December 24, 2020
By Helen Sanders, PhD
Curtain walls are ubiquitous worldwide, delivering sleek, easy to install, and highly glazed building façades. Curtain wall systems provide excellent access to daylight and views that are desired by building owners and occupants for health and productivity. However, the increasing stringency of building codes has created a conflict with delivering highly glazed buildings because glazed areas are known to be less thermally efficient than opaque walls.
Curtain wall systems present an even larger conflict because their opaque, spandrel areas cannot easily match the thermal performance of an opaque wall assembly (concrete, steel stud wall, etc.). For a curtain wall system to achieve comparable performance to a mass wall with the same window-to-wall ratio, the thermal performance of the curtain wall’s transparent areas must be substantially greater than in a mass wall system to make up for the poorer opaque wall performance. Additionally, current thermal modelling of spandrel areas over-estimates the thermal performance, leading to gaps between expected and delivered performance.
Over the years, the industry has relied heavily on increasing performance of low-emissivity (low-e) coatings to improve curtain wall U-factors (thermal transmittance). However, the centre of glass (COG) U-factor is only part of the picture.
Focusing on preventing heat flow through the COG only, rather than considering the assembly as a whole, is the thermal equivalent of damming a river at its centre, yet forgetting to block water flow all the way to its banks. In the case of the river, the water just flows around the edges of the dam and the flow is not stemmed. Similarly, in the case of the curtain wall, the energy still flows around its perimeter—through the frame and edge of glass (EOG)—even if the COG conductance is low (Figure 1).
Managing thermal bridging at interfaces and connection points is critical to delivering expected performance. Specifications can be used effectively to establish performance expectations and ensuring these are managed and delivered through the construction process.
Curtain walls are complex systems, and need to fulfil not just thermal requirements, but also those for air infiltration/exfiltration, water management, and structural performance, to name just a few. If not specified completely, including appropriate verification of design and installation performance, a significant gap between expected and as-built performance can occur.
The required high thermal performance can be achieved in curtain wall systems with enough focus on upfront design details, management of interfaces, and the execution phase.
Balanced design: Specifying the edge
The thermal transmittance (U-factor) of fenestration comprises the area weighted average of the thermal transmittances of the frame, EOG, and COG. The frame consists of the opaque elements holding the glass. EOG comprises the perimeter area of the insulating glass unit (IGU) containing the spacer and sealants, and COG refers to the vision area of the glazing.
Due to the area weighting, in curtain wall with smaller openings and more framing members, the frame and edge seal thermal transmittance dominate the overall U-factor compared to designs with larger openings where the glass area-to-metal perimeter ratio is higher. The National Fenestration Rating Council (NFRC) uses standard sizes for different fenestration types (e.g. fixed window, awning, skylight, etc.) to allow comparisons between different systems of the same fenestration type. When specifying the U-factor performance of curtain wall, it is important to specify the whole assembly U-factor and what size the required U-factor target is based on: NFRC or project size.
According to NFRC 100, Procedure for Determining Fenestration Product U-factors, thermally improved performance for aluminum fenestration requires a small, low-conductivity thermal isolator of minimum 1.6 mm (63 mils) between interior and exterior framing elements, whereas the thermally broken performance requires a significantly wider thermal barrier of at least 5.3 mm (210 mils). Thermally broken systems, therefore, deliver higher thermal performance than thermally improved systems.
Even when specifying assembly U-factor requirements, it is still important to differentiate between thermally improved and thermally broken curtain wall. Experts recommend a best practice of specifying the minimum size of the thermal break in addition to the U-factor requirement. For curtain wall, the recommendation is to require a polyamide thermal barrier with a minimum width of 19 mm (¾ in.)—and wider if higher performance is needed.
Figure 2 shows whole unit (assembly) U-factors (NFRC size) for:
Figure 4 displays a subset of the data in Figure 1 to illustrate the greater impact of improving frame and EOG performance on assembly U-factor. In this example, there are two ways to achieving an assembly U-factor of 1.99 W/m2K (0.35 BTU/F.hr.sf). In the lower performance curtain wall, triple-pane glazing is required. However, by choosing a high-performance thermally broken curtain wall and a high-performance warm-edge spacer in the IGU, only a simple, low-cost, air-filled, dual-pane, low-e IGU is needed to meet the same performance. Improving the frame and EOG first allows for greater flexibility in choice of the COG, and provides a more balanced approach.
