with chemical processes /1/. In the eighties, a working direction under the concept was developed within process engineering through the work of C. Ramshaw /2/. C. Ramshaw has intensified the heat exchange and mass transfer for gas to liquid mass transfer processes by means of centrifugal acceleration in rotary contact apparatuses, which is several times higher compared with gravitational acceleration, and has introduced the “High-g Technology” based on it. With the development of microprocess engineering in the nineties, the area of “process intensification” was further stimulated and looked at with a broader point of view. According to contemporary understanding, process intensification can be associated with - an increase of space-time yields in a system, - an improved use of energy and raw materials, - a drastic lowering of production costs (investment costs and/or operating costs), - an improvement of environmental protection (reduction of exhaust air, sewage and waste streams), - an increase of the inherent process reliability and - the generation of new or improved product quality. The difference to regular advancement of processes exists in that, on account of new approaches, significant improvements in one or several directions listed above are achieved. Such improvements can be achieved on the basis of a detailed understanding of * Prof. Dr. Siegfried Ripperger Department of Mechanical Process Engineering University of Technology, Kaiserslautern PO box 3049 67653 Kaiserslautern / Germany Tel./Fax: +49 (0) 631-205-2121/3055 E-mail: ripperger@mv.uni-kl.de

the procedures taking place in a process and the introduction of new methods and technologies. The measures, which lead to significant improvements over conventional processes, can be divided into different categories, which are considered in more detail below. 2. Measures for process intensification Process intensification can be achieved with different measures, which can be divided into the following categories. 2.1. Process enhancement The process enhancements contain an essential intensification of mixing, heat and mass transfer processes. In connection with chemical or biochemical reactions coupled thereto, the space-time yield and/or selectivity can thereby be increased substantially. Through a homogenisation of the procedures in the process space, improved and/or new product qualities can also be achieved. As an example, the development of micro-reaction technology is often named in this context, which shows that blending the reactants and the heat transfer can be greatly intensified with the minimisation of reactor dimensions. The ratio of surface to volume rises with micro-structured devices to several thousand m2/m3. Heat exchange and mass transfer through the wall of a process space can thereby be substantially strengthened, so that also the full potential of a chemical reaction is exhausted. As a result of the short distances, big gradients for momentum, heat and mass transfer are also connected with it. The necessary production rates can be achieved if one operates the required number in flow ducts and/or reactors in parallel (numbering-up). If the results achieved in the laboratory are not achieved with the large-scale system, the risk that is connected with a usual scaling-up does not exist with the concept of “numbering-up”. By contrast, measures must be taken to counteract fouling of the walls and a blockage of the small ducts. This kind of process intensification is applied advantageously, inter alia, in mix-

ing-sensitive and/or highly exothermic or endothermic homogeneous reactions. Process enhancement is often also possible if a machine is used instead of an apparatus that is characterised by the absence of driven moving parts. Apparatus are operated in connection with pumps or fans. In particular, this applies if it is possible with a machine to master difficult process conditions (e.g. high pressure, high concentrations). The High-g Technology according to Ramshaw includes such a transition from a mass transfer apparatus to a mass transfer machine. The advantages and disadvantages of mass transfer machines, compared with apparatuses, were already treated by Brauer /3/ and Mersmann /4/ in 1986, on the basis of examples. 2.2. Process integration Process integration consists of the simultaneous execution of several basic operations in one apparatuses and/or in one machine. Originally these were carried out in sequentially-connected separate apparatuses or machines. With simultaneous execution of basic processes in an apparatus and/or a machine, the transactions overlap and mutually influence each other. With exact knowledge of the transactions, these can be specifically influenced and limitations of chemical conversion can be overcome. The transactions can be “intensified”. Examples of integrated methods are reactive distillation, reactive chromatography, as well as extruders and similar screw machines, in which several transactions happen at the same time (e.g. mixing, heating, melting, reacting and giving a shape). The integration of several process stages in one apparatus and/or in one machine is often linked with advantages, in spite of necessary compromises, as a result of the mode of operation of individual basic operations not being optimal. As advantages of such integrations, there can be mentioned: increase of yields in chemical or biochemical reactions, minimisation of operation costs and investment costs, enclosed design and low handling effort. In the pharmaceutical and

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 7

Highlights 2013

food industries, in this way product contamination can also be avoided in the safest manner. 2.3. Hybrid processes With a hybrid process, similar effects can often be achieved like with process integration. With this, individual processes are operated in separate apparatuses, but which are joined together so that they mutually support each other during the chemical conversion. It is about an active group by which synergies are created that allow the limits of a single procedure to be exceeded. Hybrid processes for process intensification therefore contain configurations of process stages with new effects. 2.4. Reduction of process stages through new or improved procedures An essential efficiency increase can often also be achieved by introduction of a new or optimised process. In particular, when one succeeds with reducing altogether the process steps necessary for the production of a product. Thus, e.g. a precipitation and/or crystallisation can be operated in such a manner that the required particle size and/or particle size distribution is generated directly. Costly product processing by grinding and classification can be dispensed with. A specific influencing of elementary steps of a basic operation can often be achieved also with in-line and/or on-line process monitoring in the form of a control circuit. The precondition here is that there are sensors with which the desired control variable (e.g. the particle size) can be measured in-line and/or on-line. 2.5. Use of new auxiliary agents Process intensification can also be achieved with existing procedures through the use of a new auxiliary agent. These include, for example, improved filtration aids and flocculants, and new extraction and absorption agents (e.g. ionic liquids). During reactions, higher space-time yields can be achieved in particular by using new and/or improved catalysts. Many examples show that the introduction of a membrane process was often connected with process intensification for the purposes of one of the possibilities described above. 3. Identifying properties of the membrane process In the last five decades, membrane processes have opened new possibilities in the technology for substance separation that are increasingly used today. Decisive for this development were the following advantageous properties of the membrane processes:

F & S International Edition

No. 14/2014

- high selectivity of the substance separation, - substance separation without additives, - physical separation principle without chemical change of components, - possibility of substance separation in the molecular area without phase change, - gentle separation conditions, - simple, modular system design, - enclosed system design, - possibility of continuous operation. Often, only one of these properties is decisive for the application of a membrane process. Product saving separation is used for example in the fields of medicine and biotechnology. The possibility of selective substance separation without additives is in the foreground with applications for the specific discharge of substances from a mixture or during materials recycling. With such applications it is advantageous if the substance mixture is treated at the source of the origin. Hence, one is anxious to integrate closed and continuously working reprocessing procedures into the process. The enclosed and modular system design and also the continuous operating method of membrane systems, do justice very well to the process-integrated separation stage. Membrane systems, which allow selective substance separation in the molecular domain without phase separation, are an alternative to the thermal separation processes with additives (e.g. absorption, adsorption, extraction and chromatography). All arguments speaking in favour of the use of membrane technology measures also represent measures for process intensification in a broader sense. With the use of membrane technology, often the production of a new product quality also becomes possible. Thus, for example, sterile products were able to be produced without the negative changes of thermal sterilisation. Moreover, in the process, products are generated in which no killed cells are included. Today this method is used on a large scale for the sterilisation of beverages and other liquids. ESL milk (ESL = extended shelf life), introduced only few years ago based on microfiltration with membranes, is one example of this. In the case of a high bio-burden and the availability of other substances in the form of colloids, very dense top layers form quickly on the membrane, which strongly hinder filtration. For the treatment of such suspensions, crossflow filtration was developed, with which the membrane is flooded by the suspension to be filtered. On account of the crossflow, a back transport of separated substances away from the membrane is achieved, due to diffusion and achieved hydrodynamic forces. There are different forms of crossflow filtration

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 8

Highlights 2013

s is the thickness of the wall of the capillary membrane. εMo is the packing density of the module, i.e. the surface proportion of the sheath cross-section of the module that is filled with capillary membranes. Theoretical limit values, over which this proportion cannot be increased, arise from regular membrane arrays. The following limit values can be determined:

with a square array of the capillaries in the packing and

at the hexagonally densest packing of the capillaries. In practice, during module manufacturing with capillary membranes values are achieved in the range from 0.45 to 0.65.

Fig. 1: Specific membrane surface of a capillary module as a function of the inner diameter of the capillaries dM; ai related to the inner capillary diameter, aa related to the outer capillary diameter, εMo = 0.5; z = 1.2

known. The throughflow of membrane modules represents the most often used form. With the flow parameters and the structural design of the flow ducts mass transport processes can also be intensified and the necessary membrane area can be reduced. The sterile filtration, which is situated in the field of micro-filtration and thus still in the area of the particles and colloids, was transferred, with the development of asymmetric membranes, to membrane processes having a cut-off at a molecular level (e.g. ultrafiltration, nanofiltration, reverse osmosis, pervaporation). In this area, electrostatic effects and solution/diffusion processes in the membrane also play a crucial role in mass transfer. Some membrane processes are based on relatively slow diffusion processes, which is disadvantageous in terms of process intensification. This disadvantage can be achieved by optimisation of the membrane (e.g. very thin membrane layers with short diffusion pathway, or membrane in the form of thin capillaries or hollow fibres for the realisation of larger specific membrane surfaces), optimisation of the process manner (e.g. counterflow principle, improved mass transfer through increased overflow, use of turbulence promoters), or an optimised system configuration (e.g. multistage operating method). With such an optimisation counterpropagating influencing dependencies must be taken into account. 4. Examples of process intensification with membranes 4.1. Process enhancement 4.1.1. Use of miniaturised structures The use of miniaturised structures for process enhancement has been applied in membrane technology for a long time. Membranes in the form of the finest hollow fibres with an inner diameter of less than 0.5 mm are an example of this. The ratio of surface to volume rises with the membrane units made therefrom (membrane modules), similarly as with other micro-structured devices, to several thousand m2/m3. In Fig. 1 the specific membrane surfaces, related to the internal and external capillary diameter, of a capillary module are applied as a function of the inner diameter of the capillaries. These can be determined with the following equations:

4.1.2. Dialysis The gentle separation conditions at ambient temperature are used at hemodialysis, performed in the detoxigenation of the blood and thus the human body. Here, the transport of the substances to be removed from the body through the membrane into the dialysate (predominantly processed water) is caused primarily by a concentration difference and the diffusion coupled thereto. For the process enhancement of this, actually slowly accruing diffusion, extremely thin membranes (thickness down to 5 μm) were developed, mostly in the form of hollow fibres. The membranes can be produced with a spinning process in large quantities at a reasonable price. The substance transport could still be optimised by the fact that one has succeeded to design the membrane surface necessary for the treatment (approx. 1 m2) in a relatively small volume. This is possible because the membranes are produced with an inner diameter in the range of approx. 200 μm. The necessary throughput is achieved by the fact that the capillaries in the membrane module are connected in parallel and operated in the required number (numbering-up). For a dialysis module, if one uses the capillary dimensions listed above and a packet density of εMo = 0.5 as a basis, based on Fig. 1 it can be seen that one achieves, with respect to the capillary inner diameter, a specific membrane surface of approx. 7000 m2/m3. Similar orders of magnitude are aimed for with fibre membranes, also with gas separation. With other technical applications of capillary membranes, like for example water filtration or the gassing and degassing of liquids, specific membrane surfaces of 1000 to 2000 m2/m3 in a membrane module are absolutely usual. 4.1.3. Membrane contactors For the gassing and degassing of liquids, membrane modules are used as membrane contactors. Contactors are used if a mass transfer should be achieved across the phase boundaries between two phases. For hydrophobic microporous membranes, the interfacial forces in connection with the low pore diameters prevent the penetration of a liquid into the pore structure up to a certain pressure, so that both phases are directly in contact via the pores in the membrane surface. To this end, membranes with a surface porosity of 75 to 93% are used, with simultaneously a maximum pore diameter in the range of 0.1 to 1 μm. The mass transfer is determined primarily by the laws of diffusion. Membrane contactors are used for absorptive gas separation, for bubble-free gassing of liquids, as well as for extractive and distillative substance separation. They allow a flow guide in direct current and counter-current and allow the throughput of the discharging and absorbing phase to be varied in a wide range. In Fig. 2, the mass transfer surface with respect to the gas throughput of different contact apparatuses is applied as a function of the specific energy input resulting from the pressure drop /5/. One detects that with conventional contact apparatuses, in contrast to a membrane contactor, a high specific mass transfer surface is also always connected with an increased energy input. With a

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 9

Highlights 2013

Membrane contactor (Bubble free)

Fig. 2: Mass transfer surface area of different contact apparatuses related to the gas throughput, depending on the specific energy input due to the pressure drop; SD = jet nozzle reactor; BS = bubble column; RK = continuous stirred-tank reactor; GFK = Direct current packed column; membrane reactor with the following data: sheath diameter: 0.25 m, length: 2.00 m; capillary inner diameter A: dM = 0.6 mm, B: dM = 1.2 mm /5/

membrane contactor, on the other hand, a large specific exchange surface is also guaranteed with a low volume flow and hence also with low energy input. This one is used for benefit, inter alia, for extra-corporeal blood oxygenation with membrane oxygenators. They have extensively replaced the earlier usual bubble oxygenators. Membrane oxygenators are routinely used today during surgery in connection with heart-lung machines and take over the task to supply blood with oxygen and to remove the carbon dioxide at the same time. Another example of the application of membrane contactors is air humidification in air conditioning systems. With an appropriate membrane selection, micro-organisms are retained by the membrane and only water vapour is transferred to the air. Membrane contactors are used as a matter of routine today, also for degassing water in combination with ultra-pure water treatment. With them, vacuum degassing or degassing with a stripping gas can be implemented /6/. It can be seen from Fig. 2 that at high gas throughputs, the exchange surface area related to the gas throughput decreases and the specific energy input increases as a result of the rising pressure drop. Hence, membrane contactors have not yet asserted themselves with large gas throughputs.

Fig. 3: Rotary membrane disks filter with a radial shear gap /7/

ods of time without periodic cleaning is usually not possible to achieve. With periodic membrane backflushing, exposure of the membrane surface area cannot be guaranteed with a simultaneous overflowing of the membrane. Moreover, the slowly occurring processes of membrane fouling are to be taken into account. Given that these events do not occur uniformly but have a greater effect in some places through slight differences in the flow guidance, the non-uniformities of the ongoing processes are thereby still intensi-

4.1.4. “Numbering-up” principle With all membrane applications, the principle of numbering-up is consistently applied. It is enough for the case of capillary membranes and tubular membranes to examine the transactions in a capillary and/or a tube of the length later to be installed. Then the results can be transferred to large-scale systems through numbering-up (parallel connection of many membrane capillaries and/or tubes to a membrane module and of several membrane modules in a module block). With very big systems, the principle will be consistently accomplished by interconnecting several module blocks in parallel with the necessary systems capacity in each case. Nevertheless, in conjunction with membrane technology, the disadvantages of microstructured apparatuses and “numberingups” also become clear. A uniform inflow to a great number of membrane units and/or a great membrane surfaces over long peri-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 10

Highlights 2013

fied. Non-uniformities due to deposits and adhesions are removed by planned periodic cleaning. 4.1.5. Process intensification through moving installations or moving membranes A membrane module has no movable parts and can therefore be assigned to the apparatuses. In addition, membrane units have also been developed with which mass transfer is intensified through moving installations or moving membranes. In this case, membranes from ceramics are preferably used, often in the form of membrane disks. Particularly favourable conditions arise when, in the process, a shear gap over the membrane is realised. These membrane units have movable parts with their own driving mechanism and, hence, can be called mass transfer machines. According to their construction type, the following categories can be distinguished: - the membrane is rigid and is overflowed on account of moved installations, - the membrane moves, and is thereby overflowed and - the overflow is implemented by rigid and moving membranes. As advantages of these systems compared with systems with membrane modules, there can be mentioned: - a reduced specific energy demand since the energy being entered into the system principally serves to reinforce the mass transfer transactions into the membrane and to decrease fouling. Energy dissipation in pumps, pipes and valves, which occurs in the membrane modules, is avoided. - Membrane machines with moving parts can be operated with higher viscosity media compared with membrane modules, such as for example highly concentrated suspensions. A higher permeate yield is thereby also often possible. This applies particularly to suspensions with structurally viscous or thixotropic flow behaviour. - Advantageous with filtration processes with moving parts is also that the flow conditions on the membrane influence the transmembrane pressure difference to a lesser extent than with membrane modules. With membrane modules, the shear rates on the membrane rise with increasing throughflow speed and at the same time the pressure drop in the module increases, so that the pressure difference of the filtration across the module length changes significantly. 4.2. Process integration with membranes 4.2.1. Membrane reactors Examples of process intensification through process integration are the membrane reactors and/or bioreactors. As a rule, the membrane is arranged with them in the reaction chamber. With chemical and biochemical processes, this opens the possibility to separate intermediate products and/or by-products of the reaction selectively from the reaction mixture. This direct use of the membrane properties in reactors is still used relatively little up to now in the petrochemical industry, but is the object of numerous research projects. Because many reactions are operated in a spatially confined reaction chamber at high temperatures, selectively active membranes very resistant to temperature are mostly required. On the other hand, many biochemical reactions can be coupled with the existing polymer membranes, which is why biomembrane reactors are already used on an industrial scale in many ways. Chemical and/or biochemical reactions can be operated at optimum conditions through a continuous separation of valueadded products and/or inhibitory substances from the reactive substance mixture. The reaction equilibrium can thus be shifted towards the product side and possible subsequent reactions can be suppressed. Thus, it is often possible to increase the turnover and space/time yields substantially. One example for this is the separation of water in polycondensation reactions by means of the pervaporation.

Fig. 4: Particle produced from calcium carbonate with a membrane process that has been developed in TU Kaiserslautern

In other cases, membranes are used in reactors, e.g. as the carrier of a catalyst, so that in addition to the selective substance separation, an accelerated material transformation is achieved. The varied possibilities of the use of membranes in membrane reactors were indicated by Westermann et al. /8/ 4.2.2. Membrane bioreactors for wastewater treatment Something that has prevailed on an industrial scale are membrane bioreactors with submerged membrane units in the activated sludge basin of a biological purification stage for aerobic wastewater treatment. The membranes are overflowed through the circulation in the basin as a result of the gassing with air. The membrane units form a membrane bio-reactor (MBR) together with the basin. Membrane units with capillary membranes and with flat membranes are offered. A comparison of the two systems is made in /9/. The filtrate is sucked off from the open basin on the filtrate side by creating a negative pressure < 0.5 bar. The first municipal sewage treatment plant of this kind was put into operation in 1989 in Japan. In /10/ operational results of such facilities are reported. The permanent or cyclic air supply that is also connected with circulation in the basin allows a stable operation of the systems in conjunction with cyclic backflushing. As a rule, they are operated with specific filtrate flows < 40 l/(m2h). The specific energy input is given as 0.05 to 0.15 kWh/m3 filtrate. Through this process, integration advantages arise on account of the savings of secondary settling tanks and a higher biomass content, whereby the space-time yield is also increased. The filtrate of such systems mostly fulfils the requirements that are placed on the inlet to an RO system. Hence, an RO system can be connected directly for further water treatment. Such an application for closure of a water circuit is reported in /11/. 4.2.3. Electro deionization Another example of successful process integration with membranes is Electro deionization (EDI), which is widely used in industrial processes for ultrapure water preparation. It combines electric dialysis with ion exchangers in such a manner that the exchanger resins are continuously regenerated, allowing continuous deionization. The regeneration of the contaminated resins with salt and/or acid and alkaline solution, which is necessary with ion exchanger columns, is omitted. Different methods were developed that are described in detail by Duscher /12/. 4.2.4. Membrane precipitation With crystallization and precipitation processes, the objective is to guide the process so that the product is generated in the desired particle size distribution and morphology. The aim can be frequently achieved in continuous stirred-tank reactors on a laborato-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 11

Highlights 2013

ry scale, but the problem is to transfer the process, which is significantly influenced by micro-mixing, to continuous stirredtank reactors on a production scale. Elaborate post-processing steps, such as crushing and screening, are required when the described objective is not fulfilled. The effort increases significantly when particle systems are to be generated in the range of a few micrometres or even nanometres. In TU Kaiserslautern, a new precipitation method was developed where precipitation takes place in capillary membranes /13/. The method was applied to precipitate calcium carbonate from a calcium hydroxide solution with carbon dioxide. In the process, the gaseous component diffuses through a membrane into the boundary layer flow, so that super-saturation and particle formation take place in it. On account of hydrodynamic lift and the increasing size of the particle, this moves away from the membrane, and thus out of the reaction zone, so that it can grow further. Since it is a highly hydrophobic porous membrane, the diffusion resistance of the membrane is negligibly small. For the case of calcium carbonate precipitation, it was proved that the reaction takes place in the laminar substratum of the flow, i.e. that the concentration boundary layer for CO2 is smaller than the flow boundary layer. Furthermore the significant influence in this case of hydrodynamic lift (lift force) was shown. Experimental studies show that it is possible to generate particles in the < 1 Îźm size range in high numbers with the process. Also novel porous spherical particles could be generated from calcium carbonate (Fig. 4). 4.3. Hybrid processes For a long time it has been known that in many cases membrane processes are used especially advantageously in combination with other (conventional) separation processes (hybrid processes). Such processes can serve to increase the effectiveness of chemical conversion substantially.

F & S International Edition

No. 14/2014

Turbid liquid

Fig. 5: Purification of liquids with a hybrid process consisting of centrifugation and crossflow microfiltration

Turbid liquid

Fig. 6: Conventional multistage process for clarifying liquids with subsequent pasteurisation

Turbid liquid

Fig. 7: Conventional multistage process for clarifying beverages with sterile filtration with membranes

4.3.1. Novel purification process As an example of a hybrid process, the purification of suspensions with colloid ingredients (e.g. vinegar, oil, beer, wine) with membranes should be mentioned, in which cross-flow microfiltration is combined with a centrifuge (Fig. 5). Concentration of the turbid matter takes place through the separator and the fine filtration and sterilisation filtration is carried out with the crossflow microfiltration in a continuously operated process. It often replaces a multistage discontinuous purification process, as is illustrated as an example in Fig. 6 and 7. With the process in Fig. 6, pasteurisation (thermal process stage) is connected downstream of the purification process for the increase of the shelf-life of the product. A new product quality could already be achieved by the substitution of

pasteurisation in Fig. 5 with Dead-End filtration with membranes in the form of cartridge filters, according to Fig. 7. In addition, to guarantee a high service life of the filter cartridge, the use of upstream deep filter cartridges is recommended. The substitution of filter aids and/or sheet filters and cartridge filters that have a relatively short service life with relatively durable membrane modules is one reason that the process according to Fig. 5 can be operated economically, in spite of increased energy requirements. The combination is interesting for large and small operations if, at the same time, attention is paid to a high utilisation extent of the system. On account of the operating method, an automatic operation is easily possible, so that a relatively small system can be implemented with a high utilisation extent.

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 12

Highlights 2013

4.3.2. Combination of reactors with membrane stages The direct coupling of a reaction with a separation stage on the basis of membranes, described in 4.2.1, can also be implemented in the form of a hybrid process. Here, the chemical reactor is spatially separated from the membrane stage, but they are operated coupled. This combination is used widely in biotechnology. 4.3.3. Hybrid processes with membranes in the field of Fluid Process Engineering New approaches for hybrid processes with membranes in the field of Fluid Process Engineering are described by T. Cellar et al. /14/. They describe, inter alia, the coupling of the rectification with pervaporation or vapour permeation for overcoming the azeotropic point. One main application area of such combinations is the dehydration of bio-ethanol and other solvents. Furthermore, the reactive distillation and its combination with membrane processes are addressed. Such a combination offers the possibility to selectively improve the yield and the conversion of the reaction and simultaneously to influence favourably the distillative separation by the use of membrane separation. Other variations for the combination of membrane processes and the rectification into one hybrid process are addressed in /15, 16/. In vapour permeation, the membrane surface, in contrast to pervaporation, is directly subjected to the vaporous mixture. Here, process integration would also be suitable, i.e. the integration of the membrane surfaces into the head of the rectification column so that both processes operate in one apparatus. In /17/, the coupling of a reactor is described with a pervaporation stage for the preparation of acetals in the form of a hybrid process. Through the coupling, a significant increase in conversion and a concomitant significant energy saving could be achieved. 4.4. Reduction of process steps with dynamically operated membrane processes Some of the examples listed above show that often a clear reduction of process steps is connected with the described measures. Dynamically operated membrane processes are continuously operated and, hence, can be integrated simply into the corresponding processes. For instance, the advantages of crossflow microfiltration have been used since the early 80s of the last century for industrial vinegar production /18/. Before, vinegar was fined with bentonite and clarified with pre-coat filters and other depth filters. Today vinegar is mostly produced semicontinuously, primarily from ethanol by submerged fermentation over a period of about 30 hours and discharged directly 12

from the digester via crossflow filtration in the desired quality, in the form of a hybrid process. 4.5. Use of novel auxiliary agents Process intensification can be also achieved by novel and/or improved auxiliary agents. These include for example optimised catalysts and/or bio-catalysts (enzymes). Enzymes unfold their activity even in the low concentration range. Low application concentrations make their recovery problematic after substance conversion. Therefore, in the food industry, for example, they are no longer recovered in some cases, and are used as so-called “lost enzymes”. However, the costs are increased through the constant “enzyme consumption”. The enzyme and its attendant materials may also be considered as a disruptive component and an impurity in the product. With enzyme immobilization, the use of valuable enzymes and higher enzyme concentrations becomes economically justifiable and the advantages of continuously operated flow systems become applicable. Such systems are implemented, for example, with capillary membranes for ultrafiltration, so that the enzyme is immobilized inside the capillary membranes that represent a barrier for the enzyme. The membranes are permeable to the starting materials and products and are placed in the reactor or in a perfused membrane module associated with it. In /19/, one of the first industrial scale applications of such a system for the production of amino acids is described. Besides this “physical immobilization”, enzymes are also bound directly to the internal or external surface of a carrier membrane and are immobilized with it. In many cases, the economical use of enzymes in biochemical conversions only became possible with an appropriate membrane use, since only in this way is an economical, sufficiently long lifetime of the enzymes achieved. Membranes, preferably made of ceramic, are accordingly also suitable with chemical reactions as a carrier for catalysts. In other cases optimised catalysts can be used in the form of nanoparticles or macromolecules with an accordingly high specific surface, which can then be recovered continuously or batchwise after the reaction with a membrane process for repeated use. Such membrane applications are numerous and are operated successfully in the petrochemical industry. 5. Summary As explained, the “process intensification” concept identifies the endeavours to clearly improve chemical conversion in chemical engineering processes. It was shown that such clear improvements were often achieved with the use of membranes. On account of rising requirements, as well as rising energy and raw material prices,

the objectives linked with “process intensification” still remain topical. Accordingly, the membrane applications will increase furthermore. However, the task definitions appearing in connection with process intensification become more complex. Some require a membrane or membrane module development adapted to the task. In such cases, it needs to be considered whether the opportunities offered with it justify the often protracted development work. The fact that such opportunities exist is pointed out, for example, by a contribution in the an issue of the Journal “Chemie Ingenieur Technik” /20/, which reports about clear improvements in the depletion of LPG gas from natural gas by means of zeolite membranes. On account of the sometimes numerous coupled elementary processes, such improvements can be detected in many cases only on the basis of a “comprehensive” consideration of all processes that are coupled to each other. Often this is possible only with the help of mathematical modelling and simulation of the processes. Literature: /1/ Z. Leszcynski: Przem. Chem. 1973, 52, p. 161 /2/ C. Ramshaw: Chem. Eng. (London) 1983, 389 /3/ H. Brauer: Chem.-Ing.-Tech. 58 (1986), No. 2, pp. 97-107 /4/ A. Mersmann, H. Voit, H. Zeppenfeld: Chem.-Ing.-Tech. 58 (1986), No. 2, pp. 87-96 /5/ S. Ripperger: Die blasenfreie Be- und Entgasung von Flüssigkeiten mit mikroporösen Membranen. Chem.Ing.-Tech. 58 (1986) 4, pp. 322 - 323 /6/ St. Duscher: Zeitgemäße Wasserentgasung. Filtrieren und Separieren 24 (2010) No. 1, pp. 10-15 /7/ S. Ripperger: Mikrofiltration mit Membranen - Grundlagen, Verfahren, Anwendungen. VCH Verlagsgesellschaft mbH, Weinheim (1992), ISBN 3-527-28457-5 /8/ Th. Westermann, Th. Melin: Membranreaktoren. Chem.Ing.-Tech. 77 (2005) 11, pp. 1655-1667 /9/ St. Krause, R. Gutknecht: Filtrieren und Separieren 22 (2008) No. 4, pp. 189-190 /10/ A. Kraft, U. Mende: Filtrieren und Separieren 9 (1995) No. 6, pp. 249-254 /11/ K. Vossenkaul, Ch. Kullmann: Filtrieren und Separieren 19 (2005) No. 3, pp. 123-124 /12/ St. Duscher: Ausführungsformen und Anwendungen der Elektrodeionisation (EDI. Part 1: Anforderungen an Reinstwasser und Verfahrensweisen der EDI. Filtrieren und Separieren 24 (2010) No. 4, pp. 230-238 /13/ S. Ripperger, Ch. Schnitzer, L. Steinke: Verfahren und Vorrichtung zur Erzeugung von Mikro- und/oder Nanopartikeln. Patentschrift zum Deutschen Patent No. 10 2006 020 288 of 25.06.2009 /14/ T. Keller, T. Roth, J. F. Mackowiak, P. Kreis, A. Górak, A. Stankiewicz: Prozessintensivierung in der Fluidverfahrenstechnik. Chem.-Ing.-Tech. 83 (2011), No. 7, 935–951 /15/ U. Hömmerich: Pervaporation und Dampfpermeation mit Zeolithmembranen – Einsatzpotenzial und Verfahrensintegration. Dissertation RWTH-Aachen (1998) /16/ P. Kreis, A. Górak: Prozessanalyse hybrider Trennverfahren am Beispiel der Kopplung von Rektifikation und Membrantrennung. Chem.-Ing.-Tech. 77 (2005), No. 11, pp. 1737-1748 /17/ Th. Clavey: Prozessintensivierung mit Pervaporation zur Herstellung von Acetalen –Pilotierung. Chem.-Ing.Tech. 77 (2009), No. 10, pp. 1583-11590 /18/ H. Ebner: Die kontinuierliche Filtration von Essig. Chem.-Ing.-Tech. 53 (1981), No. 1, pp. 25-31 /19/ W. Leuchtenberger, U. Plöcker: Herstellung von Aminosäuren mit Hilfe biotechnologischer Methoden. Chem.-Ing.-Tech. 60 (1988), No. 1, pp. 16-23 /20/ K. Neubauer, U. Lubenau, C. Hecker, B. Lücke, D. Paschek, S. Wohlrab: Abreicherung von Flüssiggas aus Erdgas mittels Zeolithmembranen. Chem.-Ing.-Tech. F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 13

Highlights 2013

Ceramic membranes for the filtration of liquids: An actual overview St. Duscher*

Since nearly one century, ceramic membranes are an established component for the separation of particles out of liquids. Since the development and rollout of ceramic nanofiltration membranes, they also offered an additional method in separating dissolved solids or ions out of liquids under rough conditions. Due to constant improvements in materials and production methods, it is possible to replace established processes like e.g. evaporators by these membranes with lower investment and running costs without any disadvantages in product quality or efficiency. The following article gives an overview about the actual state of development of ceramic membranes with an emphasized focus on ceramic nanofiltration membranes. The outlook of the article gives a rough overview about some actual fields of development and optimization. they were successfully used as membranes for gas diffusion in the subproject of the uranium concentration during the â&#x20AC;&#x153;Manhattan projectâ&#x20AC;?. Shortly after, the development and modification of these membranes went on and membranes for micro- and ultrafiltration have been developed, which consisted already of a porous carrier and a thin ceramic membrane layer, their shapes have already been in single and multi-channel designs. Due to their various advantages, which are for example important for the treatment and filtration of foods and beverages, most of the ceramic membranes were used for microfiltra-

1. Development and design of ceramic membranes 1.1. Historical background of ceramic membranes While there is very diverse information concerning the time of the first ceramic membranes on a laboratory scale, the 40s mark the birth of ceramic membranes commercially produced on a large scale, as * Dipl.-Ing. Stefan Duscher Head of Sales, Inopor GmbH Industriestrasse 1, 98669 Veilsdorf / Germany www.inopor.de

tion of milk as well as (pre-)filtration of wines and juices. Next to the development of ceramic membranes for food and beverages, after the end of the second world war, companies like Carbone Lorraine and Desmarquest did a lot of research and development in ceramic micro- and ultrafiltration membranes for the enrichment of uranium. At the same time, also the development and optimization of polymeric membranes went on; the specific pricing of polymeric membranes (price per membrane area) was significantly below the pricing of ceramic membranes, but the

VISCOTHERMÂŽ Hydraulic Centrifuge Drive Systems

ROTODIFFÂŽ Technology Inherent Advantages

ROTODIFFÂŽ

Advanced design 1

Ĺ&#x2014; %HVW WRUTXH WR ZHLJKW FDSDFLW\ OHVV PRYLQJ SDUWV KLJKO\ UHOLDEOH PLQLPDO RSHUDWLQJ DQG PDLQWHQDQFH FRVWV DQG DQ RXWVWDQGLQJ OLIH WLPH

Best torque to weight ratio Ĺ&#x2014; 7KH K\GUDXOLF GULYH 5272',))ÂŽ KDV RQO\ DERXW RI WKH ZHLJKW RI D JHDU ER[ ZLWK HTXLYDOHQW WRUTXH YDOXH ZKLFK UHGXFHV WKH VWUHVV ORDG RQ PDLQ EHDULQJV DQG RWKHU FRPSRQHQWV

3 4

Overload protection 5 6

Ĺ&#x2014; $ SUHVVXUH UHOLHI YDOYH SURWHFWV WKH V\VWHP DJDLQVW VKRFN ORDGV QR PHFKDQLFDO GDPDJH Ĺ&#x2014; 1R RYHUKHDWLQJ

Quick return of investment

1. Scroll 2. Bowl 3. Scroll drive shaft 4. Rotor 5. Transfer seal 6. Connection block

Ĺ&#x2014; 5HGXFHG SRZHU LQVWDOODWLRQ DQG ORZHU HQHUJ\ FRQVXPSWLRQ Ĺ&#x2014; %HWWHU SURFHVV FRQWURO WKH PD[LPXP WRUTXH LV DYDLODEOH RYHU WKH HQWLUH ZRUNLQJ UDQJH (PHUJHQF\ FOHDQ RXW RI SOXJJHG FRQYH\RU LV SRVVLEOH Ĺ&#x2014; ([FHOOHQW WRUTXH FDSDFLW\ DOVR UHVXOWV LQ KLJKHU WKURXJK SXWV

Viscotherm AG | CH-8132 Hinteregg-ZĂźrich | +41 44 986 28 00 | info@viscotherm.com | www.viscotherm.ch F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 14

Highlights 2013

Fig. 2: Ceramic multi-channel membrane

Table 1: Properties of ceramic membranes

applications were limited due to the chemical and mechanical properties of polymeric membranes. Until the beginning of the second millennium, microfiltration and ultrafiltration was available as polymeric or ceramic membranes, but nanofiltration and reverse osmosis were still technologies which were covered exclusively by polymeric membranes. Things changed significantly, when a new ceramic nanofiltration membrane with a cut-off of 450 Dalton was developed by Hermsdorfer Institut für technische Keramik (H.I.T.K.) , which is now known as Fraunhofer IKTS and located at Dresden. While the method of reverse osmosis certainly will be operated indefinitely only with polymer membranes, a patent-protected ceramic nanofiltration membrane – based on the results of Fraunhofer IKTS - has been available since 2004 and is being sold on an industrial scale by Inopor GmbH. 1.2. Geometry of ceramic membrane carriers As mentioned at the beginning, ceramic membranes consist of a carrier - mostly also ceramic - on which the actual membrane layer is fixed. In terms of the shape and design of these carriers, there are different approaches on the market, like e.g. flat-sheet membranes, tubular membranes, capillaries and some further special designs. Next to Inopor GmbH, typical supplier of ceramic membranes on the

European market are companies like Kerafol GmbH, Tami GmbH or Atech Innovations GmbH. Because tubular geometries dominate on the market, they will be discussed primarily in the further article. Tubular ceramic membranes are made of an extruded carrier (so-called “Support”) which has one or more channels on which the membrane layers are fixed on the cannel surface by some intermediate layers. Typically, the support is also made of a ceramic material, but there are also some technical alternatives available. Fig. 1 shows some typical single- and multi-channel geometries of ceramic membranes. Today a huge number of material and membrane combinations are available on the market, like for example SiC, Al2O3, ZrO2 and TiO2. Fig. 2 shows the design of a typical multi-channel membrane, including the front-side sealing. During operation, the membrane is installed in a housing and the feed flow / raw medium flows through the channels of the ceramic carrier. The surface of the channels are coated with a ceramic membrane layer. The filtration process is done by leading liquid through the membrane layer and separating the components out of the feed medium which are not able to pass the membrane layer. Liquids and components which can pass through the membrane layer are called “permeate”, while the remaining particles,

Fig. 3: Separation principle of a multi-channel membrane

which cannot pass through the membrane layer, are called “concentrate”. For a maximum efficiency, it has to be avoided, that feed liquid gets on the permeate side without passing through the membrane layer; otherwise, this would mean a contamination of the permeate flow. So, also a frontside sealing is necessary to avoid that feed medium flows directly through the porous membrane carrier to the permeate side. As shown in Fig. 3, the filtration process is regulated by a concentrate valve, which creates a ram pressure by throttling the diameter and creating a transmembrane pressure (TMP). Concerning the porosity of the materials, the membrane layers have much smaller pores than the intermediate layers and the membrane carriers, so that nearly all of the hydraulic resistance of the flow through the membrane element is generated by the membrane layer(s) but not by the support material. Mathematically, the transport mechanism through porous membranes can be described by pore models, depending on pore sizes, driving forces and transport conditions. Nanofiltration with polymeric membranes can be described pretty good by some solution-diffusion-models which are suitable for tight membranes. Ceramic membranes – even if they are in a range of a nanofiltration membrane – are still porous membranes, so the transport mechanism cannot be described by the laws for tight membranes, but by a modification of the Nernst-Planck-Equation (NPE), which combines the effects of transport through porous media with the electrical surface effects and potentials. 1.3. Materials of ceramic membranes As already mentioned, the development of ceramic membranes partly took place at the same time in different applications, so that also different developments in materials and shapes were made with similar properties. In general ceramic membranes are distinguished by the following features: Never-the-less, also ceramic membranes are only suitable within defined ranges of chemical and physical condi-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 15

Highlights 2013

tions. A fact, which becomes more and more important for decreasing pore sizes of the membranes, because decreasing pore sizes mean also decreasing resistance. Generally, micro- and ultrafiltration membrane layers are formed by filling the channels of the membrane carrier with a slurry, drain the channels and sinter the carriers to make the slurry film becoming a membrane layer. During this production process, the average pore size of the membrane layer is related with the sinter temperature: So, higher sinter temperatures mean coarser pore sizes. Typically, a ceramic filter element does not consist of only one membrane layer but of several layers, while the first layers are intermediate layers to fix the membrane layers on the carrier. Then, one membrane layer after the other is sintered on the existing layers with decreasing pore sizes. This means, that during this production process, the sinter temperature decreases during each coating process because pore sizes become smaller. This means also a decreasing of the maximum operating temperature, because the operating temperature must not be higher than the sinter temperature of the top membrane layer. While ceramic micro- and ultrafiltration layers are sintered by a standard slurry

Tab. 1: Properties of ceramic membranes

process, ceramic nanofiltration membranes are manufactured by a polymeric sol-gel process, which means even lower max. operating temperatures and a lower chemical resistance because of a significantly increased membrane surface because of a multitude of small pores. In the end, the thermal and chemical resistance of ceramic membranes is still higher than polymeric membranes, but decreasing pore sizes mean decreasing chemical and thermal resistance.

Ceramic membranes have in common, that they are inorganic membranes, but never the less, depending on the production process and application, the ceramic materials have different physical and chemical properties. Sometimes, it is said about ceramic membranes, that they are resistant against acids and basics through the whole pH-range; unfortunately, this is not the whole truth, because different ceramic materials have different chemical properties, depending for example on the

FLOTTWEG SE – YOUR COMPETENT PARTNER FOR SEPARATION TECHNOLOGY Flottweg Decanter and Disc Stack Centrifuges embody high quality, efficiency and durability. Use our experience in the separation technology for your success!

PLEASE VISIT US AT THE BRAU BEVIALE IN NUREMBERG

Flottweg SE • Industriestraße 6-8 • 84137 Vilsbiburg • Tel.: + 49 8741 301-0 • Fax: + 49 8741 301-300 • mail@flottweg.com • www.flottweg.com

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 16

Highlights 2013

Fig. 4: Overview of the material resistance properties /1/

Fig. 5: Structure of a ceramic membrane support with membrane layer

range. In case of a phase transformation, the risk of stress cracks is very high; also the risk of cracks is given when different materials with different transformation properties are coated on each other. This risk of cracks is given during operation at high temperatures as well as during the production process. Fig. 5 shows a crosssection through channel of a ceramic membrane element taken with a SEM, including the support, the intermediate layer and the membrane layer. Depending on the different thermal and chemical properties of these different materials, the mechanical junction between the different layers has to compensate different expansions and structure conversions. While Al2O3 and SiO2 appear here to be very stable, tetragonal zirconium oxide and titanium oxide, on account of their relatively low phase transition temperature, have only a limited thermal resistance. For the selection of the suitable ceramic material, it is hence not enough to know only the medium to be filtered, also the temperature and the pH value must be taken into account. Although the prevailing proportion in ceramic tubular membranes is covered with the materials aluminium oxide and titanium oxide, reference should be made at this point also to the very good chemical resistance of zirconium oxide; as an example here, reference should be made to the work of van Gestel /2/, who succeeded in making ceramic nanofiltration membranes on the basis of ZrO2. 2. Operation of ceramic membranes

Fig. 6: Schematic structure of a system with ceramic membranes (including CIP cycles)

type of atomic bonding. Last but not least, it must not be forgotten, that the ceramic membrane layer is only one part of the membrane element or the filtration stage: To make the whole plant run properly, all materials have to resist the media, incl. the front-side sealing, the housing and the orings. Generally, metal oxids are more resistant against acids if their chemical bonding is more acidic; an increasing resistance against acids also means a decreasing resistance against basics. Fig. 4 shows his properties for some typical ceramic materials. So, for example with SiO2, this is a significant behaviour, which reduces the

hydro-thermal resistance of SiO2 significantly, because of its high affinity for water. Next to the material itself, also the allotrophy / crystal structure of the material plays an important role: For example, for example, α-Al2O3 has a good resistance against acids and basics, while γAl2O3 with tetragonal crystal structure has a worse chemical resistance, especially against basics. Next to the chemical resistance, the lattice structure of the material also influences the thermal resistance of the membrane carrier and membrane layer. A maximum thermal resistance means no changing of pore sizes, atomic matrix and chemical reactions in defined temperature

2.1. Construction and incoming flow Ceramic membranes are primarily used in the cross flow operation; time by time, there are applications – mainly in the pharmaceutical and chemical processes – where ceramic membranes are operated in “Dead end” mode, but the flow rates are very low and this are not typical installations. As far as the fundamental construction of a membrane system with ceramic membranes is concerned, this differs only slightly from the structure of a plant with polymer membranes (see Fig. 6). However, ceramic membranes require higher cross-flow velocities and larger cross-flows, compared to polymeric membranes. The practice showed and shows that especially with applications in the field of wastewater treatment, as well as the filtration of chemicals and pharmaceuticals, a cross flow of 4 m/s for ceramic membranes is an acceptable minimum cross-flow. The processes of the separation of micro-organisms (for example from fermentation broths) in this case form an exception here, since, in addition, the cross

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 17

Highlights 2013

flow is to be selected before the background that the organisms are not destroyed by the appearing shear rates. Because the necessary cross flow for ceramic membranes is clearly higher than that of polymer membranes, the higher energy entry associated with this is an argument against ceramic membranes, but only in case of that the process can also be operated with polymeric membranes. That’s why manufacturers of ceramic membranes also continuously work on reducing this disadvantage, something that, in the end, means to find a geometry where the pressure loss, the necessary oncoming flow, membrane surface and not least the manufacturability are in an optimal relationship with each other. Fig. 7 shows the cross section of such a carrier geometry. 2.2. Geometries of ceramic membrane elements Ceramic membranes are available in various geometries and shapes with various numbers of channels and channel diameter. Next to tubular membranes, also flat-sheet membranes are available with a shape of a disc and diameters between 25mm and 90mm, while diameters of 25mm, 47mm and 76mm define an internal standard, because the usual test cells for flat-sheet membranes are manufactured for discs with these diameters. One side of the disc is coated with a ceramic membrane while the carrier is made out of the same materials like tubular ceramic membranes. The advantage of test cells for flat-sheet membranes is, that initial tests can be done with a small amount of liquid: The membrane disc is installed in a housing and a liquid column is filled above the membrane. Then, the liquid column is pressed in “Dead end” mode through the membrane by filling compressed air or gas between the column and the housing of the test cell. Typically, maximum transmembrane pressures up to 6 bar / 85 PSI can be

Fig. 7: Optimized geometry of a ceramic membrane /1/

F & S International Edition

No. 14/2014

Table 2: Parameter overview of membranes commonly available on the market

realized with standard test-cells. To homogenize the boundary layer on the membrane surface, a magnetic stirrer can be used. Filtration processes with flatsheet test cells are only suitable for small amounts of liquids or for initial trials; the operating pressures of flat-sheet cells are typically lower than operating pressures of processes with tubular membranes. Neverthe less, flat-sheet test cells are invaluable for trials because they are easy to do and can be done with a minimum amount of liquid. Next to test cells for flat-sheet membranes, depending on the application chemically high resistant materials are needed which means for example metalfree housings, made of PTFE or PVDF. Fig. 8 shows an overview of lab scale equipment, incl. metal-free housings for the filtration of high-corrosive liquids. Regarding tubular geometries, the single-channel tube (EKR) forms the smallest unit, while – depending on the applications – the number of channels can be more than hundred per ceramic element. Table 2 shows an overview about typical geometries available on the market and the technical requirements to design a membrane plant with this membranes.

