ARCH 567: Advanced Architectural Structures, Winter 2012 Corey Griffin, Assistant Professor, Portland State University
Improving the integration of sustainable strategies in schematic design Developing a multi faceted tool to improve thermal resistance in architectural enclosure systems M. Boyce Postma Masters of Architecture Candidate, University of Oregon Jacob Spence Masters of Architecture Candidate, University of Oregon
1. Abstract Well designed, energy efficient buildings are no longer optional as the price of energy continues to rise, climate change becomes a real effect of increased greenhouse emissions and the realities of cheap consumable design and construction catch up to owners and occupants. Through updated codes and sustainably-minded framework systems such as the U.S. Green Building Council’s LEED rating system and Passive House Certification, comprehensive energy strategies are increasingly incorporated into the schematic design of new buildings at every scale. While many components of a building’s design are considered, the inherent multiplicity of a building’s enclosure system can present designers with opportunities to greatly improve the effectiveness of the comprehensive energy strategy. Considerations for insulation types, cladding support systems, window-to-wall ratios, window thermal resistance values and economics are crucial to the thermal effectiveness of a wall assembly and the maintenance of an effective thermally resistant wall. This study seeks to develop a tool that can be quickly and easily used by schematic designers to set a thermal resistance target for the comprehensive enclosure system and receive options in a variety of variables to achieve that goal.
2. Introduction Energy efficiency in the built environment is a well established goal of most architects, builders and clients. In response to this increased demand for high performance architecture, a wide variety of components and systems have been developed to increase energy efficiency, occupant comfort and usage longevity. As these components and systems have been developed piecemeal over time, they have historically been used as individual solutions to individual problems. These are typically tacked on at the end of the design process as a way of obtaining an energy efficiency rating that may not be true to the spirit of highly sustainable building practice. Only recently has the application of advanced energy efficiency systems been brought into the schematic design discussion in mainstream practice. Not only must designers consider the implications of individual systems in new construction, the interaction of these systems becomes a crucial design problem that must be re-evaluated throughout the design and construction timeline. In example, if you increase the ability of a wall to maintain thermal energy, the size and type of HVAC system may be reduced, which in turn may also reduce the size of ventilation systems, and ceiling heights. One energy efficiency system can greatly affect every other system, energy related or not. At about 40% of all energy consumed in the United States (40 QBtu), buildings use more energy than any other sector including transportation and industry [1]. Using this much energy, predominantly non-renewably sourced, has an extreme effect on the emissions of greenhouse gases contributing to global climate change and is a general waste of resources. This issue will become increasingly visceral as fossil fuels cease to be readily extractable from the earth. Non-sustainably sourced energy prices will most likely increase to a point where sustainably harvested energy will inevitably become an economically driven preference. Not only does our reliance on fossil fuels potentially jeopardize the balance of our atmosphere, it has the compounded effect of being highly inefficient to use and to produce. For all arguments made to reduce energy consumption in the built environment, the one most readily accepted is the reduced lifetime expenses associated with an efficient building. Over 88% of a building’s lifetime energy consumption is used in building operation; the remaining energy is used in construction and material procurement. Including the energy usage associated with the construction of buildings, currently 48% of all energy produced in the United States is used to build and inhabit buildings; 44% of that is used to heat and cool the enclosed environment [1]. This means that nearly a quarter of all energy produced is going directly into the heating and cooling of buildings, and this is only the energy used on site. Any inefficiency in the energy production and transportation further exacerbates the issue. As designers and builders, we have a responsibility to develop solutions to this extreme use of energy.