The thermal performance of curtain wall systems also can be improved by replacing the typical aluminum pressure plate with a non-metal version made from glass-reinforced polyamide (Figure 5) or fibreglass. These are widely available from multiple manufacturers in a large range of sizes. The addition of a polyamide pressure plate can decrease the system U-factor substantially. Figure 6 shows the temperature profiles for a typical thermally improved curtain wall containing a small thermal isolator with an aluminum and polyamide pressure plate, respectively.
Using a typical dual-pane, low-e, argon-filled IGU as the glazing infill, the polyamide pressure plate brings the U-factor of a thermally improved curtain wall system down from a high of 2.33 W/m2K (0.41 BTU/F.hr.sf) to 1.87 W/m2K (0.33 BTU/F.hr.sf)—an improvement of 20 per cent. Using a non-metal pressure plate in combination with wider thermal barriers within the mullions can improve thermal performance even more. The curtain wall system installed at the Georgia College of Law building in Atlanta, Ga., incorporates polyamide pressure plates in combination with wide polyamide thermal barriers that were specified to support achievement of the U.S. Green Building Council’s (USGBC’s) Leadership in Energy and Environmental Design (LEED) Silver certification.
Structural silicone glazing
Silicone structurally glazed curtain wall systems typically have better thermal performance than their captured counterparts—all other things being equal. This is due to the elimination of thermal bridging associated with metal pressure plates and related fasteners. Figure 7 compares the performance of a captured curtain wall, a two-sided structurally glazed curtain wall, and a four-sided structurally glazed system. The four-sided structurally glazed system has a U-factor 10 to 18 per cent lower than the fully captured system using aluminum pressure plates, depending on the insulating glass edge spacer used. The data in Figure 7 also demonstrates the relative impact of using a high-performance warm-edge spacer in the IGU instead of a highly conductive aluminum spacer.
In a typical captured curtain wall system, a high-performance warm-edge spacer reduces the assembly U-factor by 0.11 to 0.17 W/m2K (0.02 to 0.03 BTU/F.hr.sf) or ~5 to 10 per cent, but the impact of insulating glass edge spacer is even higher in a structurally glazed system. This is because, in the absence of an external frame, the major conduction path from inside to outside is directly through the edge of the glass. Thus, the performance of EOG is even more important in structurally glazed systems. In the example in Figure 7, a high-performance warm-edge spacer reduces the U-factor of the structurally glazed system by 14 per cent or 0.28 W/m2K (0.05 BTU/F.hr.sf). A significant amount, given the performance enhancements needed to meet more stringent code requirements with high glazing areas.
Specifying the edge of glass
Unfortunately, unlike “thermally broken” and “thermally improved,” the performance of “warm-edge” spacer in North America is not well defined. A warm-edge spacer is assumed to be any spacer with better thermal performance than the aluminum box type, which spans a wide range of performance. Figure 8 shows the assembly U-factors of a curtain wall with three warm-edge spacer types: Stainless-steel box spacer, plastic hybrid stainless steel (PHSS or hybrid) box spacer, and a flexible foam spacer—compared to using an aluminum spacer. The stainless-steel box spacer results in only slightly better performance than the aluminum box spacer, yet it is still referred to as warm-edge.
It is, therefore, important to specify more than just “warm-edge” spacer in project specifications; otherwise, what ends up installed in the project, may not meet the overall thermal performance requirements for the system. Ideally, the type of warm-edge spacer with material specifics should be specified. This not only ensures the thermal performance is consistent with delivering the performance of the assembly, but also the spacer is appropriate to the application. For example, some edge spacer systems, while having good thermal performance, are more suited for residential-type applications and are not appropriate for use with dual-seal polyisobutylene/silicone edge seal systems or in structurally glazed applications.