Similar to polymeric membrane elements, ceramic membrane elements can be installed in series or in parallel with the same advantages and disadvantages like installing polymeric membranes. While pressure vessels for polymeric membranes can capture spiral-wound elements in series installation, housings for ceramic membranes cannot capture ceramic membranes in series; to install ceramic membranes in series, housings have to be installed after each other and be coupled by piping. On the other hand, housings for ceramic membranes are available to install ceramic membranes in parallel, while spiral-wound elements cannot be installed in parallel in one pressure vessel. By installing membranes in parallel, the required cross-flow and the size of the circulation pumps grows linearly to the number of elements. Fig. 9 shows a housing for up to seven tubular ceramic membranes, while the first membrane is already installed and sealed with an o-ring against the housing. Fig. 10 shows a general cut through a membrane housing for ceramic membranes. Each membrane has to be sealed against the housing by an o-ring and a stopper plate is necessary to avoid

Fig. 8: Flat-sheet test cell and further lab equipment for ceramic membranes /1/

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 18

Highlights 2013

Tab. 3: Typical permeate performance with ceramic membrane processes

that membranes are “shot” out of the housing in case of a pressure hammer. Typically, screwed flange connections are used to connect the housing with the piping, but in pharmaceutical or food and beverages applications, also tri-clamp connections are used. 2.3. Permeate volume and transmembrane pressure Due to its properties as a pore membrane, the permeate performance of ceramic membranes within a determined window obeys quasi-linear laws, which is why the specific permeate performance of a process with ceramic membranes is often described also in the dimension, which expresses the fact that a duplication of the permeate performance entails the duplication of the transmembrane pressure. Within certain windows this can be confirmed from practise; however, it must not be concluded therefrom that, for example, the transmembrane pressure can be multiplied arbitrarily without consequences, since, the laws of transportation through the membrane and the behaviour in forming a surface layer will become more and more important with increasing pressures and potentials. Although ceramic nanofiltration will be discussed more extensively in the following chapters, it must already be mentioned that nanofiltration is especially a process that yields no satisfactory permeate qualities in the case of low transmembrane pressure, even if some scientific works argued that ceramic nanofiltration only follows the laws of porous membranes. Because the optimum working point depends in the end on a huge number of factors, such as for example viscosity,

Fig. 9: Membrane housing for up to seven ceramic membrane elements with one element installed /1/

temperature, solids cargo, particle size etc., a blanket statement about the optimum operating point to be selected is not possible. The values listed in Table 3 show some data typical from practice. In this case, it must be mentioned that, for example, an increase of the temperature of the inflow medium leads to an increase of the permeate performance with a concurrent drop of the permeate quality, while a reduction of the media temperature leads to the fact that the transmembrane pressure must be raised in order to keep the permeate flow constant. 3. Ceramic nanofiltration membranes 3.1 Membrane construction and characteristic variables With development and market launch of ceramic nanofiltration membranes with a separation limit below 1 kDa, a totally new product branch of ceramic membranes hit the market which differs in several aspects from ceramic micro and ultrafiltration membranes. Ceramic nanofiltration membranes - like ceramic micro and ultrafiltration membranes - are built up in layers, i.e. on a very porous carrier (called “Support”), separation-active membrane layers are applied in several process steps and these layers have a finer separation limit with every process step, so that at the end of the manufacturing process, the finest membrane layer - which also defines the separation limit of the ceramic membrane tube - is applied and this is also the layer that during operation receives the oncoming flow from the feed medium. An example construction is shown in Fig. 5.

It is obvious that maximum membrane permeability is to be aimed without a loss of selectivity and without using a manufacturing method that would make commercialisation impossible for cost reasons. Concerning the membrane permeability L, the law from Hagen-Poiseuille for a porous structure is:

In detail, there are the porosity of the membrane, the pore diameter, the dynamic viscosity of the medium and the membrane thickness. Essential problems already become obvious here, which one is confronted with due to increasingly smaller pore sizes. Because the pore size squarely enters in the permeability of the membrane, an increasing reduction of the pores means a disproportionate loss of permeability that one must counteract: remaining here as possible manipulated variables are the porosity and the thickness of the membrane layer, i.e. it is always aimed for that a membrane layer is very thin and has very high porosity. Unfortunately, increasing porosity goes hand in hand with a loss of chemical resistance, since increasing porosity means a surface enlargement that offers a bigger attack area for the respective chemicals. This circumstance, in turn, can be counteracted only by having starting materials as pure as possible and the cleanest production conditions. Besides the application of additional membrane layers, these requirements for cleanliness in the manufacturing process and the purities of the starting materials constitute the essential influence on the cost structure of ceramic nanofiltration membranes. In practice the law of Hagen-Poiseuille also means, for example, that the membrane layer should be very thin, but still mechanically stable enough to resist the forces occurring - for example, through incident flow of solids loaded inflow media. At the same time it can also be recognized from the law of Hagen-

Fig. 10: Membrane housing for ceramic membrane elements /1/

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 19

Highlights 2013

Tab. 4: Isoelectric points of different oxide ceramics

Fig. 11: Extrusion of ceramic membrane carriers /1/

Poiseuille that the hydraulic resistance of the ceramic support plays a minor part, since its pore size - as a function of the material and production process - is usually in the range of 3 - 5 μm, while the pore sizes of ceramic microfiltration membranes can lie approx. in the ranges 100 800 nm and the pore sizes of ceramic nanofiltration membranes can be less than 1 nm. 3.2 Manufacture of ceramic nanofiltration membranes As shown in Fig. 11, the supports of ceramic nanofiltration membranes are extruded - just like the supports of ceramic micro and ultrafiltration membranes - in order to implement tubular geometry. In this case, different oxide ceramics are used as a material, which in each case, on account of their physical and chemical properties, are of advantage or disadvantage for corresponding applications. Typical materials are, for example, aluminium oxide (Al2O3), titanium oxide (TiO2) and zirconium oxide (ZrO2), and here, aluminium oxide is used nearly exclusively as α-Al2O3, since γ-Al2O3 is not stable at the temperatures at which the respective membrane layers are sintered. The influencing of the pore size of ceramic supports is a subject of a huge number of publications and, hence, should not be discussed at this point in depth. Typically the pore size of ceramic supports can be influenced by admixing of excipients that incinerate during the combustion process and then leave behind respective pores in the matrix. A typical excipients for this is starch and also here, as a function of the starch, the pore size varies: While with potato starch pore sizes of up to 50 μm can be achieved, the pore sizes when using wheat starch lie in the range of 20 μm and with rice starch in the range of 15 μm. In addition, a crucial role is also played by the combustion temperature and burning period: While burning for too short a time or burning at too low a temperature entail an inadequate formation of the ceramic structure and therefore a poor mechanical

F & S International Edition

No. 14/2014

stability, too much time or overheating causes the loss of porosity and a decrease of the number of pores. While ceramic micro and ultrafiltration membranes are usually applied and sintered in layers with a slurry method, pore sizes for nanofiltration in a range of a pore diameter of 1 nm cannot be realized with these processes anymore. Here, a Sol gel process is used, in which either a coarseporous carrier, or a carrier that has been initially pre-coated with an ultrafiltration membrane, is coated with liquid Sol. This method - in connection with polymeric Sols - was particularly promoted by Puhlfürß et al. and optimised from the point of view that it can be used beyond just on laboratory scales /3/. The coating of the carrier can be done in different ways, for example, by means of SprayCoating (spraying on) or Dip-Coating (immersion), wherein Spray-Coating is suitable primarily for flat structures and, hence, is not applicable for using with a tubular geometry. The liquid gel is transferred into a solid gel when the gel point is exceeded. In general with this Sol-gel process there are two different methods that differ in the kind of the Sol. The prevailing part of the ceramic membranes available on the market that are produced by means of the Sol gel process are produced by means of a colloidal sol in which a metal alkoxide is hydrolyzed in an environment with a surplus of water; in this case nanodisperse particles are then formed, which can be peptized by treatment with respective electrolytes. These aqueous suspensions - which either consist of colloidal or aqueous sols - are admixed with corresponding substances which implement mechanical bonding on the surface and then are applied directly to a coarse-porous support. The adjustment of the pore size occurs during hydrolysis by varying the temperature and concentration accordingly. Because with a colloidal sol no pores can be mapped that are smaller than 2 nm, these pore sizes are implemented by

means of a Sol gel method with a polymeric sol. In this case, pore sizes can be mapped down to 0.9 nm. The method is distinguished by the fact that they are working with an alcoholic solution where a partial hydrolysis of the metal alkoxides is triggered by addition of water and the metal alkoxides then polycondense in the solution. If the dilution is corresponding, the forming oligomers will remain in solution and are allowed to gel only when applied to the membrane support, since here, a suction effect results due to capillary forces. With this method, ceramic nanofiltration membranes can be produced from titanium dioxide, whose separation limit in aqueous media is 450 Dalton. Besides ceramic nanofiltration membranes from TiO2 /2/ ceramic nanofiltration membranes with a separation limit of approx. 300 Dalton were implemented by van Gestel on the basis of ZrO2 / TiO2; for this purpose, the tetragonal zirconium oxide was stabilised with yttrium. Agoudil et al. /4/ have implemented ceramic nanofiltration membranes from a mixture of ZrO2 and SiO2 by using tetraethoxysilanes (Si(C2H5O)4) and zirconium tetrapropoxide (Zr(C3H7O)4); the transformation temperature of zirconia from the tetragonal to the monoclinic phase was able to be shifted to higher temperatures by addition of SiO2. In general, different materials and production approaches are currently to be found on the market for ceramic nanofiltration membranes, which result in corresponding advantages and disadvantages; for instance, the system SiO2 permits a good monitoring of the pore size in the manufacturing process, but at neutral and alkaline pH values has detrimental properties in the stability. However, the system γ aluminium oxide is at low pH values (pH <2) only conditionally stable while the systems ZrO2 and TiO2 have a good to acceptable stability across the whole pH value range. 3.3 Surface load of ceramic membranes Besides the pore size, the surface load plays an important role with ceramic membranes. In general the surface load can form itself with ceramic membranes on account of different mechanisms:

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 20

Highlights 2013

3.4 Material transport through ceramic nanofiltration membranes

Fig. 12: Retention of a ceramic NF membrane as a function of pressure and pH value

Fig. 13: Retention of an improved NF membrane of Inopor GmbH

- Dissolution / absorption of ions into / from the surrounding liquid - Dissociation of molecule groups on the surface - Adsorption of ions / electrolytes on the surface Because ceramic membranes typically have amphoteric behaviour, they can absorb and/or emit positive or negative ions / molecule groups as a function of the pH value. While ceramic membranes with low pH values have positive electric potential, it is negative with high pH values. This electric potential is called “Zeta potential” and can be measured by means of suitable measuring cells and electrodes. The point at which the electric potential is equal to zero is called isoelectric point. Typical values of oxide ceramics for this are listed in Table 4. As a function of the pH value of the medium to be filtered and the isoelectric point of the membrane material used, the retention of the membrane is also influenced and in general it has to be said that the retention near the isoelectric points is worse because there, the repulsion or attraction effects, which otherwise can influence the separation process positively, cancel each other out. Fig. 12 shows how the retention of a ceramic nanofiltration membrane with a pore size of 0.9 nm from TiO2 changes as a function of pressure and pH value.

An essential difference between ceramic and polymeric nanofiltration membranes is that even the ceramic nanofiltration membrane – like the micro and ultrafiltration membranes - is a pore membrane while polymer membranes are dense membranes. Therefore, the solution-diffusion model that is used in the case of dense diffusion membranes, cannot be modified for ceramic nanofiltration membranes. According to the solution-diffusion model, inter alia, it arises that with rising water flow rate and/or pressure, the retention increases and theoretically approaches a value of 100%. This is because the back diffusional proportion, with rising water flow rate, has increasingly less influence on the deterioration of the permeate; consequently it is not advisable to operate polymeric systems with wound modules with insufficient pressure, since the permeate quality noticeably decreases. The higher the proportion of the ingredients to be separated, the more clearly this effect is detected. However, with ceramic nanofiltration membranes the substance transport can be described sufficiently well for practice with the expanded Nernst-Planck equation according to Dresner. The expansion of the Nernst Planck equation according to Dresner consists of the fact that, beside the solutes and the solvent, he also included the membrane and its material in the thermodynamic consideration. To this end, Dresner introduced an Interaction parameter β /5/. The surface-related molar transport of a component j according to Dresner consists of the following terms:

Convective Transport: Diffusive Transport:

Electromigrative Transport:

Therefore the following equation arises for the surface-related molar transport of a component: . In this case, cj,M is the concentration of the component j in the membrane, vW is the velocity of the water, D(j,M) is the diffusion coefficient of the component j in the membrane, zj is the ion valency of the component j, F is the Faraday constant and is the displacement-dependent change in membrane potential. The introintroduction of the interaction parameter β changes the convective term to the effect that interactions with the solid ions in the matrix of the membrane may be considered, similar to the interaction of

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 21

Highlights 2013

solid ions in the matrix of an ion exchanger membrane. In particular, it is considered with this parameter that the convective ion transport is not necessarily carried out with the speed of the convective transport of water. Dresner looks at the parameter β, as well as the diffusion coefficient of the water, as being independent of the concentration. Tests with higher pressures and concentrations have shown that the independence from the concentration is then not totally present anymore, but for low and middle concentrations, the equation according to Dresner has still proved to be very accurate. In particular, it applies also for ceramic nanofiltration membranes that a too low pressure and therefore a too low water flow rate have a negative effect on the permeate quality. 3.5 The future of ceramic nanofiltration membranes As already described, ceramic nanofiltration membranes with a cut-off of 450 Dalton can be produced in large scale with process reliability; in laboratory and prototype scales, membranes with significant lower cut-off rates were already able to be implemented. An essential challenge will lie in the near future in also mapping these membranes with a separation limit of 300 Dalton or less with process reliability and in pipe geometry. Fig. 13 shows an example of the retention characteristics of a membrane prototype of the company Inopor GmbH, in which the membrane potential was changed specifically by trying to map electro-chemical mechanisms from ion exchangers with modified ceramic membranes. In this case, a separation limit of smaller than 300 Dalton was able to be implemented with constant porosity and constant flow rate, but currently, the long time stability of this membrane is not optimized yet. Alternatively to the optimisation of ceramic membranes, the composite of ceramic and polymeric membranes will certainly gain importance in the near future, since, here, the advantages of both membrane models can be combined in one product. In the ideal case a ceramic nanofiltration membrane would be conceivable on which there is a solidly bonded reverse osmosis membrane. In spite of intensive researches in this area - in particular in the USA - only prototypes on a laboratory scale are currently available. Beside the optimisation and improvement of the ceramic membranes as a product looked at in isolation, the best possible placement and combination of appropriate products in increasingly complex process and production chains will also increasingly be gaining in importance. This assumes, for the design and implementation of a process, that knowledge about the most different methods and products exists, in order to be able to select from among them the optimum method with the best possible suitable membranes. As an example for this, it should be mentioned that Inopor GmbH, with acetic acid produced by fermentation, could combine several methods in such a manner that the acetic acid was initially purified and then, by means of a special membrane process, was able to be concentrated to a concentration of 30-32%. Literature: /1/ Inopor GmbH, Veilsdorf /2/ T. van Gestel, H. Kruidhof, D. H.A. Blank, H. J.M. Bouwmeester: ZrO2 and TiO2 membranes for nanofiltration and pervaporation: Part 1. Preparation and characterization of a corrosion-resistant ZrO2 nanofiltration membrane with a MWCO < 300. Journal of Membrane Science 284 (2006) , No. 1–2, Pages 128-136 /3/ P. Puhlfürß, A. Voigt, R. Weber, M. Morbé: Microporous TiO2 membranes with a cut-off < 500 Da. Journal of Membrane Science 174 (2000), No. 1, Pages 123 – 133 /4/ N. Agoudjil, N. Benmouhoub, A. Larbot: Synthesis and characterization of inorganic membranes and applications. Desalination 184 (2005), No. 1-2, Pages 65 - 69 /5/ L, Dresner: Some remarks on the integration of the extended Nernst-Planck-Equations in the hyperfiltration of multicomponent solutions. Desalination 10 (1972), No. 1, Pages 27 - 46

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 22

Highlights 2013

Separation technology in winemaking H. Lyko* Wine is an ancient foodstuff and fundamental knowledge of cultivation and winery technology has a correspondingly long tradition. Nevertheless, according to the opinion of experts, never before in history was there as much good wine as today, the organisers of INTERVITIS/INTERFRUCTA 2013 stated. On the way from the grape or the apple tree to the bottle, musts and wines go through a variety of treatment steps, among them a large proportion of mechanical separation processes. Besides classical criteria like hygienic design, handling, energy efficiency, ease of maintenance and throughput/ yield, in the foreground for all processes there are always also the gentle treatment of the foodstuff and the extensive preservation of all important flavours. At the Oenology congress accompanying the trade fair, the theme of Alcohol management was the focus. Climatic changes and certain cultivation and vinification techniques sometimes led to excessive alcohol content in wine that can adversely affect the taste and regional typicity. The possibilities for adjusting alcohol content include viticultural measures like cellar management measures. The latter are also summarised in the following contribution. Grape processing The first step after the harvest is the removal of the individual grapes from the grape bunches. This happens with the help of destemmers in which the harvest is put into rotation by paddles rotating on a shaft (fingers) and is moved against the wall of a perforated basket that turns around in the same direction as the shaft. The grapes exit outwards through the holes. The length of the destemmer zone, the basket (hole diameter), the rpm and the number of rotating fingers are varied according to the size of the grapes and their sensitivity (variety and state of ripeness). For a new destemming process without rackets, shaft and rotary basket, the Swiss enterprise Bucher Vaslin received the 2013 innovation award of INTERVITIS/INTERFRUCTA in gold. The Delta Osxillys system separates the grapes from the combs in one or two inclined baskets that perform a swinging pendulum motion. The force acting on the grapes in the process is greater than the stability of the connection between grape and stalk. All components are separated from each other below the basket on a roller sifter. The quality of the destemming result by this system should be about 5 to 10 times better than with conventional destemmers, i.e. the proportion of remaining green parts among the grapes is about 0.2 to 0.5%, in contrast to 2 to 4.5% conventionally. The Selectiv‘ Process Winery system of the French manufacturer Pellenc has already been on the market a little longer, but it has now also received the innovation award in gold. This compact machine is also responsible for destemming and sorting. Also here, the destemming happens by high-frequency oscillations wherein the grapes are pushed on a lattice table between oscillating destemming modules *Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

(see Figs. 1 and 2). The harvest then falls onto the sorting table which consists of two parts. Fully toothed feed rollers separate the juice and small waste (seeds, stunted grapes, small insects...), align the stalks and prepare the harvest for sorting. The screen rollers with their studs let the sorted grapes fall through and they are collected in a trough beneath the sorting table. The screen size is adjustable to the grape size. About 95% of the leaf stalks > 35 mm are discarded and the achievable cleanliness of the grapes is stated as 99.8%. Separation processes for clarification and sterilisation For the clarification of wine and fruit juice, the user basically has the choice

between decanters, separators, depth filters and membrane filtration systems, and all systems were on exhibit with a greater selection. The GEA Westfalia Group exhibited the new GSC 110 ecoplus clarification separator for the middle and upper range of performance. The hydry machines of the ecoplus generation, with hydrostop, have a patented drum discharge system that should ensure maximum dry matter of the discharged solids and thus correspondingly low product losses for the clarification. The machines are made in a hygienic design and are provided with the possibility of automatic CIP cleaning. With the specially designed hydrohermetic inlet and outlet system, gentle and foam-free treatment of the product is achieved. Eaton demonstrated the new

Fig. 1: Lattice table and destemmer modules of the Selectiv‘ Process Winery system, from Pellenc (image: Société Pellenc SA, France)

Fig. 2: Functionality of the Selectiv‘ Process Winery destemmer and sorting machine (image: Société Pellenc SA, France)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 23

Highlights 2013

Fig. 3: New Separator GSC 110 of the ecoplus series, from GEA Westfalia Separator Group

Fig. 4: BECO® LiquiControl2 index measuring instrument to determine wine filterability, from Eaton

depth filter candle BECO PROTECT CS CellStream, which combines the advantages of the innovative BECOPAD depth filter sheet material in one filter candle. With different features, cost savings and a saving of time should be achieved, as well as an increase in productiveness. Thus, the nature of the winding causes a large filter area, the compact and backwashable design permits a higher number of steam and regeneration cycles than with prior models, as a result of which the service life is increased. The new depth filter candle can be inserted with a standardised adaptor in customary candle housings. The new BECO LiquiControl2 index measuring instrument from Eaton is used for the determination of wine filterability. In detail, a sample is filtered under defined, reproducible conditions (constant pressure) through a comparison membrane. Alternatively to the membrane, the suitability of depth filter sheets can also be determined by connecting a laboratory monolayer filter. The sample is considered to be filterable if a defined quantity of filtrate is obtained during the applied filtration pressure. Based on these test results, the winemaker can decide whether further measures to improve the filterability are necessary and coordinate the downstream filtration process perfectly. The BecoPad mentioned above is the first mineral-free depth filter medium, which consists of specially treated high-purity cellulose. From this, according to the manufacturer, an especially low content of ions and other flushable materials result. 2BFermControl GmbH, Breisach, refers, inter alia, to the necessity of minerals in its depth filter sheets with the trade name ViniFilter, made with brown diatomite and available in 5 separation efficiencies. Therefore, mineral-free sheets allow valuable ingredients to pass through almost without loss, have a problematic pressure difference profile and tend to strike through towards the end of filtration. The question of which depth filter sheet is better or worse can only be answered by a direct comparison. Such a comparison was made by employees of the Bavarian Landesanstalt für Weinbau und Gartenbau (LWG), Veitshöchheim, with white wines and red wines and published in February, 2013 in das deutsche Weinmagazin /1/. The ViniFilter depth filter sheets were not involved in this investigation, but the ones that were, besides the two qualities Beco (contains mineral filtration aids) and BecoPad, were the products of the competitors PallSeitzSchenk and Sartorius/ Filtrox, as well as sheets of the Czech enterprise HobraSˇkolník (HOBRAFILT), which were also exhibited at INTERVITIS. For the classification, the water values stated by the manufacturers were consulted for which sheet was to be assigned

respectively to which filtration sharpness. Then the filtration tests were carried out with two filter sheets in each case with graded clarification sharpness. As a conclusion of the investigations, the right choice of filter sheets or sheet combinations, as well as the right gradation of filter stages (prefiltration before sterile filtration) was considered as decisive; qualitative differences between the products of the different manufacturers have not been found. The pure cellulose sheets showed advantages concerning filtration per-

F & S International Edition

No. 14/2014

Ceramic nanofiltration from Inopor ® Inopor® is the only supplier of ceramic membranes with a cut-off of 450 Dalton. Ceramic nanoﬁltration membranes from Inopor® are known all over the world for their reliability in chemical, pharmaceutical and other high-performance applications.

Ceramic filtration and more …

Industriestrasse 1 98669 Veilsdorf www.inopor.com contact@inopor.com

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 24

Highlights 2013

Fig. 5: Combination of reverse osmosis and perstraction, Memstar method

formance after backwashing, whereby, however, the degree of clarification slightly worsened, and moreover they could score points through lower drip losses. The alternative to clarity filtration with sheet filters is crossflow membrane filtration. Its increasing use in wine cellars can be seen, for example, in the report of Pall Corp. that in May 2013 sold its 1,000th unit of the Oenoflow system, which is equipped with hollow fibre modules. With the filtration of Tank sparkling wine after tartrate stabilization, however, new ground has been broken in recent times /2/. Tartrate stabilization is particularly important in sparkling wine because the crystals otherwise cause uncontrolled foaming when you open the bottle. The filtration of the abrasively acting crystals succeeds with symmetrical PVDF membranes, while composite polymer membranes and ceramic membranes are unsuitable because the relatively thin, separationactive layer can be ablated by abrasive particles. The Italian supplier of wineries, Velo with headquarters in Treviso, is promoting the third generation of its crossflow filtration systems with a novel modular design. The polysulfone membranes, which are the centrepiece, are unchanged from the previous design, but the systems engineering should now allow maximum flexibility. All system components, including touch screen control unit, reservoir containers, pumps, containers and dosing pumps for cleaning solutions etc. are installed together on the same frame, and indeed equally for all system sizes from 70 to 350 m2 membrane area. Here, a group of 7 modules with a total of 70 m2 and two circulation pumps form the basic structure, which, for higher capacities, is supplemented by appropriate filtration units of this structure. The remote control of the whole system can optionally be carried out via a modem or the internet.

Furthermore, membranes and membrane systems are described in the following chapter for turbidity filtration, but which can be used in principle also for clarification. Filtration of sediments and fining trub Crossflow filtration systems, with which, besides wine already preclarified by sedimentation and/or flotation, also must and wine deposits with accordingly high solids contents are filtered, have only been on the market for a few years. They have the advantage that the filtration of wines and juices, and also wine extraction from the sediments, are possible with the same system. Some manufacturers also point to the lower oxygen uptake in a membrane system in contrast to a rotary drum filter. In a study that was carried out at Changins College, Switzerland /3/, the influence of such a filtration on the quality and composition of the end product was examined. Used here was an FX3 filtration system from Bucher Vaslin, France, with capillary modules adapted for this. These were capillaries made of polysulfone with 3 mm inner diameter and a nominal pore size of 0.2 μm. The system was equipped with supply lines of hot and/or cold water, as well as compressed air, and was operated with adapted software for filtration backwashing and cleaning. Residues of sedimentation were treated as a flotation of the same white wine variety of the same grower, and, for comparison, filtration was also carried out with a rotary vacuum filter with perlites. Although the residues of the sedimentation and the flotation differed concerning density and viscosity, both could be clarified to comparable, very low dry matter contents, but with significant temperature rise. With respect to the filtration process, the finished wines were not significantly different in their alcohol and acid contents, and only with the polyphe-

nols were there lower values for filtration under vacuum, when the residue of the flotation was considered. The membranes of the Oenoflow HS system by Pall consist of the same membrane material and also have the same degree of separation as the membranes for wine clarification, but the hollow fibres have a larger inner diameter and thus can process liquids with higher solids content. The wine yields should be increased by 2% with the application of this system on sedimentation trub. Besides separators and decanters, the GEA Westfalia Separator Group also presented its enlarged programme in ceramic membranes under the tradename Membraflow. Here, the configuration of the ceramic filter elements, modules and also accessories is carried out individually for the respective application. With ceramic membranes, high dry solids substances can be achieved with consistently high flux rates of the filtrate during the entire life cycle. A typical example of use is for beer recovery from surplus yeast. Thus, ceramic membranes are obviously also suited for the filtration of sedimentation trub during wine and juice production. As an essential advantage of these membranes compared with polymer membranes, the high service life is pointed out and with it the low life cycle costs. The bandwidth of the available Membraflow membrane elements consists of duct diameters of 1.3 mm to 8 mm and pore widths of 5 nm (0.005 μm) to 1.4 μm. Here, the number of ducts ranges from one up to more than 277 ducts. Also the length of the membrane elements can be produced flexibly. Furthermore the GEA Westfalia Separator Group also supplies module casings, back flushing devices and seal accessories. The Dynamos system of TMCI Padovan, Italy, contains two graded, partially interlocking stacks of rotating filter discs. It was patented in 2012 /4/ and advertised at the trade fair both for vinification and for filtration of apple juice as an alternative to vacuum and pressure filtration. An essential aspect which distinguishes this system from other dynamic membrane systems is the type of membrane lavage to remove outer layers. Besides an automatically controlled backwashing with permeate, during which the rotation of the membrane stacks is not interrupted, nozzles distributed in the housing provide for tangential flushing of the membrane surfaces. At the Oenology congress, Massimo Pivetta presented results from the comparison of the clarification of sedimentation trub from Prosecco and Merlot with a solid content of 30 Vol.-% with a vacuum filter, a tubular membrane filter and the DYNAMOS system. In all three cases, the same volume

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 25

Highlights 2013

was treated. Dynamos had the lowest energy consumption among the three process variants, a low temperature increase in comparison to the filtration in the tubular module and the least consumption of flushing water. However, the consumption of cleansing chemicals was a little higher than with both other variations, but to make up for it, Dynamos gets by without filtration aids, just like the crossflow system with tubular membranes. In light of the test results, the determined necessary separation area for 1000 l/h wine lees with 30 Vol.-% of solid content was 40 m2 for dynamic filtration, 70 m2 for the crossflow filter and 10 m2 for the vacuum filter. Technical measures for alcohol management in the wine cellar How to prevent excessive alcohol levels in white and red wines is the object of a collaborative project funded by Bundesministerium für Ernährung, Landwirtschaft und Verbraucherschutz (BMELV), which is coordinated by Prof. Monika Christmann and Volker Schaefer of the Institute of Oenology of the University of Geisenheim. In this project viticulture, microbiological and cellar measures are studied. Also the International Organisation of Vines and Wines (OIV)supports research projects on this subject that were presented within the scope of the Oenology congress. As Volker Schäfer (University of Geisenheim) explained, the currently applied cellar-technical methods for alcohol withdrawal from wine consist of distillative processes such as the Spinning Cone Column or vacuum distillation. Possible as membrane processes are osmotic distillation in the membrane contactor, or reverse osmosis with subsequent vacuum distillation or osmotic distillation of the permeate. With reverse osmosis alone, however, part of the alcohol is also taken away, but due to reconcentration, the retentate does not contain less alcohol than before in percentage terms. Therefore, the permeate alcohol is withdrawn by vacuum distillation or osmotic distillation and afterwards is mixed again with the retentate of the reverse osmosis. More rarely, and not allowed in many countries, is mixing the RO retentate with water. The two-stage process as a combination of reverse osmosis and osmotic distillation was developed by the Australian enterprise Memstar and was patented. The representation in Fig. 5 is simplified in this respect as it is not shown that the RO permeate is further degassed and heated prior to entry into the membrane contactor. Instead of the concept “osmotic distillation”, the contactor process is also called Perstraction, and it physically concerns the extraction of vaporous alcohol through a hydrophobic membrane into the extractant water. The process is carried out intermittently and the RO retentate and the dealcoholised RO permeate is circulated until the desired alcohol reduction is achieved. Besides the technical methods for the treatment of the finished wine, there exist newer methods of ultrafiltration and nanofiltration through which the sugar content is reduced in the must before fermentation. These methods generally have the disadvantage that they can only be applied very quickly, immediately after harvest and must extraction. According to the resolution of the OIV /5/, the reduction of the alcohol content of a wine is allowed by at most 20%. The end product must have a total alcohol content greater than 8.5 Vol.-% (according to the definition of wine). The removal of alcohol from the wine cannot be carried out in combination with a change in the sugar content in the must. However, in addition, the possibility also exists to produce “Grapevine products with reduced or low alcohol content” /6/ (this is by definition no longer wine). In the research projects discussed at the Oenology congress, the different ways of influencing the alcohol content were compared and their influence on the ingredients and sensory properties of the products was discussed. There is not enough space here to illustrate in detail the results that have been obtained with different methods, in different wine varieties and in different European wine regions. Therefore, some generalised information is summarised below.

F & S International Edition

No. 14/2014

The influence of dealcoholisation by up to 2 Vol.-% on the material composition of the treated wines depends on the dealcoholisation process and on the wine variety. The separation criterion in thermal separation processes is such as the volatility of the respective components in osmotic distillation, and the molecular weight in reverse osmosis. For that reason alone, the application of the RO with subsequent dilution of the retentate is an unfavourable solution, because thereby important flavouring materials also get lost with the permeate. But also the other methods of alcohol extraction are associated with low losses of highly volatile aroma compounds, whereby this has not been felt in all cases in sensory testing. If so, then alcohol reduction seems to emphasize the acidity and some bitter compounds, while fruity notes (because of being highly volatile) get somewhat lost. Worth mentioning in this context is the so-called “sweet spot”, which, as far as it is known, could be adjusted with the described methods. “Sweet spot” is the alcohol content in which the flavour intensity is particularly high and the wine is therefore judged to be qualitatively outstanding. Literature: /1/ Burkert, J.; Köhler, H.J.; Hartmann, M.: In die Tiefe gehen – Filtration mit Filterschichten, das deutsche weinmagazin 4/16, February 2013, /2/ Kahl, S.; Stoll, N.: Innovative Sektklärung – Filtration von Tanksekt nach der Weinsteinstabilisierung; Getränkeindustrie 4/2013, pp. 34 - 37 /3/ Schonenberger, P.; De Giorgi, D., Ducruet, J.: Filtration of grape juice sediments: a new application for cross-flow filters; Wine & Viticulture Journal March/April 2012, 28 – 32 /4/ Bornia, L.: Apparatus and Method for Filtering Liquids, Particularly Organic Liquids; EP 2 326 409 B1, 2012 /5/ OIV 2012: Resolution OIV-OENO 394B-2012: Correction of the alcohol content in wines, see www.oiv.int /6/ OIV 2012: Resolution OIV-OENO 294A-2012: Dealcoholisation of wines, see www.oiv.int

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 26

Highlights 2013

The paper production and pulp industry on its way to sustainable production Report from the PTS Environmental Symposium H. Lyko* The production of paper and pulp is among the most energy-intensive industrial sectors of production and also requires large amounts of water. Energy and water consumption are linked inasmuch as that the paper raw material, cellulose fibres, gets extremely diluted in the course of production and a large part of this water is expelled again using large amounts of steam. In addition a variety of additives is used in the production process, which can also be found in the wastewater. The optimisation of energy and water consumption and the recovery of recyclable materials from residues and waste streams in the paper industry are major research topics of the Papiertechnische Stiftung (PTS) in Munich. The environmental symposium, organized by it, is an important meeting place for the industry and its suppliers relating to water, energy, waste and environmental legislation, where the new developments in these fields are discussed. Subsequently, the contents of the event in November 2013 are summarized, which was run by Dr. Hans-Jürgen Öller and Holger Jung.

Environmental legislation and relevance With the use of wood as raw material in the paper industry, this industry is also among the forest destroyers. As Oliver Salge, head of the forest and marine campaign at Greenpeace e.V., explained, about 18% of annual greenhouse gas emissions can be attributed to the forest destruction. Other effects are a dramatic loss of species and the destruction of the livelihood of the directly affected local inhabitants. The focus of deforestation has shifted in recent years from Scandinavia to Canada and Asia. Salge revealed that forest conservation does not mean the end, but a change of forest management. While, in Canada, one was still unable to achieve a relenting response in discussions with Resolute Forest Products, the largest paper and pulp producer in the country, APP (Asian Pulp & Paper), referred to previously as worst paper producers in the world (see /1/), has announced its Forest Conservation Policy earlier this year, according to which no new forest areas to be cleared anymore. Forest conservation from the point of view of European paper production means, according to the elaborations of Salge, primarily the most extensive use of recycled materials and – where this is not possible the purchase of fibres from the so-called FSC production ( certificated according to the criteria of the Forest Stewardship Council® for a sustainable forest management).

*Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

The Confederation of European Paper Industry CEPI has set itself the ambitious goal, with its roadmap for the reduction of the CO2 emissions (inspired by the corresponding „Energy Roadmap“ of the European Commission), to reduce CO2emissions until 2050 by 80 % of the value of 1990 while increasing the yield of recyclables by 50 %. Dr. Johannes Kappen, PTS, described the current state (cf. Fig. 1, /2,3,4/) from which it arises which reductions of the environmental impact were already possible in spite of an increase in production. He pointed out that, beside the possibilities already discussed up to now of using regenerative energy sources and above all, to “decarbonise” electricity generation as well as to operate all pro-

duction parts according to the BVTDirectives, so-called „breakthrough technologies“ are required, that revolutionise the paper production industry, from substance treatment up to the product delivery. To develop ideas for this, two independent teams with experts representing producers, suppliers, research institutes and associations were formed, including the speaker, which were supposed to find the most appropriate strategies for this. Unfortunately, the results of this teamwork, a total of eight different concepts , were subject to confidentiality at that point of time. Afterwards, the usage rights of these concepts will be assigned by the CEPI to the European paper industry in the form of licences.