Of the many building systems that can reduce energy needs of a building, the enclosure system is typically the most impactful. Many changes in construction methods and assembly chronology have taken place in the last thirty years as building science engineering continues to increase our understanding of the interaction of heat, water, wind and earth, and the built environment. Of those external forces (or loads), the control of thermal radiation can have the greatest impact on reducing energy needs. Regardless of improvements to the insulation and assembly, thermal bridging remains an issue. This occurs when energy travels via thermal conduction from a place of high energy to low energy, or rather from a warm place to a cold place through the material of least resistance. One recent change in basic construction methodology is the location of insulation in an enclosure assembly. It is becoming common practice to locate insulation on the exterior of the primary structure. This is in contrast to previous building practices that predominantly located insulation within the primary structure. This improved method of insulating reduces thermal bridging between the exterior and interior environments — that is to say it reduces the amount of thermal energy that flows through walls, and specifically the primary structural support. Though this method has greatly reduced thermal bridging in most cases, the exterior cladding in this system is typically supported through the insulation and maintains this crucial design problem of thermal bridging. A thought experiment : A perfect thermal barrier would be composed of adequate insulation (as determined by climate) wrapping a completely sealed and puncture-free enclosure. Imagine how this building would be assembled. A primary structure would be built, covered in a sheathing material such as OSB or plywood and then wrapped in vapor, water and air control layer such as Tyvek polyolefin building wrap. Beyond this, a layer of insulation is installed and then a cladding material to prevent water, weather and UV from penetrating all that is underneath. There is typically a 2” or greater gap between the cladding material (also known as a rainscreen) and the insulation that provides drainage for any water that makes it past the initial cladding surface material. A building using this “perfect wall” system would preserve its interior environment with very little need for heating or cooling until a series of systematic failures occur over the course of time due to the limited, though intended lifespan of the included materials [2]. Despite the great efficiency of the system, thermal imaging on the exterior surface of the
building would reveal lines (or at minimum, points) of warm temperature over the entire surface. This is due to the physical realities of utilizing multiple materials in a building assembly. Materials must be held in position securely and in some instances spaced away from each other. The system of substructure that holds materials together in an assembly leads to thermal bridging [3]. Windows create the most dominant thermal bridges; however, they are necessary parts of a typical building assembly and are crucial to a well developed comprehensive energy strategy as they can greatly reduce dependence on electricity for lighting. In a perfect wall, a single substructure support holding cladding through thermal insulation does not allow very much energy to pass through the enclosure, but hundreds or thousands of points or lengths of contact will very quickly erode the thermal barrier which once was so strong. Thermal bridging is more obvious in the traditional building method of placing insulation in the cavities of a common softwood or light gauge steel studs. In this assembly, thermal energy from the interior will travel through the most conductive material in the wall assembly, which will be the structural studs. Wood as a framing material can cause a reduction in the insulation efficiency of up to 10%, whereas in light gauge steel framing, the reduction can be as much as 50% due to its high conductivity [3]. To the layman, these seemingly sparse points of contact on a frame, either from structural studs or cladding support may seem inconsequential; however, studies done on common housing typologies in southern California have shown that up to 27% of walls in standard balloon or platform framing systems are structural studs [4]. Using standard cavity insulation, that means that 27% of the walls are uninsulated. Assuming that externally insulated systems will only increase in prevalence in the next decade, we seek to develop a tool that will allow designers to set a comprehensive wall assembly R-Value target, a number that describes the thermal resistance of a material or assembly, and evaluate a variety of insulation types, cladding support systems, window-to-wall ratios, and window R-Values that meet that goal. Ultimately, results will be converted into relative prices to establish the most economical way to build a highly insulated wall. The most accessible way to make buildings more efficient is through a highly effective building enclosure. Not only can such an enclosure reduce the dependence on energy, but it can have the compounded effect of reducing the size or even need of a mechanical control system, providing more finances for use elsewhere [5].