It is important to require insulating glass certification from a reputable certification program, such as through the joint Insulating Glass Certification Council (IGCC)/Insulating Glass Manufacturers Alliance (IGCC/IGMA) program or the Insulating Glass Manufacturers Association of Canada (IGMAC) program, which is part of the Fenestration and Glazing Industry Alliance (FGIA). It is also important to require submittals of ASTM E2190, Standard Specification for Insulating Glass Unit Performance and Evaluation, test reports for the spacer and sealant system that are proposed for the project. These reports must demonstrate high performance for durability and gas retention. The most stringent certification programs require test units to be made on the production line in the presence of an auditor annually and then sent for testing according to ASTM E2190.
Another conclusion to be drawn from the data in Figure 8 is a PHSS spacer results in the same thermal performance improvement as non-metal types. The high desiccant carrying capacity box profile and thin solid stainless steel envelope around the back and sides of the hybrid spacer, delivers the same benchmark durability as the stainless steel and aluminum box spacer with the thermal performance of non-metal spacer. In recognition of this, leading insulating glass certification bodies allow the substitution of hybrid box spacers for stainless steel without additional testing.
Figure 9 shows images of the structural glazing in the newly renovated Space Needle, which required the use of a high-performance hybrid warm-edge spacer to meet the stringent Seattle energy codes. The Atmos level’s (observation deck) classic reverse curtain wall had to be preserved. Therefore, enhanced performance at EOG was needed to control the flow of heat from the large external mullion (acting as a heat sink) to the inside. On the Loupe (restaurant) level, a high-performance warm-edge hybrid box spacer was specified again for the oversized IGUs with unsupported edges to block heat flow at the edges. The hybrid system was called out in the specification because of the thermal performance and the rigidity needed to manage the torsional stresses at the head and sill edge support conditions.
It is especially true in curtain wall systems U-factor and the condensation resistance are not correlated well. As a result, one cannot assume a system with a low U-factor will deliver the expected condensation resistance. Improving the U-factor of a curtain wall system means reducing the amount of metal framing and increasing the use of engineered plastic thermal barriers, yet condensation resistance can improve with increasing metal mass on the interior mullion. As a result, it is important to specify condensation resistance along with U-factor performance.
There are several measures of condensation resistance (read the article, “Effectively Specifying Fenestration; Managing thermal, structural, and durability performance,” published in the November 2017 issue of The Construction Specifier). The most common are condensation resistance (CR), as defined by NFRC—a calculated value—and condensation resistance factor (CRF), as defined by the American Architectural Manufacturers Association’s (AAMA) 1503, Thermal Transmittance and Condensation Resistance of Windows, Doors and Glazed Wall Sections, an FGIA standard, and based on measured values. Even though they share the same scale from zero to 100, the systems are not the same and cannot be compared. It is important to be clear on which metric is specified, and ensure submittals provided to demonstrate compliance are consistent with the chosen performance system.
Research has demonstrated there are two major performance issues related to opaque spandrel areas in curtain wall systems (consult Stephane Hoffman’s presentation on “Energy Code Implications for Spandrel Design; Quantifying and Mitigating the Impact of Thermal Bridging” in the Façade Tectonics Institute’s Proceedings of 2016 World Congress).
First, heat flows around the insulation in the spandrel because the position of the insulation is not aligned with the position of the thermal breaks in the metal framing. This allows heat to bypass the insulation via the aluminum frame. Even if the curtain wall is thermally broken, the framing still has less thermal resistance than insulation. Moreover, conventional 2D modelling, per NFRC 100, over-estimates the thermal performance of spandrel by up to 33 per cent because it does not capture this heat flow effectively (consult Stephane Hoffman’s presentation on “Energy Code Implications for Spandrel Design; Quantifying and Mitigating the Impact of Thermal Bridging” in the Façade Tectonics Institute’s Proceedings of 2016 World Congress).
Second, heat flows unhindered to and from the transparent areas flanking the spandrel through long, continuous vertical mullions. Current simulation methods, including NFRC 100, do not include the impact of thermal bridging to transparent areas, which also causes an over-estimation of thermal performance. Research shows thermal bridging to flanking modules can reduce performance by up to 30 per cent (consult Stephane Hoffman’s presentation on “Energy Code Implications for Spandrel Design; Quantifying and Mitigating the Impact of Thermal Bridging” in the Façade Tectonics Institute’s Proceedings of 2016 World Congress).