Fig. 1: Development of the production volume, the energy and water consumption and emissions in the European paper and pulp industry since 1990 (Source: CEPI Annual Statistics 2012 /2/)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 27

Highlights 2013

In the area of environmental legislation, the paper industry must obey, primarily, the Directive about industrial emissions (IED) and the water framework directive (WRRL). Norbert Suritsch und Walter Grotz, MĂźller-BBM GmbH, Planegg, described which effects arise from its national implementation. All paper manufacturers from a production volume of 20 t/d onwards are affected by the IED. Compared with the predecessor Directives, increased demands are being made in this respect on operators since stricter application of the best available technologies (BVT) is required. In future, within 4 years after publication of new BVT conclusions the implementation by means of approval conditions, is required also for existing plants. The latest document on the BVT for the paper and pulp industry was published as a draft in July, 2013 /5/. Further requirements are being made on the monitoring practise and these also depend on the federal state concerned. For the new establishment or essential modification of an IED plant, a report on the source state of soil and ground water is part of the approval documents. For the compliance with the water framework directive, above all two principles are to be followed, namely the deterioration ban as well as the improvement order concerning discharge into surface waters. While the deterioration ban is a strict law and, accordingly, the effects of withdrawal and discharge in view of contaminants and temperature must be also monitored (see further below: Monitoring of the introduction quality and risk analysis), the improvement order has a goal character. Energy management and power supply Paper manufacturers, as energy-intensive enterprises, can apply for a compensating regulation according to the renewable energy law (EEG) or for a peak equalization scheme according to the Electricity Tax Act. For this, the introduction of an energy management system according to ISO 50001 is mandatory . The details of the legal framework conditions and time limits, the approach for the introduction as well as the advantages of such a system were described by Maria Cordes-Tolle, M Consult GmbH, and Dieter Pfaff, dp consulting. It was pointed out among other things that, in spite of all present endeavours of the enterprises to reduce the specific energy consumption, unused potentials still exist in many places that could be grasped and be exhausted by a management system. Here, a great importance is being attached to the integration of all employees of an enterprise. Instead of the appointing an individual person responsible, so-called energy teams are formed that meet regularly, take part in internal and external audits, are informed about all data relevant for energy and thus can act as multipliers for the remaining employees. Concerning the data collection, the trend is towards the continuous recording of all energy data within the production facilities. The influences of the energy transition of the energy supply of the paper industry were the subject of the presentation by Dr. Bernd GĂśtz of the Verband Deutscher Papierfabriken e.V., Bonn. He pointed out the significance of the supply security for the industrial location of Germany and with it also for the paper industry. Additionally he made reference to the contribution that some energy-intensive industrial branches made, through provisions of systems and materials, to the development of renewable energies. Due to the expiration of the operating licenses for 8 nuclear power plants, the disorganised, too rapid expansion of volatile renewable energies(electricity from the sun and wind has priority in the grid) and expansion of the grids that has been implemented only at 15% of the demand, it has come, since 2011, to a tripling of the number of necessary interventions to maintain the stability of power grids and hence of the supply of the consumers. A stabilisation of the power supply is only possible through the building of new base load power stations and, here, the existing gas power stations are currently operating uneconomically because of the priority of

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 28

Highlights 2013

renewable energies. To become independent of possible grid variations, energy self-supply using CHP plants in island mode would be suitable. In this case, it is beneficial to the enterprises that own electricity utilization is currently exempt from the EEG apportionment. The practical implementation of energy self-supply through a plant for the combined production of steam and power (CHP (combined heat and power) plant) was demonstrated by Oliver Blum of ECM Ingenieruunternehmen für Energie- und Umwelttechnik GmbH in Munich on the basis of the design and construction of such a facility for the medium sized paper manufacturer Hans Kolb in Kaufbeuren. The enterprise has been operating, since 2008, a new paper machine with an annual production of 65,000 t in paper of 90 to 170 g/m2, the total capacity, however, could be expanded to 100,000 t/a. The existing steam generation was on the capacity limit also with the current production volume. A further motivation for the construction of the new CHP plant, with a power output of 5.7 MW, an electric efficiency of 31 %, a power performance indicator of 0.61 and a maximum steam production of 24 t/h, was the independence from any grid variations and the possibility to generate an added value via the feed-in of surplus electricity. The saving in primary energy with the system was about 16.5% within the operating year under consideration. An increase of the economic efficiency and efficiency of CHP plants is expected from the decoupling of electricity and heat consumption by the fact that surplus heat is stored in suitable heat accumulators. Residual and recyclable materials In the paper and pulp industry, large amounts of damp residual materials occur, the disposal of which in waste landfills has become considerably more difficult and/or more expensive after introduction of the European Waste Directive. The alternative is waste incineration on site for the generation of electricity and heat. Dr. Hermann Schwarz of Siemens AG described the socalled SIPAPER Reject-Power-Technology and its application in the paper mill of the company Saica in the British town of Partington. This paper mill processes 420,000t/h of waste paper. In the process, about 80,000 t/a in residual materials originate, among these also coarse-grained residues with high plastic content from waste paper treatment, fibre residues from the papermaking machine and sewage sludge from the in-house sewage treatment. The latter also contains an anaerobic treatment stage, from which biogas for cofiring is gained. The waste has a mean

water content of about 45 %, a mean calorific value of ≥ 6,500 kJ/kg as well as an ash content of up to 35 % In Partington, high-pressure steam is generated with the combustion plant with throw feeding and dual burner (also for gaseous fuels). Combined heat and power generation is also possible in principle and is also being implemented in other plants, partly with biomass combustion. Beyond all that is currently understood as a recycling management (including also the waste-to-energy-process, as described to begin with), the so-called “cradle to cradle” principle is currently being developed. For a deeper understanding of this principle of 100% circuit closing, reference is made to the websites of the cofounder, Prof. Michael Braungart (www.braungart.com) or of the consulting firm EPEA Internationale Umweltforschung, whose Managing Director of the Swiss company headquarters, Albin Kälin, presented the principle. At the core, this concerns the observance of two circuits, one each for biological and for technical systems, in which the circuit closing can happen without origination of waste or quality loss. For this purpose, all materials are divided into four categories: A, B and C stand for optimum, optimiseable and tolerable for the purposes of either the recovery or ultimate biodegradability, X stands for not acceptable. According to the cradle-to-cradle- principle, material composites from substances of the classes A-C with a substance of the class X are excluded. What this means for a paper product was illustrated by Eric van Hoof of the Dutch tissue manufacturer Van Houtum B.V. Tissue paper is produced under use of about 25 chemicals in order to obtain the desired material cohesion, wet strength and other properties. The development and finding of alternative additives, such as, for example, of a starch-based wet strength agent, has taken a total of two years. Dr. Kurt Schloffer, Added Value GmbH, dealt with the recovery of the organic substances from the bleachery waste water from pulp production and their subsequent utilisation, in his function as a coordinator of a cooperation project in which the bleachery waste water is supposed to be concentrated by means of ultrafiltration, then evaporated, and the thus gained lignin derivatives are supposed to be activated and be used as an adhesive for chipboard production or vacuum impregnation. An essential step in the treatment chain is the dynamic membrane filtration with the TRF system (Turbolation Rotation Filtration) that was developed and is being manufactured and distributed by the Austrian firm Minerwa Umwelttechnik GmbH & Co. KG. Organic ingredients like lignin, sugar and big metal ions were concentrat-

ed with ultrafiltration by more than 65%. Afterwards, the retentate was evaporated to about 55% TS. The residue was activated with different chemicals, the chipboards made with it were examined for their mechanical properties. The suitability of the bleachery water concentrate for wood impregnation turned out to be dependent on the kind of wood concerned. Industry focus: Basic Award Criteria for the “Blue Angel” eco label The “Blue Angel” was introduced as the first environmental label worldwide in 1978. Because the runtime of the six environmental labels currently to be awarded for the product groups hygiene paper, recycled paper, wallpaper and ingrain wallpaper, recycled cardboard, unbleached cooking filter papers and hot filter papers as well as printing papers and press papers ends on 31/12/ 2014, the basic award criteria are reworked according to the progress with the state of the art. Almut Reichart of the Federal Environmental Agency (Umweltbundesamt UBA) in Dessau reported about the status of the discussion. In detail, for example, the pentachlorophenol content need not be monitored anymore, while the substance Bisphenol A is being monitored furthermore. It is present in all waste paper, however, a limit value cannot be introduced as long as there is no method around to reduce this substance or to remove it completely. Hygiene papers are limited in the degree of whiteness and must not contain any additives like lotions, nanosilver, odoriferous substances or micro-organisms. Ingrain wallpaper with the Blue Angel must in future to 100% consist of recycling paper. For printer colours, the approved mineral oil hydrocarbons are limited to saturated hydrocarbons ranging from C10 to C20. In addition, later in 2013, a project funded by the UBA for the development of the basic award criteria for environmentally friendly printed products (green printing) would be launched. This environmental label newly to be awarded in 2015 includes the paper production, including raw material selection and printing. Particularly, a high standard for printing inks would also reduce the proportion of undesirable substances in the waste paper or in the then originating deinking sludges. Waste water treatment The reduction of the water consumption of the paper industry is an essential component for achieving the objectives of CEPI roadmap for reducing CO2 emissions. Here, thus far, the aspects of energy recovery and recovery of valuable sub-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 23:34 Seite 29

Highlights 2013

stances from waste water are still at the beginning stages of their large-scale introduction. Holger Jung and Dr. Benjamin Simstich, PTS, provided an insight into recent research projects on these topics. A typical paper production loses about 300kWh/t of heat with the effluent, 30 kg/t of organic ingredients, measured as COD and 1 kg/t of fibre. For the recovery of heat, heat exchangers for preheating of the process water are suitable, moreover, the example of the installation of a sludge turbine was named, where one uses the gradient between the clarifier and secondary clarifier of a paper mill with hillside location. Part of the energy bound in the waste water is also gained back through anaerobic decomposition of the organic load resulting in the generation of biogas. A promising and less energy-intensive alternative to the conventional circuit water treatment has been developed with the aerobic waste water treatment at increased temperature under use of a submerged membrane module, (see /6,7/). A variation that will be described in more detail in the further course of this article is algae supported aerobic wastewater treatment. Among the organic ingredients of paper waste waters, starch, acetic acid, carbohydrates, fillers and fibres have a market value that is rather high in comparison to the wastewater treatment costs, so that recovery instead of biological degradation is a worthwhile alternative. As an example of this, starch should be named, which enters the process water in appreciable quantities during waste paper dissolution, where it metabolises relatively quickly. In a research project together with the Fachagentur Nachwachsender Rohstoffe, FNR, one is looking for ways to prevent or limit metabolisation in order to retain the starch as a means for increasing the strength and be able to use it again. As an other initiative for recyclables recovery, the project LignoBoost of the Finnish paper manufacturer METSO was named. There, they succeeded this year in commissioning a 30,000 t/a-pilot plant, after several yearsâ&#x20AC;&#x2122; development work, which allows the recovery of lignin with concurrent separation of the other ingredients from the black liquor originating during cellulose production. Herbert Buchinger reported about operational experiences with anaerobicaerobic sewage treatment at Tissuefabrik Ortmann (SCA Hygiene Products). About 80% of recycled paper are used for production, in addition there is new fibre material. The original aerobic sewage treatment plant dates back to the middle of the 1980-ies, has been implemented in tower configuration manner and also processes 10% of municipal sewage, besides the process waste waters. Owing

F & S International Edition

No. 14/2014

Fig. 2: System diagram for paper wastewater treatment in an AnMBR process (Image: Bauer Water GmbH)

Fig. 3: Functional principle of a microbial desalination cell (according to the lecture by Antje Kersten, Technical University of Darmstadt)

to the increasing waste-paper recycling, wastewater volume and load has risen continuously, moreover, foam problems have appeared. In order to provide relief to the aerobic stage, a BiobedÂŽ-EGSB-Reaktor (EGSB = expanded granular sludge bed) was added upstream, that was meant to stabilise aerobic sewage treatment, should reduce the excess sludge occurrence and produce biogas, moreover, in addition. In addition to this reactor, the wastewater treatment plant has been equipped with a new pumping station and a cooling tower. Since the commissioning in 2011, expectations were be extensively met. It is important for the design of the anaerobic waste water treatment to know with which waste water composition, at which reactor size, one receives a certain gas yield. Richard Moosbrugger, IB-MR in Bizau, Austria, has developed a simulation programme for this purpose, and he explained its foundations, structure and results to be achieved.

Dario Gallotini of KUBOTA Membrane Europe described the submerged disk module of his company for microfiltration in MBR-plants and, among the rest, reported about operational experiences in the Spanish tissue factory Goma Camps. The decisive factor for the conversion of a conventional sewage treatment plant into an MBR plant was the endeavour to increase the capacity at a constant space offer, moreover, the sludge sedimentation caused difficulties. The membrane modules were installed into the newly built aeration tank without prior piloting. The first cleaning with sodium hypochlorite solution was performed after 12 months. An intensive cleaning (in a separate cleansing basin) has not been necessary yet up to now (after two operational years). Daphne Hortensilla of the Department of Chemical Engineering at the Universidad Complutense in Madrid described the conversion of the water supply for the paper mill Holmen Paper in Madrid from urban water to treated waste-

02_fs_international_2014_seite_03-98__ 21.05.14 23:34 Seite 30

Highlights 2013

Fig. 4: Diagram of the laboratory facility for algae assisted biological wastewater treatment (Image: PTS)

water of the municipal sewage treatment plant. Already before, they had one of three most modern manufacture plants for newspaper printing paper in the world, the water consumption of which, at 7m3/t , is below the value stated in the BVT documents. The treatment of secondary effluent was given preference because the water recirculation was limited within the operation through the enrichment of silicates taking place. The NEWater concept in Singapore acted as a model for the reuse of purified waste water. With test plants for in each case 1.5 m3/h of the prefiltered secondary effluent, three different multi barrier concepts were tested, one of these with microfiltration and two different submerged UF-systems, in each case followed by a retentate-graded reverse osmosis and UV disinfection. The results of the preliminary tests can be read up in /8/. Meanwhile, the large-scale plant with an ultrafiltration as a pretreatment for reverse osmosis has been built and in summer, 2012, it was commissioned. Up to now, no negative effects on the individual production processes using water were able to be established. The waste water of the paper

mill is supplied to the sewage treatment plant, from which the enterprise draws its raw water. Since anaerobic waste water treatment is already being considered as a suitable method for waste waters from the paper industry, they are dealing with its optimisation at BAUER Water GmbH, among the rest in the form of an anaerobic membrane biology reactor. Herbert Bassek informed about the status of the process development carried out together with the PTS. The process diagram can be seen in Fig. 2. In the filtration tank behind the pellet sludge reactor, two different small modules with flat membranes were inserted. Since no crossflow ventilation is possible in an anaerobic system, the respective flow along the membranes was implemented by means of circulation pumps. If, with the pellet sludge reactor, the sludge globules, if applicable, can be discharged under impact load and also be crushed through the circulation, the membrane offers a safe retention. The method is still optimiseable in this respect as that one has not achieved the objective of implementing similar permeate flow rates as with membrane stimu-

Fig. 5a) Set-up of the experiment facility for algae assisted wastewater treatment in a Slovenian paper mill (Image: PTS)

lation processes and the circulation, up to now, still constitutes insufficiently effective fouling control at the membrane. As Daphne Hortensilla had already mentioned, the high salt load, particularly a high calcium concentration, complicates the treatment and recovery of paper waster water. If one succeeded in developing a simple and effective method with which organic ingredients can be degraded, as well as the salt load be reduced, the water loop closure would be easier. At the University of Technology of Darmstadt, under the direction of Antje Kersten, they are experimenting with a microbial desalination cell, originally developed for sea and brackish water desalination (see, among others/9/). At the source, micro organisms at the anode are meant to facilitate the transfer of electrodes, and hence the transport of the ions through the ion exchange membranes (see Fig. 3). The thought is at hand that the micro-organisms are processing the organic ingredients of the waste water. In laboratory tests with a cell with 100 ml capacity, desalination of a NaCl solution was tested first with the help of glucose as a nutrient for the micro-organisms. However, in a further step, paper waste water was also used. Up to the large-scale technical implementation, a lot of research work is still to be performed. For example, it has to be clarified how one can reduce biofouling on the membrane and with which membranes the transport of divalent ions can be improved. Algae have the property to take up CO2 and to release oxygen via photosynthesis. The specific production of algae as a biomass for the recovery of valuable chemicals or for the production of biofuels under use of CO2 from industrial or combustion processes is a strategy that has been known of for some time. The European cooperation projects ALBAQUA and the succession project ALBAPRO (s. www.cornetalbapro.eu) do not aim at a maximum

Fig. 5b) Close-up of the biomass during algae assisted wastewater treatment (Image: PTS)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 23:34 Seite 31

Highlights 2013

algae yield but at the generation and preservation of a stable symbiosis between algae and micro-organisms for the aerobic sewage treatment. Gabriele Weinberger of PTS is involved in both projects and reported about the most important results of ALBAQUA and the progress of ALBAPRO. For instance, in the first project, they had successfully generated a stable algae biomass and use it in the waste water treatment. After the first tests in a laboratory system (see. Fig. 4) external tests were carried out in a Slovenian paper mill (see. Fig. 5 a) and b). With these, it could be shown that the solar irradiation is sufficient for an open-topped container only. The biomass, even after a 10-hour dark phase, did not yet come within the anaerobic range. An even bigger challenge than the incident light irradiation is presented by the water temperature. The algae need the observance of a temperature range of from 15 to 35 °C. The algae composition adapts itself to the waste water composition, and so does the sedimentation capability of the biomass. For the operation in outdoor complexes, a strategy must be found to protect the algae against natural predators (insect larvae). Another problem to be solved concerns a fast analysis of the biomass that makes a distinction between algae and bacteria. For the future, this wastewater treatment process will be optimised and stabilized and, moreover, an energetic utilisation will be found for the surplus biomass obtained. Monitoring of discharge quality and risk analysis The compliance with the water framework directive and other ordinances for the preservation of the quality of waters requires the monitoring or reduction of the introduction of critical trace substances. The following are considered to be possible trace contaminants in waste waters of the paper industry: DTPA, Bisphenol A, PAK, PFC and others. One possibility of the monitoring of the water quality without individual substance determination is the duckweed growth inhibition test, which is listed in the revised Annex 28 of the Waste Water Ordinance. During the duckweed growth inhibition test, the poisonous effect of contaminants on the minor duckweed (Lemna minor ) is determined, by the fact that differently diluted water samples are spiked with these duckweeds and their growth is determined by image analysis after 7 days incubation under defined conditions. As a result, a dilution factor Dw is then stated, with which the growth inhibition, compared with unexposed comparative samples, undercuts the value of 10 %. The revised Annex 28 states a limit value Dw = 8 , which will have to be observed eventually after three years after the entry into force. Dr. Jürgen Öller showed the results of the F & S International Edition

No. 14/2014

sampling of waste waters from a total of 7 paper mills that distinguish themselves in the product range, the raw materials used (original fibres, waste paper with or without deinking) and in their wastewater treatment. Here samples were drawn at the outlet of the respective treatment plants, but also between different purification steps. Overall, it was found that the limit value stated above, at the outlet of most paper mills, was clearly fallen short of. As problematic, with drainage values of Dw = 48 or higher, the outlets of plants turned out in which graphical papers are produced using waste paper, which has been subjected to a deinking process. The individual analysis of trace substances was carried out at the Department of Paper Technology and Mechanical Process Engineering at the Technical University of Darmstadt by Antje Kersten. The substances examined were all within the order of magnitude of the detection limits during the analysis. With the data obtained, the initiating volume flow and the average flow rates of running waters, into which introduction took place in each case, a simplistic risk analysis was carried out. The environmental concentrations calculated from the data (PEC = predicted environmental concentration) were compared to the respective PNEC values (predicted no-effect concentration). As long as the PEC value is smaller than the PNEC value, there is no risk present. However, the requirements for the observance and monitoring of the waste water quality increase when multiple dischargers discharge similar trace substances into the same water bodies. Literature: /1/ Greenpeace International: How Sinar Mas is pulping the planet; published in June 2010 and revised September 2010 /2/ Confederation of European Paper Industry (CEPI): Sustainability Report 2011, www.cepi.org (retrieved in November 2013) /3/ CEPI Annual Statistics 2012, s. www.cepi.org /4/ Confederation of European Paper Industry (CEPI): Sustainability Report 2013, www.cepi.org (retrieved in April 2014) /5/ European Commission: Best Available Techniques (BAT) Reference Document for the Production of Pulp, Paper and Board (draft July 2013) /6/ Simstich, B: Einsatz der Getauchten MBR-Technologie zur thermophil aeroben Behandlung von Kreislaufwässern der Papierherstellung; Helmreich, B.(Editors) Reports from Urban Water Management Technical University of Munich, Report Book No. 207, 2012 /7/ Simstich, B.; Beimfohr, C.; Lyko, M.; Horn, H.: Thermophiler Betrieb eines getauchten MBR bei 50°C zur Prozesswasserreinigung in der Papierindustrie; KA Korrespondenz Abwasser, Abfall 2012(59), No.5; p. 465 – 472 /8/ Ordóñez, R.; Hermonsilla, D.; San Pio, I., Blanco, Á.: Evaluation of MF and UF as pretreatments prior to RO applied to reclaim municipal wastewater for freshwater substitution in a paper mill: A practical experience; Chemical Engineering Journal Vo. 166, Issue 1(2011), S. 88-98 /9/ Cao, X.; Huang, X.; Liang, P.; Xiao, K.; Zhou, Y.; Zhang, X.; Logan, B.E.: A New Method for Water Desalination Using Microbial Desalination Cells; Environmental Science & Technology, 2009, 43 (18), p. 7148 - 7152

02_fs_international_2014_seite_03-98__ 21.05.14 20:09 Seite 32

Highlights 2013

Centrifuge technology: Materials, machine design and applications Report from the 4th International Separation Technology Symposium of the Schmidt + Clemens Group H. Lyko* There are industrial centrifuges with a huge number of construction types and specifications, which are optimally adapted to the separation task to be solved, the prevailing ambient conditions and the substance properties of the mixtures to be separated. The bandwidth in single components and materials from which these are made is accordingly large. Important centrifuge components, like drum bodies and screw bodies, are produced with centrifugal casting methods. Furthermore, there is a great need for customized components as mould or precision castings. The Special products business unit of the Schmidt + Clemens Group with the Department of separation technology specializes in the production of all relevant stainless steel components that are used in centrifuges. The separation technology symposium offers to all parties involved a good opportunity for communication on one hand about new materials, as well as about important application fields, and on the other hand about their requirements, machine models, design methods and operating conditions. In June in Cologne, Wolfgang Unterbusch, Vice President of Sales, and Christian Lenz, Sales manager of the Special products business unit, welcomed about 50 participants of all the leading centrifuge manufacturers of Europe and of important supplier companies of the centrifuge industry. In addition, a representative of a municipal wastewater disposal company, as well as one from an enterprise which deals with sewage sludge drying, respectively supplied interesting information from the environment of a classical application area for centrifuges.

New stainless steels for casting, not only for centrifuges Many centrifuge components are produced of so-called duplex steels. These are steels with a 2-phase texture in which the properties of stainless ferritic chromium steels are combined with those of stainless austenitic chromium-nickel steels. Duplex steels thus have both high corrosion resistance and also high ductility and strength /1/. The essential alloy elements of duplex steels are nickel and molybdenum besides chrome. The contents of these three elements are summarised as a measure of the corrosion resistance into the so-called PRE number (%Cr + 3.3%Mo +16%Ni). As Georg Wilhelm Overbeck, Schmidt + Clemens, explained, the prices of duplex steel are determined particularly by the prices for molybdenum and nickel, which, in the past, were not only high, but were also subject to large fluctuations. The endeavour to produce cost-effective, yet firm and corrosion-resistant materials has led to the development of the so-called Lean Duplex and/or Ultra-Lean-Duplex grade. Already in May, 2012, the casting material Centralloy® G21-02 (GX3CrMnNiN21-5-2, an Ultra-LeanDuplex steel) was launched on the market. The material has good strengths similar to conventional austenitic Cr-Ni steels, which are used in pumps, centrifuges and related equipment, as well as very good resistance to intergranular corrosion. *Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

Dr. Henrik Astemann presented Centralloy® G45Mo (G-NiCr35MoCuN), an austenitic cast steel, as the latest material development, which has not, like many other cast steels, been developed through modification of a comparable rolled and forged steel, but rather, the special requirements for processing form casting were taken into consideration in the development. In addition, certain requirements for mechanical strength and corrosion stability also had to be fulfilled, and, moreover, the material should be cheaper than existing nickel-based steels. G45 Mo contains 45 wt.% of nickel and 35 wt.% of chrome, and hence is planned for use in highly corrosive surroundings. Spin, shape and precision castings can be produced from G45 Mo. The steel was developed within the

scope of a BMU research project for the realisation of geothermal power plants, which is being coordinated by the Georesearch centre in Potsdam. The potential to gain geothermal energy and to produce electric power from it, too, is harboured by territories with aquifers whose water temperature is in excess of 100°C. The Gross-Schönebeck research station lies in such an area. There, water horizons are being opened at about 4000 m depth. The hot liquid which is thereby extracted to the surface is a salt brine with an overall concentration of about 300 mg/l and a pH value of 5.3 - 5.8. With this medium, different materials that are known as being corrosion-resistant were and are being tested in a specially installed bypass pipe, among them also the new Centralloy® G45

Fig. 1: Sewage sludge treatment in the 2.7 million PE sewage treatment plant of Hamburg (image: Hamburg Wasser)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 33

Highlights 2013

Mo in the form of pipes, flanges, etc. The probability of general corrosion or pitting corrosion is determined on-line with electrolytic methods. In parallel, additional laboratory measurements on corrosion resistance are carried out in the Federal Institute for Materials Testing (BAM). The results of both measures will lead in future to finding the most suitable materials and possibly operating parameters for such demanding applications. An important market: Sewage sludge treatment Sewage sludge dewatering is a standard application for decanters. Here, dewatering performance (throughput and achievable solid matter content) and energy consumption are generally in the focus of the users. How sewage sludge treatment is handled in Germany’s largest municipal sewage treatment plant in Hamburg, was described by Hendrik Schurig of Hamburg Wasser. The sewage treatment plant is designed for 2.7 million PE (population equivalents) and consists of two biological treatment stages. The whole system of sludge treatment is shown schematically in Fig. 1. In this sewage treatment plant, about 1.5 million m3 fouled sludge originate during the cleaning of about 150 million m3 of wastewater annually that is dewatered with 5 decanters and afterwards is dried and combusted on site. In addition, 8 decanters provide for the thickening of the surplus sludge since static thickening produces insufficient dewatering. Moreover, the sludge combustion accepts dry sludges from other plants. Altogether the energy requirements of the sewage treatment plant can be covered up to 75%, and the remaining 25% are provided by wind power. The biogas originating with the sludge digestion is processed and fed into the gas grid. The low solid matter content of the dewatered sludge of 21% before the dryer is noteworthy here. Because of the combustion on site, there are no transport costs for the sludge, and therefore polymer consumption for dewatering was crucial for the treatment costs. A stronger dewatering would have led here to the increase of polymer consumption. Sewage sludge dewatering with decanters down to only 21% TS is far below what these machines are able to perform and is also not economical, since for disposal one has to factor in about 40 to 80 euros per ton of dewatered sludge, dependent on location. Here, one calculates for the reduction of water content by 1%, with a reduction of the sludge volume to be treated and to be transported by 4 to 5%. This was also highlighted by Wolfgang Steiger of the company Flottweg, which was also the supplier for the Hamburg sewage treatment plant. His lecture was about the design of decanters for sludge dewatering, inter alia. The basic principle of the design is based on the calculation of the required clarification areaaccording to the so-called Sigma principle. The equivalent clarification area Σ is the area for static dewatering multiplied by the acceleration factor (expressed as a multiple of gravitational acceleration g). In case of consideration of the centrifuge bowl as a cylinder, one can accordingly convert to the diameter and the length and/or the centrifugal acceleration - the same value Σ will then always mean the same dewatering performance. That is the theory, because decanter drums consist not only of a cylindrical part but have, according to make and construction form, different long and steep conical elements (see Fig. 2). How one includes the not cylindrical parts of the drum in the calculation of Σ is not established uniformly and, much to the regret of the lecturer, is solved by every manufacturer with another formula. Into the formulae, the length of the conical part and the diameter of the solids discharge are included with different factors. If one carries out the Scale-Up of the pilot up to industrial scale in each case with the same formula, one receives similar magnification factors. It becomes difficult for various manufacturers in preparing their tenders, because the different calculation methods mean that different enterprises offer machines of different sizes for the same dewatering performance, whereby distortions of competition can occur. With the same external dimensions, dewatering performance is increased

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 34

Highlights 2013

Fig. 2: Dimensioning of a decanter (image provided by Flottweg GmbH)

through a steeper cone (synonymous with a longer cylindrical zone) and through a deep pond (corresponding to a bigger process volume). The calculation of the equivalent clarification area does not include any specific design parameters of a machine like the screw design, the inlet geometry or the drive system. However, these play also an important role, inter alia, for the compaction pressure and hence the achievable dewatering. In practice scaling-up is performed according to the required maximum capacity. For applications on wastewater and sewage sludge, Flottweg offers its standardised decanter series C4E. It was pointed out that maximum dewatering does not always mean the most efficient operating mode, but rather one must consider the relationship between water content of the solid matter and the throughput capacity. The dryness of the solid matter drops approximately linearly with increasing throughput. For further use of the sewage sludge as fuel, this is dried. In the Hamburg sewage treatment plant, this happens in disc dryers, which are heated with steam generated during the sludge combustion. Also here there are now alternative drying systems on the market, like solar drying alone or in combination with waste heat recovery. Dr. Steffen Ritterbusch, Thermo-System GmbH, reported about such methods. The drying happens in climate- controlled halls of glass or PE sheeting, which resemble greenhouses and, when required, are also equipped with a waste air treatment facility. According to sludge volume, the process is carried out intermittently (loading and unloading with a wheel loader), semi-continuously (automatic loading, unloading with a wheel loader) or continuously (automatic filling and emptying). In the meantime, a large number of plants exist worldwide. The biggest solar sludge drying plant, with a capacity of 140,000 t/a, is currently in construction in Dubai. The fact that solar sludge drying can be operated not only in southern countries is shown by the example of Oldenburg, Lower Saxony. Here, the sludge of a 400 - PE sewage treatment plant (corresponding to 40,000 t/a sludge) is dried by utilisation of solar heat and waste heat. The use of waste heat reduces the space requirements of the plant to about 20% of that for pure solar drying, for which one calculates about 1m2 surface area for 1 t of sludge. The supply of the waste heat is done either via the supply of warm air or via underfloor heating. A component especially worth mentioning for small and medium-sized sludge drying plants is the Electric Mole™, an automatically controlled appliance, made completely from stainless steel, for turning and aerating the sludge during drying.

Fig. 3: Diagram of a vertical decanter (image: provided by Pieralisi Deutschland GmbH)

Other construction type: Vertical decanter As a known manufacturer of decanters and separators, the Pieralisi Group has significantly expanded its product portfolio in recent times. On one hand, the company’s own centrifuge programme is complemented by horizontal screen screw centrifuges. On the other hand, through cooperation with Kyffhäuser Maschinenfabrik Arten GmbH (KMA) and the Japanese enterprise TOMOE Engineering Co. Ltd., other machine models are available. KMA is a manufacturer of vertical disc centrifuges and TOMOE is regarded as a world leader in large decanters for the process industry and wastewater treatment. Gert Bergjohann, Pieralisi Deutschland GmbH, presented the vertical decanter built by TOMOE (see Fig. 3) as a special machine model. This construction type was originally developed more than 50 years ago at Sharples in the USA and was for a long time the only form in which a decanter was feasibly gastight. Nowadays, vertical decanters are used for separation tasks that take place at high temperatures and under high pressure. These include coal liquefaction and applications in the presence of organic solvents, such as for example the recovery of catalysts. As is evident from Fig. 3, the spindle area and all technically sophisticated equipment, such as the gearing and the main bearing for the drum, are outside the

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 35

Highlights 2013

product chamber. Besides their suitability for temperatures up to 400 °C and pressures up to over 30 bar, the vertical construction form is fascinating because of the relatively small installation surface. In Asia, this machine model takes on an important role in the production of PET (Polyethylene terephthalate). The centrifugal acceleration in vertical decanters is normally 2000 - 3000 *g. They are made with diameters of 150 to 900 mm and can process 100 kg/h up to 20 t/h of dry solids. In addition, there are special designs with 5000 - 10,000 *g centrifugal acceleration that are not suitable, however, for high solids concentrations. Centrifuge drives Different drive systems are available to put the drum and screw of a centrifuge into rotation at different speeds. Nils Zieglgänsberger from Sumitomo (SHI) Cyclo Drive Germany GmbH presented the Cyclo drive whose operating principle had been invented in the 1920s and had been used first in 1949 in a centrifuge built by the manufacturer Siebtechnik. A Cyclo drive is an excentric drive whose outer profile describes a cycloidal curve path. It consists of the components: high speed shaft with excentric bearing, cycloid discs, ring gear housing with pins and rollers, as well as the slow speed shaft with pins and rollers. The single-stage cycloid drive here creates a transmission ratio of 1:87 (see Fig. 4). Higher transmission ratios are feasible in multistage construction methods, inter alia, in combination with a planetary gear in which mechanical transmission takes place via gear wheels in a toothed ring gear member. Different construction forms permit different operating modes of screw centrifuges, depending on whether the screw should turn more slowly or faster than the drum. With a two-stage gear unit with differential planetary gear stage, the setting of a variable relative rotational speed with pre-triggered and also with retarding screw is possible in the entire control range of the engine. The alternative to the energy transfer to the screw of a decanter is a hydraulic driving mechanism whose advantages compared with a mechanical transmission had already been shown by Marco Metzger of the Swiss firm Viscotherm AG at the 3rd Separation Technology symposium /1/. The heart of this driving mechanism is a so-called radial piston machine whose stator has a cam disc with cylindrical winding. The rotor consists of a cylinder block in whose boreholes pistons are pressurized alternately with the pressure medium and return medium via a distribution system. The radial forces are converted into tangential forces, through the incline of the cam disc over the piston parts as a joint, and generate therefore the necessary

F & S International Edition

No. 14/2014

Fig. 4: Structure of a Cyclo drive (images: SUMITOMO (SHI) Cyclo Drive Germany GmbH)

torque between rotor and stator. This year Metzger presented, inter alia, the most recently developed stepped piston with which an even higher torque can be generated at the same dimensions because more liquid is moved per stroke of each of these pistons. Screen screw centrifuges and hybrid forms Screen screw centrifuges are screen centrifuges with forced transport of solids /2/. They are used for dewatering coarse material like coal or minerals (grain size for coal about 0.25 to 2 mm). Bernward Feldkamp, Andritz Separation GmbH, showed details of the design of this type of centrifuge. The cone angle of the screw, the height and spacing of the filaments and their rotation speed determine the dwell time of the solid matter in the machine and with it the levels of achievable dehumidification. With the help of these parameters, the speed that the solids cake onto the basket, or the time-dependent position of the solid particles on the transport path until discharge, are calculated. The result is the curve of the saturation of the filter cake above the basket diameter, which is hyperbolic. This curve path helps in the design of a screen screw centrifuge for a separation process, but cannot be carried out for every substance system, which is why one ventures to dare a transmission. The differential speed between the screen basket and screw results from the speed of the screen basket and the transmission of the gearbox. Here, Cyclo or planetary gear drives are interchangeable because of their similar external dimensions. Sedimentation and filtration centrifuges in many special designs make up the product portfolio of Siebtechnik GmbH, and in most applications solid matter is the product. As Andre Adam elaborated, this mostly concerns crystalline particulate solids, which sediment rapidly. In his presentation Adam drew special attention to the machines in which sedimentation, as happens in a solid bowl centrifuge, and filtration in the centrifugal field, take place side

by side or successively with a drum designed as a screen basket. The motivation for these combination machines lies, inter alia, in that a pure screen centrifuge, depending on the degree of separation and selectivity of the screen, generates particle losses and that one wants to lower the humidity of the dewatered solid matter and, nevertheless, one wants to get a clear centrate. In the Conthick combination centrifuge, the feed suspension enters first into a conical bowl portion and is then transported into the cylindrical screen member. This arrangement allows a smaller separation limit compared to a pure screen centrifuge (as the minimum, still separable particle diameter at which no clogging of the screen occurs). The Turboscreen model is a combination of decanter and screen centrifuge and the diameter of the screen member is increased abruptly relative to the decanter member. A higher centrifugal acceleration is thereby achieved in this area and a thinner filter cake is generated. In both combinations one obtains a clear centrate of the decanting part and a filtrate conditioned by the separation limit of the screen, in which particles might still be present. The TwinCone machine is a decanter with two separation stages of different diameters. The dewatering degree and the solids capacity can be increased by this arrangement. As was shown on the basis of test results from comparisons of a standard decanter with Turboscreen and/ or Twin Cone, the improvement of the material separation through the combination machines is not equally high for every product, so that one must produce the justification of the higher investment costs for a combination machine in individual cases by tests. Literature: /1/ Lyko, H.: Industrial centrifuges for different applications: Challenges for manufacturers and suppliers – Report from the 3rd International Separation Technology Symposium of the Schmidt + Clemens Group F&S International Edition No. 11 (2011) pp. 27 -31 /2/ Benz, M.: Siebschneckenzentrifugen, in: Stahl, W.: Industriezentrifugen – Band II: Maschinen- und Verfahrenstechnik; DrM Press, ISBN 2-9522794-0-4, 2004

02_fs_international_2014_seite_03-98__ 21.05.14 20:34 Seite 36

Highlights 2013

Systems and processes for drinking water and process water treatment H. Lyko* The re-use of water and water ingredients taking into account energy efficiency and resource conservation is deemed a decisive criterion for the selection of systems and processes. This was stressed by Gottlieb Hupfer, Chairman of the VDMA specialist Water and Wastewater Technology department, who was able to welcome about 100 participants from authorities, industry and engineering firms to the 20th anniversary of the series of events “From practical experience for practical application”. Prof. Sven-Uwe Geissen of TU Berlin, who moderated the two-day event, in his introduction pointed out the trend towards decreasing water consumption in industry. The use of processes for recycling water is increasing, as well as the use of alternative resources. The use of the heat released with wastewater streams is also gaining greater importance. Newly originating industrial branches that are based on the use of biomass for the most part consume a lot of water. With new processes, new questions also arise about the required water quality and sustainable treatment processes.

Industrial water technology in Europe

Tab. 1: Overview of European initiatives and activities in industrial water technology

The urgent need for sustainable management of increasingly strained water resources was recognised years ago in Europe and, inter alia, was formulated within the scope of Strategy Europe 2020 in the flagship initiative “Resource-efficient Europe”. Dr. Thomas Track, who represents Dechema and is also involved in the EU project E4Water as project coordinator, outlined the present water situation and the development expected on account of climate change. Therefore, challenges for water management arise from the already prevailing water stress in some regions of Europe. In some coastal regions, there exists the danger of salinisation of ground water and in southern Europe there is a demonstrable shortage of fresh water. The need to close water cycles also causes pressure for treatment of highly concentrated waste streams. However, with this there also arises the chance to recover recyclables. Track gave an overview about European initiatives and ongoing or recently completed research projects on water issues. They are summarised in Table 1, and detailed information about this can be found on the stated websites. Surface water pollution control and handling of priority substances The definition, achievement and verification of the ecological and chemical good state of surface waters also depends, inter alia, on the knowledge of the substances interfering with this state and the possibilities of gathering the data by appropriate *Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 37

Highlights 2013

analytical methods. Currenta GmbH in Leverkusen specialises in analytic services and also deals with the recording of the so-called priority substances in surface waters. As Dr. Günther Müller demonstrated, the list of priority substances maintained in the revised Water Framework Directive (2000/60/EC) and/or UQN Directive (2008/105/EC) currently contains 33 substances/substance groups. These are divided into three categories “Priority substances”, “Priority hazardous substances” and “Priority hazardous substances for testing”. One finds an exact listing with accompanying data sheets on the website of the Federal Environment Agency (www.uba.de). The list is not closed, but should be checked and, if necessary, supplemented every four years. Thus an expansion by 15 substances was suggested in January, 2012 by the European Commission. Currently 88% of waters are found as being in a chemically good state, but for 10 of the 33 listed substances, there are currently no analytical methods by which compliance with the low limits of the environmental quality standards can be checked. Moreover, with some priority substances one knows that they either enter into the aquatic environment from natural sources (e.g. mercury), or that their concentration cannot be influenced via improved wastewater treatment because they reach the water through the air (e.g. PAK from combustion processes). Cooling water treatment In open cooling circuits, the formation of biofilms is the essential cause for the reduction of the cooling effect and for corrosion. The enterprise MOL Katalysatorentechnik GmbH, Schkopau, offers two catalytic processes by which cell components contained in the water are partially oxidised, whereby so-called biosurfactants arise with which biofilms are detached from surfaces and die. Dr. Jürgen Koppe reported about several projects successfully carried out with the MOL® Clean method in refineries and the petrochemical industry in Russia. The application of this method also turned out to be transferable to reverse osmosis systems in which the pre-cleaning filter is fitted with a corresponding catalyst. The catalyst is installed as a wound package in the filter housing for the candle filter and thus prevents fouling of the reverse osmosis modules. At DAT Dynamic Aquabion Tower GmbH in Roßdorf, chemical-free treatment of cooling water is relied on. Dr. Marco Reiche pointed out that chemicals only served symptom control, but did not eliminate ingredients. According to him, this globally unique DAT concept provides a combination of several processes that are shown in Figure 1. In a bypass system, mechanical, physical and biological technologies are integrated. In detail, this concerns a particle filter, a biologically active filter that deprives the water of dissolved nutrients such as phosphate, nitrate, ammonium and organic matter, and hardness stabilisation on the basis of sacrificial

Fig. 1: Integration of a combination of mechanical, physical and biological technologies for cooling water treatment in a bypass system (image: DAT Dynamic Aquabion Tower GmbH)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 38

Highlights 2013

Tab. 2: Comparison of different mechanical processes for dewatering of drinking water sludges according to the presentation of Kühner and Schmid, Gebr. Bellmer GmbH

anode technology, where crystallisation nuclei originate so that crystallising salts do not deposit on the walls but are carried along by the flow. An intelligent plant control system completes the system, which is illustrated in Fig. 2 and of which, up to 2013, 35 facilities had been installed worldwide. If one does not want to renounce chemicals for cooling water treatment, the efficiency of these treatment processes can be increased, and hence the chemical consumption can be reduced, by optimising the measuring and control technology. The dosage of water chemicals and its monitoring with innovative sensor technology were the subject of the lecture of Dr. Thomas Winkler, ProMinent Dosiertechnik GmbH. For an automated dosage of water chemicals in the closed control circuit, actual and target values must be measurable on a real-time basis without the measured results being affected by other water ingredients. For the recording of oxidative disinfectants, a new resistant sensor series was developed whose measuring principle consists of the electrochemical amperometric method with two membrane-covered electrodes dipped in electrolyte. Thus the DULCOTEST® CBR1 sensor is supposedly the only sensor worldwide with which free chlorine is measurable up to a pH value of 9.5 with no cross sensitivity towards combined chlorine. A sensor working on the same principle was developed for the measurement of free bromine from the agent also used in

cooling water circuits, BCDMH (1bromo-3-chloro-5, 5-dimethyl-hydantoin). The sensors work together with a measuring and control unit especially coordinated to cooling water systems. Drinking water treatment Climate change not only affects the availability of fresh water and the need for water treatment and boundary conditions for water treatment - water treatment plants also cause a CO2 footprint themselves that increasingly must be discussed and considered in plant design. The CO2 footprint is composed of the emissions from the operating equipment and investments (appliances and system components) and the emissions of the process under consideration. In his presentation, Dr. Klaus Hagen, VWS Deutschland GmbH (Krüger WABAG), compared an example of the PCF (Product Carbon Footprint) from systems for partial desalination. In detail this concerned the CARIX method (an ion exchange method with CO2 regeneration), nanofiltration and reverse osmosis with which water of initial hardness 25° should be brought to the target of 12°. With all three processes the biggest proportion of CO2 emissions originated from the power consumption during the operation. The CARIX method came out on top and moreover, in spite of higher investment costs, it causes the lowest production costs for the delivered drinking water.

Fig. 2: Complete system for chemical-free cooling water treatment (photo: Dynamic Aquabion Tower GmbH)

For drinking water disinfection ozone is used, inter alia. If bromide ions are present in the water to be treated, this treatment causes its oxidation to bromate. If this is ingested with the drinking water, it reacts in the stomach with the gastric acid (HCl) into bromine and chlorine. The latter is considered to be toxic and carcinogenic. Therefore, in the EU the concentration of bromate and bromoform compounds must not exceed limit values of 3 and/or 1 μg/l. Prof. Klaus Nonnenmacher, ANSEROS GmbH, described the relationship between the course of ozone concentration over time and bromate formation. On account of the reaction kinetics, bromate formation can be suppressed if ozone is dosed in a very short time with very high concentration. This is implemented by the patented ozone system PAP30000, with which the disinfectant is introduced in a 10 times higher concentration than with other ozone generators. The mass transfer rate thereby doubles. An accordingly short dwell time is implemented in so-called Plug-FlowTank reactors (PTFR). In drinking water treatment, processes are common like the precipitation with iron or aluminium salts to remove dissolved and colloidal organic matter, as well as iron and manganese removal by oxidation with ozone. With these processes, sludges develop that have to be thickened and/or dewatered and disposed of (see Fig. 2). Florian Schmid of Gebr. Bellmer GmbH, that is represented on the market with belt filter presses, introduced the different methods for mechanical dewatering of drinking water sludges. Their typical performance data and consumption data are summarised in Table 2. The relatively large value ranges for the achievable dry matter contents and polymer consumption arise from the differences of the dewaterability of iron and aluminium sludges (aluminium sludges are more difficult to dewater, see Figs. 3 and 4),

Fig. 3: Dewatered pressed cake at the discharge of a belt filter press (photo: Gebr. Bellmer GmbH)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 39

Highlights 2013

Fig. 4: Iron drinking water sludge in the gravity dewatering zone of a belt filter press (photo: Gebr. Bellmer GmbH)

Fig. 5: Aluminium drinking water sludge in the gravity dewatering zone of a belt filter press (image: Gebr. Bellmer GmbH)

as well as from the dependence of dewaterability on the concentration of organic ingredients.

Hager+Elsässer GmbH, Stuttgart, supplies reverse osmosis systems for process water treatment, inter alia. Cyrus Ardjomandi explained the saving potential through the so-called “Suck-back effect” which, in a two-stage reverse osmosis system, causes membrane cleaning in the second stage and therefore increases the service life of the membrane. Suck-back means a short-term interruption of the operation at this stage, causing an osmotic backflow of permeate, which releases any surface layers from the upstream side of the membrane. This membrane cleaning takes place during the ongoing operation by the fact that a three-way valve stops the inflow before the second stage. On the basis of case studies, it was

Process water treatment Boiler feed water for steam generation as process steam, or for electricity generation, must be largely treated so that boiler scale formation and corrosion are prevented. In addition, water with low salt content leads to lower desalination losses and with it to lower water consumption altogether. After pre-treatment adapted to the raw water composition, which provides particle removal and deferrization/manganese removal and possibly removal of other contaminants, water softening or desalination is carried out. Magnus Häbel of Enviro Falk GmbH described and compared the two processes, ion exchange and reverse osmosis. The essential difference in the effect of the two processes is that, with ion exchange, the salt content is not reduced per se, but only the hard water salts are removed. These salts are concentrated in the liquid phase during steam generation, which is why blowdown must be done more often. Blowdown water is wastewater, and also the heat contained therein gets lost. On the other hand, reverse osmosis reduces the salinity of the water, thereby reducing costs and energy losses for blowdown and additional water treatment. However, also here, a concentrate results that must be disposed of. Martin Glittenberg of LAR Process Analysers GmbH focussed in his lecture on the monitoring of TOC contaminations in process water treatment and drinking water treatment. In the presence of dissolved CO2, organic matter can be oxidised into organic acids and cause corrosion. An essential criterion for the quality of the monitoring is the response time, so that the loss of high-purity water and hot water can be minimised. Moreover, a satisfactory sample speed must be implemented so that low concentrated substances, easily adhering to container and pipe walls, even arrive at the measurement location at all. As measurement methods for TOC concentration determination, on one hand there is UV absorption with 254 nm, as well as different processes with which the organic matter is converted to CO2, which is subsequently detected. These include oxidation in combination with persulfate and ultraviolet light, catalytic high temperature combustion at 680 -1,100°C and patented high temperature combustion at 1200°C. At this temperature a catalyst is not necessary, because the reaction of organic substances is complete. Moreover, the response time at about 1 min. should be shortest here. LAR AG offers such a system under the name QuickTOCpurity together with a patented calibration, which is carried out with a specified test gas, resulting in the fact that the production of an aqueous calibration standard is unnecessary.