3. Methodology
Our research began with a conversation with practicing architects. The goal of their firm was to create a comprehensive building enclosure that reaches an R-30 thermal resistance. Our original intent was to try to reach this by eliminating the cladding support system thermal bridging and increasing insulation as needed. Early into the research, we realized that though we could potentially find a way to do this, it would be much more beneficial to the firm and the design community to create a tool that would allow the user to quickly analyze variables that have the greatest effect on the thermal resistance of the comprehensive enclosure system during schematic design. Those six primary variables are: Window to Wall Ratio R-Value of the Window R-Value of the comprehensive enclosure Cladding Support System Cost Insulation Type
WWR Rwindow Rtarget CSS $ VIP, XPS, EPS, SF, MW
Other Variables introduced in calculations include: R-Value adjusted for thermal bridging R-Value adjusted for thermal bridging and windows 1 / R-Value U-Value for comprehensive enclosure system
RadjustedTB RadjustedW U Uwall
We analyzed 100 WWR’s, to account for all possibilities rounding to the nearest integer. Most tests were done at the 60% / 40% range as this is an established baseline for the tradeoff between energy losses through thermal bridging in windows and energy savings possible through daylighting. We analyzed windows ranging from R-1 — R-8. Single pane glazing typically has an R-Value around 1, double and triple glaze windows typically range from R-2—R-5 and some very expensive and more experimental gas filled or film layered windows can reach R-8 or even higher in laboratory settings. Though an R-8 window is unaffordable in the traditional market, we believe standard insulation values in windows will continue to rise and it won't be long before it is economically feasible. The R-Value of the comprehensive enclosure is the target set by
the designer. This number is based on necessity of climate and client/designer energy goals, and is typically set high to begin with and lowered as cost analysis ensues. Three cladding support systems (CSS) were analyzed. The first is a fiberglass and steel bolt system developed by Cascadia Windows Ltd. These come in sizes ranging from 3.5” 6” in depth and have a sheathing contact surface area of 3 in2. They are held onto the wall by 2” x ¼” steel bolts. The second system is typical z-shaped steel furring. In order to hold variables constant, we assumed the contact area to also be 3 in2 per unit. Since the contact area is already steel, we did not consider the steel nails or bolts holding the z-channels in place. The third system is a hypothetical all-fiberglass cladding support that works by creating a thermal bridge between two pieces of fiberglass or other material with low conductivity. One of the pieces would be bolted to the sheathing and have a place for another piece to hook onto it. This cladding support system was also held at a hypothetical 3 in2 contact area. The insulation types considered are Vacuum Insulated Panels (VIP), Extruded Polystyrene (XPS), Expanded Polystyrene (EPS), Spray Foam (SF), and Mineral Wool (MW). We were most interested in the variance in the three CSSs so we started with developing an adjusted Rtarget to account for thermal energy lost in the CCS. The formula below takes any WWR and the components of the CSS and gives an output of a new RadjustedTB that must be met in the opaque wall without the inclusion of the CSS to bring the comprehensive enclosure R-Value to the Rtarget. In order to find the adjusted R-Value for a wall assembly using each of these CSSs, we utilized a formula commonly used to find the adjusted R-Value of the WWR. The R-Value describes the thermal resistance of a material or assembly. The U-Value is the reciprocal of the R-Value and describes the thermal conductivities of a material or assembly. In order to find the Rtarget we must begin with the known conductivity of the materials. Uwall = (%opaque area * Uopaque wall) + (%glaze area * Uglaze) For a CSS with multiple materials like the Cascadia CSS, the formula corresponds: U wall = (%opaque area * Uopaque wall) + (%fiberglass area * Ufiberglass) + (%steel bolt * Usteel) Converting all U-values to R-values and rearranging the equation to solve for the adjusted R-Value of the comprehensive wall assembly after thermal bridging changes the Uwall to the Rtarget and the Uopaque wall to the Radjusted. Radjusted = %opaque area / ( ( 1 / Rtarget ) - ( ( %fiberglass + %steel ) / (Rfiberglass + Rsteel ) ) )
As seen above, the relationship between the Rtarget and the difference between the Rtarget and RadjustedTB is linear and increases as the target increases. This means that as you increase the target R-Value for the whole wall, the impact of the thermal bridge increases. The CSS with the highest R-Value in its components will always create the smallest change in the Rtarget and RadjustedTB. We use the RadjustedTB as our Rtarget for all further calculations. t A table of all 100 WWR’s was developed using whole integers. The comprehensive enclosure system U-value was found using the RadjustedTB, each WWR, and Rwindow values R-1—R-8. The formula to find this value is derived from the same formula as above: Utotal wall = (%window / Rwindow) + (%opaque + RadjustedTB) This gave us a table of 800 U-values, which were subsequently turned into 800 Rvalues using: Utotal wall = 1 / Rtotal wall Rtotal wall = 1 / Utotal wall Similar to our CSS process, we again found the necessary R-Value to achieve the RadjustedW after R-Value reduction from windows. RadjustedW = %opaque / ( (1 / Ropaque(adjustedTB)) - (%window / Rwindow) ) At this point we have an RadjustedW for every WWR accounting for Rwindow. The next step was a conversion of this table to five tables of insulation thickness’. We now have data for the thickness of five types of insulation needed to reach an RadjustedTB for 100 WWRs and 8 Rwindow values using a specific CSS. The next table converted this data into approximate cost per ft2. All of this data is in itself useful, but potentially difficult to use without a system to sift through it; a way to input parameters and receive the most cost effective combination of wall assembly to reach a goal. Therefore we developed interface allows schematic
designers to input data that we believe would most likely be known from the initial stages of the design process. The output available from the interface includes the RadjustedTB, and the price, type and thickness of the most cost effective insulation type for each Rwindow. We picked these variables as input and output because we believe they are most relevant to a schematic designer. A schematic designer will have a rough idea of the WWR, and hopefully after a conversation with the client and users, an Rtarget can be established based on the expected longevity of the building and other aspects of the comprehensive energy strategy. The schematic designer can very quickly find the limits of the established Rtarget and design around the resultant WWR. Because this tool is geared towards increasing the R-Value of the comprehensive enclosure system, an increase in wall thickness is a likely outcome of the design investigation and can be a powerful design consideration in the earliest phase of the project.