In fact, not only do spandrel areas have worse performance than opaque mass walls, the standard 2D simulations (such as NFRC 100) used to determine U-factor over-estimate thermal performance. This means curtain wall performance is actually worse than expected unless these heat transfer mechanisms are accounted for and controlled.
To address heat flow around the spandrel insulation, the insulation ideally should be aligned with the glazing and thermal barrier. One example attempting to do this uses a vacuum insulating panel (VIP) within an IGU in the spandrel area (Figure 10). It needs to be accompanied by a high-performance warm-edge spacer to reduce thermal bridging at the edges and wide thermal barriers in the curtain wall framing, which presents best-case spandrel performance (consult Stephane Hoffman’s presentation on “Energy Code Implications for Spandrel Design; Quantifying and Mitigating the Impact of Thermal Bridging” in the Façade Tectonics Institute’s Proceedings of 2016 World Congress).
It is recommended specifications for spandrel assemblies, where the spandrel height to floor height is more than 40 per cent, require 3D modelling to figure out thermal performance or physical guarded hot box testing according to NFRC 100 to determine performance. In the absence of both, it is best to use the 2D NFRC 100 calculated U-factor increased by 30 per cent (referenced from Yvon Chiasson’s oral presentation on “Building Material: Performance and Geometry Driving Expansion of the Façade Materials Palette” at a Façade Tectonics Institute forum in Toronto in May 2019).
Research shows that in spandrel to floor heights of less than 40 per cent, the impact of the spandrel will not be significantly different than of the transparent areas because heat loss at the edge of frame significantly impacts the value of the insulation (consult Stephane Hoffman’s presentation on “Energy Code Implications for Spandrel Design; Quantifying and Mitigating the Impact of Thermal Bridging” in the Façade Tectonics Institute’s Proceedings of 2016 World Congress).
To address the degradation in performance due to thermal bridging to flanking systems, more creativity is required, and may change the look of the curtain wall. Various strategies have been discussed within the façade design community including:
This last option would certainly change the esthetics of the curtain wall, but the verticals would no longer run continuously through the glazed and spandrel areas.
As an additional resource, the Building Envelope Thermal Bridging Guide (BETB Guide) commissioned by B.C. Hydro and built on the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) Research Project Report RP 1365, Thermal Performance of Buildings Envelope Details for Mid and High Rise Buildings, provides a large catalogue of many different wall assemblies, including spandrel panels, ranging from poor to high-performance.
The BETB Guide provides performance values derived from 3D thermal simulations for each assembly, which accounts for thermal bridging effects, and which can be referenced for design and specification purposes.
Thermal bridges at attachments
Another thermal bridging area to watch is through curtain wall attachments, such as sun shades. Specifying thermally broken attachment mechanisms and requiring thermal simulations to determine the impact on assembly U-factor is important.
Thermal bridges at transitions: Beware of “by others”
Façade engineers often warn of the dangers of ignoring the interfaces of the curtain wall with adjacent building elements and the unintentional thermal bridging that can occur if not managed effectively. The widespread use of the “by others” clause can cause the responsibility for managing the important details of the transitions to be missed. Who has responsibility for tying into the adjacent systems must be specified.
Additionally, sequencing of work around the transitions should be laid out in design documents to the extent possible. This becomes apparent at the mockup stage, but this is often too late and ideally should be laid out in design documents and reviewed in the shop drawings in advance. The following are recommendations for the execution section:
Getting what is expected: Quality assurance
It is important to recognize field testing or commissioning of curtain wall is not a replacement for the upfront verification the design itself will meet needed performance. It is important to require submittals of laboratory test reports to validate the curtain wall being proposed meets the specified air, water, and structural performance, according to AAMA/the Window and Door Manufacturers Association (WDMA)/the Canadian Standards Association (CSA) 101/I.S. 2/A440, NAFS – North American Fenestration Standard/Specification for windows, doors and skylights, before the start of fabrication.