F & S International Edition

No. 14/2014

DIACEL® organic filter aids

eco-friendly & sustainable cost-efficient process optimisation easy waste management safe & non-hazardous to health

DIACEL®- reasonable & beneficial choice for your pre-coat filtration! CFF GmbH & Co. KG Arnstaedter Str. 2 D-98708 Gehren

www.cff.de Tel.: +49 36783 882 - 0 Fax: +49 36783 882 - 252

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 40

Highlights 2013

Fig. 6: Operating principle of the VOICE® system control system for an ultrafiltration system (image: Mainsite Technologies GmbH)

shown that the yield was increased with this method and therefore the concentrate volume to be disposed of was reduced. The optimum operating parameters for the operation of a membrane system depend on the composition in the inflow. During ultrafiltration of water with fluctuating compositions, designed to help in the near future is VOICE® (Virtual Operator Intelligent Control Equipment), which is an innovative plant control system for membrane systems. This system is a Scada-System (Supervisory Control and Data Acquisition), superseding the PLC System, which was developed primarily for use in freshwater treatment by ultrafiltration, nanofiltration and reverse osmosis. Its creator is Mainsite Technologies GmbH. Dr. Marco Estermann presented

the system, which not only constantly monitors, but also processes both water and system parameters, and integrates them into the control loop of the system control (see Fig. 5). Both online measured parameters are processed, as well as values that are obtained at longer intervals in laboratory analyses. With the system, according to customer premises, the target values for parameters such as dosing concentration for flocculation aids or cleansing chemicals, cleaning intervals etc. can be adapted to the raw water composition. The control and automation of water treatment systems was also the focus of the contribution of Karl-Heinz Fischer of Siemens AG. When planning the system, the measurement and control equipment needs to be designed as well and selected

Fig. 7: Process diagram for advanced wastewater treatment with the yields indicated in the presentation of A. Eisenhofer (Grünbeck Wasseraufbereitung GmbH).

from a plurality of available products on the market. To facilitate this, Siemens publishes a consultation DVD as a planning aid, with examples of use for different system models and design examples. All products to be used are already deposited with RfP texts. With this, time and costs should be saved for system planning, the technical risk should be minimised and a higher standardisation degree for water treatment systems should be achieved. Application examples from various industrial sectors The possibilities for the realisation of process water circuits in breweries were examined by Grünbeck Wasseraufbereitung GmbH. Armin Eisenhofer presented the types of wastewater incurred there, the amounts per hl of beer, their characteristics and the principle of more extensive wastewater treatment as is outlined in Fig. 6. Here, Grünbeck handles the treatment part behind the membrane bioreactor. From the discussion, the question arose as to why another ultrafiltration plant is connected downstream of the membrane bioreactor. The extension of the service life of the reverse osmosis membranes that can thus be achieved was named as a reason for this. Reverse osmosis reduces the conductivity of the purified water, so that, inter alia, it can also be used in CIP systems and as boiler feed water.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 41

Highlights 2013

Fig. 8: Diagram of the LOPROX method (image: Bayer Technology Services)

The LOPROX method (Low Pressure Wet Oxidation) for the treatment of highly contaminated, toxic wastewater streams was already introduced in the 1980s by Bayer and has since been continuously further developed. Dr. Johannes Leonhäuser, Bayer Technology Services, reported about the largest ever facility at a manufacturer of pesticides in Israel. The method outlined in Fig. 7 is a wet air oxidation taking place at moderate temperature (up to 200°C) in a multi-stage bubble column, which is catalysed by Fe2 + -ions and quinone-forming substances. Poorly biodegradable and toxic substances are degraded to the extent that they can be treated in a downstream biological wastewater purification treatment. In the discussion, the suitability of this and similar processes was also addressed for the treatment of concentrates after membrane filtration. In the automotive industry, different compositions of wastewater are produced at various production processes. For Volkswagen AG, Rüdiger Eppers reported about the history of the development of process water treatment and the comprehensive programme “Think Blue. Factory.”. The latter has the objective to reduce the environmental impact of all VW plants by 25%. For this purpose, energy consumption, the amount of waste, CO2 emissions and solvent emissions are also looked at, besides the water consumption. The improvement measures implemented in the different plants during recent years are also adapted worldwide to the circumstances of the respective locations, inter alia, with rainwater usage and a reforestation project in the low water environment of the factory in Puebla, Mexico. The engine production plant in Salzgitter is a zero industrial water discharge factory and in the new factory in Foshan (China), the largest MBR system in an automobile factory worldwide is being created.

The contribution by Günter Müller-Czygan, HST Hydrosystemtechnik, was about the production-adapted system control in industrial wastewater treatment and the importance of information flow from the customer to the planner and, if applicable, also to the component supplier. This also includes comprehensive data collection at the customer before system planning. This should be an iterative process, wherein the final process solution is developed from the first draft through constant comparison with expected production aspects. This also includes that any “failures”, such as for example a sudden change of the composition of the inflow, are looked at as an operating mode to be taken into consideration in the system control. In this context, the wish was submitted to the VDMA to carry out training courses on the subject of system engineering according to ISO 15288. Dr. Klaus Wolfer, Würth Elektronik GmbH & Co. KG, described the mode of action of wastewater treatment in PCB manufacturing with an economical, especially successful recyclables recovery. Printed circuit board manufacturing in itself is a very complex process with a large number of wet process steps. In the Würth Elektronik Niedernhall factory, about 200-300 m3 of wastewater result per day for the production of 700 - 1000m2 of printed circuit boards. According to the state of the technology, the various effluents are purified in partial streams in batch treatment plants, because the different wastewater ingredients in part require different methods. In detail, a method was described for the simultaneous recovery of etching solution and metallic copper after alkaline corroding (MECER method). The etching solution contains ammonium chloride and copper chloride. For an optimum etching process, copper-free solution is replenished, the density is regulated and saturated etching solution is withdrawn. From this, copper is transferred through liquid-liquid extraction into an organic phase, extracted into dilute sulphuric acid and from there is deposited electrolytically as pure metal. With the present copper prices, the relatively complex treatment plant will have amortised after 1 to 1.5 years.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 42

Highlights 2013

Nanotechnology in practical water application: Efficient solutions for the future Report of the event nano meets water IV at Fraunhofer UMSICHT H. Lyko* At the end of 2012, scientists and professionals from the industry met in the Fraunhofer Institute for Environmental, Safety, and Energy Technology UMSICHT in Oberhausen to discuss new developments, opportunities and risks of nanotechnology for practical water application. In this, it was not only about scientific discoveries and technological developments, but also about the right market launch. For as the Deputy Head of the Institute, Prof. GĂśrge Deerberg, pointed out by way of introduction, the market launch of technologies that involve risks only succeeds if one discloses and explains these risks to society. 1. New membranes for industrial uses The concept of nanotechnology in the treatment of water also always implies the development of nanostructured membranes or of membranes equipped with nanoelements. Prof. Claudia Staudt of the BASF Chemical group, which for decades, inter alia, has dealt with the development and production of typical membrane polymers, showed more recent developments of new polymer membranes with defined nanopores or integrated nanoparticles. For instance, already in 2007, in the Helmholtzzentrum Geesthacht, a new method for the production of membranes was developed, where a thin separation-active layer forms within a few seconds by selforganisation from amphiphilic block copolymers from styrene and 4-vinylpyridine, so that a very regular structure results, made up of parallel, cylindrical ducts, oriented perpendicularly to the surface with an effective pore width of about 8 nm /1/. The spongy understructure originates from subsequent phase separation in a non-solvent (water). At university of Minnesota, USA, a mixture of polymers was polymerised in the presence of a molecule that, although also being soluble in the monomer solution, is not, however, compatible with the forming polymer. From the cross-linked, uniformly structured layer originating from microphase separation, the non-compatible material is extracted again by means of a chemical etching process. Continuous nanopores originate from it /4/. At University of Alberta in Canada, thin film composite membranes were produced for water uses, *Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

where multi-walled carbon nanotubes (MWNT) treated with acid are integrated in the substructure and silver nanoparticles are integrated in the thin separation-active layer. The MWNT in a mass concentration of 5% cause a good 20% higher water permeability and the silver nanoparticles cause a reduction of the adsorbing tendency for bacteria so that the danger of biofouling is limited /3/. Thin film composite membranes made from polyamides, on a supporting layer made from polysulfone, in which nanoscale zeolite particles are placed in the separation-active layer, were introduced in 2010 in the market for sea and brackish water desalination by the Californian enterprise nanoH2O /4/. 2. Nanotechnical materials and methods Many technical innovations in the areas of nanotechnology are copied from nature, or one uses, as described below, natural phenomena in a technological manner. An example of this is given by types of bacteria, which have a specific, grid-like cell wall structure, so-called S-layer. In the Helmholtzzentrum Dresden Rossendorf (HZDR), Dr. J. Raff and employees examined the microbiological diversity in a uranium rubbish waste dump in order to ascertain interaction of bacteria with the uranium. There, one also found bacteria that have layers made from proteins and polysaccharides or glycoproteins, whose sub-units are assembled into a tetra or hexagonal structure. In the HZDR, one can make these structures visible with the help of transmission electron microscopy (see Fig. 1). With bacteria, the S-layer apparently serves as the protection against aggressive environmental conditions. They

are able to hold back certain ions or molecules or to bind certain enzymes to themselves. One wants to take advantage of these special characteristics in water treatment by immobilizing such bacteria on support materials. For instance, metalselective filter materials, metal or metaloxide catalysts or biosensors can be implemented with this. An example is the binding of TiO2 or ZnO nanoparticles with the help of an S-Layer on a carrier, for example, an expanded metal mesh, in order to carry out photocatalytic water purification with this. At the symposium â&#x20AC;&#x153;nano meets waterâ&#x20AC;? III, Volkmar Keuter from UMSICHT reported about the modification of metallic microsieves with TiO2 and silver nanoparticles for the combination of filtration and photo-catalytic degradation of dissolved water contaminants /5/. The works that were carried out within the scope of the NanoPurification project have led to the construction of a pilot system. Dr. Christoph Bohner, from the project partner Enviro Chemie, confirmed that the manufacturing of the microsieves, as well as technologies for coating with nanoparticles, have meanwhile matured. The first combination modules that incorporate the microsieve and in which LEDs supply the ultraviolet light for the photo-catalytic degradation, have been built and are in the testing phase. The system shown in Fig. 2, which is also still in the testing phase, was designed for a permeate flow of about 200 l/h. 3. Nanostructuring of surfaces Nanostructured surfaces which find application in water technology may have different properties. As Dr. Ilka Gehrke, UMSICHT, pointed out, chemically active

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 43

Highlights 2013

as well as chemically inactive surfaces are possible. The chemically inactive surfaces cause, for example, an increased wear protection, reduce friction, have a water or oil repellent effect or offer a mechanical barrier against corrosive attacks. Chemically active surfaces can have absorptive or catalytic properties or kill bacteria. Various nanostructured surfaces are now being implemented in marketable products, including in waterproof surfaces on airplane wings or wind turbines to reduce the tendency for ice to form, in anti-friction surfaces (shark skin), for wall paints that mimic the lotus effect, but also in textile filter media coated with nanoparticles. The latest developments at UMSICHT play an important role with the availability of microsieves, which were originally used for filtration or emulsification, or can be modified with nanoparticles, as mentioned to begin with. This is a new production method by means of ultra-short pulse lasers through which the microstructures, some of which are smaller than 1 μm, can be produced in relevant processing sizes and process times. Fig. 3 shows a microsieve made in this manner. The essential effect of the greatly reduced pulse duration of the laser is because no heat conduction occurs anymore into the surrounding material and thus also no material is damaged in the environment of the treatment zone. Smaller dimensions having a high surface quality can thereby be implemented. Because the material erosion happens by evaporation, the treatment of materials is also possible that up to now could not be processed with laser systems or only badly so. Also, composite materials can be “drilled” without influencing the texture. Axel Rosenhahn, professor for analytic chemistry at Ruhr university of Bochum, is researching with his group the interactions between technical surfaces and marine organisms and cells. From this, measures will be derived for the containment and/or monitoring of the biofouling on ship hulls and other surfaces located underwater. Marine organisms have different strategies “to select” surfaces for colonisation and to stick to these. An overview of these mechanisms, illustrated in detail on zoospores, which are the asexual propagules of many algae and barnacle larvae, is given in the freely available article /6/. The colonisation mechanisms were precisely analysed, for example, with holographic 3D images. Inter alia, different patterns of movement of organisms in the vicinity of the surface or on it were revealed, which depend on the surface characteristics. Fouling minimisation can happen in three ways: by a toxic coating which kills the organism when it strikes the surface, by a surface modification which hinders adhesion and/or by a surface on which the adhesive forces are so low that deposited organisms are detached again immediately by flow forces (the motion of a ship). With the latter one speaks of self-cleaning surfaces. Nanostructures and microstructures can prevent the settlement of organisms, as well as facilitate their detachment. Thin films with such surfaces were achieved by layered spraying of oppositely charged polyelectrolytes (polyacrylic acid and polyethylenimine) (see /7/). By varying the pH value, the structures have been varied and by chemical modification with polyethylene glycol and tridecafluorooctyl triethoxysilane, the wettability is changed. The structures that one generates in this way resemble those of the skin of whales. It showed that colonisation is made particularly difficult by structures whose dimensions are scarcely below those of the colonizing organism, while especially easy detachment is caused by nanostructures. The latter is also caused by so-called hydrogels that form a micrometer-thin layer between the surface to be protected and the sea water. 4. Risk assessment and toxicity analysis The Federal Environment Agency in Dessau is involved in various fields with the analysis and evaluation of nanotechnological processes and products. Dr. Wolfgang Dubbert presented excerpts from recent national and international research projects in which

F & S International Edition

No. 14/2014

A: The bacterium Lysinibacillus sphaericus JG-A12 on a scanning force microscope image. The diagram shows the square array of its S-layer proteins on the cell surface. B: Isolated S-Layer in the light microscope C: Transmission electron microscopy (TEM) image of the S-Layer. The small image shows a computed reconstruction of the protein grid. D: Diagram of the S-layer of Lysinibacillus sphaericus JG-A12 Fig. 1: Structure of the bacterial S-Layer (image: Helmholtzzentrum Dresden-Rossendorf)

the sustainability of nanotechnology is examined and/or is supposed to be partially funded. In the recently completed Project Sustainability Check (for details see /8/), an instrument was developed with which, within the scope of an internal self-evaluation, risks and challenges for the market launch of nanoproducts can be identified early on. In this, for certain case studies, a variety of key factors, which include for example, the CO2 footprint, the life cycle costs, a risk assessment, but also societal factors such as the possibility of research funding, are entered into a SWOT matrix. At international level, an OECD Working Party on Nanomanufactured Materials has been formed, which also examines and evaluates known applications of nanomaterials according to a similar pattern. For the consideration of the toxicity of nanoparticles, multiple studies on absorption through the respiratory tract have already been carried out, namely in vivo, i.e. in creatures (very often in rats, see also/9/) as well as in vitro, namely on macrophages. Macrophages are the cells in the respiratory tract that absorb these nanoparticles. IBE R&D GmbH in Münster, represented by Prof. Martin Wiemann, deals with the biological-toxicological evaluation of micro-scale and nanoscale particles, namely with airborne ones as well as with nanoparticles in the aquatic environment. For the investigation of the toxicity of water contaminants, there are a number of tests on algae, daphnia (water fleas), duckweed, fish and spawn, whose suitability was considered for studies with nanoparticles. Thus, the survival rate of daphnia and zebrafish steeply decreases with increasing concentration of nanosilver, nanocopper and cerium oxide. With daphnia, one observes a hindering effect on skinning with the offspring in the presence of Nano-TiO2, because the nanoparticles deposit on the outer skin of the animals /10/. 5. Nanotechnically optimised components “Membranes with pores like pearls on a string”, this was the promising characterisation of microsieves at a membrane conference some years ago. The extremely narrow pore size distribution in connection with low material strength and low flow resistance promises varied application possibilities and a drastic reduction of energy requirements and space requirements compared with customary membrane systems. However, in practice microsieves have prevailed in only a few applications. On one hand this is due to the fact that their pores are still more than 1 micrometer and that their

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 44

Highlights 2013

Fig. 2: Pilot plant for combined water filtration with photocatalytic degradation of pollutants (image: EnviroChemie GmbH)

Fig. 3: Microsieve produced by ultra short, pulsed laser (photo: Fraunhofer UMSICHT)

production by moulding or a photolithographic process is still relatively expensive. Prof. Werner Goedel of the University of Technology of Chemnitz showed how one produces microsieves from a hardenable oil with submicron pore diameters. The process during production that has already been reported on here in the magazine /11/ is shown schematically in Fig. 4. Silica gel beads were used as particles. While the advantage of the sieves made in this manner is in the submicron pores that are very closely juxtaposed, the disadvantage is in the low mechanical stability of the layer that is only about 300 nm thick. To address this problem, the microsieves produced according to the principle shown in Fig. 4

Literature: /1/ Peinemann, K.-V.; Abetz, V.; Simon, P.F.: Asymmetric superstructure formed in a block copolymer via phase separation; nature materials Vol. 6 (2007), 992 – 996; DOI: 10.1038/nmat2038 /2/ Seo, M.; Hillmyer, M.A.: Reticulated Nanoporous Polymers by Controlled Polymerization-Induced Microphase Separation; Science 336 (2012), 1422 – 1425. DOI: 10.1126/science.1221383 /3/ Kim, E.-S.; Hwang, G.; El-Din, M.G.; Liu, Y.: Development of nanosilver and multi-walled carbon nanotubes thin-film nanocomposite membrane for enhanced water treatment; Journal of Membrane Science 394-395 (2012), 37 – 48. DOI: 10.1016/j.memsci.2011.11041 /4/ Kurth, C.J.; Burk, R.L.; Green, J.: Improving Seawater Desalination with Nanocomposite Membranes; IDA Journal, 3rd quarter 2010, 26 -31 /5/ Lyko, H.: Nanotechnologie in der Wassertechnik – report from the III symposium “nano meets water” at Fraunhofer UMSICHT; F&S Filtrieren und Separieren 26 (2012) No. 1, pp 11- 15 /6/ Rosenhahn, A.; Sendra, G.H.: Surface Sensing and Settlement Strategies of Marine Biofouling Organisms; Biointerphases (2012) Vol.7, 63; DOI 10.1007/s13758012-0063-5 /7/ Cao, X.; Pettitt, M.E.; Wode, F.; Arpa Sancet, M.P. Fu, J.; Ji, J.; Callow, M.E., Callow, J.A., Rosenhahn, A.; Grunze, M.: Interaction of Zoospores of the Green Alga Ulva with Bioinspired Micro- and Nanostructured Surfaces Prepared by Polyelectrolyte Layer-by-Layer SelfAssembly; Advanced Functional Materials 2010, 20, 1984-1993. DOI 10.1002/adfm.201000242 /8/ Möller, M.; Groß, R.; Moch, K.; Prakash, Pistner, C.; Küppers, P.; Spieth-Achtnich, A.; Hermann: Analyse und strategisches Management der Nachhaltigkeitspotenziale von Nanoprodukten – NachhaltigkeitsCheck von Nanoprodukten, final report Ecological institute texts 15/2012, available as a download at http://www.uba.de/uba-info-medien/4276.html /9/ Lyko, H.: Nanomaterialien am Arbeitsplatz: Herausforderung für den Arbeitsschutz?; F&S Filtrieren und Separieren 25(2011) No.4, pp 230 – 232 /10/ Dabrunz, A.; Duester, L.; Prasse, C.; Seitz, F.; Rosenfeldt, R.; Schilde, C.; Schaumann, G.E., Schulz, R.: Biological Surface Coating and Molting Inhibition as mechanisms of TiO2 Nanoparticel Toxicity in Daphnia magna; PLoS ONE 6(5) 2011, e20112. DOI: 10.1371/journal.pone.0020112 /11/ Goedel, W.: Neue Art von Mikrosieben; F&S Filtrieren und Separieren 26 (2012) No. 4, 253 /12/ Goedel, W.: Von partikelassistierter Benetzung zu porösen Membranen; Nachrichten aus der Chemie 54 July/August 2006 /13/ Yan, F.; Ding, A.; Gironès, M.; Lammerding, R.G.H.; Wessling, M.; Börger, L.; Vilsmeier, K.; Goedel, W.: Hierarchically Structured Assembly of Polymer Microsieves, Made by a Combination of Phase Separation Micromoulding and Float-Casting; adv. MAterilas 24 (2012), 1551 – 1557. DOI: 10.1002/adma.201104642

were applied onto coarser microsieves so that this compound is stable enough for further handling /13/. In the Fraunhofer Institute for Applied Solid State Physics, micro-sensors and nanosensors are developed for a variety of measurement tasks in nature conservation and environmental protection, in the gas and fluids analysis field and for medical diagnostics. Björn Albrecht presented narrow-band UV detectors for the monitoring of UV lamps on the basis of aluminium gallium nitride (AlGaN). These semiconductor detectors are highly sensitive in the UV spectral range and the band gap, i.e. the wavelength range in which light is absorbed, can be regulated via the aluminium proportion. Such sensors are used, for example, for spectrally resolved monitoring of UV lamps for water disinfection. 6. Nanotechnology from the point of view of the investor

Fig. 4: Diagram of the production of porous membranes via particle-assisted wetting, hardening of the oil, transferring the composite layer and removal of the particles (according to /12/)

Nanotechnology is a branch of industry with a huge market increase potential. According to the analyses of different institutes, the market volume of products manufactured using nanotechnology has increased from about 40 billion US$ in 2000 to about 250 billion US$ in 2010, and for 2020 an increase is expected to about 3 trillion US$. This development is reason enough for a specialised sponsor like Nanostart AG to invest in young enterprises using and further developing such technologies. Dr. Hans Joachim Dürr explained what the advantages are in financing young enterprises with venture capital rather than credits and what risks still exist for the investor. Venture capital is equity capital, i.e. the investor becomes co-owner of the enterprise. There is thus an urgent interest in the economic success of the enterprise. Therefore, before an investment decision, the venture capital investor judges not only the technology and the market to be opened up, but, above all, also the management of the enterprise to be supported. The portfolio of Nanostart AG also includes, inter alia, the specialist company for nanoceramic flat membrane systems, ItN Nanovation AG.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 45

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 46

Highlights 2013

Status and perspectives of organophilic nanofiltration

H. Lyko*

Separation processes in the chemical and pharmaceutical industry make up a high proportion of plant and operating costs for production processes; this is true for water-based substance systems, as well as for those that are based on organic solvents. While membrane technology is established for aqueous substance mixtures and its advantages are indisputable compared with thermal separation processes, particularly with respect to energy consumption and space requirements, membranes and membrane separation processes in conjunction with organic solvents are often still in an early stage of market development. Especially interesting for a large number of separation tasks, that up to now can only be handled thermally, is the nanofiltration of organic solvents. At the 4th International Conference on Organic Solvent Nanofiltration (OSN) /1/, that was held in March 2013 by the Department of Chemical Process Engineering (Aachener Verfahrenstechnik) and the DWI (German Wool Research Institute, now the Institute for the Development of â&#x20AC;&#x153;interactiveâ&#x20AC;? materials) of the Aachen RWTH under the direction of Prof. Matthias Wessling, about 60 participants discussed the newest results from research and development, as well as experiences with industrial processes of OSN. In the course of this, it seems that all research institutes in Europe specializing in this area were represented, as well as scientists from Russia, Asia and the USA, and also industrial enterprises active in this sector. 1. Introduction

Tab. 1: Overview of different membrane polymers that are suitable for OSN (Basic structures, partly still cross-linked by specific molecules at the time of membrane production)

A big potential attributed to organophilic nanofiltration with separation limits in the range from about 150 to 1000, is to overcome the previous limits of separability of substance mixtures, to regenerate and clean solvents, as well as to carefully extract temperature-sensitive recyclables. A number of industrial processes were already able to be implemented during the last few years. The basis for them is formed by a range of commercially available membranes and module configurations. However, in contrast to many applications with aqueous substance systems, OSN systems are never to be looked at as turnkey systems. This, above all, was emphasized by Youri Bouwhius, Andrew Boam and Axel Kobus of Evonik Industries AG in their introductory lecture. Differently from the water market, it is impossible to act simply as a supplier for OEMs without having contact with end customers oneself. Rather, the enterprise gets involved itself in the development of a solution for special separation tasks. This is also important in the respect that organophilic nanofiltration is not just replacing another mostly thermal separation process, but often also means a new process procedure and/or obtaining new products. Up to now, 32 different applications for OSN have thus been identified and of these, 20 have been judged as commercially viable. At the time of the lecture, three of these were in the technical stage, eight in the laboratory or pilot phase and the remaining nine were being looked at as case studies. To persuade an end customer of the advantages of a process transformation, or the production of a new product, as well as also to implement this, then *Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 47

Highlights 2013

Fig. 1: Hybrid NF membrane for organophilic nanofiltration - ceramic coating (approx. 200 nm) on a robust supporting structure made of polymer (image: ECN, SolSep BV, HybPON-Konsortium NL)

Fig. 2: Test cell for performing spectroscopic ellipsometry under filtration conditions (image: Twente University see /7/)

requires time and also a different business model than that of being a pure component supplier. Therefore, the benefit from the introduction of OSN, on one hand through saving energy and material compared with thermal processes, and on the other hand as increased added value through possibly better products, should also be reflected in attainable prices for the membranes and modules. Besides the development, production and characterisation of new membrane materials and the realisation of new processes, many presentations were on the agenda in this conference that also dealt with methods of advance calculation of separation efficiencies, membrane selection and process design.

generated PEEK variations, with different solvent compositions for the drawing solution, and with different crosslinkers and other parameters of phase inversion, membranes were produced and characterised with different methods. Inter alia, a clear influence of the solvent combination on the thickness of the separationactive layer, the permeability and the retention of the test substance rose bengal (a dye) was found. Polymers with intrinsic microporosity have solid molecular scaffolds, whose interstices can be changed neither by mechanical compression during filtration, nor by temperature influences. One example of this is the polymer shown in Table 1, PIM-1 (PIM stands for: polymer of intrinsic microporosity), with which

2. Membrane materials and their properties The number of solvent-stable membrane materials available so far in industrial orders of magnitude is relatively limited. Indeed, in addition, various substance classes or modifications were still presented that, up to now, were manufactured and characterized on a laboratory scale and have been tested in filtration tests. Some polymers form the basis of the most important membrane developments for OSN, which, as an overview, are shown in Table 1 in their basic structure, most notably polyimides and polydimethylsiloxane (PDMS). Thus, the products by Evonik, Duramem and Puramem, which were also described several times at this event, consist of such polymers. Duramem is an ASN (asymmetric integrally skinned) membrane made from polyimide. Puramem is a composite membrane with a silicone layer. The Borsig group, too, to which GMT Membrantechnik GmbH belongs, uses siliconebased composite membranes for organophilic nanofiltration. Katrin Ebert showed this in her report on the activities of GMT in this area. However, in her explanation, module selection also played an important role. Spiral wound modules and pocket modules are being used. On the question of a performance and cost comparison of the two types of module, the pocket module was shown to be the more expensive one, but possibly better suited for higher flow rates. Vitreous polyetheretherketone (PEEK) is also considered to be especially chemically stable. Gas separation and fuel cell membranes are also made from this. Katrien Hendrix of the Centre for Surface Chemistry and Catalysis at KU Leuven has produced membranes from polymers of this construction type and has characterised them. Here, the standard polymer, which is produced by polycondensation of difluorobenzophenone and hydroquinone, was modified to make it soluble for the manufacturing process of OSN membranes. Thus, the monomer hydroquinone was replaced by another one from the group of diphenones. The molecule illustrated in Table 1 contains tertiary butylhydroquinone. With the

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 48

Highlights 2013

Fig. 3, a: Isolines of permeability for three different solvents, a so-called Membrane Permeability Map (MPM) for commercial OSN membrane Starmem 122 (image: University of Technology of Dortmund, Department of Biochemical and Chemical Engineering, Institute of Fluid Process Engineering)

Patricia Gorgojo et al., from Imperial College in London, experimented. The essential advantage of intrinsic microporosity (in accordance with IUPAC, pores less than 2 nm are understood by this) exists in the clearly increased permeability in comparison with customary TFC membranes. Extraordinary stability in organic solvents, as well as in highly corrosive environments, is also displayed by Polybenzimidazole. This type of membranes was also made and tested in Imperial College under the direction of Prof. Andrew Livingston and a patent was applied for /5/. Silsesquioxanes (Organosilicates) unite the stability of silicates with the flexibility of rubbery polymer. Toshinori Tsuru et al. from Hiroshima University make membranes from materials that can be represented by the sum formula RSiO3/2. The type and length of hydrocarbons hiding behind the abbreviation “R” cause the flexibility / strength of the originating structures. If alkanes are used, one receives flexible, dense membranes; when inserting aromatic compounds, rigid porous structures originate. One material type from this group, from which one expects superior properties, are so-called polyhedral oligomeric silsesquioxanes (POSS, see Table 1). How to produce ultrathin POSS-polyamide films from these building blocks is described in the thesis of M. Dalwani from Twente /4/. In Aachen, Yali Zhang, University of Technology of Twente, presented a process in which these films are implemented in situ in microchannels. They are supposed to be used for solvent exchange in microfluidic systems. In the course of this, the selectivity of these membranes is not as important as their permeability and their mechanical stability, since the solvent exchange takes place via a displacement process.

Fig. 3, b: Isolines for the retention of Phenyldodecane in a ternary solvent mixture (Membrane Rejection Map (MRM)) by the OSN membrane Starmem 122 (image: University of Technology of Dortmund, Department of Biochemical and Chemical Engineering, Institute of Fluid Process Engineering)

The Dutch company SolSep BV, under the trade name HybSi®, produces pervaporation membranes which consist of a microporous layer on a ceramic substrate (as capillaries). For OSN, Managing Director Peter Cuperus presented a new concept for hybrid membranes made from a polymer substrate with an approx. 200 nm thick ceramic coating (see Fig. 1), which also allows for the production of flat sheet membranes and with it for the assembly to spiral-wound modules. With purely ceramic membranes the application limits still to be overcome lie in the achievable separation limit and in the hydrophilicity of the ceramics, so that they must be modified for sufficient wetting with organic solvents. As Ingolf Voigt of the Fraunhofer Institute for Ceramic Technologies and Systems (IKTS) demonstrated, commercial ceramic membranes with separation limits of 450 have been available there for ten years. Hydrophobization of the ceramics is achieved by modification with silanes. However, their molecular dimensions are so large that one does not take nanofiltration membranes as a starting material, but instead one takes ultra-filtration membranes in order to allow sufficient penetration of the silanes. The lower separation limit of a silanised UF membrane is approximately 600 Da. More recent developments within the scope of the nano-membrane project, in which, inter alia, the Merck chemicals company is also involved, aim at the production of ceramics by means of Sol-gel processes and in situ hydrophobicisation. The new membranes have a substructure made from TiO2, an intermediate layer made from ZrO2 as well as a separationactive layer made from both oxides. The production is done using complex-forming agents, which support gelation and also have an influence on the freedom from

defects and therefore also on the real separation limit. Membranes were produced which, in aqueous solutions, presented MWCOs of 200 - 300 Da, and the cut-off of the finest membrane in organic solvents was approx. 350 Da. The recoverable retention of organic substances depended strongly on the solvent used. As a special, also mechanically very stable class of membranes, Santanu Karan (Imperial college of London UK) described diamond-like, ultra thin hydrocarbon layers (10 - 40 nm, called DLC layers (diamond-like carbon)) on porous ceramics, which allow a high, viscositydependent flow of organic solvents with concurrent retention of dissolved organic molecules (for details see /6/). The ultrathin layer is applied by plasma deposition of a vaporous hydrocarbon, diluted with an inert gas. Here, not a pure carbon skeleton is formed, but a self-supporting amorphous layer of very regular and highly cross-linked organic structures with a high carbon content, between which there are hydrophobic nanopores with a total porosity of about 12%. Before the plasma deposition of the hydrocarbon, the substrate is covered with an approx. 80 nm thin layer made of cadmium hydroxide nanofibres, which is subsequently dissolved again, using an ethanolic solution of hydrochloric acid. So that the nanofibres are not damaged during the plasma coating, the coating takes place at a temperature of -20°C. 3. New method of membrane production A large part of the membranes used for OSN are thin-film composite membranes in which the separation-active layer is deposited on a porous substructure, for example a UF membrane. A common process for applying the separation-active layer is interfacial polymerisation. For

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 49

Highlights 2013

4. Characterisation of membrane materials One of the most important interactions between solvent and polymer membranes that affect the separation behaviour is the swelling behaviour, i.e. the absorption of solvent by the membrane material. The more fluid the membrane absorbs, the more flexible it will become and with this, stability, permeability and retention capacity change. The swelling behaviour also depends therefore on the membrane thickness. Nieck Benes, University of Twente, demonstrated in his Keynote Speech how this behaviour is analysed with the help of spectroscopic ellipsometry. With spectral ellipsometry, thickness, refractive index and surface roughness of thin layers of material can be measured in a contactless manner. In the present case, this measurement method was applied in a special test cell under filtration conditions, i.e. under increased pressure with permeation through the membrane. The course of the thickness across temperature and hence the glass transition temperature of the polymer used can be determined by heating the film to be examined. Through the course of the change in thickness in the presence of various solvents, through the time and depending on the pressure, mechanisms of transport through the examined membrane are illustrated. An exact description of the method and the results obtained with the permeation of n-hexane by means of differently cross-linked PDMS films is provided in /7/. The sorption capacity of membranes is also measured using gravimetric steam sorption measurements or by gas chromatography. The latter means the exposure of the material to a test gas mixture and the analysis of the transmitted components with a flame ionisation detector. Thomas Schmid, as a representative of the Measurement Systems manufacturer Surface Measurement Systems Inc, presented measurements on a poster that had already been carried out with different membranes. One can determine the essential features of membrane permeability and selectivity only in filtration tests. Here, there is the endeavour to achieve a general comparability of materials. Patrizia Marchetti et al., from Imperial College in London, inter alia asked the question about the feasibility of a “Robeson plot” for OSN membranes. The original Robeson plot illustrates the selectivity of gas separation membranes in mixtures of two gases depending on the permeability of the faster permeating gas /8/ and has an obvious upper limit of the achievable performance numbers. For OSN membranes, a reasonable number of common solvents and solutes such as polystyrene, linear and aromatic alkanes, as well as dyes, were suggested, as well as possibilities for a uniform mode of representation. Analogous to the representation of gas separation

F & S International Edition

No. 14/2014

membranes, the application of selectivity (solvent compared with solute) over the permeability is an option for the solvent. In this representation, one can put either different solvents, different solutes or different membranes in a diagram and compare them. 5. Membrane selection, modelling of the separation performance and process design Because of the huge number of different solvents and the complexity of the interaction between solvents and membranes, the pre-calculation of typical data on process design is difficult. Thus, methods are investigated in different projects with which, on the basis of a limited number of measured values, one can make statements about expected separation performance. In TU Dortmund, together with the chemicals company Merck, a study on membrane selection was carried out by means of heuristic methods. As Stefanie Zeidler explained, series of measurements involving selected solvents and solutes were carried out, which were supposed to be sufficiently representative that one can judge the suitability of OSN for a separation task problem to be solved, given knowledge of easily accessible parameters of the substances involved. As a graphic means for the assessment of permeability and the retention capacity of membranes, Patrick Schmidt et al. (also TU Dortmund) created ternary diagrams. The graphics illustration shown in Figures 3 a and b, which depict the results of measurements on the commercial membrane Starmem™122, illustrate a clear dependence of the retention of the dissolved component on the composition of the solvent mixture. With the help of the isolines, which can also be represented as selectivities (Membrane

w w w . w e i s s e . d e

this, the substrate is saturated with an aqueous solution of the monomer concerned (possibly after pretreatment for the improvement of wettability), and afterwards the system is immersed in the organic solution containing the second reactant of the polymerisation. At the interface of the two immiscible solvents, the two monomers meet with each other and polymerise in a very thin layer. For this production process, simpler and faster alternatives are sought. For instance, Elke Dom of the Catholic University of Leuven in Belgium suggested combining the production of the UF membrane used as a substrate with the first step of the wetting with the aqueous solution. In the examined case, the last step of the manufacture of a UF membrane was carried out by phase inversion with an aqueous solution, which at the same time already contained the monomer for interfacial polymerisation. Besides the saving of a production step, one expects from this method also an easier and more complete impregnation of the substructure with the monomer, because it is precisely already present in the coagulation bath. If the substructure consists of solvent-stable, crosslinked polyimide, then the crosslinkers are identical for the substructure and interfacial polymerisation.

Weisse & Escherich Innovation + Precision

= Expertise in Wire Mesh for:

■

Machine-building Screen printing industry ■ Automobile industry and component suppliers ■ Chemical industry ■ Electronics industry ■ Plastics industry ■ Oil pumping industry ■ Aerospace ■ Radiation protection ■ Food industry ■ Recycling and enviromental protection ■ Tire industry ■

Weisse & Escherich GmbH & Co. KG · Wire Mesh manufacturers P.O.Box 1261 · D-96334 Ludwigsstadt Phone +49-92 63/9 46-0 · Fax +49-92 63/9 46-40 info@weisse.de · http://www.weisse.de Ask for our free leaflet.

02_fs_international_2014_seite_03-98__ 21.05.14 20:10 Seite 50

Highlights 2013

transport models (solution permeability model, pore flow model or another phenomenological model that describes the permeation and the retention by the membrane), as well as models for the process procedure (batch operation, semi-batch diafiltration, stationary continuous filtration) had to be integrated in the simulation surroundings. As an application example, the concentration of rosemary essence in ethanol was named among other things /10/. Here the total energy consumption of a membrane system could be determined as a function of reconcentration and compared with respective values for distillation in order to assess which processing is more economical with what concentration. 6. Phenomenon negative retention Fig. 4: Test facility for the characterisation of OSN membranes (by courtesy of SIMA-tec GmbH, Hürth)

Fig. 5: Process diagram for hydroformylation with micellar systems, followed by an OSN for purifying the organic phase (image: Department of Process and Chemical Engineering of the University of Technology of Berlin)

Selectivity Maps, MSM, not shown here), one can select the solvent composition suitable for a desired filtration result (see also /9/). Patrizia Marchetti ventured to take on the forecast of permeability for different solvents by means of OSN membranes with the help of a relatively simple model. Thus the viscosity is identified as an important parameter of the solvents used. According to the Hagen Poiseuille model for a viscous pore flow, there is proportionality between the permeability and the inverse of the viscosity. Miscellaneous correlation factors dependent on the pore size consider surface effects in the nanopores, and the source behaviour of flexible polymer membranes is taken into account through the Hansen solubility parameters (parameters calculated from the energy densities of the different intermolecular interactions). Other membranespecific model parameters were determined by permeability measurements with from four to five different solvents in each case. With the developed model, correspondence could be received successful-

ly between measured and calculated flow rates through ceramic as well as through polymer membranes. The permeation behaviour of solvents through ceramic membranes is described by Anita Bueckenhoudt of VITO, Belgium, by a phenomenological model. Here, the flow rate of an organic solvent through the membrane, its viscosity as well as the overall Hansen solubility parameter are brought into context with the corresponding data for the solvent water, namely with the interposition of a membrane-specific parameter that depends on pore size and the hydrophilicity / hydrophobicity of the membrane. With this phenomenological model, one should be able to calculate the flow rate of any solvent or mixture through a membrane if the substance data and the permeability for water are known. Dimitar Peshev of Imperial College London presented his simulation tool “OSN Designer” with which the performance of a membrane system can be simulated by using commercial simulation programs like Aspen One, MATLAB and CAPE OPEN. To achieve this, existing

Stefanie Postel, RWTH Aachen, showed that the retention of dissolved components by a certain OSN membrane depends on the affinity of the solute, on one hand relative to the membrane, but also relative to the solvent (mixture) used. In certain cases, it comes even to negative retention of solutes. This was proved with the help of the three solvents toluene, methanol and isopropanol with different solutes to be retained. Therefore, negative retention can appear if the solubility of a component in a solvent is especially high. Here, negative retention of carboxylic acid in methanol and of n-alkanes in methanol and isopropanol were measured. The technical usability of this phenomenon was discussed, but also the objection was brought forth that, in such a case, classical extraction is the more suitable separation process. 7. New processes and products with OSN The development of hybrid membranes at SolSep described in section 2 happened within the scope of the EU project SOLVER (Solvent Purification and Recycling in the Process Industry Using Innovative Membrane Technology) with the objective to clean solvent streams of various industries and return them to respective processes. As project coordinator Peter Vandezande, of the Belgian research institute VITO, presented, about 85-90% of material consumption in the production of a pharmaceutical active ingredient is attributable to a solvent. For the electronics industries, the market volume lies at about 2 billion €/a for the solvent requirement. Here alcohols are used mainly for cleaning and conditioning of components, whose content in metals after the treatment must be below 10 ppb.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 51

Highlights 2013

Meanwhile flat sheet membranes were tested on different scales and the retention of a total of about 35 metals from alcohols was determined. A pilot plant for 4” spiral wound modules or for ceramic modules allows nanofiltration with up to 45 bar pressure at a maximum of 50°C. With optimum setting of the operating parameters, as high degrees of purity of the solvents could be achieved as are required in the electronic industry. Ionic liquids, which are used for extraction of recyclables from biomass, also play a special role in the area of solvent recycling. Olli Nakari of the University of Technology in Lappeenranta, Finland, examined retention of hemicellulose from ionic liquid [EMIM]Oac (1-ethyl-3methyl-imidazol-acetate). Because these liquids do not behave like organic solvents, no organophilic membranes are actually necessary. However, in practice one often mixes the liquids with organic solvents (e.g. methanol) in order to reduce the viscosity and achieve satisfactory flow rates through the membrane without increasing temperature. Thomas Fahrenwaldt of the Institute of Chemistry of Rostock University is investigating OSN as an alternative to chromatography for the retention of small organic molecules, like different amino acids, urea derivatives or alkaloids, which, as catalysts, support enantio-selective reactions. As an example reaction for the preparation of such an organic catalyst, the derivatisation of quinine with the chlorides of various organic acids was implemented. The chlorides were dosed in excess, to ensure a complete reaction and, therefore, were still present as residues in the product solution. For the purification of the target product with simultaneous removal of the impurities, a multi-stage diafiltration with ethanol was tested. After 5 diafiltration cycles, a much purer product was received than with chromatography. In the Department of Process and Chemical Engineering of TU Berlin, in cooperation with the Department of Life Science Engineering of Berlin University of Technology and Economics, one also deals with the extraction of base materials from renewable raw materials. In the catalytic conversion of long-chain olefins from renewable raw materials, these are mixed with the help of surfactants with the aqueous phase containing the catalyst. After the reaction, there is still a lesser portion of the surfactant in the organic product solution. Daniel Zedel showed that with the help of the surfactant Marlipal 24/70 in Dodekan, this can be removed with the help of OSN. However,

F & S International Edition

No. 14/2014

it has been established in long-term experiments that both permeability as well as surfactant retention decreased with time. The possibilities to simplify the production of peptides in the liquid phase with the aid of OSN, and render it more economical, are investigated by Wenquian Chen et al. from Imperial College, London, in cooperation with Schweizer Lonza AG. The peptides, which are in high demand in the pharmaceuticals market with huge growth rates, are “tailor-made” from individual amino acids in the conventional production process in recurring reaction steps. Between the reaction steps, the excess amino acid must be separated, something that happens up to now by extraction and precipitation. Diafiltration with organophilic nanofiltration membranes should serve as a substitute for these complex separation processes. The feasibility has been demonstrated, but the requirements for the chemical resistance of the membranes are very high for this process. Therefore, one works currently with ceramic membranes. Literature: /1/ VIVTA e.V. (Hrsg.): 4th International Conference on Organic Solvent Nanofiltration, 12. – 14.03.2013, Aachen, Book of Abstracts /2/ Dutczak, S.M.; Tanardi, C.R.; Kopéc, K.K., Wessling, M.; Stamatialis, D.: “Chemistry in a spinneret” to fabricate hollow fibre for organic solvent filtration; Separation and Purification Technology 86(2012), pp. 183 – 189 /3/ Volkov, A.V.; Parashchuk, V.V. Stamatialis, D.F.; Khotimsky, V.S.: High permeable PTMSP/PAN composite membranes for solvent nanofiltration; Journal of Membrane Science 333(2009), pp. 88-93; DOI: 10.1016/j.memsci.2009.01.050 /4/ Dalwani, M.: Thin film composite nanofiltration membranes for extreme conditions; Dissertation, Twente University, 2011, ISBN: 978-90-365-3276-1;DOI: http://dx.doi.org/10.3990/1.9789036532761 /5/ Livingston, A.G.; Bhole, Y.S.: Asymmetric membranes for use in nanofiltration; WO 2012/010886 A1, 2012 /6/ Karan, S.; Samitsu, S.; Peng, X.; Kurashima, K.; Ichinose, I.: Ultrafast viscous permeation of organic solvents through diamont-like carbon nanosheets; Science Vol 335(2012), S. 444-447. DOI: 10.1126/science.1212101 /7/ Ogielo, W.; van der Werf, H.; Tempelmann, K.; Wormeester, H.; Wessling, M.; Nijmeier, A.; benes, N.E.: n-Hexane induced swelling of thin PDMS films under non-quilibrium nanofiltration conditions, resolved by spectroscopic ellipsometry; Journal of Membrane Science 437 (2013), pp. 313-323. DOI: 10.1016/j.memsci.2013.04.039 /8/ Robeson, L.M.: The upper bound revisited, Journal of Membrane Science 320 (2008) No. 1-2, pp. 390 -400. DOI 10.1016/j.memsci.2008.04.030 /9/ Schmidt, P.; Köse, T.; Lutze, P.: Characterisation of organic solvent nanofiltration membranes in multi-component mixtures: Membrane rejection maps and membrane selectivity maps for conceptual process design, Journal of Membrane Science, 429 (2013), 103-120. DOI: 10.1016/j.memsci.2012.11.031 /10/ Peshev, D., Peeva, L.G.; Peev, G.; Baptista, I.I.R., Boam, A.T.: Application of organic solvent nanofiltration for concentration of antioxidant extracts of rosemary (Rosmarinus officiallis L.); Chemical Engineering Research and Design 89 (2011) no. 3, 318-327. DOI: 10.1016/j.cherd.2010.07.002

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 52

Highlights 2013

Chemical-free treatment of cooling water in open and half-open circuits H. Lyko*

In a huge number of industrial processes, heat is used for the treatment of materials, or a treatment process or conversion process is connected with heat generation. In many of these cases, substance flows, workpieces and semi-finished products are cooled by water. The water consumption in industrial cooling plants (except power stations) amounts to about 5 billion m3, of these, about 376 million m3 come from plants with open circuit cooling (according to the state of 2002, p. /1/). In open or halfopen cooling circuits, the heat is removed from the water by evaporation in a cooling tower. Here, dissolved or suspended substances are concentrated. Moreover, direct contact of the water with the outside air leads to the entry of airborne particles such as dust or pollen as well as to the absorption of gaseous impurities, such as SO2. Cooling water contains, according to the raw water source and the materials with which it is in contact in the cooling system, particulate contaminants, salts, dissolved organic matter and micro-organisms in different proportions. Moreover, the cooling water is (almost) saturated with dissolved oxygen through the intimate contact with the air in the cooling tower. To limit the concentration of the impurities, in particular the salinity, a part of the water is regularly discharged from the circuit and replaced with clean fresh water (makeup water). In addition, measures are required with which lime deposits, corrosion, the formation of biofilms (fouling) and an impermissible contamination of the water (for example with legionellae) can be prevented. It is common practice to carry out the conditioning of cooling water with a cocktail of chemical hardness stabilizers, corrosion inhibitors and biocides. Beside the high costs for the chemicals to be used, an ever increasing ecological awareness and the pursuit of more sustainable production processes lead to the search for alternative methods of treatment. In the following sections, previously common processes and a relatively new, almost chemical-free cooling water treatment process will be considered.