4. Results
After running multiple iterations of an R-30 target, we began to fully understand the implications of high R-Value systems. The most striking revelation, and perhaps the most obvious is that even with the best windows, an R-20+ enclosure becomes extremely hard to accomplish with anything less than an 80% opaque wall. We also quickly realized that the difference between a highly conductive CSS and a less conductive CSS is much greater than we anticipated. With standard metal furring using only 3 in2 of contact per unit, at 60% opaque wall, the best Rtarget a project of typical budgetary means can hope to achieve is around R-6. Throughout this research, we have continually asked ourselves what the real effects an increased R-Value will have on an enclosure system. We asked at what point do the benefits of a high R-Value enclosure system begin to be outweighed by the size of the thermal insulation, the restrictions on windows and the price of windows with high R-Values, and the size of the framing to accommodate these increasingly complicated systems. The Rtarget of the enclosure is just one factor to consider in the design of any building, and it is often highly influenced by climate, budget and the willingness of designer and client to experiment with potentially untested building systems. We hope that this tool will allow designers to make better decisions earlier in their design process. It was developed to give designers a better grasp of the implications of their designs and to inspire a conversation between the architect and client on energy goals for the project at its inception. The values should not be taken as actual values for particular assemblies. It has been created to help designers make more responsible
decisions when adding glazing, cladding, insulation and the cladding support systems. Though currently a viable tool for the industry at large, there are many improvements planned to increase its functionality. The first improvement is the addition of window pricing to the comprehensive output. Because the difference in pricing between an R-1 and R-8 window is so great, neglecting to include this as a design consideration significantly alters the user’s perception of project economy. Consideration for increased cost of framing also needs to be addressed. Further details such as projected air gaps in window systems and types, permeability of weather barrier and an increased variety of cladding support systems will continue to improve functionality. There is also an issue with the way the tool selects the best possible insulation type and thickness. Currently it will always use the least expensive insulation material, regardless of thickness, however a realistic limit on the thickness of insulation needs to be applied so that in the event the least expensive insulation does work at an exorbitant thickness, the output will move to an insulation within a reasonable thickness range. Other factors that could potentially add to the accuracy of results include consideration of isolating roof insulation from comprehensive enclosure insulation as well as building orientation and shading.
*All R-Values in United States Customary Units: ft2 * T * h/Btu *This tool was designed as a schematic design aid and is no way a true representation of actual possible enclosure R-Values. An issue with rounding will affect all resulting Radjusted values due to the very small fractions created by multiplying small percentages of building materials and very small U-Values and then finding the reciprocal. Microsoft Excel is limited to 15 decimal places, which will be rounded when it converts the actual number into binary coding. Further improvements to this tool will include increasing the number of decimals tracked and looking to alternative data management software.
References 1. Crabtree, George, et. al., “How America an Look Within to Achieve Energy
2. 3. 4.
5. 6.
Security and Reduce Global Warming,” Energy Future: Think Efficiency, American Physical Society, p. 21, Sep. 2008. Lstiburek, J. W, “BSI-001: The Perfect Wall,” Building Science Insights, May 2008. Straube, J, “011—Thermal Control in Buildings,” Building Science Digest, 11 Nov. 2006. Syed, A. M., and J. Kosny, “Effect of Framing Factor on Clear Wall R-vValue for Wood and Steel Framed Walls,” Journal of Building Physics, vol. 30, no. 2, pp. 163-180, Oct. 2006. Straube, J, “011—Thermal Control in Buildings,” Building Science Digest, 11 Nov. 2006. Brohard, G. J., et al., “Advanced Customer Technology Test for Maximum Energy Efficiency (ACT2) Project: The Final Report,” 1997.