Cooling water properties Cooling water is no ultrapure water but, rather, it is sufficient to keep the impurities mentioned within limits in order to prevent the malfunctions named to begin with and to avoid pollution of the environment from

germs. The selection of the cooling water treatment depends on the one hand on the quality of the raw water used and, on the other hand, on the materials used in the circulation system that come into contact with the water. Table 1 gives an overview

Tab. 1: Recommended limit values for the condition of the recooling circulation water according to VDI 3803

Tab. 2: Overview of commonly used chemicals for cooling water treatment (s /1/)

*Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 53

Highlights 2013

about the recommended limit values of recooling circulation water according to VDI 3803. With these limit values and the composition of the raw water, the maximum concentration factor on the cooling tower is determined. If one succeeds in maintaining the appropriate concentrations low with a treatment in the circuits, the concentration factor EZ increases, resulting in an overall water savings. Chemicals for cooling water conditioning The essential malfunctions that may occur through water constituents include precipitation of calcium carbonate and phosphate as fixed deposits on internal surfaces, biofilms and corrosion. The latter is not only caused by dissolved oxygen, chloride or a low pH value but also by metabolites of micro-organisms present. These survive and multiply in the circulation water because, here, they encounter dissolved organic matter as food. Table 1 gives an overview about common substances or groups of substances which are added to cooling water circuits in order to prevent these malfunctions. The designation â&#x20AC;&#x153;hardness stabilizerâ&#x20AC;? here does not necessarily mean that the crystallisation of calcium carbonate or - sulphate does not take place. Rather, the additives hinder the formation of the stable calcite lattice, for example, by adsorption on the surface of micro-crystals. In some cases, the hardeners therefore manifest as sludge flocs. By using of biocides that contain chlorine or bromine, corrosive substances are in turn introduced into the water circuit, whose effect must then be curbed by inhibitors. An extrapolation of consumption data for Germany showed total loads of approx. 4,000 t/a of oxidative (mainly chlorine, chlorine-releasing agents, BCDMH and hydrogen peroxide) as well as 125 t/a of non-oxidative biocides (mainly isothiazolinones, DBNPA and QAV). In addition, approx.1,500-2,200 t/a of phosphonic acids, 45-135 t of molybdate and 113-216 t of zinc. These data were all determined in the same study commissioned by the Federal Environmental Agency /1/. The dosing of chemicals requires a certain construction and equipment effort in the form of storage capacities, reservoir containers, supply lines, dosing pumps and the accompanying measurement and control technology. Also, the proper safety guidelines must be observed for storage and use of the chemicals. To guarantee the optimum effect of conditioning chemicals and hence a minimum consumption, automated dosing systems are being used, with specified on-line sensors for certain water constituents /2/. For facilities that are equipped with the most modern measuring and dosing technology, one can assume that their chemical consumption compared to the period of the study is lower now, but the basic problem persists.

The reverse osmosis performs a nearly entire desalination and a relatively high working pressure must be applied for this purpose. Moreover, in the presence of organic and particulate contaminants, a pretreatment of the water would have to be done in order to avoid fouling of the membrane. The concentrate of the reverse osmosis must be otherwise disposed of or be treated. In researching reverse osmosis processes for cooling water treatment, it was found that this stage of treatment is usually applied in combination with various other conditioning methods, including a softening or deferrisation / demanganisation as well as the dosing of conditioning chemicals. A local disinfection is effected by high-energy UV-C radiation, which irreversibly damages the genetic material of micro-organisms, therefore preventing their propagation. However, this method of treatment works only in the plant component in which the radiation encounters micro-organisms. Biofilms with living microorganisms capable of reproduction at more distant sites are not affected by this treatment. Alternative hardness stabilization As with the effect of many hardness stabilizers, it often suffices if hardeners are not removed from the water but rather are prevented from forming insoluble deposits on plant walls. A variant of this is the selective formation of seed crystals through local displacement of solution parameters (the pH value or the temperature), so that precipitating calcium carbonate attaches to the crystals and not to the walls. The zinc sacrificial anode principle has a similar effect. Here, a local element is formed by the combination with the electrochemically higher quality brass. In the process, zinc ions are released, which react into zinc carbonate with the

Physical methods for cooling water conditioning Established physical methods which are also considered for cooling water include softening by ion exchangers, the removal of suspended and dissolved substances by means of membrane processes with different degrees of separation (ultrafiltration, nanofiltration, reverse osmosis) as well as disinfection by means of UV irradiation. In the ion exchanger, the hardeners calcium and magnesium are replaced by sodium. This must be regularly replenished during the regeneration of the ion exchanger. The demand per m3 auxiliary water is stated as approx. 0.7 kg salt /3/. As is apparent from table 1, while calcium carbonate is unwanted as a crystalline deposit, however, it does also have a corrosion-inhibiting effect through the formation of passivating surfaces. Accordingly, a complete removal is not desired, but rather a hardness of 4Â°dH is considered as optimum. In the case of complete softening by ion exchange, this hardness must be readjusted again by blending with untreated make-up water.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 54

Highlights 2013

Tab. 3: Data from reference plants of the DAT- Cooling water treatment system

by a new one without the whole galvanic element having to be replaced /5/. The entire treatment system is dimensioned such that about 20% of the whole cooling water in circulation is always treated. This is sufficient in order to keep the concentrations of all contaminants within the recommended limits (see. Tab. 2). Presently plants are available with capacities from 12 to 1000 m3. Examples of applications

Fig. 1: DAT- Cooling water treatment system in the container

hydrogen carbonate present in the water. Thus, crystallisation nuclei are provided on the one hand, on the other hand, hydrogen carbonate is removed from the water. The freely floating crystals can be separated at another place of the circuit through a particle filter and be discharged. Alternative method for prevention of biofouling The formation of biofilms on heat exchanger surfaces and other plant components requires the unrestricted multiplication of micro-organisms. This biomass grows on only when it is supplied with nutrients in the form of organic substances contained in water. A novel biofilter, in which the micro-organisms are immobilized on a special packed bed, causes the local biological degradation of organic water constituents, so that the nutrients for a biofouling are continuously removed from the circulation water at another plant site.

Combination of chemical-free treatment options With a novel, patented combination of different methods, the young enterprise DAT Dynamic Aquabion Tower GmbH, Roßdorf, implements a cooling water conditioning without addition of hardness stabilisers, corrosion inhibitors and biocides. The treatment zone, which is arranged in the bypass to the cooling water circuit, comprises an automatic backflush filter for particle removal, a biofilter that, with the help of immobilised micro-organisms, degrades the dissolved organic water constituents, a UV sterilisation unit arranged behind the biofilter, as well as a galvanic element with zinc-sacrificial anode. The only substance which is still consumed in this method is zinc in the form of the sacrificial anode. Here, a special construction of the galvanic element is used, with which the sacrificial anode that is consumed during the operation, can be removed from the pipework and replaced

The chemical-free cooling water treatment system can cause, in addition to savings in chemical conditioning agents, an optimisation of the water consumption and the wastewater volume (in the form of make-up water and blowdown brine). Under certain circumstances, the softening of the make-up water by means of ion exchangers can be omitted, whereby large quantities of salt can be reduced. This was, for example, the case for the installation of the plant to treat the water of a cooling system in industrial milk processing. In addition to savings in terms of salt and conditioning chemical, the concentration factor could be increased from 3 to 4, through which the consumption of makeup water and waste water was reduced accordingly. In one enterprise for special chemicals, the cooling system is being operated with well water. The treatment system was mounted fully assembled on a frame rack and was lifted via a heavy lift crane into the upper floor of the production hall. In one enterprise, in which natural oils and greases are further processed into oleo chemical products, the system was installed, for lack of space, outside of the production buildings in a weatherproof container (see. Fig. 1). The savings in these three applications are summarised in table 3. Lierature: /1/ Gartiser, S., Ulrich, E.( Edited by UBA): Einsatz umweltverträglicher Chemikalien in der Kühlwasserkonditionierung; Research Report 200 24 233 UBA-FB 000298 (2002) /2/ Winkler, T.: Mit effizienter Kühlwasseraufbereitung Energie und Wasser sparen und die Umwelt schonen – Innovative Sensorik für optimale Kühlwasserdesinfektion; Paper presented at the 12. VDMA Wasser- und Abwassertagung, 28./29.November 2012, Frankfurt am Main /3/ Reiche, M.: Neue Trends in der Kühlwasseraufbereitung – Chemikalienfreie Alternativen; Paper presented at the 12. VDMA Wasser- und Abwassertagung, 28./29. November 2012, Frankfurt am Main /4/ Flettner, M.: Wasserbehandlungsvorrichtung; Patent WO 2005/042803 A1, 2005 /5/ DAT Dynamic Aquabion Tower GmbH: Kühlanlage und Verfahren zu deren Betrieb; Patent WO 2011/045071, 2011

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 55

Highlights 2013

Filters for clarification of auxiliary liquids R. Berndt*

Clarifying filtration is used mainly to remove solid contaminants from a liquid or emulsion in order to reuse it, or to use it as a product. Often the contaminating particles are not suited to building up a sufficiently permeable filter cake with satisfactory thickness and strength so that it can simply be removed from the filter media by mechanical means. In many of those cases, the solid particles are collected on the filter or in its depth, respectively, and disposed of together with it. For the purification of liquid manufacturing auxiliaries (e.g. cooling-lubricating agents, hydraulic fluids, gear oils, scrubbing liquids), preferably band filters, cartridge filters or precoat filters are used to separate particles in the typical size range of 5…100 μm. These filters are characterized by specific advantages and disadvantages, and are suitable for different fields of application. Especially when small particles of about 5 μm at higher concentration are to be removed from liquids with viscosities above that of water, filtration becomes critical. A diagram of a filter is shown here, which can advantageously complement the stock of filters for such tasks.

1. Targets, working fields and peculiarities of clarifying filtration Clarifying filtration is typically employed to remove solid impurities from liquids or emulsions, so that these media can be reused, or used as a product. Often the particles separated in this way do not form a sufficiently permeable, uniform and solid filter cake, for which reason cake growth and removal of the filter cake from the filter medium are obstructed. Typical working fields of clarifying filtration are the clarification of liquid end products (e.g. beverages, fuels, chemicals), the clarification of liquid intermediates (e.g. reactants, product components, solvents), and * Prof. Dr.-Ing. habil. Rolf Berndt (VDI) RBFM Consulting Eichstraße 16, D-01309 Dresden Tel.: 0351-31778056 Fax: 0351-31778057 rbfm_consulting@t-online.de

the clarification of auxiliary media used in manufacturing. Clarifying filtration of such auxiliary liquids (e.g. cooling-lubricating agents, hydraulic fluids, gear oils, scrubbing liquids) is typically applied to limit the concentration of solid particles that could interfere with the reuse of the liquid. Therefore, owing to the recirculation of the auxiliary medium, concentration of solids in the filter inlet is often comparatively low. Low concentrations of solids, however, complicate the cake formation and the cake growth on the filter media, in particular in combination with small particle sizes and higher viscosity. The cake formation rate κ is defined according to eq. (1) as the increase of the filtered solids mass ms;K over time relative to the filter area AF, (1)

in which ms;K is obtained as the product of the filtrate volume Vf and a modified solids concentration cs. By inserting the well-known Carman equation for cake filtration /1/ (2) into eq. (1), eq. (3) for the cake formation rate κ is obtained. (3) The mass-specific filter cake resistance rs is approximately inversely proportional to the square of the mean size of solid particles. Small particle sizes, high viscosities η relative to the filtration pressure difference Δp, and low solids concentrations cs thus constrain the formation of a sufficiently thick and heavy filter cake. A particular disadvantage of low solids concentrations is the low probability of bridge formation. Therefore finer filters must be

Synthetic nonwovens for air, gas and liquid ﬁltration. Textile solutions in HVAC, automotive and industrial applications as well as vacuum cleaner bags

Sandler & Filtration: PURE PROTECTION Q efficient and long-lasting

Q progressively structured and uniform

Q moisture-resistant and bacteriostatic

Q pleatable and easy-to-process

Q shedding-free and recyclable

From the ﬁrst developmental idea to the ﬁnished product – Sandler is your reliable partner Sandler AG • Germany • Phone: +49 92 84/60-0 • Fax: +49 92 84/60-269 • E-mail: filtration @ sandler.de • w w w.sandler.de

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 56

Highlights 2013

Tab. 1: Overview and evaluation of filter equipment for auxiliary liquids

Fig. 1: Hydrostatic inclined bed band filter (schematically)

used in order to avoid turbidity, which, in turn, have a higher filter medium resistance R. Thereby, according to eq. (3), cake growth κ is slowed even more. Especially with wide particle size distributions, there is also the risk of irreversible filter media clogging. Certain features of the particles (shape, deformation resistance, surface properties) and the particle size distribution can in addition affect the cake formation and solids removal from the filter medium. Therefore, two strategies are used. Both principally renounce the formation of a sufficiently thick, filtering, completely removable filter cake. In the first case, the solid matter is retained on or in the filter medium, and discarded at least partially together with the filter medium. Precoat filters, filter sheets, disposable filter car-

tridges and nonwoven band filters work according to this principle. The second strategy uses frequent hydro-mechanical cleaning of the filter medium, in particular periodic backwashing. Back flush cartridge filters and constructed edge filters are typical representatives of this. 2. Clarifying filters for auxiliary liquids The term “auxiliary liquid” in this paper summarizes those liquids and emulsions that do not become part of the final product but are essential for the correct functioning of a production process or a technical system. Inter alia, these include: - Cooling lubricants for the manufacturing of components - Scrubbing liquids for the cleaning of components after machining

- Hydraulic fluids for energy transfer in hydraulic systems - Lubricating oils for gear units and similar assemblies with wear-sensitive components. Such fluids are generally circulated and they require loop maintenance that often includes filtration. This ensures the limitation of the solids concentrations and/or the maximum particle size according to the needs of the respective application. Particles that entered in the fluid beforehand as chips or swarf or surface contaminants must therefore be constantly removed from the fluid. The typical size range of particles to be filtered out is 5 … 100 μm; larger particles (e.g. chippings) are often separated upstream to filtration by means of sedimentation and/or magnetic separation. Besides filters in isolated operation, centralized systems and mobile filters are utilized, especially for coolants. Both full-flow filtration and bypass filtration is practised. Therefore the filtration of auxiliary media covers immense capacity ranges, from several liters per hour up to several hundred cubic meters per hour and more. The filtration of auxiliary media is mainly applied in the metal-working industry and in the past has not attracted much attention among chemical and process engineers. This fact, together with the application-specific requirements, led to independent trends and developments, which in some cases differed from those in the chemical and process industry. Robustness, low service and maintenance costs with little capital investment and low running costs are expected to a large extent; specialized process engineering staff are rarely available. Several band filters with disposable nonwoven filter bands are widely used for the filtration of coolant solutions and emulsions. In basic flat-bed band filters or inclined bed band filters (Fig. 1), nonwoven filter bands from polymeric or cellulose fibers are used as disposable filter media for purely hydrostatic filtration. The filter band forms a trough into which the product to be filtered is fed. While the filtrate is collected below the band, the solid matter that was retained in and on the media reduces its permeability, causing the liquid level in the pool to rise. A float switch then triggers the band drive, whereby fresh fleece is unwound from the roll and dirty nonwoven - if necessary after separate solids removal - is deposited in a sludge collection box, or is wound up on a reel. The simple design and the availability of modular systems for different variations allow cost-effective production and relatively low acquisition costs. Because the band feed is activated depending on load, such a filter adapts itself to variable

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 57

Highlights 2013

Fig. 2: Cartridge filter and constructed edge filter (schematically)

Fig. 3: Precoat filtration by means of a cartridge filter (schematically)

conditions and is therefore robust in this respect and less susceptible to interference. The relatively slow filtration is accompanied by sedimentation of fast sinking particles that form a precoat, improving the particle retention. However, this segregation makes the filter cake nonhomogeneous and may increase the cake resistance, resulting in more frequent band feed and increased filter medium consumption. The fundamental disadvantages of band filters include small usable pressure difference, low compactness AF/Ast (large footprint relative to the filter area) and consumption and waste of filter media. Numerous variations of the basic design try to limit these disadvantages. These include, inter alia, the application of vacuum beneath the filter band, or excess pressure above the band, during the filtration (pressure differences approx. 0.6 bar and/or 1 bar). The sealing of the underpressure or overpressure zones against atmosphere must be resolved during the band feed – which is a technical challenge, basically prone to wear and failure and significantly increasing manufacturing and acquisition costs. Other derivative designs can be made more compact by using several stacked filter beds /2/, or wrapping the band partially around the outer periphery of a submerged, perforated drum that is operated in an inside-out mode /3/. The smallest economically removable particle sizes range - depending on design and application case - from 10 to 100 μm. In the upper particle size range, endless bands with cake removal and band washing instead of disposable non-woven bands are used as well. Cartridge filters and constructed edge filters /4/ are preferably used for pressure filtration where the liquid flows - typically outside-in - through the filter material of filter elements that are arranged in a filter housing (Fig. 2). The solid matter is separated on and in the filter material. Disposable cartridges and backwashable filter elements are available in a large variety and with typical separation limits down to approx. 1 μm. Constructed edge filters have defined slot-shaped channels for the filtrate flow which widen in the main flow direction. Their lower separation limit is about 20 μm. They are regenerated by backwashing, which is supported by the rotation of the filter element and/or temporary slot width enlargement. Cartridge filters and constructed edge filters allow medium to large filter areas AF with low floor space Ast. Disposable filter cartridges cause considerable filter replacement costs (material, disposal, staff costs). The concentrate of backflush filters creates increased disposal expenditure if it is not further treated (e.g. by mechanical/ thermal drying). The use of granular or fibrous filter aids is characteristic of precoat filtration /5/. Filter aids are inert, fine-grained solids with large specific surface, e.g. kieselguhr (diatomaceous earth), perlite, or cellulose fibers. They are admixed to the suspension to be filtered (continuous precoating, “body feed”) and/or initially - prior to the filtration itself – precoated onto a filter medium (preliminary precoating, “precoat”), where they form a temporary filter sheet

(Fig. 3). For clarifying filtration of products that are very difficult to filter (e.g. fermentation products), body feed is traditionally used on pressure filters for cake filtration. Cartridge backflushable filters and centrifugal discharge plate filters are the main types. For precoat filtration, rotary vacuum drum filters are also used, in which the upper, soiled filter aid layer is peeled off pseudo-continuously. The lower separation limit is approx. 0.2 μm /6/. In relation to the floor space Ast, medium to large filter areas AF are possible. Disadvantages are the complexity of equipment and the space requirements for precoating and body feed, plus costs for the procurement and disposal of the filtration aids, which also bind additional quantities of the liquid product and hence reduce the filtrate yield.

F & S International Edition

No. 14/2014

• • • • •

G. BOPP + CO. AG www www.bopp.com .bopp.com Bachmannweg 21 . CH-8046 Zurich Phone +41 44 377 66 66 . info@bopp.ch

THE ART OF SWISS PRECISION 57

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 58

Highlights 2013

4. A suggested solution

Fig. 4: Step-band drum filter (schematically)

3. Evaluation of clarifying filters The filter principles as described differ substantially in their advantages and disadvantages and their preferential application areas. These differences are illustrated below by means of a fictitious application case. The suitability of the filter principles is verified for compact, isolated “standalone” units (i.e. small and medium-sized throughput rates) at filtrate viscosities of 5 cP and higher, solid-volume concentrations ϕs ≤ 0.1% and particle sizes d <20 μm. These conditions may appear isolated or combined and are typical for auxiliary liquids that are difficult to filter. Table 1 overviews the results of evaluation. The viscosity, and the high filter cake resistance, owing to (small) particle sizes, disqualifies hydrostatic band filters; high equipment expenses are arguments against vacuum-operated and pressure-operated band filters, and band filters in general are unfavorable due to large space requirements. Disposable cartridge filters meet almost all requirements but are disqualified by the high expense and workload for cartridge replacement. The rating of backflush filters is basically moderate; the large concentrate volume, however, is unfavorable. Precoat filters are not particularly suitable for somewhat small “stand-alone” filter systems because of the large floor

space required and the effort needed for precoating. Thus the filter principles as evaluated cannot be considered as ideal solutions. The filter systems widespread in the classical process industry, especially in the chemical industry, (e.g. rotary vacuum drum cell filters, horizontal vacuum band filters), might offer advantages, but they often encounter price reservations. A search should therefore be made for filter principles that are easier to implement. On closer inspection, it becomes obvious that disposable cartridge filters are unsuitable for medium and high solids concentrations, whereas vacuum and pressure-operated band filters can handle them. It seems obvious to maintain the specific advantages of the two filter classes, and to eliminate their disadvantages, through the combination of the principles (see Table 1). A cartridge filter would therefore be desirable whose filter material is renewed automatically and without any cartridge replacement when its limit load is obtained. From another point of view, it would be advantageous to use a band filter in a closed container, designed in a more compact manner than the standard market models and without complicated seals to be released during the band feed. These requirements represent a task for engineering.

Fig. 4 shows schematically how the foregoing task could be performed /7, 8/. The suggested filter system is based on a filter drum whose shell surface is perforated with the exception of the peripheral zones and a bridge connecting them. This filter drum is rotatably placed in a housing or container; the drum axis can be arranged horizontally or vertically. The filter drum is connected with the environment by a hollow rotatable shaft that is guided through the housing wall. The perforated part of the drum shell is covered by a filter band (nonwoven) that reels off a dispenser roll and ends up on a collector roll. The drum stands still during filtration, which can be pressure or vacuum filtration. The filtrate flows through the filter band into the drum and leaves it via the hollow shaft. When the filter resistance has been reached a limit value, the pressure levels in the housing and the interior of the drum are equalized. Afterwards, the filter drum is turned through one full revolution, whereby the spent leg of the filter band is wound up on the collector roll and is replaced by new material unwound from the dispenser roll. Then filtration can be continued. This cycle of filtration and band replacement can be repeated until the dispenser roll is empty. Not till then do the dispenser roll and the collector roll need to be replaced. Based on this basic principle, numerous individual designs are conceivable and advantageous (Fig. 5), which can be supplemented by optional additional components. Inter alia, these include - dispenser roll and collector roll mounted in a submerged or dry position - peripheral sealing by means of endless sealing strips and deflecting pulleys - cake scraper blades and other band cleaning devices - band pretensioners - band stock detectors. A filter system designed in such a way cannot and should not replace the available choice of filters, but it may offer a reasonable extension. Its application range is the filtration of particle-containing liquids and emulsions, in particular in cases where the particles clog up the filter medium instead of building up a thick, well-filtering filter cake. The typical particle size ranges from 5 to 50 μm. Some of the fundamental advantages over the conventional band filters are - more compact design; AF/Ast ≈ 2 (i.e. 3…7 x greater) - 10…300 times greater applicable pressure difference

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 59

Highlights 2013

- better filter medium utilization, i.e. lower media consumption - applicable with finer media and at higher viscosities - complete encapsulation extremely simplified - easier structural design compared to conventional pressure and suction band filters. Advantages compared to disposable cartridge filters are - applicable at higher solids concentrations - filter media replacement frequency lowered by a factor of 10…30 for equal exposure to solid matter - considerable reduction of filter media replacement costs (labor, material, disposal). Based on the specific features and benefits, the potential application fields include the maintenance of cooling lubricant circuits that are difficult to filter, clarification filtration of washing fluids, and particle removal from lubricating oil circuits and hydraulic circuits. General application possibilities, apart from auxiliary liquids, are pre-filtration upstream of ultrafiltration or microfiltration systems, as well as polishing filtration downstream of filters and centrifuges. 5. Summary and conclusions For clarifying filtration of auxiliary liquids, somewhat basic filtering equipment is used, whose principal drawbacks become obvious especially in difficult separation jobs with disabled cake formation and complicated conditions. Common solutions reach their limits especially at rising viscosities of the filtrate, increased solids concentrations and small particle sizes. Increasing use of the more efficient filter equipment widely used in classical process engineering applications will be an alternative. In addition, a simple filter system which combines the characteristics and advantages of band filters and cartridge filters would be able to complete and enrich the range of filter equipment.

Fig. 5: Versions of step-band drum filters (schematically)

List of Symbols AF Ast cs ms ms;K Δp R rs t Vf η κ

filter area footprint ratio of solid mass ms in the feed to filtrate volume VF solids mass in the feed solids mass in the filter cake filtration pressure difference filter media resistance mass-specific filter cake resistance filtration time filtrate volume dynamic viscosity of the filtrate cake formation rate

Literature: /1/ Gösele, W.: Rechnerische Behandlung der Filtration (Chap. 8.2). In: K. Luckert (Hrsg.): Handbuch der mechanischen Fest-Flüssig-Trennung, 1st edition, Vulkan-Verlag, Essen (2004) /2/ Hoffmann Maschinen- und Apparatebau GmbH: Saugbandfilter SF3/SFA3 – Schleifpräzision durch Feinstfiltration. Company prospectus, as of June 2013, http://www.hoffmannfilter.de/deutsch/filter.htm /3/ Reber Systematic GmbH + Co. KG: RESY KBF Funktionsprinzip. Company prospectus, as of June 2013, http://www.resy-filtration.com /4/ K. Sutherland: Filters and Filtration Handbook, 5th edition, Elsevier, Amsterdam, Boston, Heidelberg, London, New York (2008), Chap. 3K-M /5/ Gösele, W.: Filtermittel und Filterhilfsmittel (Chap. 8.4). In: K. Luckert (Hrsg.): Handbuch der mechanischen Fest-Flüssig-Trennung, 1st edition, Vulkan-Verlag, Essen (2004) /6/ Berndt, R.: Filter zur Klärung von Flüssigkeiten – Prinzipien, Klassifizierung, Auswahl. F&S Filtrieren und Separieren 26(2012)4, 236-241 /7/ Utility model DE 20 2013 101 108.8, 14.03.2013 /8/ Patent Application DE 10 2013 101 984.6, 28.02.2013

F & S International Edition

No. 14/2014

Experts in sheet metal Processing Filter tubes, per forated metal components, containers and housings manufac ture, design in steel and stainless steel

Franz-Méguin-Str. 20 66763 Dillingen GERMANY Phone Fax

+49 68 31 / 700 33 00 +49 68 31 / 700 33 50 w w w.preziehs.de info@preziehs.de

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 60

Highlights 2013

Universal similarity laws for the description of the initial pressure loss of metal woven wire cloths during the filtration of non-Newtonian fluids M. Müller, M. Piesche* In numerous processes of the plastics, chemicals and food industry, the filtration of highly viscous liquids is an important process step. The fluids which have to be filtered have not only a purely shear thinning, but usually also a viscoelastic flow behaviour. On account of their high thermal, mechanical and chemical robustness, the use of metal woven wire cloths is widespread in such filtration processes. In addition to the separation performance, the initial pressure loss in the unloaded state must also be determined for the design of these filtration processes. For this, universal similarity laws are developed using characteristic dimensionless numbers. With the acquired similarity laws, it makes possible to the users to estimate the initial pressure loss of the filtration process of highly viscous liquids and thus to assess the process in terms of power demand already previously. 1. Introduction

2. Derivation of the similarity laws

Nearly the whole spectrum of filtration tasks can be covered by the use of metal filter media. The media selection is caused by the process tasks, specifications and requirements. In particular, if filtration processes with high demands on quality are used, metal woven wire cloths are selected as the filter medium. Thus metal woven wire cloths are used on industrial scale, for example for the separation filtration respectively the solid matter extraction, for the clarifying filtration, for the purification of fuels and hydraulic fluids, for water treatment as well as for the separation of dirt particles from polymer melts and polymer solutions. Metal woven wire cloths derive their wide area of application from their excellent separation and grading efficiencies, on account of their dense pore size distribution, their high thermal and chemical robustness, their high mechanical strength and at the same time good ductility properties, as well as the possibility to clean off the woven wire cloths. The flow-induced pressure loss as well as the separation efficiency of particles are the essential quality features of a filter medium and cause the selection of a medium suitable for a special application case. For a reliable prediction of the through-flow conditions of the metal woven wire cloths suitable models are required. In the literature, design bases for metal woven wire cloths exist up to now only for the filtration of low-viscous Newtonian liquids and air. Kopf /3/ summarises the different models and indicates their advantages and disadvantages. Moreover, he develops his own similarity laws for the description of the pressure loss of metal woven wire cloths which he ascertained experimentally and numerically. Universally valid model equations for the description of the initial pressure loss of unloaded metal woven wire cloths during the filtration of non-Newtonian fluids are not known up to now. For this reason similarity laws in the form of characteristic dimensionless numbers for the description of the pressure loss of metal woven wire cloths during the filtration of highly viscous, shear thinning and viscoelastic fluids are developed within the scope of this work.

2.1 Basics

* M.Sc. Martin Müller Prof. Dr.-Ing. habil. Manfred Piesche University of Stuttgart Institute of Mechanical Process Engineering Böblinger Straße 72, 70199 Stuttgart / Germany Martin.Mueller@imvt.uni-stuttgart.de Phone: +49 (0) 711-685-85386 Fax: +49 (0) 711-685-85390

The flow in a metal woven wire cloth can be considered as a flow in a group of parallel tubes and consequently for the calculation of the differential pressure on a woven wire cloth you can resort to the knowledge of the pressure loss calculation of a pipe flow. In the case of a stationary, isothermal and fully developed flow, the following applies: (1) From this it is clear that the pressure loss is on the one hand caused by viscous forces (the first addend on the right side of equation (1)) and on the other hand caused by inertia forces (the second addend on the right side of equation (1)). By the filtration of highly viscous fluids, inertia forces can be neglected compared to viscous forces /5/ and equation (1) can be simplified to: (2) The equivalent velocity veq is the mean flow velocity in the cloth. Due to continuity reasons, the following is valid: (3) In this, v0 is the undisturbed filtration velocity and ε is the porosity of the filter medium. This is defined as the ratio of the perfused volume Vfree relative to the total volume Vtot: (4) Vw is the volume of the weft and warp wire of the metal woven wire cloth. Since the cloth height δ multiplied with the characteristic perfused surface deq2 is equal to the perfused volume ε·Vtot, the following results for the equivalent diameter:

(5) In a flow channel with the cross section deq2 and if the specific surface ϕ is defined as the ratio of the wire surface relative to the total volume according to (6)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 61

Highlights 2013

Fig. 1: Structural parameter of a square mesh in plain weave.

the characteristic length leq multiplied by the circumference, which is proportional to deq, is proportional to the wetted surface ϕ·Vtot: (7) With the equations (3) to (7) it is possible to transfer the wellknown approach for the pressure loss of a pipe flow (equation (2)) to the conditions in metal woven wire cloths. Another task is now to describe the newly defined geometrical and dynamic equivalent variables as functions of the cloth structure parameters of the metal woven wire cloths of different types of weave. 2.2 Similarity laws of the different types of weave 2.2.1 Similarity law for square meshes Starting from Figure 1 in which a section of a square mesh in plain weave is illustrated, the still unknown cloth height δ, the porosity ε, the specific surface ϕ, as well as the equivalent diameter deq and the equivalent length leq, can be expressed as functions of the cloth structure parameters, i.e. the mesh size w and the wire diameter d of the square mesh. If the deformation through the weaving process of the weft and warp wire is neglected, the cloth height δ is equal to the double of the wire diameter: (8) The total volume of a mesh amounts to: (9) Within this mesh of the cloth, the wires take up a volume of

More than 50 years experience in the process industry filtration

(10) where the wire length lw can be calculated according to equation (11), disregarding the enlacement of the wires. (11) Starting from the definition equations in section 2.1, the following arises for the porosity of a square mesh (12) and for the equivalent diameter:

Sefar AG Hinterbissaustrasse 12, CH-9410 Heiden Tel. +41 71 898 57 00, Fax +41 71 898 57 21 ﬁltration@sefar.com, www.sefar.com

(13)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 62

Highlights 2013

In an opaque Dutch Weave the weft wires are in contact, so that d1=t1 and consequently for the total volume of a mesh you get: (21) Within this volume, the warp and weft wires take up a volume of (22) wherein the wire length lw can be determined from (23) Starting from equation (4) the following arises for the porosity ε: (24) In this, D12 is the ratio of the weft wire diameter to the warp wire diameter (25) and T2 is the lattice spacing t2 transformed by the warp wire diameter d2 to a dimensionless expression: (26) The equivalent diameter is therefore determined from: (27)

Fig. 2: Structural parameter of a Single Plain Dutch Weave.

Taking into account that the wire surface Aw within a mesh is Within the total volume of a mesh, the wire surface Aw amounts to: (14) By multiplying the specific surface ϕ (equation (6)) with the wire diameter d, this can be transferred to the dimensionless specific surface φ. Taking into account the equations (9) and (14), the following arises for a square mesh:

(15) The still unknown equivalent length can be therefore estimated from: (16) Starting from the approach put up in equation (2) for the pressure loss and with the relations presented above, the in equation (17) outlined similarity law for square meshes in plain weave arises. (17) Here, Eu is the Euler number formed with the filtration velocity v0 (18) and Re is the Reynolds number defined with the wire diameter d as characteristic length: (19) 2.2.2 Similarity law for Dutch Weaves To be able to determine the geometric and dynamic equivalent variables still unknown from section 2.1, a Single Plain Dutch Weave according to Figure 2 is used as a basis. The following implementations are valid analogously for BetaMesh-weaves and Reverse Plain Dutch Weaves. Single Plain Dutch Weaves can be described geometrically with the two lattice spacings t1 and t2 of the weft and warp wire, as well as with their diameter d1 and d2. If the deformation due to the weaving process of the weft and warp wire is neglected, the cloth height δ arises from:

(28) you get for the dimensionless specific surface: (29) Starting from equation (2) and with the help of the equations (20) to (29) the similarity law for the initial pressure loss can be depicted as: (30) Here, the Euler number is defined according to equation (18) and the Reynolds number according to equation (31). (31) 2.3 Transfer of the similarity laws to non-Newtonian fluids by means of the principle of the effective viscosity In case of a non-Newtonian fluid it is not immediately known, on account of its shear rate-dependent viscosity, with which viscosity μ the Reynolds number should be formed in the similarity laws for the initial pressure loss of the metal woven wire cloths. To find a universal conditional equation also for this further degree of freedom, one resorts to the principle of the effective viscosity, which is already known in literature for a pipe flow, and this methodology is transferred to the conditions in metal woven wire cloths. For the stationary flow of a shear thinning medium in a circularly cylindrical pipe with the diameter d, Chmiel and Schümmer /2,6/ _ derive a conditional equation for the representative shear rate γ· (32) with which it is possible to transfer the known pressure loss characteristics of a Newtonian medium to a non-Newtonian fluid. To this end, the Reynolds number Re is replaced in the similarity law by an effective Reynolds number Reeff, formed with the effective viscosity μeff. Here, the effective viscosity μeff arises according to (33)

(20)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 63

Highlights 2013

from the flow curve γ· = f(τ) of the considered shear thinning medium. Within the scope of this work, the conditional equation for the representative shear rate (equation (32)) is transferred to the conditions in metal woven wire cloths, whereby it is then possible to apply the similarity laws developed in section 2.2 also with nonNewtonian fluids. In equation (32) v is the mean flow velocity in the pipe. In case of the through-flow of a metal woven wire cloth, this has to be replaced with the mean flow velocity within the cloth. Due to continuity reasons, the following applies: (34) The studies carried out showed, that the hydraulic diameter dh of the projected cross-section with the highest flow velocity within the cloth must be used as diameter d in the equation of definition of the representative shear rate. This is due to the fact that in this cross-section, on account of the high flow velocity, also the largest proportion of the pressure loss is induced. Therefore in case of the through-flow of a metal woven wire cloth with a non-Newtonian medium, the representative shear rate is determined from equation (35). (35) If the effective viscosity is determined from the flow behaviour of the medium under consideration of equation (33), the Reynolds number

Fig. 3: Characteristic cross-section of a Dutch Weave.

Starting from the definition of the hydraulic diameter the following arises for the passed through area (cf. Figure 3):

(36) for square meshes respectively Therein is (37) for Dutch Weaves can be calculated and the pressure loss behaviour of the metal woven wire cloths can be described also for nonNewtonian media with the similarity laws from section 2.2. Characteristic cross-section of square meshes As already illustrated in Figure 1, the narrowest cross-section of a square mesh in plain weave is in the middle of a mesh. Therefore for the representative shear rate of this weave type, the following results: (38) Characteristic cross-section of Dutch Weaves In a Dutch Weave the cross-section with the highest flow velocity is identical with the triangular-shaped pore passageway between two weft wires and the warp wire, as it is marked in blue in Figure 3.

(39) (40)

and

(41) For the calculation of the wetted perimeter U, a case distinction is required: If the perimeter arises from:

(42) Here, sinα and sinβ are calculated according to equation (40) respectively (41). If the following applies: (43)

Filter Testing conforming to standards

Tests of particulate air filters according to EN 779 / ASHRAE 52.2

Filter Test System ALF 114 Topas GmbH Oskar-Röder-Str. 12 D-01237 Dresden

Phone +49 (351) 21 66 43 - 0 Fax +49 (351) 21 66 43 55

F & S International Edition

No. 14/2014

E-Mail office@topas-gmbh.de Internet www.topas-gmbh.de

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 64

Highlights 2013

behaviours were modelled with suitable rheology models and implemented by using a “user-defined function” in the flow simulations with Fluent /5/. Figure 5 illustrates the viscosity curves of the modelled media.

Fig. 4: Representative section of a square mesh (left) and simulation model (right).

3. Numerical studies 3.1 Modelling and simulation The numerical investigations of the flow phenomena through metal woven wire cloths were done with the help of the commercial flow simulation software ANSYS Fluent. This flow solver is based on a finite volume formulation of the fundamental fluid mechanic equations for the total mass, momentum, energy and the various species in a multicomponent system. For their detailed representation see for example /1/. The spatial discretization of the solution domain, which is necessary for the flow simulations, was done with the help of the mesh generator Gambit and is represented for a square mesh in the following exemplarily. For Dutch Weaves the procedure was analogously /5/. To make the task accessible to a numerical flow simulation and to reduce the computational effort, a whole filter element was not modelled as a simulation model but a representative section of four adjacent meshes, as they are marked in red on the left in Figure 4. This geometry reduction is permissible since the four meshes repeat periodically in the cloth. The simulation area in which the flow is numerically calculated is thus limited by a cuboid volume enveloping the wire geometry (cf. Figure 4 on the right). On account of this reduction of the simulation model, the side faces of the cuboid have to be assumed as “periodic” boundary conditions. At the “velocity-inlet”, the undisturbed inflow velocity v0 can be specified. Within the scope of the numerical investigations, Newtonian highly viscous media with viscosities in the range of 0.1 to 1000 Pas and with a density of ρ = 998,2 kg_m were considered as well as shear thinning fluids in which on the one hand a 0.75 wt.% carboxymethyl cellulose (CMC) dilution and on the other hand different silicone oils of the series KORASILON oils M of the company Kurt Obermeier GmbH were considered. Their flow 3

Fig. 5: Modelled viscosity curves in the numerical investigations.

3.2 Results With the help of the differential pressures on square meshes, determined by means of the numerical flow simulations, the in equation (17) still unknown constant of proportionality could be determined to K1 =14. Furthermore, the principle of the effective viscosity, only known up to now from the pipe flow, was transferred to metal woven wire cloths (cf. section 2.3). Therefore the pressure loss of an unloaded square mesh can be described universally with the similarity law shown in equation (44). (44) In the case of the use of a Newtonian medium the Reynolds number Re according to equation (19) has to be inserted into the similarity law and in the case of the use of a non-Newtonian fluid the effective Reynolds number Re_eff according to equation (36) has to be inserted. Figure 6 illustrates the conformity of the similarity law with the numerically calculated differential pressures of different square meshes. The pressure loss on Dutch Weaves can be described with the following similarity law: (45) This is the principle derived according to equation (30). Due to the differential pressures calculated by means of numerical flow simulations it became clear, that the left side of the similarity law must be modified with the empirical factor T2-0,5 D12-0,1. In the diagram in Figure 7 the left side of equation (45) is plotted on the ordinate as a cloth-specific modified Euler number and on the abscissa the Reynolds number respectively the effective Reynolds number is plotted. The determined pressure losses for different Dutch Weaves are grouped around the curve path of the similarity law. 4. Experimental studies For the experimental validation and for the expansion of the similarity laws for the pressure loss of Newtonian and shear thinning fluids in case of the through-flow of a metal woven wire cloth as shown in section 3.2 onto viscoelastic media, experimental investigations were also carried out besides the numerical investigations. 4.1 Test rig and test medium Figure 8 shows the process engineering flow diagram of the test rig. The test medium is provided in the vacuum-steady receiver

Fig. 6: Similarity law for the initial pressure loss in the unloaded state on square meshes during the filtration of Newtonian and shear thinning fluids.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 65

Highlights 2013

Fig. 7: Similarity law for the initial pressure loss in the unloaded state on Dutch Weaves during the filtration of Newtonian and shear thinning fluids.

tank B1 to allow the degassing of interfering air pockets. The gear pump P1 pumps the medium and is operated with a frequency converter for an infinitely variable volume flow control. The measurement of the volume flow is done with an ultrasonic flow rate measuring system in a clamp-on design, based on the runtime difference method in line with VDI/VDE 2642. Using this measuring system, a non-contact measurement of the flow velocity in a pipe is possible. The consideration of the temperature influence on the viscosity of the test medium is done with the thermocouples upstream and downstream the filter holder F1, in which the metal woven wire cloths are inserted. As a test medium the KORASILON M10,000 silicone oil from the company Kurt Obermeier GmbH was used. Figure 9 above illustrates the viscosity curve of the medium at different temperatures measured by means of the shear stress-controlled rotational rheometer Haake RheoStress 600, Thermo Electron GmbH, in connection with a peltier temperature-controlled cone-plate geometry as well as by means of the capillary rheometer Rheograph 2000 from the company Göttfert. The first normal stress difference N1 and the shear stress τ as functions of the shear rate γ· are mentioned in Figure 9 below. With increasing shear rate, the influence of the first normal stress difference, i.e. the viscoelastic substance properties, increases. The appearing normal stress differences cause forces cross to the direction of flow and cause an additional pressure loss.

Fig. 8: Process engineering flow diagram of the test rig.

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 66

Highlights 2013

Fig. 10: Similarity law for the initial pressure loss in the unloaded state on square meshes during the filtration of viscoelastic fluids.

(48) As becomes clear, for the consideration of the additional pressure loss on account of the normal stress differences, on the right side of the similarity law, the further addend was introduced. Figure 10 shows how the experimentally determined differential pressures of different square meshes can be described with the similarity law according to equation (48). If one then proceeds analogously to square meshes for Dutch Weaves (Single Plain Dutch Weave, BetaMesh-weave and Reverse Plain Dutch Weave), the similarity law shown in equation (49) results for the description of the initial pressure loss using a viscoelastic medium.

Fig. 9: Viscosity curve at different temperatures (on top) as well as the first normal stress difference N1 and shear stress τ as functions of the shear rate γ· (below) of the M10,000 KORASILON oil.

4.2 Results To consider the additional pressure loss on account of the viscoelastic material behaviour in the already demonstrated similarity laws, these laws have to be extended. The viscoelasticity can be detected by implementing the Deborah number De. The Deborah number is the relationship between the relaxation time of the medium trelax and a characteristic observation time tobserve: (46) The relaxation time is the material-specific time which a medium requires to respond to a deformation by viscous flow. For the M10,000 silicone oil trelax was determined in /4/ to trelax =7.2 · 10 –4 s. The observation time is the time, which the fluid spends within the metal woven wire cloth during the passage of the cloth. In the case of a square mesh the following applies: (47) Starting from the similarity law shown in section 3.2 for the filtration of purely viscous media and with the expansion described before, the following similarity law results for the initial pressure loss of square meshes:

(49) Starting from the definition of the Deborah number (equation (46)), the following applies for Dutch Weaves: (50) Figure 11 shows how the measured pressure losses of different Dutch Weaves with different operating conditions group around the curve path of the similarity law according to equation (49). 5. Summary and outlook Based on the description of the pressure loss in a circular cylindrical pipe, universal similarity laws were derived for the description of the flow resistance of metal woven wire cloths during the filtration of highly viscous Newtonian and non-Newtonian fluids. In the scope of this work square meshes as well as Dutch Weaves (Single Plain Dutch Weaves, BetaMesh-weaves and Reverse Plain Dutch Weaves) were considered. Through the application of the principle of the effective viscosity it is possible to transfer the similarity laws to shear thinning media. Through the definition of a – representative shear rate γ· , an effective viscosity μeff can be determined from the flow behaviour of the medium under consideration, with which then an effective Reynolds number Reeff can be defined. This effective Reynolds number is used in place of the Reynolds number Re in the similarity laws. Moreover, the similarity laws have been extended to viscoelastic media, in which, on account of the occurring normal stress differences, an additional pressure loss appears compared to the purely viscous fluids. This is taken into account through the intro-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 67

Highlights 2013

Fig. 11: Similarity law for the initial pressure loss in the unloaded state on Dutch Weaves during the filtration of viscoelastic fluids.

duction of the Deborah number De in the similarity laws. With the developed similarity laws it is possible to pre-compute the differential pressure on metal woven wire cloths in the unloaded state for the filtration of highly viscous, purely viscous as well as viscoelastic media (for example polymer melts) with the help of the operating conditions and the structural parameters of the cloths. Up to now, the results are limited to single-layer metal woven wire cloths. The pressure loss of multi-layered filter media is also of interest and the investigations are extended to this. Furthermore, the main filtration process is being considered in more detail to be able to make a statement also about the dirt holding capacity another evaluation variable of a filtration process besides the initial pressure loss. Acknowledgements The results were developed within the scope of the research project “Untersuchungen zum Einsatz von metallischen Drahtgeweben bei der Filtration hochviskoser Fluide”. The IGF plan 16370 N/1 of the Forschungsvereinigung ForschungsGesellschaft Verfahrens-Technik e.V. - GVT, Theodor-Heuss-Allee 25, 60486 Frankfurt am Main, was funded via the AiF within the scope of the programme for the promotion of joint industrial research and development (IGF) by the Federal Ministry for Economic Affairs and Energy on the basis of a decision by the German Bundestag.

A filtration conference & exhibition

1st > 2nd October 2014 Steigenberger Hotel Berlin I Germany

Including speakers and topics tailored to the needs of key professionals from the filtration industry : — Development in clean water — Energy efficiency — Innovation in filtration — Indoor air quality — Pollution & emission control For exhibition opportunities contact Delphine Rens | delphine.rens@edana.org

New Literature: /1/ Bird, R. B.; Stewart, W. E.; Lightfoot, E. N.: Transport phenomena. John Wiley & Sons, Inc., 2 ed., 2007. /2/ Chmiel, H.; Schümmer, P.: Eine neue Methode zur Auswertung von Rohrrheometer-Daten. Chem.-Ing.-Techn. 43 (1971), pp. 1257-1259. /3/ Kopf, P.; Piesche, M.; Schütz, S.: Beschreibung des Druckverlusts von Drahtgeweben mit Hilfe von Ähnlichkeitsgesetzen. Filtrieren und Separieren 21 (2007), pp. 330-335. /4/ Michele, J.: Die Messung der ersten Normalspannungsdifferenz mit dem „Mechanischen Spektrometer“ – das Normalspannungsverhalten von linearen Polysiloxanen und Polyacrylamidlösungen. Rheol. Acta 15 (1976), pp. 15-22. /5/ Müller, M.; Piesche, M.: Untersuchungen zum Einsatz von metallischen Drahtgeweben bei der Filtration hochviskoser Fluide. Abschlussbericht zum Forschungsprojekt IGF 16370N der Forschungs-Gesellschaft Verfahrens-Technik e.V., Institut für Mechanische Verfahrenstechnik, Universität Stuttgart, 2013. /6/ Schümmer, P.: Zur Darstellung der Durchflusscharakteristik viskoelastischer Flüssigkeiten in Rohrleitungen. Chem.-Ing.-Techn. 41 (1969), pp. 1020-1022.

Panel discussions on filtration future trends and table tops exhibition

FO R INFO RMATIO N VISIT

www.edana.org

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 68

Highlights 2013

Optimized filtration mesh for ballast water management systems Perfectly designed filtration media and filtration packages to meet increasing demands M. Knefel* For maritime shipping, ballast water is a necessary evil. It is used as compensation for missing or shifted cargo and empty fuel tanks to maintain the required draught of the vessel and ensure the stability required for safe passage across the world’s oceans. According to the World Wide Fund For Nature (WWF), up to 7,000 different organisms are drawn on board with ballast water and discharged into new habitats at the port of destination. There, they may cause massive damage to the economy, the ecology and to human health. To put an effective stop to these unwanted hitchhikers, in 2004 the International Maritime Organization (IMO), a sub-organization of the UNO, agreed on a globally applicable Ballast Water Management Convention. Regulation D-2 of the convention prescribes the treatment required on every ship during intake and discharge of ballast water. One of the stages of effective treatment is fine filtration in the 10 to 50 μm range, and this is where Optimized Dutch Weaves (ODW) and Optimized Reverse Dutch Weaves (ORDW) made by GKD – Gebr. Kufferath AG play a key role. In a globalized world economy, shipping is the most important mode of transport. Two thirds of all goods are transported by sea, with an upward trend. Ships are getting bigger and, with speeds of 25 nau* Dipl.-Ing., MBA Markus Knefel GKD – Gebr. Kufferath AG Metallweberstraße 46, 52353 Düren / Germany Tel. +49 (0) 2421-803-0, Fax: +49 (0) 2421-803-227 www.gkd.de

tical miles (approx. 50 km/h), faster, too. Docking times for offloading and loading are getting shorter, and the frequency of port calls is increasing constantly. According to the UNO, this implies an increase in one of the four major threats to the marine environment: the infiltration of ecosystems with alien organisms through ballast water. The WWF estimates that ten to twelve billion tons of ballast water a year are transported across the oceans and discharged in ports of destination during the loading of new cargo. Countless alien animal and plant organisms are released along with the ballast water. Without natural enemies they spread rapidly in their new habitat, doing irreversible damage to the local aquatic flora and fauna. This leads to considerable economic losses through destruction of fish stocks, damage to harbor installations and blocking of industrial conduits. The WWF puts the value of the documented damage worldwide at more than 11 billion euro per year. Cholera bacteria and toxin deposits in common mussels pose a direct threat to human health. To protect against these dangers, the IMO laid down strict regulations in its BWM convention in 2004 for control and treatment of ballast water in ships with over 400 gross tonnage. This so-called D-2 standard will apply from 2016 at the latest to around 50,000 ships in the world’s merchant fleet. In the USA, the IMO limit values, in a lot more stringent form, have been law since 2012. In their territorial waters, ships are only allowed to operate if they have been certified by the US Coast Guard (USCG).

Fig. 1: According to the WWF, about twelve million tons of ballast water are transported annually by ships over the oceans and are discharged at the destination ports

Two-stage treatment To comply with the D-2 standard, manufacturers of onboard ballast water treatment systems use a two-stage approach consisting of mechanical separation followed by chemical or physical disinfection. After preliminary removal of larger organisms using a hydrocyclone, a fine filtration stage removes all organisms with sizes in the range of 10 to 50 μm. A subsequent stage of chemical disinfection is performed either with chlorine produced from the seawater by electrolysis, through addition of biocides like peracetic acid or hydrogen peroxide, or with ozone produced directly on site. The options for physical disinfection are heat treatment, ultrasound or UV irradiation. The expectations ship operators have on these extremely expensive ballast water treatment systems are correspondingly high. In addition to high throughput rates to ensure fast deballasting times in harbors, they want the systems to have small footprints, low operating costs, easy handling and retrofitting, long service life and low maintenance requirements. Reliable treatment of different water qualities, e.g. seawater, freshwater, brackish water, without environmental impact through chemical residues rounds out the high performance profile the ship operators expect these systems to deliver. High demands on filter media For the separation of organisms and particles > 50 μm, ballast water treatment

Fig. 2: Multi-layer Dutch Weave from GKD for ballast water treatment

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 69

Highlights 2013

Fig. 3: Illustration of a CFD flow simulation of the flow through a Dutch Weave

Fig. 4: Disc filter for ballast water treatment for the separation of particles greater than 50 µm

systems usually employ filter cartridges and discs with pore sizes between 10 and 50 μm. Depending on the manufacturer, they are cleaned automatically via back-flushing, back-pulsing , scraping or suction, so the filter media have to have lasting mechanical strength. This application also calls for filter media that combine the required fineness with high throughput rates and long service life, have a low tendency to clog, and are easy to clean. Another crucial requirement for operational reliability is an absolutely dependable filtration rate. However, in practice, it appears that not all filter media available on the market meet these diverse requirements. With their Optimized Dutch Weaves and Optimized Reverse Dutch Weaves, GKD offers a solution that fits the bill in all respects. To meet the customer’s specific requirements, GKD has a range of mesh constructions (laid, wrapped or sintered) and seam designs (welded, folded or clamped) available. Aggressive seawater and the type of disinfection stage used in the system put demands on the filter media materials. With Super Duplex (1.4410), Hastelloy (2.4602), Monel 400 (2.4360), SS 904L (1.4539) or SS 316L (1.4404), GKD has a wide range of materials to cover all needs. This allows perfect tailoring of the filter media materials to the particular system in terms of resistance to corrosion, chloride ions and heat, as well as non-stick and anti-fouling properties. Optimized Dutch Weaves and Optimized Reverse Dutch Weaves owe their outstandingly high performance to the special design of the stainless steel wire mesh. The slot-shaped pore geometries on the mesh surface are smaller than the pores inside the mesh. This design ensures that particles of the specified separation range are reliably separated on the surface of the wire mesh, while smaller particles pass through the larger internal pores

without clogging them up. The slight pressure increase at a constant contamination level makes for a higher dirt-holding capacity than other mesh constructions of the same fineness. In practice, this means longer filtration times without having to regenerate the filter media. Precise design and layout through simulations Woven wire meshes as filter media in ballast water treatment have an edge over other media, for example higher throughput rates at defined pore sizes, depending on the weave pattern used. To maintain this lead, the focus of research and development activities at GKD is firmly on continuous optimization of mesh geome-

WOVEN STRUCTURES FOR INDUSTRIAL APPLICATIONS

FILTERMEDIA + COMPONENTS

– technical woven wire meshes – 2D + 3D filter components – semifinished wire mesh products

GKD – GEBR. KUFFERATH AG | BUSINESS UNIT SOLIDWEAVE Metallweberstraße 46 | D – 52353 Düren | www.gkd.de Fig. 5: Candle filter for the separation of particles greater than 50 µm

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 70

Highlights 2013

tries and filter package layouts. Single-layer and multi-layer wire meshes are constantly being analyzed to improve their permeability and mechanical strength. In the process, experimental and computational simulation methods like the Finite Element Method (FEM) or Computational Fluid Dynamics (CFD) are used. FEM enables GKD to localize critical weak points inside the woven wire meshes or filter elements. Filtration tests in GKD’s own laboratory allow exact specification of retention rates for particles and organisms of defined sizes – including for soft or elastic particles – as well as prognoses of flow rates under various operational pressures. In this way, GKD defines the absolute mesh openings needed to guarantee the required cut-point accuracy. In the specific example shown here, particle measurement of a sample of seawater containing artemia was carried out, before and after filtration, according to the ISO 13319 standard. Analysis and measurement confirmed that all particles > 49μm were retained. In addition to selecting the appropriate filter media with defined absolute pore

sizes, GKD also determines the optimal layout of the various layers inside the filter package. Options are constructions with or without woven wire drainage mesh between the perforated base plate and the fine filtration mesh. Integrating a layer of drainage mesh increases the flow rate. GKD also uses CFD simulations to provide recommendations the best way to clean the filter media, for example through back-flushing or back-pulsing. Thanks to their sophisticated construction and customized design, Optimized Dutch Weaves and Optimized Reverse Dutch Weaves by GKD are a key factor in the full compliance of ballast water treatment with the strict regulations of the IMO convention. Compared to other filter media on the market, they offer the advantage of substantially higher flow rates, absolutely precise separation and reliable efficiency in a demanding physical environment. For this reason, they are already in use in the filter discs and cartridges of numerous ballast water treatment systems, with great success.

Automatic backwash filter improves performance in pharmaceutical process St. Schöpf, St. Strasser, L. Ertl* Due to the patented backwash principle, the Lenzing Technik OptiFil ® has the ability to filter down to very small particle sizes, while having lowest amounts of reject losses. This fact makes it the perfect choice, whenever valuable products are filtered. The Lenzing OptiFil ® : Originally, the technology was developed for high-viscosity spinning solutions and has recently been redesigned for the microfiltration of water and other low-viscosity fluids. The Lenzing OptiFil® is a fully automatic, continuous system that works according to the principle of depth, surface or cake filtration, depending on the selected type of filter material. A metal or synthetic fiber fabric or fleece is used as filter media, retaining particles of different sizes either inside or on its surface. After the pre-determined degree of contamination has been reached, the filter material is cleaned by backwashing a small quantity of filtered medium, with continuous filtration during backwashing. In detail, the filter material of the Lenzing OptiFil® is installed outside a perforated supporting structure (“perforated drum”). In case of cake filtration, a very thin filter cake (of typically 0.5 – 2mm) is formed inside the holes of the perforated drum during the filtration from the inside (Room P1) to the outside (Room P2). During the partial backwash from “Room P2” (Filtrate) to “Room P3” (Concentrate or Reject), the cake is completely discharged within a few seconds, using a small amount of filtrate to force it out of the filter. New cake formation already starts during backwash and is typically finished resulting in clear filtrate within less than 10 seconds. * Stefan Schöpf, Stefan Strasser, Lisa Ertl Lenzing Technik GmbH Werksstraße 2, 4860 Lenzing, Austria Tel.: +43 (7672) 701-2702 Fax: +43 (7672) 918-2702 E-mail: s.strasser@lenzing.com

Figure 1 shows the operating principle and setup of the filter in detail. In applications with valuable base materials, the Lenzing OptiFil® reveals its big advantage of low reject amounts due to the patented backwash principle. The Application: The low reject amount was the major reason for a renowned multinational pharmaceutical company to install the Lenzing OptiFil® in the production of a drug for treatment of high blood pressure. In this process, a fermentation broth is put into a reaction tank together with enzymes and mixed with a kieselguhr type of

Fig. 1: Operating principle Lenzing OptiFil®

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 71

Highlights 2013

filter aid. The distribution density of this filter aid is shown in Figure 2, which illustrates that it contains very fine particles, even smaller than ten microns. After finishing the reaction, the mixture is pumped into a centrifuge where the enzyme solids are filtered. However, a certain amount of filter aid, typically the fraction of the smaller particles, always migrates into the filtrate. Those filter aid residues need to be removed prior to the downstream ultrafiltration. Previously, disposable bag filters out of high quality 10 μm monofilament were used for this prefiltration step. In Figure 3: Process Flow Diagram, the process implementation of the Lenzing OptiFil® is shown. It replaces the previously used bag filters, positioned between the centrifuge and the ultrafiltration unit.

Fig. 2: Distribution density of the filter aid

Improvements: Cost for filtration: With each batch produced, up to ten bag changes were necessary. The previously installed 10μm monofilament bags have already been made out of a rather costly material. However, much more crucial was the fact that with each bag change about 30 liters of the very expensive base material were lost. Extrapolated to the batch volume, the product losses amounted to between 5 and 10%. Through implementation of the Lenzing OptiFil®, product losses for a whole batch were reduced to less than 1%. This means higher yield and therefore more product output with each batch. Additionally, no manual filter material change occurred due to the automatic cleaning.

F & S International Edition

No. 14/2014

Fig. 3: Process Flow Diagram

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 72

Highlights 2013

Filtrate quality: Even though premium quality filter material was used for the bag filter, the filtrate quality was fairly poor for two reasons: (1) For reaching reasonable life times and change intervals (high dirt holding capacity), a rather large filter area per flow volume had to be installed. This required filter area led to sedimentation effects in the bag, resulting in a non-uniform cake formation in the filter bag. (2) Furthermore, even the best monofilament material made of polymer filaments has a rather high variation in pore sizes, meaning that there are many pores being larger than 10 μm.

Both effects led to a poor performance of the downstream ultrafiltration unit, resulting in a low flow rate through the membrane, so that it became a bottleneck. As the Lenzing OptiFil® is operating an automatic backwash system, it was designed for achieving the highest flow/time instead of focussing on the dirt holding capacity. This led to a much smaller filter area (only about 10% of the bag filter system) and hence a uniform cake formation as well as high quality filtrate very shortly after backwash. Additionally, a special stainless steel weave was used, also with 10 μm pores, but with a much more uniform pore size distribution. Therefore the actual filtration performance is close to 1 μm! By using the Lenzing OptiFil®, the flow through the ultrafiltration system and the

module life time could be increased significantly so it does no longer represent a bottleneck in the process. Workplace, Health and Convenience: The fermentation broth contains ammoniac, which leads to high odour nuisance along with each bag change. The Lenzing OptiFil® is a completely closed system, using a double acting mechanical seal with a thermosyphone system to seal the rotating shaft to the outside, leading to zero emssions during operation. Productivity: Since the application of Lenzing OptiFil®, the company has been able to finish a batch in much less time, leading to a significant increase in production efficiency.

Aerosol technology is flying high Report from the 27. Palas-Aerosol Technology Seminar H. Lyko* Karlsruhe-based Palas GmbH, specialist for aerosol technology and filter testing, celebrated 30 years of existence during the 27th Aerosol technology seminar in September 2013. The intensive exchange between all players in the field of aerosol technology, already cultivated since the 4th year of existence, has certainly also made a contribution to the large number of new technical developments that have been established in the market through the years. The enterprise success was particularly emphasised by the awarding of the SEED and GROW AWARD of IHK-Technologiefabrik to Palas GmbH, during the evening event of the seminar.

Resonance as a part of corporate culture A jubilee always gives a cause for review. This initially happened here through the head of the Business Reframing Institute for Organisation and Humane management, Prof. Dr. Dr. Wolfgang Berger. He pointed out that the company foundation fell in the same period as the establishment of particle measurement technology as a business field of the VDI. He compared Palas to electromagnetic fields, in which resonance creates a positive working atmosphere as a measure of the direct vibration between fields transferred onto corporate culture. Because for complex systems, it is not momentary survival that is crucial but the constant adaptation to changing conditions, the success of the company Palas also lies in the fact that new strategies and decisions are communicated in a way that the people in the enterprise responsible for implementing them can identify with. *Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

Past and future of Palas In a two-part presentation, Managing Director Leander Mölter and authorised signatory Dr. Maximilian Weiß described the history of the enterprise and its products from 1983 till today, as well as the planned developments for the future. The spectrum of the aerosol generators, dilution systems, aerosol spectrometers and filter test systems was extended more and more over the course of the years. Technological development is particularly evident through the changes that the aerosol spectrometer has undergone. The principle of the Particle Counter Sizer (PCS) was invented in the beginning of the 1970s by Dr. Umhauer in the Institute of Mechanical Process Engineering at the University of Karlsruhe. The first PCS that Palas installed in a filter test system came from Russia, and afterwards the enterprise got the rights for in-house production. The original PCS sensor weighed 28 kg and still had peripheral zone errors that had to be corrected. With the development and patenting of T-aperture technology, the

peripheral zone error correction became superfluous, and also, the system allows the identification of coincidences. The welas® sensors designed with the T-aperture have become lighter and lighter and also less expensive since then. While the first welas® sensor nevertheless still weighed 11 kg, the smallest and lightest version, the Mini-sensor, now only weighs 200g. Dr. Maximilian Weiß, with the help of the so-called Ansoff matrix, illustrated in what direction the enterprise should grow in future. In the last five years alone, particularly the demand for high-quality devices with high flexibility and high measuring accuracy was satisfied with eight new device models. The future development aims on one hand at smaller lighter systems like mobile aerosol sensors triggered via wireless radio. On the other hand, one also wants to offer devices for the low price sector for applications where compromising on the measuring accuracy can be tolerated. Furthermore, through participation in different research projects with cooperation partners, one also wants

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 73

Highlights 2013

Fig. 1: Schematic diagram of nanoparticle production by spark discharge

to open new markets with new products. The Institut fĂźr Energie- und Umwelttechnik (IUTA) e.V. is a long-standing cooperation partner of Palas. The Director of the institute, Dr. Stefan Haep, reported on the fertile cooperation within the scope of the SME development BMWi IGF/ ZIM-KF. Since 1998, various different filter test benches have been developed and built jointly or implemented through the supply of components and with consultation by Palas. The more recent test systems, among them the test bench for large compressed air filters which has been reported on at this point several times already and still is being reported on (see below), originated within the scope of the ZF3 (Centre for filtration research and functional surfaces). Within the scope of IGF and other projects, numerous questions were worked on, like for example those concerning the behaviour of vehicle interior filters, the characterisation of filters for heating, ventilation and air-conditioning systems, for cooling lubricant deposition and for detection of fluorescent particles. The latter was implemented for the assessment of particle emission (and/or for the prevention of the same) through safety workbenches. A national and European Patent Application arose from this project (/1/). Generation of nanoparticles Nanoparticle aerosol generators for measuring tasks and testing tasks have already been available for some time in the market. Their output capacity lies in the order of magnitude of 5-7 mg/h. For the use of nanoparticles as a catalyst or chemical sensors, for the use in materials for hydrogen storage or electrodes for batteries or fuel cells, however, also for research into atomic clusters or the investigation of their toxicity, there is a need for larger production quantities. Prof. Andreas SchmittOtt of the University of Technology of Delft specialises in the spark generation of nanoparticles and explained more recent developments in the context of the EU project BUONAPART-E (Better Upscaling

F & S International Edition

No. 14/2014

Fig. 2: Schematic diagram of TXRF- method with a standing wave field

and Optimisation of Nanoparticle and Nanostructure Production by Electrical Discharges), in which he works together with Palas and the IUTA, inter alia. The function principle of electrical discharge is schematically shown in Fig. 1. The spark that originates between two electrodes, on account of the high temperature at this point, leads to the evaporation of electrode material and the vapour is driven from the gap through the accompanying gas stream and condenses to particles in the order of magnitude of a few nm, which, in the further flow course assemble into agglomerates. As materials, in principle all electrically conductive materials are acceptable. Through the use of mixtures of materials as the electrodes (alloyed or sintered, two different electrodes) mixed particles can also be produced. The size of the primary particles is adjusted via the spark energy, the gap width between the electrodes and the inert gas flow rate. If one wants to prevent agglomeration, one must dilute the particle cloud. However, one can also produce spherical particles to 100 nm diameter by aggregation and fusion of the primary particles in a heating track. The objective of BUONAPART-E were systems for the production of particle volumes in the order of magnitude of 5 kg per day. One increases the production volume by an increase of the repetition frequency of the electric spark per electrode pair and by parallel connection of several electrode pairs. Moreover, the frequency increase has no influence on the resulting particles as long as it is ensured that successive particle clouds do not touch. The repetition frequency could be increased within the scope of the EU project by the factor of a 100. The decisive factor was the development of new electronic switches and appropriate switching circuits. Analysis of nanoparticle aerosols As already explained in reports of past ATS, nanoparticles are counted by increasing the particle size by condensation of a working fluid and afterwards supplying them to an optical sensor. Dr. Maximilian

WeiĂ&#x; explained the special features of the Palas UF-CPC in comparison to condensation nucleus counters of other manufacturers that consist, in detail, in the feed of the working fluid, the special features of the optical sensor as well as in the volume flow control. The working fluid, Butanol or water, is actively being transported into the saturator via a spiralling U-channel, with other devices, the aerosol flows along a saturated fleece. In the first case, the exchange of the working fluid is simply possible because no residues can remain in the saturator. The optical sensor is operated with a LED as a light source, the measuring volume is limited thus that a single particle counting is possible also at high concentrations. This is not the case with sensors with which the entire aerosol volume passing through a laser light sheet is recorded. Via the variation of the volume flow, the Cut-Off (this is, the diameter with which the countable efficiency amounts to just about still 50%) can be shifted. Depending on the surface properties of the particles, different particle numbers can be measured. While Butanol condenses on nearly every particle, this is valid with water only in a limited manner, however, this is the more ecologically friendly working means, particularly for application of the measurement technology in interior settings. Depending on particle material, working fluid, the temperatures in the saturator and capacitor and the volume flow, one expects different diameters of the condensed drops and an influence on countable efficiency course and CutOff diameter. The experimental determination of these contexts i.e. the thermodynamic characterisation of a condensation nucleus particle counter by Palas, was carried out in a work by Susanne Baltzer at the Institute for Mechanical Process Engineering and Mechanics at the Karlsruhe Institute of Technology. For this, an experimental plant was used, in which a neutralised aerosol generated by a spark generator or an atomizer is classified in a DEMC and is supplied in parallel, either as a mono-or - polydisperse aerosol,

02_fs_international_2014_seite_03-98__ 21.05.14 20:11 Seite 74

Highlights 2013

Tab. 1: Examples of the replacement of pathogenic micro organisms by harmless species for test bioaerosols

Tab. 2: National and European measurement campaigns for the detection of aerosols in ambient air

to the CPC and a Faraday cup electrometer. In the tests described by Mrs. Baltzer it was found that the drop diameter increased with increase of the difference in temperature between saturator and capacitor, reduction of the aerosol volume flow and with increase of the aerosol particle diameter in the Cut-Off area. At the same time, the increase of the difference in temperature and the reduction of the volume flow cause a reduction of the Cut-Off diameter i.e. the measuring range is shifted towards smaller particle diameters. An improved chemical compatibility of the material combination particles-working fluid works

along the same lines. Overall, such studies allow a better adaptation of the measurement system to certain materials. Angle-dependent X-ray fluorescence analytics (TXRF) is a measuring method with which one can also determine the chemical composition of substance samples, besides particle size and particle concentration. It was explained by Dr. Stefan Seeger, Federal Institute for Materials Research and Testing, who had further developed this method for automated insitu-characterisation of aerosols, together with the Physikalisch Technische Bundesanstalt and the measurement instrument

Fig. 3: Post-colorized electron microscope photograph of a native bio-aerosol sample (Image: von-ThĂźnen-Institut, Braunschweig)

manufacturer Bruker. If one irradiates material samples with X-rays, the radiation is reflected with an energy typical for the respective chemical element. By varying the angle of incidence of such a ray gradually in the range below the critical angle of the total reflection, a standing wave field originates (see. Fig. 2). For a particle within this wave field, the course of the excitation intensity over the angle of incidence is dependent on the size i.e. how many antinodes and nodes of the wave â&#x20AC;&#x153;gridâ&#x20AC;? made from strips of high and low radiation intensity are covered by this particle. With the angle scan, one can derive conclusions about the particle size from the steepness of the leading edge and from the position of the first intensity maximum and conclusions about the volume of the detected samples can be drawn from the surface underneath the curve. Since this process is not a method for analysing a particle collective flying past in sampling pipes insitu and in real time, a suitable system was developed with which the sampling on a substrate in which the TXRF method is carried out is automated. Moreover, models had to be developed for the interpretation of the signal courses and be verified on the basis of reference samples collected by impaction. This part of the development was undertaken by PTB and BAM on the measuring station BESSY II of PTB. Bruker was responsible for development of the hardware for a mobile TXRF device for which - beside the components for the actual measurement as previously described - a gas purge for the measurement and a sample changer for a maximum of 96 aerosol samples were provided. Important applications for this measurement technology are the analysis of complicated ambient aerosols and workspace aerosols as are emitted, for example, by laser printers or welding machines. Supply and analysis of bioaerosols Bioaerosols contain biological material such as endotoxins, viruses, bacteria, fungal spores, pollen and plant residues and are identified - beside with the parameters

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 75

Highlights 2013

also decisive in inorganic aerosols - also by the biological activity of their components. Because biological activity is essential for different questions, bioaerosols often cannot be replaced by inorganic aerosols for research purposes and test purposes. Dr. Marcus Clauß from the Thünen-Institute in Braunschweig deals with the requirements for test bioaerosols, their production and dispersion as well as the applicable measurement technology. One example for a bioaerosol is shown in Fig. 3. Components of bioaerosols for test purposes should have no sensitising or toxic effect, however, they should have similar properties like their pathogenic relatives. Table 1 shows a selection of organisms which would be suitable for this. In addition, processed dust from henhouses is also used. For the application of the bioaerosols, the tenacity (survival ability) of the organisms is vital. Because these also depend on the measures for cultivation, production of suspensions or piles and for their transportation in the airborne state, these steps must be documented and be carried out according to the standard protocols as far as possible. Brush dosers are in principle also suited for the dispersing, if this concerns dry particulate materials, or compressed air atomizers are suited for the dispersing of suspensions; in the process, however, stress is exercised on the organisms. A more recent development for the nebulization of suspensions is a modified “Bubbling Aerosol Generator“, which is described in Simon et al. /2/ . For particle measurement, it is also possible to resort to established methods, wherein the optical particle counting, however, does not detect the individual species. The fastest detection, here, is in the collection on adhesive silicone surfaces and subsequent fluorescence microscopy with which the activity of the cells is detectable in addition. The measurement of the concentrations of germs in living rooms and in workspaces for risk assessment and verification of the observance of any limit values, according to the statements of Anja Konlechner, University of Agricultural Sciences Vienna, is not as unambiguous as would be necessary. Since one receives, according to collector device and analysis method, different results, whereby the actual particle number cannot be determined because of the unknown devicespecific losses, moreover, during the cultivation on a culture medium, only living germs are detected. To be able to compare collector devices with each other a calibrating chamber was designed together with Palas and then built, in which, ideally, a consistent, uniform particle concentration should be present. An aerosol is dosed into the accordingly purified supply

F & S International Edition

No. 14/2014

Fig. 4: Octocopter HORUS with a Fidas Ultra-Light Sensor (Image: Palas GmbH)

air, the concentration of which was measured on the inside with particle counters in 3 different heights in nine different positions. For this validation, an inorganic test dust was transferred into the air flow with a brush disperser. The particle concentration in the chamber was not equal in all positions but increased from top to bottom and with increasing distance from the aerosol dispenser, it decreased slightly, while the particle size distribution was the same in all positions. Taking into account these concentration profiles, one still deemed the chamber suitable for comparing germ collection systems. However, for the re-enactment of real bioaerosols, lower number concentrations would have to be generated than is possible with the generators used. Particle analytics in the ambient air Dr. Norbert Höfert, Commission on Air Pollution Prevention in the VDI, reported about current developments and open questions of particle measurement in the outside air as well as about current measuring campaigns. While the Directive series VDI 3867 about counting measuring methods was completed in 2013 with the publication of the Folio 6 about the electrical low pressure impactor, the works on the standardization of aerosol generators still continue in the Directive series VDI 3491 (cf. also /3/). However, the technological developments led to the fact that the first folios of the VDI 3867 (from 2008) have become outdated again. Currently, technical specifications for counting measuring methods are being promoted at European level , among the rest, on condensation nucleus counting. To this end, there is already a draft, publication is aimed at for 2014. Such technical specifications define minimum requirements for devices. They have a validity of a few years and can, if necessary, be converted into a European Standard. The

development of standards happens in parallel with different national and international measuring campaigns in which the possibilities and the suitability of counting measuring methods are demonstrated for a longer period, and the effects of ultrafine particles are examined (see Table 2). Beside the particle counting, the composition of the particles will also be in the focus in the subsequent programmes. Not recorded in the table are further local measuring campaigns dealing, for example, with the effect of environmental zones on air quality. New possibilities are also being opened up by unmanned aerial vehicles, which are equipped with measurement sensors. They serve the clarification of the aerosol emissions of certain sources and can elucidate their propagation paths . For this purpose, two aerial vehicles were also demonstrated directly at the conference venue. Firstly, an aerosol sensor was sent into the air, attached to a weather balloon , and the data collected by it could be traced on the screen in the conference room. On the other hand, the drone HORUS from Dresdner AirClip GmbH (see Fig. 4) was demonstrated, an ultralight, remotely controlled element made of carbon fibre composite material, which, according to load capacity, is built as quadrocopter, octocopter or dodecacopter. As a complete system for particle counting with recording of all size fractions occurring in the outside air or even indoors from 8nm to 40 μm, Jürgen Spielvogel, Palas GmbH, presented the system U-Range. In this system, the two equipment units U-SMPS with DEMCsize grader and the condensation nucleus counter are combined with the fine dust measurement system Fidas. In each of the 3.5 decades, up to 64 size channels can be resolved, moreover, the classical PM fractions can be detected in addition. Limit values for the fine dust pollution in the ambient air are still considered as mass concentrations and in spite of all

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 76

Highlights 2013

Fig. 5a) Test bench MMTC 200 for cleanable filter media, horizontal raw gas conduit (Image: Palas GmbH)

Fig. 5b) Test bench MMTC 3000 for cleanable filter media, vertical raw gas conduit (Image: Palas GmbH)

progress in the development of counting methods, the particle concentration in the many environmental monitoring stations is always also detected gravimetrically. For the practical execution in the state of Baden-WĂźrttemberg , this means that, in more than 50 measuring stations of the State measurement network, filters have to be collected daily and be equilibrated in the laboratory according to a routine required by the prEN12341 2012-02-10 (setting of rel. humidity and temperature) and they also must be weighed. Dr. Harald Creutznacher of the responsible State Environmental Agency pointed out the increase of the samples to be evaluated, to nearly 35,000 by 2011. To tackle them, an automatic weighing system was purchased and commissioned, with which 320 filters can be weighed successively on 20 plates and a total of 640 filters can be equilibrated at the same time. Every individual filter is automatically encoded so that it is assigned to a date and measuring station. By manual and automatic parallel weighings, it was shown that the failure rate of the machine with about 2%, by faulty coding / detection, is slight and that these errors can be avoided through plausibility checks by the laboratory staff.

Anja Baum of the Federal Highway Research Instiute reported about a special application case of the pollutant detection and measurement in ambient air. The object of the investigation was the length-dependent aerosol composition in road tunnels. The measurements were carried out with the background that later on, interior walls or internal fittings provided with TiO2- nanoparticles would be able to be implemented, through which, under the effect of a suitable light source, the catalytic degradation of the nitrogen oxides emitted by motor vehicles can be effected. As a preparation for this, the pollutant distribution was determined in the Rudower HĂśhe tunnel in Berlin, using an especially developed robot. This is a remotely steerable vehicle which is equipped with an aerosol collecting system and can take up different 19â&#x20AC;&#x2DC;- measuring instruments. The first measurements showed that the nitric oxide concentration throughout the tunnel length increases by a factor of up

Fig. 6: Schematic diagram of drum filter (Image: Institute of Process Engineering, Environmental Technology and Technical Biosciences of the Vienna University of Technology)

Fig. 7: Drum filter test bench at the Institute of Process Engineering, Environmental Technology and Technical Biosciences of the Vienna University of Technology

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 77

Highlights 2013

to 15, while the particle concentration increases by a factor of 11, however, for particles > 1μm more prominently than for smaller ones. Systems and examples for filter testing The testing of filter media and complete filters according to the standards applying for the respective product groups is implemented in different filter media test systems and filter test systems. Martin Schmidt, Palas, presented the individual versions and reported about new developments. As an example of the change of filter media testing with a changing standard, testing cleanable filter media may be mentioned. VDI Directive 3926 describes two test bench models, one each with horizontal and vertical raw gas conduit (cf. fig. 5a) and b)). The DIN ISO 11057, applying since 2011, describes the test bench with vertical raw gas conduit as a reference system, other measuring methods not explicitly excluded, however, they must fulfil equivalence criteria. In an especially designed test bench MMTC 3000 with two raw gas conduits, the differences are being examined together with Saxony Textile Research Institute (STFI) within the scope of a ZIM project. Another new development is a small compressed air filter test bench, the only measuring station where particle sizes and numbers can be measured at a pressure of up to 7 bar. Besides finished compressed air filters, flat media, such as they as used in compressed air filters, can also be tested. The small compressed air filter test system is constructed analogously to the large measurement path at the IUTA, on which Dr. Wolfgang Mölter-Siemens has been conducting research on large compressed air filters for some years. The object of the tests described here was the temperature influence on the filtration properties. The biggest part of the particles deposited are oil droplets collecting in the filter medium, which are discharged into a reservoir through a drainage layer. Because the viscosity of certain oils, with the temperature change by 30 – 40 K, changes by one order of magnitude and the surface tension also drops clearly, a significant temperature influence was expected. The temperature dependence of the filtration properties was determined with immersion saturated or aerosol-saturated filters with in each case fluctuating or steady temperature. It was found that the pressure loss of the filters correlated with the saturation degree and accordingly is not dependent on temperature, however, that the aerosol concentration in the pure gas behind a saturated filter increases with temperature, something that one attributes to reentrainment (detachment and entrainment of already separated oil). Dr. Frank Schmidt, University of Duisburg-Essen spoke on the influence of the relative humidity and the discrepancy between the filtration output of room air filters or vehicle-interior filters if they are tested at low (according to standard) or high relative humidity. The fact that high relative humidities significantly increase the pressure loss in the case of NaCl as a loading aerosol, the application of water droplets causes an extreme increase of the pressure loss and also the detachment of already deposited particles, has already been published in /5,6/. On account of these preliminary experiments, a new research project will start in 2014, which deals with the filtration behaviour of gas turbine filters for offshore operation. Gas turbine filters must demonstrate high filtration and separation efficiencies, even in harsh surroundings. The proof of the filter quality is provided according to the Saudi Aramco Standard “32-SAMSS-008”. Dr. Dirk Renschen reported about construction and commissioning of such a filter test bench. The details are published here on pages 85 - 88. Detecting the fading behaviour of an aerosol in an indoor space caused by a indoor air cleaner is to be regarded in principle as a filter test, too. Such indoor air cleaner are used, for example, in China for the reduction of exposure to both particles and gaseous air pollutants in indoor spaces. Hartmut Finger, IUTA, conducted

F & S International Edition

No. 14/2014

the testing according to the Chinese standard GB/T 18801-2008, and used cigarette smoke as a test aerosol. The particle concentration in the test chamber decays exponentially and the fading coefficient, with air cleaner in operation, as expected, is higher than the natural the fading coefficient (natural decrease of particles or adhering to walls). For an indoor air filter equipped with a certain filter, the generateable clean air volume flow CADR (for: clean air delivery rate) is measured, into which the fading coefficient and the volume of the test room are integrated. These coefficients and hence also the CADR value can also be determined separately for certain particle size fractions . With respect to the electrical output of the device, its cleansing performance arises from this as a quality feature for comparing different devices. Development of filter media and filters For the application of filter media, not only pressure loss and separation efficiencies play a role but also the operational safety. Thorsten Stoffel, Gea Heat Exchangers, described the disastrous consequences of flying sparks, as arises, for example, when welding, for filter systems. Synthetic standard filter media are considered to be highly flammable construction materials and conventional flame retardant equipment is often toxic. With GEAFireTex, he presented a new filter medium that has been classified into class B1according to a fire shaft test pursuant to DIN 4102-1:1998-05 (Standard for the fire behaviour of building materials). This classifies the medium as flame retardant, that is the material does not burn completely after edge ignition and no burning ingredients will drip or fall off. According to the DIN 5510-2-2009-5 for preventive fire protection in railway vehicles, the material is characterized with the flammability class S4, the drop forming capacity class ST2 and the smoke development class SR2. In addition, the FireTex®- filter mats have been certified according to the latest standard EN 45545-2:2013-08 (fire protection of railway vehicles) in accordance with requirements set R5 category “HL3”and therefore suitable without restriction, according to the latest standard, for air conditioning systems of railway wagons. The requirements made there are even clearly undercut. Johannes Wolfslehner, Vienna University of Technology, described his works on the modification of a drum filter for dust separation. The function principle of such a filter, which is used particularly in the textile industry and tobacco industry for high volume flows with low concentration of fibrous dusts, is evident from Fig. 6. In its original design with flat sheet filter media with high air permeability, the deposition of nanoparticles is not sufficient. Therefore, filtration experiments with pleated filter media which have lower air permeability, however, higher separation

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 78

Highlights 2013

efficiencies of nanoparticles - were carried out on the test bench shown in Fig. 7. As filter media , one medium with aluminium fibre web and one medium with ePTFE membrane were selected. Because of the clearly bigger filter area, the filter face velocity could be reduced so that the same pressure loss level could be kept as with a flat sheet medium. Here, the biggest challenge arose in the implementation of the cleaning through the externally mounted extraction nozzles. This was reflected in the substantial increase of the residual pressure loss causes by insufficient cleaning in the pleated depths. Here, measures are planned for the adaptation of the nozzle geometry and the cleansing strategy. Literature: /1/ Bankodad, A. et al.: Verfahren zum Bestimmen des Eindringens von Prüfpartikeln in einen Messbereich; DE102008029700A1 (2010) /2/ Simon, X. et al.; Aerolisation of Escherichiacoli and associated endotoxin using an improved bubbling bioaerosol generator; Journal of Aerosol Science 42 (2011) 517 -531

/3/ Lyko, H.: Filterprüfung, Emissionsüberwachung, Luftreinhaltung, Wolken- und Klimaforschung: vielfältige Einsatzmöglichkeiten für Aerosoltechnologie – Report on 26. Palas ATS; F&S Filtrieren und Separieren 26(2012) No. 6, p. 409 – 414 /4/ Birmili, A. et al: Atmospheric aerosol measurements in the German Ultrafine Aerosol Network (GUAN) – Part 1: Soot and particle number size distributions; Gefahrstoffe – Reinhaltung der Luft 69 (2009) No. 4, p. 137 – 145 /5/ Schmidt, F; Breidenbach, A; Suhartiningsih: Der Druckverlust von Luftfiltern bei hohen relativen Feuchten und bei Beaufschlagung mit Wassertröpfchen; F&S Filtrieren und Separieren 27 (2013) No.1, p. 6-9 /6/ Schmidt f.; Breidenbach, A.: Vergleichende Prüfung von KFZ-Innenraumfitlern, F&S Global Guide of the Filtration and Separation Industry (2012-2014) ISBN: 978-3-00-037568-2; p. 256 – 262 /7/ Renschen, D.; Schamberg, J.; Guttenbrunner, N.; Schneider, N.: Halbautomatische Prüfung von Gasturbinenansaugfiltern nach dem ARAMCO-Standard und zur Leistungsoptimierung; F&S Filtrieren und Separieren 27(2013) No. 6, pp. 374 – 378 /8/ Renschen, D.; Schamberg, J.; Guttenbrunner, N.; Schneider, N.: Semi-Automated Gas Turbine Inlet Filter Testing According ARAMCO Standard or for Performance Optimization; Presentation at Filtech 2013, October 22-24th, Wiesbaden, Germany

Energy-efficient generation and treatment of compressed air Report from COMVAC 2013

H. Lyko*

As an international, leading trade fair for compressed air and vacuum technology, ComVac has been an independent part of the Hanover Trade Fair since 2005. Both compressed air and vacuum technology are technologies in which energy and resource efficiency are an important trend, and one that is crucial in competition. This is partly because, for example, the use of compressed air is indispensable in almost any industrial plant. The VDMA Blue Competence sustainability initiative is also considering the trend towards the increase of energy and resource efficiency of this technology. The VDMA Association Compressors, Compressed Air and Vacuum Technology, as conceptual sponsor of the trade fair, supports users in this area, among others, with tools such as the EcoLexikon and the Compressed air model computer. The EcoLexikon offers an informal introduction to the comprehensive operation of compressed air and vacuum technology to the users. With the interactive Compressed air model computer, the energy consumption and saving potentials of a compressed air system are determined. A large number of components illustrated, especially compressors, but also filters and other processing equipment units, should conform to the requirements of greater efficiency and sustainability. Improved Process Control and Instrumentation also make their contributions. Compressed air qualities Since 2010, the new version of ISO 8573-1 is valid, in which the quality classes are defined for compressed air, namely separately for contaminants particles, water (vapour) and oil (vapour). Accordingly, the quality classes are also specified as the combination of three digits (2.3.1 means: Class 2 for particulate pollution, Class 3 for humidity and Class 1 for oil). In comparison to the previously applicable version ISO 8573-1:2001 (see /1/), it is noticeable that the limit values for particles in Classes 1 - 5 have not fallen, but risen. This development is contrary to the experience often observed that newer standards also include stricter limit values according to the technological advance*Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

ment of components and test instruments. In the new standard, on the other hand, more stringent requirements without general specification are categorized in Class 0. As the enterprise Parker Domnick Hunter advises in a brochure, the requirements for Class 0 must be defined in writing between the operator and device manufacturer. Moreover, they must be verifiable with the test methods or measurement methods described in Parts 2 to 9 of ISO 8573. This will provide users with high quality standards the possibility of getting compressed air quality tailor-made for their application. The compressed air quality classes > 5 (apart from humidity) are rather uninteresting in view of the application of treatment techniques, and they are not to be found in more demanding applications. Table 2 gives an overview of which quali-

ty classes are possible for what kind of application, namely as is communicated by Omega Air GmbH in Moers. To give consumers of compressed air a possible orientation about the qualities required for their applications, at the VDMA they are currently working on the creation of a new Standards Sheet /3/, in which, on the basis of the revised ISO 8573-1:2010, typical compressed air qualities for different uses are indicated, as well as measures to achieve these qualities, to monitor them and to uphold them by means of maintenance works. In addition, attention is paid to energy efficiency. Efficiency of compressed air systems On average, an efficiency increase of 33% and an effectiveness increase of 25% are considered feasible for industrial com-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 79

Highlights 2013

Tab. 1: Compressed air quality classes according to ISO standard 8573-1:2010, Reference conditions: 20°C, 1bar

pressed air systems /2/. To be able to exhaust this potential, Postberg +Co. Druckluftcontrolling GmbH developed, within the scope of a DBU project, a “Compressed Air Efficiency Manager” (CAE manager), with which, on the basis of traceable measuring data, the system efficiency of compressed air efficiency is supposed to be determined and feedback should be given to the user. Important characteristics of this manager were compiled to begin with in talks with users from different SME (small and mediumsized enterprises). This CAE manager was installed in a compressed air system established in cooperation with the Institute of Electrical Engineering of the University of Kassel. In this pilot plant compressed air is guided in closed circuits and is generated by the combustion of natural gas or diesel. The knowledge won from the project culminated in a training concept on the subject of compressed air efficiency for engineers and technicians. In the current year, these training sessions, 2 days in each case, are offered and carried out by Postberg. In large, technical compressed air stations, several compressors and treatment plants often work in parallel. For the setting of an optimum operation, i.e. energy and cost-effective, the manufacturer of compressors and processing technology, KAESER, offers a superseding system control (Sigma Air Manager). This serves the coordination of compressors and other system components with each other so that the ultimate pressure active on the consumer is not higher than necessary.

Tab. 2: Typical areas of application of different compressed air qualities

first task of this project was an extensive stock-taking and performance judgement of compressor types /4/. This document gives a detailed overview about compressor construction types and compressor functionalities, as well as accompanying standards. According to the required compressed air quality, displacement compressors (screw or piston compressors) or turbo compressors are used. As a coolant, water is often injected beside oil. In addition, dry-running compressors are being offered that will, however, be able to achieve lower end pressures because of the missing cooling medium. ALMIG Kompressoren GmbH offers watercooled compressors for the performance range of 15 to 110 kW. Besides the freedom from oil of the compressed air, as a side effect dust particles that are suctioned in are washed out by the water and the condensate that is caught does not need to be processed. In addition, the water guarantees higher heat dissipation in comparison to the oil. With water-injected screw compressors, the final compression temperatures lie at 50°C (cf. VDMA-EcoLexikon). The air-cooled screw compressors of Aerzener Maschinenfabrik are called Delta Screw. Their overall size VM 45 (differential pressure up to 3.5 bar, volume flows up to 3,200 m3/h and drive power to 250 kW) has now been reworked. A reduction of energy consumption of up to 6% should thereby be achieved compared with the preceding version. Other revised models of this series should follow. The improvements are achieved by a constructive advancement of the suction area, optimised flow channels and a new silencer on the pressure side that has technically improved flow and is free of absorbent. If higher pressure ratings than in the remaining compressed air network are needed for certain uses, a compressor for local pressure boosting is necessary. For this purpose, HAUG Kompressoren AG, for example, offers a compressed air booster that is free of oil, and with which the air can be increased from an initial pressure of between 5 and 9 bar up to 25 bar. The compressor works very quietly and is low in vibration, and so can be installed directly in the workplace.

Compressors In the consideration of the energy efficiency of compressed air systems, compressors play a central role. As was determined within the DBU project /2/, the proportion of useful energy, i.e. the energy that is really needed to compress the air volume flow, is only about 3% of the used primary energy for an electric compressed air compressor with 25 kW of electric output. This is based on an efficiency factor of just about 33% for the electricity generation from primary energy. Therefore, in the aforementioned DBU project, the compressed air is generated in a combustion process within a CHP system (combined heat and power generation). Also with electrically operated compressors, the potential for increasing the energy efficiency exists, for example, through frequency-regulated drives and heat recovery. To exhaust these potential factors is the objective of the EU LOT31 initiative (Ecodesign Preparatory Study on Electric Motor Systems, ENER LOT31). The

F & S International Edition

No. 14/2014

Fig.1: Compressor stage “effilience” by BOGE (Image: BOGE KOMPRESSOREN Otto Boge GmbH & Co. KG)

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 80

Highlights 2013

The compressor stage “effilience” (for efficiency & silence) (see Fig. 1) is at the heart of the screw compressor series of BOGE with the designation S-3. The selfwound rotors with a 5:6- screw profile work especially efficiently and quietly. Inter alia, this is also caused by a clearly lower rotating speed in comparison to present compressor stages. Filtration before and directly behind the compressor Some of the particles available in the outside air, as well as the greater part of the oil droplets entered in the case of oilcooled compressors, can be removed directly on the compressor. Mann+ Hummel have presented new filter elements for both functions, as well as oil filters for compressors. The ENTARON XD series (see Fig. 2) and IQORON series are intended for the air filtration in compressors. ENTARON XD is a two-stage air filter with a star-folded main element and an integrated pre-separator, which, at the moment, is available in overall sizes from 7 to 28 m3/min. In the second half of in the second half of 2013, an element for up to 40 m3/min was supposed to be added. Through their oval construction form and their design, the IQORON filters are particularly suitable for narrow installation rooms and cover air requirements from 4 to 14 m3/min. According to required pressure loss and available space, they are available in single stage and two stage implementation. The new Starbox2 (see Fig. 3) is an element for de-oiling the compressed air escaping from the compressor, which in comparison to conventional de-oiling elements, should have a pressure loss that is lower by up to 15%. This improvement is achieved by an optimised internal construction and an amended design of the connection area. This provides a reduction

in energy costs of about 400 € across the life span of the element (calculated with electricity prices of 8ct/kWh). Methods for advanced compressed air treatment According to the required compressed air quality, certain treatment equipment units connect to the compressor and here, for uses with air qualities of Class 0, it always seems to make sense to generate this quality directly at the point of use. The individual treatment elements are filters that retain both solid matter particles as well as liquid droplets, condensate separators, dryers, absorption systems (also in combination with aerosol filtration), as well as oil-water separators. Compressed air filters are central elements in nearly all treatment plants. According to the place of action and the required compressed air quality, they are produced with different filter ratings, also partially with activated charcoal or a catalytically acting layer. However, it always concerns cylindrical filter elements with a relatively similar construction. A protective fleece is pulled over an internal support cylinder of expanded metal (mostly stainless steel), and then the actual pleated filter medium follows, which again is surrounded by an external stainless steel protection cylinder. Completely outside, there is a drainage medium for the discharge of the separated liquid phase. In the stationary operation of the compressed air treatment, in this filter there is always a certain level of separated liquid /5/. Quality characteristics of newer compressed air filter elements are the very high filter surface density and a filter medium made from a fine fibre structure, because this means high separation efficiency or absorption capacity with relatively big volumes flowing through. BEKO Technologies promotes its new CLEARPOINT filter series

Fig. 2: Air filter ENTARON XD 40 for compressors from Mann+Hummel (Image: Mann+Hummel GmbH)

according to the 3E filter concept (3E stands for Energy Efficient Element) with these characteristics. These filters are flange filters where the user can decide whether he wants to use the filter with 100% volume flow and the lowest differential pressure, or with volume flow increased up to 11% in a performanceoptimised manner. According to the operating mode, volume flows of 1,420 m3/h to 31,240 m3/h and or 1,580 m3/h to 34.680 m3/h are implemented. The filter medium consists of a poly-fibre compound with thermally fused borosilicate and polyester fibres, which are equipped with an oil-resistant and water-resistant coating. Especially fine fibres are also processed in the compressed air filters of Ultrafilter GmbH. The nanofilter elements have been on the market for some years already. The pleating of the filter medium causes a usable filter surface that is about 4.5 times greater for the same element size than in wound filter elements. The compressed air filters of Ultrafilter are always provided with end caps of anodised aluminium. In other makes, plastic end caps are also installed depending on the application. For the further drying, i.e. for the reduction of the proportion of water vapour, both modes of action, cooling (condensation) and adsorption, are applied. The Hiross Zander Division of Parker Hannifin presented the newly developed “Antares Tandem Technology” (ATT), in which both drying principles are integrated. In this, the compressed air that is entering into the dryer flows first through a pre-filter and afterwards through a patented cooling circuit that extracts from it much of its moisture. This is followed by a high performance filter of the Zander GL series, which separates the other oil proportions and water proportions and particles. The filtered air is escorted into an ultra-com-

Fig. 3: StarBox2 deoiling element (Image: Mann+Hummel GmbH)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 81

Highlights 2013

tents of O2, CO, CO2, SO2 and NOX and water vapour. For the monitoring of these parameters, BEKO Technologies has introduced a new stationary system called METPOINT MMA (Monitoring Medical Air). With it, all parameters mentioned above are documented consistently and exceeding the limit values triggers an alarm. The treatment of compressed air for medical uses is performed, for example, in the ECOTROC system of KSI Filtertechnik. Here, several treatment components are combined with each other into a station. In detail it concerns a filter combination (1 μm/0.1 μm), a fully automatic, heatless, desiccant dryer with special desiccant filling, an activated charcoal filter, a catalyst, and a high-performance post filter (0.01 μm). Moreover, an electronic control with compressor synchronization circuit belongs to it.

Fig. 4: METPOINT MMA system for the monitoring of medical compressed air (Image: BEKO Technologies GmbH)

pact adsorption dryer. Before this equipment unit, the air has a pressure dew point of +3 to +5°C, and the adsorption dryer guarantees a standard pressure dew point of -40°C and is flexibly adjustable between +3 and -70°C. Afterwards the exiting air is warmed up by a patented air / air heat exchanger with the incoming air. The advantage of ATT technology compared with a series connection of desiccant and adsorption dryers lies in that the air leaves the refrigeration circuit in the saturated state. The subsequent adsorber needs only about one fifth of the regeneration air volume of a traditional heatless, desiccant dryer. All in all, ATT technology, which was announced to be introduced on the market with seven models, after a three year testing phase, should bring energy savings of 44% compared with competition systems. In the compression of outside or ambient air suctioned in, considerable volumes in condensate originate according to the relative dampness of the suctioned air and ambient temperature of the compressor, namely once on the compressor and on the other hand in the compressed-air drying. Thus one 10 m3/min compressed air station running under full load produces about 10 l/h condensate (according to Wortmann Druckluft GmbH). With oil-injected compressors, the accumulated condensate also contains dissolved and dispersed oil in such high concentration that the water must be degreased prior to discharge. Wortmann Druckluft GmbH has built oil-water separators for this purpose for over 35 years. In the standard system, Druckomat passes the condensate through an expansion and deaeration chamber before it enters into a deflecting and settling chamber, in which the majority of the oil floats up. The residues are removed by coalescing and activated carbon filters in a series connection. The newer Drukomat Plus system takes into account that modern compressor oils are often mixed with additives that hinder coalescence. Therefore, in this system, the socalled self-cleaning filter is arranged in between deflecting and settling chambers, which force the coalescence of the oil, so that the downstream filter must act only as a police filter.

Literature: /1/ Lyko, H.: Druckluft für hygienisch sensible Bereiche, F&S Filtrieren und Separieren 21 (2007) No.4, pp 245 – 248 /2/ Otto, P.: Entwicklung eines Systemhardwaremoduls “Compressed-Air-Efficiency-Manager” zur dauerhaften Senkung des Energieeinsatzes bei der Druckluftnutzung und leicht interpretierbares/nutzerfreundliches Nutzer-Feedback; final report on DBU project 28777, 2012 /3/ VDMA standard sheet 15390 Compressed air qualities - typical purity classes according to ISO 8573-1 and instructions on the generation treatment and monitoring of compressed air; draft /4/ Van Holsteijn en Kemna B.V. (VHK): Ecodesign Preparatory Study on Electric motor systems / Compressors ENER Lot 31-Task 1 Product Definition, Standards and Legislation, prepared for the European Commission, contract no. ENER/C3/413-2010 /5/ Mölter-Siemens, W.; Lauber, G.; Kerßenboom, A.; Lindermann, J.; Finger, H.; Haep, S.: Separation of ultrafine droplets with multi-layer fibroid filtermedia using the example of compressed air filtration F&S Filtrieren und Separieren International Edition (2012) pp 56 - 60

Medical compressed air Medical compressed air cannot be described only with the simple parameters of particles, water and oil, and moreover, limit values for various of its parameters are also specified in the European Pharmacopoeia. In detail, these are the required or maximum con-

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 82

Highlights 2013

Pressure loss of air filters for general ventilation at high relative humidity or exposed to water droplets F. Schmidt, A. Breidenbach, Suhartiningsih* In anticipation of DIN EN 779:2012 (particulate air filters for general ventilation) and the associated introduction of a minimum efficiency, once more a intense discussion was triggered about the benefit of the performance data of filters determined in the laboratory, in particular compared to operational data. For example, the suitability of the test dust and the decreasing electret effect of synthetic filter materials [1] were discussed. In addition, essential differences exist also with the climate conditions, comparing the standardized conditions according to DIN EN 779 or also ISO 11155-1 (vehicle interior filters) with the actual operating conditions. Therefore, some first systematic attempts were carried out to indicate the effect of high humidity on pressure loss during the exposure to different dusts, as well as the influence of water droplets on the pressure loss. Furthermore, fraction separation efficiencies were measured for relevant operating conditions.

1. Motivation The basic sense and purpose of standardization is illustrated by an overview article by Ripperger [2]: “Standards are means for the unification of tangible and intangible assets. So they also offer recognized solutions for constantly recurring tasks. One goal of standardization is to promote the national and international exchange of goods and services...” So it is not the primary purpose to depict “the reality”. On the other hand, the standard-compliant measurement may not be seen completely independently from the actual conditions in practical applications, since new media and filter elements are being developed specifically in view of the classification of the standard. * Prof. Dr.- Ing. Frank Schmidt Dipl.- Ing. Achim Breidenbach Ms. Sc. Suhartiningsih Universität Duisburg- Essen Institut für Verbrennung und Gasdynamik Lehrstuhl: Nanopartikel- Prozesstechnik Bismarckstr. 90, D-47057 Duisburg

Current standards of air filtration provide rather non-specific guidelines for the climate conditions in the test channel: DIN EN 779:2012 (filters for general ventilation) > Room air or outside air; no information about the temperature > Relative humidity should be < 75% ISO 11155-1 (vehicle interior filters) > Climate conditions: temperature 23°C +/- 5°C; humidity 55% +/- 15% Fig. 1 shows the measured values of relative humidity (location Mülheim/ Germany in 2011) in a cumulative representation. It can be seen that values of relative humidity ≤ 30% are only present approx. 1% of the year and values of relative humidity ≤ 70% only up to 30% of the year. This means on the one hand that a test channel without climate control would only be allowed to be used at certain times and, on the other hand, that for 70% of the year, conditions prevail that the standard does not allow for. Moreover, with condensed humidity (fog), sea spray or rain, the problem can

Fig. 1: Measured values of relative humidity (location Mülheim in 2011, provided by the Institute for Energy and Environmental Technology, Duisburg)

appear that water droplets themselves deposit on the filters. In this case, it results in a sticking together of the already deposited dust and leachings, which partially leads to a drastic increase of pressure loss. In particular in coastal regions and with offshore locations of gas turbines, crystalline salt particles or salt dissolved in water droplets are transported in the air and can also reach the turbine blades [3]. Fouling and corrosion thereby caused lead to a reduction in the performance and/or of the life-time of the gas turbine. Due to the good water solubility of sea salt, water that passes through the filter can also wash off already deposited particles again and transport them into the turbine [4]. These effects are not taken into account in EN 779:2012, which is used also to test coarse and fine dust filters for the supply air of gas turbines. 2. Test procedure Exposure of the filter media to test dust A2 and NaCl is done in a test channel according to DIN 71460, already

Fig. 2: Extended test channel at the University of Duisburg-Essen

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 83

Highlights 2013

Fig. 3: Media test (filter surface area: 600 mm x 300 mm, volume flow: 80 m3/h, T: 23°C)

described several times, at the Department of Nanoparticle Process Engineering at the University of Duisburg-Essen [6] (see Fig. 2). For the supply of A2, a rotating brush aerosol generator (RBG 1000, from Palas GmbH) is used. The generation of the NaCl aerosol is done by atomisation of an aqueous salt solution with an aerosol generator AGK 2000 (from Palas) with downstream diffusion dryer. A Welas 2300 aerosol spectrometer (from Palas), operating according to the scattered light principle, is used as the particle measuring instrument (Welas measurement => scattered light - latex spheres - equivalent diameter). The temperature of the volumetric air flow during the measurements is 23°C, and the relative humidity varies between 30% and 90%. Through the existing climate components, the temperature can be adjusted precisely to +/- 1.0 K and the relative humidity to +/- 2%. The mean velocity is approximately 15 cm/s.

Fig. 4: Media test (filter surface area: 600 mm x 300 mm, volume flow: 80 m3/h, T: 23°C)

loss according to variation of the humidity is also to be found after concurrent loading with A2 and NaCl, as well as in the case of higher NaCl loading. In Fig. 4 the hysteresis dependent on variation of humidity can be seen more clearly. Here the increase of the mass deposited on the filter is shown as a function of the rel. humidity. At a rel. humidity of 30%, approx. 1 g of dry salt particles are separated at first. From 70% onwards, the mass on the filter considerably increases and above the deliquescence point of 76% rel. humidity, solution droplets are present. After the reduction of the rel. humidity to 30%, the same mass is measured again as directly after the separation of salt particles. To explain the identified changes of the pressure loss and in particular to be able to

recognize the influence of humidity on the separated particles, SEM images have also been executed (Dr. Notthoff). Fig. 5 shows on the left side the formation of dendrites between the fibres through small salt crystals. These are no longer visible after moisture variation. Big solution droplets have formed predominantly at the fibre crossing points and after renewed drying, big salt crystals were present (Fig. 5 right side). Fig. 6 depicts the result if, at the same time, A2 and NaCl are separated. In the dry state (30% rel. humidity) A2 and NaCl particles are present side by side. After moisture variation, predominantly bigger mixing particles appear at the fibre crossing points. Though the present test results show a pressure dependence of the loaded filters

3. Test results In Fig. 3 the pressure loss is shown as a function of rel. humidity for different loading situations of the filter medium. Here, the pressure loss of the filter medium without any loading shows no dependence on humidity. In the second test series the unloaded filter medium is exposed to the A2 test dust until an additional pressure loss of approx. 20 Pa is present (new => 30%); the rel. humidity is slowly increased in steps of 10%. From 90% onwards it is reduced stepwise again. Also here, no dependence on the rel. humidity can be identified. If this test is carried out after separating NaCl particles in the filter media, one obtains first a continuous reduction of the pressure loss up to values of rel. humidity of approx. 70%. With a further increase of humidity to values of 90%, there is a substantial increase of pressure loss to more than 60 Pa, which, however, decrease to the original value by reducing the rel. humidity back to values of 30%. This behaviour of the pressure

F & S International Edition

No. 14/2014

Fig. 5: Media test + NaCl exposure without/with moisture variation

Fig. 6: Media test + NaCl/A2 exposure without/with moisture variation

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 84

Highlights 2013

Fig. 7: Media test with water droplets (volume flow: 80 m3/h, T: 23°C, filter surface area: 600 mm x 300 mm)

depending on the particle characteristics, extremely high pressure losses (known from some operating data) have not been ascertained. Therefore, another test series has been carried out, where finely dispersed water droplets are added into the test channel. The results are summarised in Fig. 7. After a very short time, the pressure loss rises from 25 Pa to a maximum value of 1800 Pa. This pressure value can be considered as a saturation state of the filter. It is independent of the rel. humidity in the channel and also does not change with a significant increase of the fed droplet number concentration (realised by an increase to an inlet pressure of 2 bar for the atomizer). Fig. 8 once more indicates the pressure loss as a function of the rel. humidity. After loading with NaCl and spraying with water droplets (4th test), the pressure loss shows no more dependence on humidity. As the weighing of the filter has also illustrated, the salt is flushed out of the filter by the first application with water. The pressure loss and the filter mass achieve nearly the initial value. In Fig. 9 the pressure loss is displayed as a function of rel. humidity. These tests have been carried

out on compact filters of fibreglass materials already used to protect gas turbines at locations close to the ocean. The pressure loss at this compact filter was similar to those measured at the filter media mentioned above. Fig. 10 represents exemplarily the resulting fractional separation efficiency (with DEHS as test aerosol) in new condition, after variation of rel. humidity or application with Isopropanol (IPA). A series of individual tests have been carried out on a filter medium. It was pre-treated with Isopropanol (IPA) and charged with NaCl particles or water droplets. The results of these individual measurements can be summarised as follows: The fractional separation efficiency in new condition decreases significantly after IPA treatment. After humidity treatment or charging with water droplets the curves of the resulting fractional separation efficiencies lie between those of the curves for material in new condition and after IPA treatment and is dependent on the exposure time at higher values of rel. humidity or the intensity of the droplet exposure, respectively.

Fig. 9: Pressure loss measurements of used gas turbine filters (carried out on the HVAC test bench according to DIN EN779-2012 [5] and provided by the Institute for Energy and Environmental Technology, Duisburg 84

Fig. 8: Media test with water droplets

Conclusion: Initial test data show: > The variation of relative humidity causes a significant change of pressure loss if NaCl is present as loading aerosol. > If additional water droplets are applied, then the pressure loss rises extremely. Existing particle layers may be washed away. => The transferability of performance data according to standardized tests is also limited concerning the influence of the humidity. Literature: [1] Schmidt, F.; Breidenbach, A.; Däuber, E.; Ergebnisse der RLT-Filterprüfung nach EN779 im Vergleich zu Messungen an Filtern aus dem Betrieb; Chem. Ing. Techn.; 84, No. 6, 808-812 (2012) [2] Ripperger, S.; Sinn und Zweck der Normung und Standardisierung auf dem Gebiet der Separationstechnik, Filtrieren und Separieren 25 (2011) No. 2 [3] R. L. Loud, A. A. Slaterpryce, Gas Turbine Inlet Air Treatment; Schenectady, General Electric Company publication, New York 1991 [4] T. Schroth, M. Cagna, in Proc. of GT2008, Gas Turbine Technical Congress & Exposition, Berlin 2008 [5] Schmidt, F.; Däuber, E.; Lindermann, J.; Praxisnahe Tests der Tropfenabscheidung von Koaleszern, Chem. Ing.Techn. 82, No. 5, pp. 651-656, 2010 [6] Breidenbach, A.; Schmidt, F.; Finger, H.; Haep, S.: Prüfung von KFZ-Innenraumfiltern - Dieselruß als Prüfund Beladungsaerosol; Gefahrstoffe –Reinhaltung der Luft 69(5), 2009, 189-193

Fig. 10: Fractional separation efficiencies of the tested filter media (DEHS test aerosol)

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 85

Highlights 2013

Semi-automated Gas Turbine Inlet Filter testing according ARAMCO standard or for performance optimization D. Renschen*, J. Schamberg*, N. Guttenbrunner**, N. Schneider** There is a growing demand for improved gas turbine air filters with regard to the energy efficiency in operation and filtration efficiency in view of the increasing demands of modern gas turbine for inlet air quality. Therefore, supported by DMT as air filter testing specialist, Bilfinger Gerber set up a special test rig for testing semi-automatically gas turbine inlet filters according to ARAMCO standard (32-SAMSS-008) which describes a harsh gas turbine filter testing program. The process measuring and control technology allows an automatic run of the different test phases as required by the ARAMCO standard. Only for dust sampling for the determination of the filtration efficiency manual support is needed. This facilitates the sequential tests of different air inlet filter types e.g. for optimization of the filter performance within R&D. 1. Introduction New construction plus the growing replacement market were supposed to boost turbine inlet air filter sales to over $400 million in 2013. This segment would account for 6% of the $7 billion air filter market, as claimed by the McIlvaine Company in its Air Filtration and Purification World Markets report. There is a growing demand for improved gas turbine air filters with regard to the energy efficiency in operation and filtration efficiency in view of the increasing demands of modern gas turbine for inlet air quality. A lot of the existing and future gas turbine power plants are erected in arid or semiarid zones with potential high dust exposure. Usually air inlet filter for gas turbine applications are tested for classification according EN 779. This gives a clear indication about performance in view of initial pressure drop behavior and efficiency against DEHS and ASHRAE test dust. But this standard gives no information about behavior about cleanable filter systems and its performance within operation especially under severe operation conditions. For the filter media of such air inlet filter pretests can be made by the VDI 39261 (“Testing of cleanable filter media” or ISO 11057) to receive information about the ageing behavior and performance against test dust from the application. The so called Saudi Aramco DeskTop Standard “32-SAMSS-O08” is a “Materials System Specification” including a test specification. The purpose of the “Speci* D. Renschen, J. Schamberg, DMT GmbH & Co. KG Product Assessment Refrigeration & Air Quality Am Technologiepark 1, D-45307 Essen / Germany ** N. Guttenbrunner, N. Schneider Bilfinger-Gerber GmbH, D-44379 Dortmund / Germany

F & S International Edition

No. 14/2014

fication” is to ensure the ability of an inlet air filtration system for combustion gas turbines to withstand the rigid environmental conditions in a high ambient dust desert area, with relative humidity ranging from zero to 100% (coastal zones). The intent of such an inlet air filtration system is to limit the ingestion of airborne contaminants such as sand, dust and alkali metal salts (sea salt aerosol) so as to prevent gas turbine compressor erosion, fouling, cooling air passage blocking and hot corrosion of the turbine section components. The system specified is intended for continuous duty, high efficiency air filtration. Therefore, the air filtration systems covered by this specification are limited to the automatic pulse-jet self cleaning type. For other environmental conditions with much less dust like in Central Europe the test standard “Air intake filter systems for rotary machinery - Test methods - Part 1: Static filter elements” (ISO/DIS 294611:2010) was just developed for filter which are not cleanable. DMT as air filter testing specialist and third party testing institution supported Bilfinger Gerber in setting up a special test rig for testing semi-automatically gas turbine inlet filters according to the Saudi Aramco Material System Specification 32SAMSS- 008 (ARAMCO standard). Appendix II of the Aramco Specification describes a test procedure for a harsh filter testing program divided in several phases to detect the capability of resistance against desert sand storms. 2. Test Standard Saudi Aramco Material System Specification 32-SAMSS-008 (Aramco Standard) Most chapters of the specification are dealing with material and system require-

ments for an inlet air filtration system for a gas turbine installed in a filter house. It includes requirements for the filter element (structure). Further on the specification describes requirements for the system performance, mechanical and electrical instrumentation and the filter house itsself. In the appendix information about protective coatings are summarized and a qualification and performance tests for materials e.g. against salt fog is defined. For planning a test (rig) it is important to consider the requirements for the pulse air system (Chap. 5). All other important figures for the test equipment, test materials (test dust), test conditions and the test procedure of filter elements and the reporting of the results are included in Appendix II (Filtration Testing Requirements). 3. Test Rig Design and Validation In spite of the requirements and descriptions for the appropriate equipement for the test rig according appendix II of the Aramco Standard a lot of details had to be fixed. By the target to set up a semi-automatically working test rig further requirements had to be defined especially for process measuring and control technology devices and the software for data aquisition and control. Other requirements resulted from the necessity to set up a test system which generates data which are valid in view of a certification of filter elements which shall be tested by means of this system (DIN EN ISO/IEC 17065). Therefore, all these measurement devices, which are integrated in the test rig, must have a precise specification and be able to be calibrated. After finalization the test rig as complete testing system has to be validaded (EN 45002 “General criteria for the assessment of testing laboratories”). The

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 86

Highlights 2013

Tab. 1: Survey of measurement points within the Aramco standard test rig

Tab. 2: Basic modules of the Aramco standard test rig

1 Intermediate compressed air tank

Compresso with air drier

Compressed air tank with diaphragm valves, blow tubes connection and manometer

Pressure regulator valve

3 Controller blow tube diaphragme valve

Diaphragme valve

2 blow tubes with 6 outlet nozzles (each)

5 + 6 Test duct Section for dust sampling

validation test method characterizes the method features of the Aramco Standard test procedures. The further target of validating the test rig and the test procedure (by DMT) is to ensure that testing of gas turbine filter cartridges (which is performed at a testing site of Bilfinger Gerber), is according to the Aramco Standard. For certification of tested filter types the validated test procedure has to be secured even against manipulation within data acquisition. DMT acts as mandated neutral evaluation service. Bilfinger Gerber as testing laboratory. Therefore, DMT has to ensure the integrity of the measured data when measured in the external laboratory. The testing procedure of filters according the Aramco standard is diveded in six phases during which sets of different parameters and measurements are carried out. Essentially, for the Aramco test procedures (phases 0-5), the following data have to be measured continuously: - Differential pressure on the Wilson flow grid (volume rate test duct – up to 12750 m3/h): Δpflow grid - Ambient pressure (0 – 1100 mbar): pbarom. - Ambient temperature: Tambient - Relative humidity: ϕambient - Differential pressure inlet air filter (3 pressure transmitters: I 0 to 300 Pa, II 0 to 1000 Pa, III 0 – 10000 Pa): Δpfilter - Compressed air vessel pressure (up to 10 bar): pcomair - Valve opening time (is adjusted to the specified value, not measured) - Volume rate at sampling point 1 (Mass · flow controller 0 – 200 l/min): V PN1 - Volume rate at sampling point 2(Mass · flow controller 0 – 200 l/min): V PN2 - Dust concentration raw gas (upstream, via dust generator, gravimetric dosage, balance 0 to 50 kg): m· test dust The data acquisition and transmission of measured values to the central computer (control station) have to be ensured. All measurement parameters are summarized in table 1. 4. Semi-automatically gas turbine inlet filters test rig The new test rig of Bilfinger Gerber is composed of several different modules which are suitable for a testing according the technical specifcation “32-SAMSSO08”. The test setup is divided into 10 modules: 1) Compressor (supply) for compressed air tank (vessel) 2) Compressed air tank 3) Diaphragm angle valve

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 87

Highlights 2013

4) 5) 6)

Blow tube/pipes Test duct Test section for dust sampling/dust sampler system 7) Dust Injection System (ejector) 8) Differential pressure measurement system (for the filter) 9) Fan (suction mode) 10) Steam vaporizer for humidification The different modules are shown in the table 2 (below) in more detail. All components of the resulting test rig are connected to the central control computer. The software program aquires all data and controls different components like the frequency controlled fan for automatic adjustment of the volume rate to the specified value. Further on test dust generation, pressure pulsing and sampling of the dust on the upstream side (clean gas) of the test filters is done automatically within one test phase according the Aramco standard. Only the change of the sampled filters for the gravimetrical determination of the test dust in the clean gas has to be done manually at the defined time intervalls by the operator. By the central computer the measured data are controlled against special limits (for instance upper pressure drop). If this limit value is reached the test is stopped automatically and the operator is informed by an alarm signal. A picture of the screen of the software control program is shown in figure 1.

Tab. 2: Basic modules of the Aramco standard test rig

Dust sampling system

Test dust containment

Dust injection system (nozzels)

F & S International Edition

No. 14/2014

Controller unit

8 Differential pressure transmitter

Measuring points

Radial fan unit axial-flow fan

Ventilation flap with actuator unit

5. Validation of the test rig and the test procedure To ensure a correct test procedure as base of a certification the whole test rig was thoroughly examined. The examination was performed in particular with regard to the measuring sensors, in order to inspect the proper positioning and calibration of the measuring equipment. First the status of calibration of the different measurement devices was controlled. This means as part of the test rig validation, the calibration certificates were checked and included in a test rig manual. All sensors of the test equipment have to be calibrated periodically according to the requirements of the agreed quality control system (tab. 1). By testing inlet air filters according to ARAMCO standard under supervision of DMT the validity of the calibration certificates will be inspected. For the dust generation system the mass flow of dust was examined by DMT and the system calibrated accordingly a internal QC procedure. The quality of dust dispersion of the dust generation system was examined by measurement with a DMT modified Berner type impactor in comparision with the particle size distribution of the used test dust (fig. 2).

Mass flow controller (dust sampl.)

Steam vaporizer (front outside of test chamber)

The data measured by the different sensors in the test bench are acquired by the central control computer at a rate of about 5/min. A resulting table is arranged to a continous time scale to which the data of each sensor are included in the according columns. The data are broken down by each of the five test phases and stored in a “csv-file”, which is transferred into an “Excel” file-format which can easily be edited for the following data analysis. To validate the semi-automated filter test procedure adapted according the standard 32-SAMSS-O08, two equal tests for the same type of an inlet air filter were carried out at the DMT and at the Bilfinger Gerber test rig. This first validation test on

Humidity and temperature sensor (inside of test chamber)

the test bench at Bilfinger Gerber was accompanied by a test engineer of DMT. Within this test method validation following work packages have been made by DMT: - Definition of the requirements for this external certification testing - Definition of the validation procedure for the test method - Round Robin test between DMT and Bilfinger Gerber - Check of the integrity of the raw data and the data transfer This whole validation process was fullfilled by Bilfinger Gerber successfully.

02_fs_international_2014_seite_03-98__ 21.05.14 20:24 Seite 88

Highlights 2013

DMT gains access to the acquired data at the external test rig within an ongoing measurement and to complete data sets after completion of the tests. The intrusion by DMT on the computer system of the external test rig is made via Internet access about the program “TeamViewer”. Thus, it is possible during the experiments to monitor the current operating state. By comparing data documented during monitoring via “TeamViewer” with the raw data set after finshed testing validity of data can be verified. 7. Capabilities of this unique test system

Fig.1: A picture of the screen of the software control program

Fig. 2: Comparison particle size distibution of the test dust with the dispersed test dust sampled from the raw gas in the test rig

6. Testing and Certification of inlet air filter elements The testing of filter cartridges according the Aramco standard at the new test rig will be performed at the site of Bilfinger Gerber. Nevertheless, the test results will be certified by DMT. This assumes that

Bilfinger Gerber and DMT closely cooperate in the examination of filter performance for other customer. Therefore, the cooperation is based on a secrecy agreement between both parties which secures confidentiality of all information within third party testing.

The semi-automated “Aramco” test rig of Bilfinger Gerber provides results which are plausible and trustworthy. It is designed appropriately for the inlet air filter test procedure within the specification 32-SAMSS-O08. By this semi-automated filter system testing can be done much easier and with less man-power. Beneath these basic features the test rig also incorporates a steam humidification and an air circulation system. Therefore, different other testing options are possible due to an “open” air inlet which is integrated in a special test room. This test rig facilitates the sequential test of different air inlet filter types (not only pulse filters) and is suitable for optimization of the filter performance within R&D. Therefore, beneath the testing options according the standard further tests can be made like optimization of the distance between blow pipe and filter cartridge or of the consumption of compressed air, determination of the efficiency of coalescer filter for mist elimination etc.. Literature: /1/ 32-SAMSS-O08, Materials System Specification for Inlet Air Filtration Systems for Combustion Gas Turbines, Saudi Aramco, 2008 /2/ DIN EN ISO/IEC 17065, “Conformity assessment Requirements for bodies certifying products, processes and services” /3/ EN 45002, “General criteria for the assessment of testing laboratories” /4/ ISO/DIS 29461-1:2010, “Air intake filter systems for rotary machinery - Test methods - Part 1: Static filter elements” /5/ VDI 3926-1, “Testing of cleanable filter media Standard test for the evaluation of cleanable filter media”

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 89

Highlights 2013

Secure waste air filtration in laboratory autoclaves Integrated water intrusion test to fulfil the requirements in the area of biotechnology C. Grumbach, P. Czermak* The thermal inactivation of natural or genetically modified micro-organisms is usually done in biological and medical laboratories in a steam sterilizer (autoclave) under defined parameters for pressure, temperature and time. The discharge of these organisms from the autoclave chamber - for example on the waste air or waste water path - must be prevented as a function of the risk group of the organisms [1, 2]. In particular in the course of the heating phase of the autoclaves, when the air is removed from the chamber and steam flows in, biological working substances can get into the workspace as aerosols. In practice, sterile filtration is often used for the inactivation of this waste air. For this purpose, the waste air is routed through hydrophobic membrane filters, completely removing its germ load from the waste air flow with proper handling. Sterile filtration can be associated with risks under certain conditions [3]. The filters have hitherto not been tested for their functionality (integrity), although this is technically basically possible and must be considered as necessary in terms of a risk assessment. On account of recent developments in the field of sub-statutory regulatory frameworks, changes are to be expected here [4]. The possible risks are hereinafter illustrated by means of practical example and the system design of a new test system, which was developed on the basis of the special requirements for integrity testing in laboratory autoclaves, will be presented. Introduction The methods for the inactivation of micro-organisms by means of saturated steam have been successfully carried out in laboratories of the pharmaceutical and medical industry in autoclaves for a long time. The sterilisation success of such processes causes complete removal of the air from the autoclave chamber as well as from the autoclaving material/items to be sterilised. The waste air removed from the autoclave in the process is normally discharged from the system into the environment. In biological laboratories, contaminated process waste air of autoclaves, from the protection and safety level 2 onwards, must be purified according to the Genetic Engineering Safety Ordinance (GenTSV) [2]. The removal stage of the air is followed by a downstream sterile filtration, in order to prevent around a discharge of airborne pathogenic germs into the environment. For this application, membrane filters are being used, through the narrow-pore membranes of which (as a rule 0.2 Îźm diameter) bacteria are reliably retained [5]. However, long-standing experience shows that sterile filtration can be associated with risks under certain conditions [3]. The filtrate is sterile for as long as the integrity (i.e. the intactness) of the * Dipl.-Ing. Carsten Grumbach Prof. Dr.-Ing. Peter Czermak Technische Hochschule Mittelhessen Institut fĂźr Bioverfahrenstechnik und Pharmazeutische Technologie www.thm.de/ibpt peter.czermak@kmub.thm.de

F & S International Edition

No. 14/2014

filter is guaranteed. If bioaerosols are thrown up into the environment from the waste air path of the autoclaves in the event of a failure of the filter, this creates a significant risk to humans and the environment. The manufacturer specifications for the service lives of the filters ordinarily refer to ideal filtration conditions. The selection of a filter suitable for the process, as well as its coordination with the manufacturersided specifications on maximum flow rate, pressure load and operating temperature, are decisive for its integrity and service life [6]. Up to now, the functional capability of the autoclave waste air filters is not routinely checked in the areas mentioned above, although this is technically basically possible. Recent developments in the field of sub-statutory regulatory frameworks name the water intrusion test procedure as appropriate safety measure [4]. Risks of sterile filtration on autoclaves The qualification of sterile filters is an essential task of the filter manufacturer. However, potential damage in transit, damage by improper handling or, for example, through steam sterilisation, can never be excluded completely [7]. Also, damage to the filter through the filtration process is possible. A great risk is already present with the installation of the sterile filter. If the O-ring seal necessary for the adaptation to the filter seat is damaged in the process, sterile filtration is no longer given for the subsequent sterilization

cycles. Loss of the integrity of a sterile filter can already appear on account of microscopic damages or changes of the pore structure on the filter, which are macroscopically not to be identified. Therefore, verification is only possible by means of sensitive measuring methods, which are called filter integrity tests. With such integrity tests, a statement with regard to the filter integrity becomes possible with the aid of non-destructive technologies [7]. In addition to a verification of the filter integrity, an integrity test, besides, allows the inspection of the waste air filter system for possible leakages which can appear after its installation. Integrity testing The integrity testing of sterile filters supplies the microbiological proof that a sterile filter delivers a sterile filtrate when undergoing exposure to a certain number of bacteria. It thus constitutes an essentially necessary quality assurance measure for sterile filtration. Only through a test will it be directly demonstrated that the test organisms used are reliably being retained by the filter [8]. However, this test kind concerns destructive testing, through which the filters become useless for their proper application. The filter elements are subjected, in addition, within the scope of a product / process specific validation, together with the respective product, to a bacterial load test under process conditions. According to the ASTM Guideline [9] a filter is only allowed to be classified as a sterile filter if it, upon bacterial expo-

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 90

Highlights 2013

pressure [mbar] flow rate [ml/min] flow pressure pressure rate [mbar] [ml/min] build up

stabilisation phase

measure phase

time [min]

Fig. 1: Waste air filter system of a laboratory autoclave with the measurement path of an adapted water intrusion test

sure with 107 of a model organism of the species Brevundimonas diminuta, delivers a sterile filtrate per square centimetre of filter area. This test constitutes a â&#x20AC;&#x153;WorstCaseâ&#x20AC;? scenario with an immensely high bacterial density and substantiates the high level of safety. The respective manufacturers of the filter elements are responsible within the scope of their product validation for the correlation of the test values with microbiological examinations. With the help of such destructive load testing, defined limit values are specified for the non-destructive integrity test by means of suitable model organisms [7]. Water intrusion test To decontaminate the waste gas stream of laboratory autoclaves, gas-permeable hydrophobic membrane filter elements are usually used. The water intrusion test is a suitable integrity test method for testing hydrophobic membrane filters. Here, the hydrophobic filter element is flooded with water and a test pressure, which is below the penetration pressure of the filter, is applied to the measurement system. The water steam originating on the filter membrane by evaporation passes through the filter membrane. A measurement of the pressure drop provides the water intrusion rate (WIR), which can be correlated with the bacterial retention rate of the filter [10, 11]. A breakthrough or failure of the filter elements can be detected at an early stage in this way and be prevented. The static pressure drop method is the technically simplest test method. The test run is divided into three phases (see Figure 2): a pressure build-up phase, which serves for the establishment and for the stabilisation of the critical constant pressure conditions necessary for the test. Only under these conditions can the intrusion flow to be measured be stabilised, via a stabilisation time typical for the filters, to a steady flow rate. By means of the subse-

Fig. 2: Pressure drop method (schematically)

quent measuring phase, the intrusion flow can be determined from the pressure drop [12]. Factors that influence the integrity Beside the selection of a filter suitable for the process, the process terms are also decisive for its integrity and service life [6]. A rapid failure of the filter elements can appear in the autoclave when large amounts of steam are extracted via the vacuum pump under high temperatures and the filter, under these conditions, is subject to an increased differential pressure. The process parameters for temperature, autoclaving time and the amount of the drawn vacuum arise from the chosen autoclaving method and the contamination to be sterilised of the autoclave. Hence, the autoclaving method is of central significance for the filter integrity because the waste air filters are polluted through an evacuation of the autoclave chamber in connection with increased temperatures or through the permeation by means of saturated steam from the chamber. With regard to the mechanical stability of filter cartridges, it should be considered which maximum pressures and above all pressure fluctuations can appear in the running system. Extreme pressure pulses can be caused here, systems-specifically, by quick-closing valves or equipment units downstream of the filter (for example vacuum pumps). The temperature problem Basic requirements for the reliable execution of a water intrusion tests are the availability of stable ambient conditions. Above all, the temperature stability is important because a temperature variation leads to a corresponding pressure fluctuation in the measurement system which is then falsely evaluated as water intrusion. The temperature problem gains further

complexity still for the measurement in that the surface tension of the wetting liquid is also dependent on temperature, so that a temperature change can also have a direct influence on the flow rate to be measured [13]. Requirements description Many test systems on the market are designed very generically for filter measurements and are on a par with the investment costs of laboratory autoclaves [14]. Such test systems are seldom suited for small plants, on account of the additional high qualification costs. Such costs can be reduced by the fact that a test is designed and qualified for only certain filters types and sizes. Besides, a test system should be able to be easily integrated into various plants. Besides, different technologies have established themselves on the market in order to determine the flow rate across the filter membrane. A very inexpensive variation is the static pressure drop method which allows the determination of water intrusion from the measurement of an individual pressure drop within the measurement system. The water intrusion rate can be calculated from this, given knowledge of the exact boundary conditions. An essential role, here, is being played by the knowledge of the exact net volume of the gas quantity in the measuring system [12]. Many test systems automatically determine the net volume of the filter candle housing before the actual intrusion measurement. However, the basic test procedures can already be implemented by means of an individual pressure transducer and a fine pressure reducer [15], if one succeeds in allowing a reproducible setting of the test gas above the filter element. In this way, the relatively complex net volume determination can be avoided. Another important requirement is the already depicted temperature problem. If

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 91

Highlights 2013

System description The overall configuration (Figure 3) of the test system consists of three main components: Measuring equipment with pressure sensor for the measurement of the intrusion flow across the filter element. A valve system that switches the test system into the positions of the individual test phases. An exact pressure controller which regulates the test pressures necessary for the test and, furthermore, allows automated drying and draining of the test system. The valve system controls the steps necessary for the test execution like filling, pres-

F & S International Edition

No. 14/2014

water

compressed air

measuring equipment

filter + housing

exhaust air

drain

Fig. 3: Flow chart of the measurement system adapted to the waste air filter system of an autoclave

pressure [mbar]

flow rate [g/min] pressure [mbar]

flow rate [g/min]

the thermal radiation of the autoclave chamber has an influence on the thermodynamics of the test gas, this can lead to a considerable impairment of the measured results [8]. That is why special measuring equipment was developed which shields the test gas against thermal radiation from the outside. Long test runtimes and cooling periods can be avoided in this manner. Moreover, a defined gas volume in the measuring system can only be set through the filling process of the test system. For the measurement of filters of different sizes the measuring equipment can be designed and varied accordingly. Besides the requirements for the test system, special demands are also to be made on the waste air filter system of the autoclave, to which the test is adapted. The essential test requirements include a successful condensate return from the waste air filter housing, as well as a successful inline sterilisation of the waste air filter. The sterilisation conditions prevailing in the autoclave chamber must also be transported into the waste air filter system for a successful sterilisation. Condensate occurring in the filter housing must be discharged into the autoclave chamber to avoid a condensate accumulation which would lead to a collapse of the sterilisation conditions [3]. The success of the sterilisation of the filter element has to be monitored here, so as not to contaminate the test water necessary for the integrity test, which is conducted after the test in the wastewater path. Moreover, care has to be taken during the installation of the waste air filter system to ensure that this is filled bubblefree for the test procedures and that there are therefore no dead spaces in the system which are not flooded during the filling. Otherwise, this would lead to a falsification of the total net volume of the gas volume. In particular, this concerns the measurement path of the measurement system. On account of these requirements, a special valve circuit was developed which eliminates dead spaces both in the test system and also in the exhaust air filter system of the target plants.

time [min.]

Fig. 4: Automated calibration sequence of the test system

sure build-up, stabilisation measuring and final draining of the system. Within the stabilisation phase, stabilisation of the intrusion flow takes place at exactly constant pressure conditions. The pressure drop generated as a result of the flow rate must be regulated in this phase to the exact test pressure via a fine pressure reducer. After the stabilisation phase, the measuring system is closed up and a pressure drop across the now stabilized flow is measured, which supplies the water intrusion rate. Integration into the target systems The test system (Figure 3) can be adapted to the waste air filter system of autoclaves via two lines. For this purpose,

installation of a valve (V-6) is necessary in the waste air filter system of the autoclave. The inlet and outlet line of the test water to the test system is enabled via the valve V5. The sterilisation of the filter element is monitored via a temperature sensor during the autoclaving. The control of all the components depicted, as well as the readout of the sensors, is done via integrated controller hardware of the test system. Calibration The flow rate references required for calibration are usually generated by means of a needle valve and are recorded gravimetrically via laboratory scales. Alternatively, mass flow meters are also suitable for accurate determination of the flow. Moreover, these allow automation of

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 92

Highlights 2013

Tab. 1: Results of the integrity determination of a High Flow Tetpor of II sterile filter (Parker Domnick Hunter), effective filter area 0.1 m2, pore size 0.2 µm, limit flow rate 0.06 ml/min

filter type

ser. no.

filter loading non-loaded after 1 autoclavation after 2 autoclavation after 3 autoclavation

measurement 1

measurement 2

filter test device Sartocheck 4 new test system measure- measurement 3 ment 4 measurement 5

Tab. 2: Results of the integrity determination of a High Flow Tetpor of II sterile filter (Parker Domnick Hunter), effective filter area 0.2 m2, pore size 0.2 µm, limit flow rate 0.13 ml/min

filter type

ser. no.

filter loading non-loaded after 1 autoclavation after 2 autoclavation after 3 autoclavation

measurement 1

measurement 2

filter test device Sartocheck 4 new test system measure- measurement 3 ment 4 measurement 5

Tab. 3: Results of the integrity determination of a Sartofluor® GA Mini Cartridge (Sartorius Stedim Biotech), effective filter area 0.1 m2, pore size 0.2 µm, limit flow rate 0.09 ml/min

filter type

ser. no.

filter loading non-loaded after 1 autoclavation after 2 autoclavation after 3 autoclavation

measurement 1

measurement 2

filter test device Sartocheck 4 new test system measure- measurement 3 ment 4 measurement 5

Tab. 4: Results of the integrity determination of a Sartofluor® GA Mini Cartridge (Sartorius Stedim Biotech), effective filter area 0.2 m2, pore size 0.2 µm, limit flow rate 0.11 ml/min

filter type

ser. no.

filter loading non-loaded after 1 autoclavation after 2 autoclavation after 3 autoclavation

measurement 1

the calibration procedure, as these also allow precise control and monitoring of the flow. The following Figure 4 shows the calibration by means of a calibration kit specially developed for the new test system. The test system is filled once more with water for every single measurement and the desired target flow is set by means of a mass flowmeter/controller (Bronkhorst Mini Cori-Flow model: M12V14IAGD-11-K-S). The system is afterwards stabilised, whereupon a determination of

measurement 2

filter test device Sartocheck 4 new test system measure- measurement 3 ment 4 measurement 5

the flow rate is done via a pressure drop measurement. The test system is drained afterwards and once more filled, whereupon the next measurement can be done. The flow rate values determined by the test system can be correlated with the flow values generated by the flow controller. Figure 5 confirms a correlation of both flow values and, in addition, confirms reproducibility of the volume setting via the newly developed measuring equipment.

WIR [ml/min]

flow regression

flow (reference) [ml/min]

Fig. 5: Calibration curve (flow calibration)

Pre-treatment and post-treatment of the filter A problem frequently appearing in practice is a collapse of the sterilisation conditions in the waste air filter systems of autoclaves during filter sterilisation [3]. To remove these condensate problems, drying the filter element after sterilisation has a supportive effect, in order to prevent water residues in the drainage layers of the filter hindering the sterilisation process of the filter. Also, for the pre-treatment of the test procedure, drying the filter can prevent a falsification of the measured results. If a filter still contains portions of water in its outer supporting layers after sterilisation, the test water can wet the actually hydrophobic membrane during a subsequent autoclave run, as a result of the increased temperature, and disturb the filter measurement. Hence, the new test system allows automated drying of the filter element before and after the water intrusion test. In addition, a cooling of the filter element via the test system was installed, which cools down the filter to the test temperature after the sterilisation. Filter measurement in autoclaves Figure 6 shows the entire sterilisation course (gravitational method) of the filter element of an ordinary laboratory autoclave (upright autoclave; chamber volume 85 L) with subsequent water intrusion test. The filter was sterilised in the function of a waste air filter by means of the sterilisation run of the autoclave and afterwards was tested by the test system. The software of the test system continuously records pressure values from the test system, as well as temperature values of the filter in the waste air filter system of the autoclave. The sterilisation run in its simplest form consists of a heating phase, the holding phase and a cooling phase. Only the holding phase is substantially decisive for the sterilisation success and it must hold the given temperature for a fixed period. The test run consists of drying, cooling, water intrusion test, draining and repeated drying of the filter element. After the end of the sterilisation, the test run begins with the drying of the filter element. The drying process uses the residual heat still available in the filter for the removal of condensate from the drainage layer of the membrane filter. The evaporating condensate leads to a strong temperature drop and works as an additional cooling for the filter element. The residual heat still available in the filter is discharged by the subsequent cooling step by means of water. The cooling for this is done in intervals in order to exploit the heat capacity of the water optimally. After the integrity test, draining and complete drying of the filter is done in order to prepare the filter for the next sterilisation cycle. F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 93

Highlights 2013

Comparative measurements

Conclusion The test procedures for integrity testing of membrane filters seem to work very simply at first glance metrologically. However, in the end, it is the interplay of many factors that leads to complexity in the matter and makes their development and integration into appropriate target systems anything but trivial. A hindrance also appears to be the low investment costs for such test systems that are necessary for

F & S International Edition

No. 14/2014

pressure [mbar] temperature [°C]

heating phase

pressure [mbar]

holding phase

filter temperature [°C]

For the evaluation of the new measurement system, it was compared to two commercially available measuring instruments (Palltronic® Flowstar XC and Sartocheck 4). The comparative measurements were done in four different sterile filter models of the manufacturers Parker Domnick Hunter (High Flow Tetpor II) and Sartorius Stedim Biotech (Sartofluor® GA Mini cartridge). The filters were then inserted as waste air filters in the waste air filter system of a laboratory autoclave (upright autoclave; chamber volume 85 L) and exposed under normal process conditions (sterilisation temperature 121 °C, holding phase: 20 minutes, gravitational operation). The measurements of the individual filters were made both before and also after the stresses of the autoclaving process. The sterilisation procedure corresponds to the picture described in the previous section and the concluding test phase of the filter element was done as multiple determinations through the three test systems. In Tables 1-4, the measured results of the comparative measurements are shown. With multiple determinations, the flow rates drop according to experience. The causes of this effect can be derived only from observations so far which show that, through the pressurisation, air bubbles on the filter surface dissolve in the water phase. This leads to a volume loss which is interpreted by the test system as a flow rate. With the pressure relief carried out at the end of the test, these bubbles are partially removed from the filter by ascending. However, the actual causes can be more complex, which makes it difficult to compare the flow values of measurement systems of different test systems. But the comparative measurements can be assessed under the criterion of whether an integrity test has been evaluated by the measurement system as having passed or as having failed. With all four filter models, no critical changes in the flow values can be observed through exceeding the limit flow due to exposure, something which can be confirmed with all three test systems. The integrity of the filter can be proved by the new test system without influence from external temperature.

cooling down phase

WIT

1st drying

cooling

time [min.]

Fig. 6: Sterilisation procedure

small plants. However, with an accurate definition of requirements, these can be adapted to the corresponding target systems and to the filter elements to be tested. Under these conditions it is also possible to develop safe test systems for small plants, which, in addition to the easy detection of filter failure through to finding leaks in the system, also allow a supporting effect on the sterilisation method of the filter elements by means of their drying. Expression of thanks We thank the Hessischen Ministerium für Wissenschaft und Kunst for financial support of our development project: Development of a modular, universally applicable Water intrusion test (WIT) for small plants. In addition, we thank our project partner biomedis Laborservice GmbH for the fruitful and trusting cooperation. Literature: [1] Technische Regeln für biologische Arbeitsstoffe 100, Schutzmaßnahmen für gezielte und nicht gezielte Tätigkeiten mit biologischen Arbeitsstoffen in Laboratorien (Technical rules for biological agents 100, Preventive measures of specific and non-specific activities with biological agents in laboratories), 2006, ABAS. [2] Verordnung über die Sicherheitsstufen und Sicherheitsmaßnahmen bei gentechnischen Arbeiten in gentechnischen Anlagen (Gentechniksicherheitsverordnung) (Ordinance about the safety stages and safety measures for genetic engineering operations in genetic engineering facilities (Genetic engineering safety ordinance)) in the version published on 14th March, 1995 (BGBI. I p. 2768). [3] C.Grumbach, H. Schulte-Lünzum, P. Czermak (2013): Abluftfiltration unter der Lupe, Risiken bei der Sterilfiltration der Autoklavenabluft (Exhaust air filtration under the microscope and risks with the sterile filtration of autoclave exhaust air): Techno Pharm, in printing [4] Einbauempfehlung für Neuanlagen, Nachrüstung oder Ergänzung, zur Wahl der Abluftbehandlung von Autoklaven (Recommended installation for new sys-

tems, retrofit or addition, for the selection of the exhaust air treatment of autoclaves), Resolution 3/2009 of ABAS, ELATEC. [5] ASTM Standard F838-83, American Society For Testing Material (ASTM), ASTM 1983, amended 1988. [6] U. Brendel-Thimmel, R. Jaenchen, F. Schlamp (2006), Chemie Ingenieur Technik (Chemical engineering) 78:11, 1655-1665 [7] J. Meyer, Einführung in die Integritätstestung von Membranfiltern (Introduction to the integrity testing of membrane filters), SPI8000-d11041, Sartorius Stedim Biotech, 2011. [8] M.W. Jornitz, T. H. Meltzer, Filtration and Purification in the Biopharmaceutical Industry, Informa Healthcare USA, Inc., New York, 2008. [9] Standard Test Method for Determing Bacterial Retention of Membrane Filters Utilized for Liquid Filtration, Committee D19.F838-05, ASTM, 2005. [10] P. Czermak, G. Catapano (2003), PDA Pharmaceutical Science and Technology, 57:4, 277-286 [11] P. Czermak, G. Catapano (2000), European Journal of Parenteral Sciences 5:3, 59-63. [12] M.W. Jornitz, T. H. Meltzer, Sterile Filtration a Practical Approach, Marcel Dekker, Inc., New York 2011. [13] R. Jaenchen, J. Schubert, S. Jafari, A. West (1997), European Journal of Parenteral Sciences, 2:2, 39-45 [14] C. Grumbach, P. Grace, P. Czermak, T.Pillich, C. Rühl, K. Fey (2011): Einbauempfehlung für Neuanlagen, Sichere Überwachung der Abluftbehandlung von Autoklaven mittels Wasserintrusionstest mit hoher Temperaturstabilität (Recommended installation for new systems, Safe monitoring of exhaust air treatment of an autoclave by means of water intrusion test with high temperature stability); GIT Laborfachzeitschrift 55: 8, 540-542 [15] Filterelemente - Membranfilterelemente - Teil 12: Integritätsprüfung von hydrophoben Membranfilter?elementen mit Wasser (Wasserintrusionstest) (Filter elements - Membrane filter elements - Part 12: Integrity testing of hydrophobic membrane filter elements with water (Water Intrusion Test)), DIN 58356 – 12, NA 063 Normenausschuss Medizin (NaMed). [16] Kradolfer M, R Jaenchen (2008), P&A Select Biotech 2: 28-30

Note: This project (HA Project No.: 307/11-52) ) is funded within the scope of Hessen ModellProjekte from funds of the LOEWE State Offensive for the Development of Scientific and Economic Excellence, funding line 3: KMU Collaborative Project.

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 94

Highlights 2013

Product development challenge S. Ripperger* Today products and processes are more than ever in demand which optimally fulfil the needs and desires of society. The contented customer is a goal on which the company’s activities must be oriented in a buyer’s market. The market offers the chance to be successful with an excellent product range. The competition of the offers promotes developments and with them also the changes over time associated with them, to which in turn the enterprises must react. Here, in the field of separation technology, attention is being increasingly paid to energy efficiency and resource conservation. Hereinafter, some basic contexts and trends are shown that must be observed with product development. 1. Need for ongoing product development The economic system of western countries is based on the ideas of the market economy and free trade. It is typical for this that - a big number of products are offered freely in the market, - the consumers are free in the choice of the products and - the prices are determined by supply and demand. Moreover, the producers can freely decide, in compliance with statutory regulations, about the type and design of products and the deployment of the means of production. Hence, for an enterprise that offers tangible or intangible products, the answer to the following questions is in the foreground: - What should be produced? - For whom should it be produced? - What volume should be produced? - How should it be produced? In the market economy these questions are in principle not answered by a governmental authority but the enterprises must find for themselves the answers and, however, they must respect the laws of the market and of the States in doing so. Economic transactions are in a market economy the result of the meeting of supply and demand. In accordance with supply and demand, the market regulates the pricing. State interventions should ensure the maintenance of competition and hence serve the function of the market. For this purpose the “Act Against Restraints of Competition” was passed in 1958 in Germany, which is also called “Anti-trust law”. Furthermore state interventions ensure the safety of products, workplace safety during their production and environmental protection. State interventions * Prof. Dr.-Ing. Siegfried Ripperger Department of Mechanical Process Engineering, Technische Universität Kaiserslautern Gottlieb-Daimler-Straße 67663 Kaiserslautern / Germany Tel.: +49 (0) 631 205 -2122/2121 E-Mail: ripperger@mv.uni-kl.de

should be limited to a minimum in the market economy. Today, the movement of goods across national borders is influenced by the ideas of free trade. According to that, this movement of goods should not be limited by any dirigistic intervention. Numerous international agreements were concluded to expand and to guarantee free trade and to curb protectionism and foreign trade monopolies. In Europe in 1957, the idea of free trade resulted in the foundation of the “European Economic Community (EEC)”, the predecessor organisation of the “European Union (EU)” and in 1960 the “European Free Trade association (EFTA)”. Also in other areas of the world, states have united according to the European model. Internationally, the World Trade Organisation (WTO) endeavours to dismantle trade barriers. A result of this development was that many different products are offered in larger markets, as a result of which competition is substantially increased. To be successful, the enterprises must constantly further develop existing products and offer new products to the customers. This constant competition promotes innovations

and ensures rapid technical development in many areas. In the bigger markets, larger quantities can be sold, as a result of which reasonable “mass production” became possible with many products. All these measures led in past decades to never known prosperity in industrialized countries. Most products are offered in industrialized countries in a buyer’s market, i.e. compared to the installed production capacities, there is a demand deficit for many products. Accordingly, enterprises must align their strategy in terms of supply and product development. In a buyer’s market, customers are at the core of enterprise activity. Their behaviour determines the economic success or failure of a newly developed product. The customers compare the products offered in the market and buy those which offer the biggest benefit. They hence form the basis of existence of enterprises and also ultimately determine which products are produced and offered. Today customers are more critical than ever before. More knowledge changes the requirements and increases the demands concerning quality, environmental protection and energy efficiency. The quick

Fig. 1: Causes for increasing requirements for products, systems and processes

F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 95

Highlights 2013

development of Information Technology (IT) and the spreading of technical knowledge that is possible through this, allows new forms of competition to originate, as a result of which the profit contributions of a product during its life span decreases faster than earlier. Moreover, the significance of commodity products decreases, i.e. of products with stationary specifications that are produced in large quantities. On the other hand, the number of specialities increases, i.e. products that are developed and are produced for definite uses. Accordingly the number of implemented product ideas has also risen rapidly. The technological changes cause a shorter life cycle of individual products, an increasing number of short production runs, as well as an increase of the research proportion in production costs. The development times for new products had to be accordingly shortened and flexible and at the same time efficient production processes had to be introduced. At the same time the requirements for products were specified more numerously and much more precisely. Causes for this are listed in Fig. 1. Essential requirements are formulated by the needs and desires of the customers. Others are defined by provisions or technical regulations. On account of increasing competition between enterprises, consumers could bring their influence to bear more and more. In addition to that, services were increasingly demanded together with products. These include, for example, proper consultation and/or support through application technology, or a guarantee to take back the products and/or individual components after use. All this meant that task definitions in enterprises in connection with product development became substantially more complicated and more difficult. On account of the growing variety of the market, the tasks became more specific and more versatile. A uniform approach across all products is not possible anymore. Moreover, the development times with some products increasingly developed into a critical success factor in relation to the competition. In terms of market penetration, the high growth rates internationally and also the emerging market saturation within Germany must be considered for many products. So at the moment more than 60% of appliances, machines and equipment produced in Germany are being exported. Two thirds of the turnover of the chemical industry is the result of activities abroad. Altogether, Germany is one of the leading export nations. In connection with globalisation, the advancement of the European Union into a common market

F & S International Edition

No. 14/2014

with the same currency and harmonised rules is a logical consequence. 2. Requirements within process engineering With process engineering procedures and equipment, intermediate, semi-finished or finished products are produced in many industrial branches. They have to meet the criteria of - product quality, - economic efficiency, - environmental sustainability and environmental protection, - efficient use of raw materials, - sparing energy utilization and - safety Accordingly, the development of new or improved processes, systems or system components is a complex task for which the aspects listed above cannot be looked at in isolation. The number of materials treated and products produced with process engineering methods cannot be overlooked. They are solid, liquid or gaseous and can be present as highly pure substances or as complex mixtures. During the production process, the desired properties of the products are adjusted and/or produced. Hence, for products produced with process engineering methods, their development is coupled closely with the process and equipment development. The products made with this must be competitive. Energy-saving and resource-saving procedures, high product yields, high and consistent product quality and economically favourable process guidance are necessary requirements for this. The following requirements, inter alia, arise for process engineering and the accompanying process stages: - automated operational flow, - online control of material transformation, - online control of essential product specifications (e.g. the particle size and/or the particle size distribution), - reduction of the appliances and machines for production of a product (several process steps on one appliance and/or on one machine), - allocation of process parameters to product characteristics, - model-based simulation of the process sequencing for rapid product and process development and for continuous process optimisation. In the area of separation and solids process engineering, many of the requirements still cause big difficulties. The processes for solid matter processing are, as a rule, more difficult to automate than fluid processes. There is for this, inter alia,

a lack of proven online sensor systems by means of which particle dimensions or other solid matter properties in the process can be monitored. Also the models for the simulation of solid matter processes mostly do not yet fulfil the requirements. They are predominantly missing the relation to relevant product characteristics. 3. Resource management In their purchase decision, many customers increasingly consider effects relevant for the environment and ethical points of view that are connected with the purchase and the use of a product. In this, the demands of some customers often go well beyond what is dictated by existing legislation and ordinances. It is assumed nowadays that environmentally sound disposal routes for consumable products are indicated. The use of recyclable materials, the use of renewable raw materials and the minimisation of energy requirements in the production and use of the products is a trend that is not yet completed. It was triggered by a change of values that has taken place in broad sections of the population in the last decades and which is supported increasingly also by the political stakeholders. The customers would like to buy and use a product with a “clear conscience”. Some customers are guided in this by comparative environmental analysis and life cycle assessments, which are developed for individual products. In an environmental analysis, the “ecological rucksack” (see for example /1/) of a product is investigated, i.e. the sum of all natural raw materials used that are invisibly dormant in a product (less the dead weight). This approach and the concept were introduced in 1994 by Friedrich SCHMIDT-BLEEK in the Wuppertal Institute for Climate, Environment and Energy. The amounts recognized in the ecological rucksack are a measure of the consumption of resources for the production of a product. It covers thereby the Material Input per Service Unit expended over the product lifetime (MIPS concept). Today analogous considerations are also done concerning the water consumption or the CO2 emission. These considerations are a result of the knowledge about the huge material and energy requirements of our economic activity and the environmentally destructive effects of some extraction and production methods, as well as the aforementioned change in values. The quick consumption of many products in industrial societies ensures a growth of material supplies and rubbish. One has been becoming increasingly aware in recent years that, for the maintenance of the economic system in industrialized countries,

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 96

Highlights 2013

man interferes heavily in the ecosystem worldwide. The interventions of man in the ecosystem are so big that many researchers consider it justified, to speak of a new geological era, the “Anthropocene”, in which life on Earth is affected in all areas by the activity of man (see for example /2/). This knowledge, which is also shared in the last decades by a large part of the population, supports for many the desire and efforts for a sustainable economy. The goal here is to sustain the natural resources on which our existence is based across as many generations as possible and if necessary even to improve them again. To achieve this aim, the environmental aspects must already also be taken into account during the decisions on material and energy use i.e. in the development of the products. To achieve this, legal foundations were created in Germany in the last decades, as well as in the EU. This also affects numerous statutory regulations which must be observed during product development. In accordance with the call to preserve human livelihoods and living conditions, there was an industrial branch established, for example in the field of waste management in recent years, that ensures that a large proportion of products is again disposed of in an “environmentally friendly” manner. Environmental protection is thereby pursued beyond the limits of production facilities, so that, for example, the avoidance and/or reusing of waste in closed circuit systems are aimed for. With the Materials Recycling Act currently in force, (KrWG), the objective is pursued to achieve a sustainable improve-

ment of environmental and climate protection, as well as of resource efficiency through the reinforcement of waste prevention and the recycling of waste. 4. Implications for product development In order to best meet the environmental objectives, long lasting products with a low consumption of consumables and short-lived replacement parts would be asked for. On one hand this is in contrast to technical progress, which is a cause for short product life cycles and results in new and improved products in short time intervals, and on the other hand long lasting products are often more expensive as a result of a higher cost of materials and a more solid design and configuration, as a result of which, for example, investment costs can be increased. The first problem can often be counteracted during product development by the scheduled possibility of retrofitting. This is not easy, since future developments must already be considered and taken into account at the stage of product development. In this way at least the life cycle can be extended of many parts and components. The second problem can often be diminished by a consideration of life cycle costs (LCC). In this, the ascertained expenditure for the useful life of the equipment is compared to the savings to be expected and other benefits. Cost advantages that are essential for a purchase decision often arise in spite of increased investment costs for long lasting products with a low need for consumable materials. Hence, the monitoring of life cycle costs

plays a bigger and bigger role with cost management and delivers essential criteria for product comparison and decision-making. A problem in this case is that a complicated calculation is necessary that needs to be explained to the customer, which must be backed by convincing data on the lifetime and operating costs. This challenge must already be faced at the stage of product development. In connection with the development of environmental legislation and the objective to recycle valuable materials again as much as possible, the field of “recyclingoriented product development and product design” was developed at the end of the eighties in the 20th century. In the course of this, recycling is already considered as a component of the technical product life cycle during product development and is planned as far as possible. In 1991, VDI Directive 2243 /3/ came out, of which a revised version has meanwhile become available. The Directive pursues the objective “in particular to give the developer and the designer information, guidance and decision support for the various phases of product development, in order to develop and select technical and economic possibilities and alternatives for improving the recyclability of technical products.” Literature: /1/ Encyclopaedia of Sustainability, at www.nachhaltigkeit.info /2/ P. J. Crutzen, M. Davis, M. D. Mastrandrea, St. H. Schneider, P. Sloterdijk: Das Raumschiff Erde hat keinen Notausgang. Energie und Politik im Anthropozän. Suhrkamp, Berlin 2011, ISBN 978-3-518-06176-3. /3/ VDI Directive 2243: Recycling-oriented Product Development, Beuth Verlag, Berlin, July 2002

Fine dust pollution decreases slightly A report by the Federal Environmental Agency According to information from the Federal Environmental Agency, Nitrogen dioxide and fine dust still affect the air quality in Germany. Temporary measurement data from the states and the Federal Environment Agency prove this for 2012. With Nitrogen dioxide, the pollution compared with the previous years remains consistently high. The average concentrations of fine dust - measured in PM10 - in 2012 in Germany lies at the level of 2008 and with it clearly under that of the preceding three years. Above all, in cities and metropolitan areas the limit values for fine dust and Nitrogen dioxide are still exceeded too frequently in the immediate vicinity of roads. According to UBA President Flasbarth, the new figures on air quality in Germany point out that the emission of 96

pollutants into the air absolutely must be reduced further. The fine dust pollution in Central Europe still reduces the average life expectancy by almost half a year. Although air pollution in Germany has evidently decreased since 1990, further efforts are still needed. The focus is primarily on fine dust and nitrogen oxide. They arise from car exhausts but also in combustion processes in industry and above all in households from energy and heat generation. The increasing number of fireplaces meanwhile contributes, as can be proved by the fine dust pollution during the winter months. Studies point to an increase of up to five exceedances of the daily values. Moreover, the EU target value for benzopyrene is exceeded by sev-

eral times. This is 1 nanogram per cubic metre, ng/m3. The substance - a polyaromatic hydrocarbon (PAK) – originates, inter alia, from the burning of wood. The states currently record air quality nationwide at 643 stations. 52 percent of the urban, traffic-oriented stations exceeded the allowable annual average value for Nitrogen dioxide (NO2) of 40 micrograms per cubic metre of air, μg/m3. On the other hand, the PM10 annual mean limit value for fine dust, which is also 40 μg/m3, was adhered to in Germany in 2012. The pollution of the air with fine dust last year was one of the lowest ever recorded. Exceedances of the daily average value especially occurred again at traffic-oriented stations. Even under the favourable weather conditions of last year, F & S International Edition

No. 14/2014

02_fs_international_2014_seite_03-98__ 21.05.14 20:24 Seite 97

Highlights 2013

the measured values at nearly 10 percent of the traffic-oriented stations, however, lay above the fine dust daily limit value. This is exceeded if more than 50 μg/m3 of fine dust are detected on more than 35 days in the year. “The development away from district heating to more decentralised, small combustion systems in households and industry must not lead to higher dust levels than are produced by today’s main power stations. Here, with stricter rules for small combustion systems, Germany has made an important step”, says Flasbarth. How emissions can be specifically decreased in wood-fired, small combustion systems, is something that the Federal Environment Agency reveals in its brochure “Heating with Wood”.

Other needs for action exist where the breathing air of many people is polluted with fine dust and Nitrogen dioxide: in the cities and metropolitan areas. There, the emissions from transport can be reduced by environmental zones. If they are implemented with few exceptions and strictly with Stage 3, which only allows entrance with a green badge, they demonstrably have a positive impact on air quality. Meanwhile, since more than 80 percent of passenger cars and 50-60 percent of commercial vehicles get a green badge, environmental zones that still allow a yellow badge cause no essential improvement. So that the air can improve further in metropolitan areas, it is necessary to reduce the large-scale pollution from agriculture. For this, the UBA is counting on a

revision of the approval factors for intensive farming, as well as on the implementation of good agricultural practice, which includes, inter alia, the rapid incorporation of farmyard manure. For further information - Temporary evaluation of air quality values of 2012: http://bit.ly/fs0213-01 Year of the air 2013: http://bit.ly/fs0213-02 - Current air quality data: http://bit.ly/fs0213-03 - UBA websites “Air and Air Pollution Control”: http://bit.ly/fs0213-04 - Brochure “Heating with Wood”: http://bit.ly/fs0213-05

Prevention of the spreading of legionella H. Lyko*

Measures for prevention and monitoring Inspired by the legionella infections in August that were triggered in the North Rhine-Westphalian town of Warstein by these pathogens spreading through the cooling system of an industrial enterprise, the VDI referred to the need of measures for prevention and monitoring in open recooling systems. Up to now, open recooling units in Germany - differently from abroad - have not been subject to any registration. Dr. Irene Tesseraux, chairperson of the Directive Panel “Measurement and evaluation of legionella” of the Commission on Air Pollution Prevention (KRdL) in the VDI, considers a register for evaporative cooling systems, such as are found on larger administrative and industrial buildings, to be absolutely necessary, so as not to waste valuable time searching via helicopters and aerial maps on the occurrence of infectious cases. Rainer Kryschi, chairperson of the VDI Directive Panel “Hygiene requirements for evaporation cooling plants”, demanded the inclusion of such systems in the technical inspection regulations of the federal states. Legionella in drinking water constitutes no health risk for people with intact immune systems. It can, however, become dangerous to people when they penetrate as a fine spray into the lungs. This was already shown by the legionella epidemic in January, 2010 in the Ulm area, where five people died. Alarming is the fact that the spreading of legionella can happen through the air across large distances - up to 10 km has been reported. With various directives on air pollution and on the topic of construction and building technology, the VDI offers assistance for preventing the spread of legionella into the ambient air through water-bearing systems (see following Table). In Warstein, an evaporative cooling system was the trigger for the spread of the dangerous pathogen, but finally it was found further ahead, in a water treatment chain. As is presented by evaluating the various press releases on root cause analysis, the pathogens appeared in the effluent of the wastewater treatment plant of a brewery, which was conducted into the municipal wastewater treat*Dr.-Ing. Hildegard Lyko Dortmund / Germany, Phone +49 (0) 231-730696

F & S International Edition

No. 14/2014

Table: Directives on air pollution and on the topic of construction and building technology that are helpful in the prevention of the spread of legionella

ment plant. The effluent of this wastewater treatment plant enters the surface water bodies from which the above-mentioned industrial enterprise draws its cooling water. Procedures for maintenance and inspection of air conditioners and air conduits According to a report of TSI GmbH, the incident shows that either omitted or non-professional servicing of air conditioning equipment means that this can develop into a dangerous breeding ground for bacteria. Within an air-conditioning unit, there are many places that are well-suited for growth due to condensation and heat. Antibacterial treatments are part of regular professional maintenance just as much as measurements that point to leakage in the pipe system. Also, measuring instruments like the TSI Biotrak, with which viable strains in the air can be determined, can contribute to this. With a vacuum blower that can be connected in a mobile manner to air conditioning lines, and also with a handheld meter, conduit systems and bleeds are tested for leaks. To verify the air diffusion, anemometers or volumetric flow hoods are suitable. Thus, the exit velocity of the air can be measured on outlets. For this purpose, the measuring hood TSI AccuBalance uses a differential pressure measuring method. After the source is discovered, above all the protection of the employees is vital. For this, filter masks are used for respiratory protection, inter alia. Training is vital for the use of masks. With the TSI PortaCount, TSI offers a globally used device with which the face seal of breathing masks can be quantitatively verified.

02_fs_international_2014_seite_03-98__ 21.05.14 20:12 Seite 98

Highlights 2013

IMPRINT Publishing house: VDL-Verlag GmbH Verlag & DienstLeistungen Address: F&S - Filtrieren und Separieren VDL-Verlag GmbH Verlag & DienstLeistungen Heinrich-Heine-Straße 5 63322 Rödermark/Germany Phone: +49 (0) 6074 92 08 80 Fax: + 49 (0) 6074 9 33 34 e-mail: vdl-verlag@t-online.de www.fs-journal.de Editor: Prof. Dr.-Ing. Siegfried Ripperger Birkenstraße 1a 67724 Gonbach/Germany Phone: +49 (0) 6302 57 07 Fax: +49 (0) 6302 57 08 e-mail: SRipperger@t-online.de Dr.-Ing. Hildegard Lyko Dortmund/Germany Publisher: Eckhard von der Lühe Advertising department: Eckhard von der Lühe Phone: +49 (0) 6074 92 08 80 Fax: + 49 (0) 6074 9 33 34 e-mail: vdl-verlag@t-online.de International Sales Manager: Margot Görzel Phone: +49 (0) 6196 65 32 11 e-mail: fs-journal@mgo-communications.de Printing Office: Strube OHG 34587 Felsberg/Germany Layout: Ralf Stutz, Gestaltung Hainer Hof 1 60311 Frankfurt am Main/Germany Nicola Holtkamp

Eleven requirements for a future environmental policy The German Association for Water, Wastewater and Waste e. V. (DWA) has summarized its current positions on environmental policy in a policy memorandum. It was presented to the public on 24th April, 2013 within the scope of the event “Wasser Berlin International”. The main points were summarized in the following eleven theses: - Energy transition To take more into account the potential of water management The creation of suitable boundary conditions for secure energy supply from renewable energy sources must be a focus in the 18th legislative period. For this, it is a matter of incorporating the water management issues at an early stage. Storage capacities, also in the area of water management, are an essential factor for a secure energy supply. Moreover, the DWA expects from the political process a removal of unequal treatment in the funding of measures under the Renewable Energies Act (EEG). - Fracking not at the expense of the environment Fracking technology considerably influences the soil and groundwater. The environmental impact and risks of a fracking measure must be scientifically determined and evaluated. Because of the water management risks, fracking must not be permitted in sensitive areas, such as wetlands and drinking water catchment areas. Fracking measures need an assessment of environmental effects and must be subject to regulatory approval and supervision. The DWA expects there to be full disclosure to the authorities of the materials used and the composition of the liquids used.

- Anthropogenic micropollutants act with a sense of proportion With the updating at EU level of the list of priority substances, i.e. of substances which are particularly harmful to the aquatic environment and humans, a close coordination with the REACH method for risk-based substance evaluation is necessary. The REACH Ordinance is an ordinance of the EU on the handling of chemicals. The inclusion of drugs into this list should currently be deferred, because specifications for such substances should be left to the approval regulations for drugs at the EU level - Wastewater Ordinance Amend in an EU-compliant manner It is necessary to further develop the Wastewater Ordinance also for the implementation of EU specifications. Nevertheless, the introduction of a comprehensive fourth purification stage for municipal sewage treatment for the elimination of trace substances would not be appropriate in this case. - Wastewater charge continue to develop constructively The wastewater charge should be subjected to an intensive reform. In the course of this, the number of parameters is to be substantially reduced. The charge has to be calculated according to the actually discharged cargo. Ecological, not fiscal criteria should be decisive for the amendment. - Phosphorus ensure recovery Legislation to protect the phosphorus resources is necessary. With the amendment of the Sewage Sludge Ordinance, this requirement must be taken into account.

www.fs-journal.de - Agriculture Reduce substance discharges into the aquatic environment

FILTRIEREN UND SEPARIEREN

Diffuse nutrient emissions into ground and surface water have to be decreased by appropriate measures, including by the legislature. The Fertiliser Ordinance is to be amended accordingly. Biomass production must be targeted to be more environmentally friendly. The cooperation of agriculture and water management should be supported by the legislator.

- Federal Decree design in practical form The harmonisation of the requirements between groundwater protection, soil conservation and the requirements for replacement of construction materials must be done in practical form with a new Federal Decree. With the theses, inter alia, environmental problems are addressed for which a solution must be found. See also www.dwa.de.

F & S International Edition

No. 14/2014

00_fs_international_2014_umschlag__ 21.05.14 20:18 Seite 99

00_fs_international_2014_umschlag__ 21.05.14 20:14 Seite 100