The same recommendation applies to requiring submission of test or simulation reports demonstrating the system meets thermal performance according to NFRC 100, providing clear assumptions for glazing infill configuration and components, including EOG details. These details must be consistent with the glass specification or strict rules for evaluation substitution risk must be applied.
These test data will demonstrate if the system performs under perfect conditions. Additional field testing for air, water, structural, and thermal performance can be specified to control for installation quality.
For larger projects, design concepts also could be required to be presented during the bid process, along with performance validation of the proposed solution, to prevent the systems with poorer performance being inadvertently substituted.
In addition to specifying the size of the thermal barrier (as identified above), it is important to include requirements to help ensure the appropriate quality of the thermal barrier material and its integration into the aluminum extrusion.
Specifications should include a requirement for fabricators to follow AAMA QAG-2-12, Voluntary Quality Assurance Processing Guide for Polyamide Thermal Barriers, and the quality tests referenced in AAMA TIR A8-16, Structural Performance of Composite Thermal Barrier Framing Systems, which details a range of flexural, shear, and tension tests that ensure the required structural performance is achieved. These routine structural tests are important to control the risk of lower quality material substitutions (e.g. using polyamide with lower glass fibre content, or with uni-directional fibres rather than high-quality 3D fibres, or with impurities or voids) or poor process control, which can cause degradation to structural performance.
Curtain wall systems are complex. The considerations for delivering thermal performance have been discussed here, but there are many other considerations to be balanced. Finding a trusted partner is highly recommended to support the development of robust specifications that are consistent across the curtain wall and glazing sections and do not let important items fall through the cracks. It is often useful to ask a third-party to provide a specification review to help catch potential issues before they negatively impact the performance of the as-built structure.
To create a robust specification for a high-performance curtain wall, it is important to focus on several areas. First, improve the thermal performance of the frame and EOG, and then specify COG. Specifying an assembly U-factor performance in the curtain wall section is a must, as is a separate specification for CR, since the two are not correlated.
These assembly-level specifications also should be augmented with a minimum thermal break dimension and an EOG performance, which is consistent with achieving that assembly performance. For EOG, provide details of the edge spacer material and type rather than using the generic “warm-edge” term to achieve the thermal performance and durability desired. The “warm-edge” spacer performance definition is too broad to be effective in specifying high-performance fenestration systems and will often deliver a stainless-steel spacer, which does not match the performance required.
High-performance warm-edge spacers, such as PHSS spacers, can deliver a five to 15 per cent U-factor improvement in the assembly U-factor depending on the curtain wall system. They have the highest impact with structurally glazed systems. EOG, COG, and frame performance specifications should be consistent with achieving the overall assembly U-factor requirement.
Next, focus on the spandrel areas. Recognize and adjust for the performance degraders and the lack of accuracy in regular 2D thermal simulation. If possible, find ways to line up the insulation with the insulating glass and thermal barriers. Implement strategies to minimize the thermal bridging from transparent to spandrel areas through vertical mullions.
Actively manage thermal bridging at the interfaces through the specification execution sections. Avoid accepting “by others” by specifying which party is responsible for the interface connections and require a roadmap for sequencing and interface management.
Finally, remember to specify submittal requirements ensuring the contractor’s proposed curtain wall assemblies meet the design requirements. Do not rely solely on mockups and commissioning tests. By then, it is too late to discover the system itself does not perform as required. Mockups and site testing can only manage installation quality, they cannot address design gaps. Trusted partners are invaluable in supporting the development of high-quality specifications that deliver high-performance curtain walls.
Helen Sanders, PhD, manages strategic business development at Technoform. She has 25 years of experience in glass technology and manufacturing, with expertise in functional coatings, insulating glass, and thermal zone technologies for fenestration. She is the president of the Façade Tectonics Institute and a Fenestration and Glazing Industry Alliance (FGIA) board member. She has a master’s degree in natural sciences and a doctorate in surface science from the University of Cambridge, England. She can be reached at firstname.lastname@example.org.
Source URL: https://www.constructioncanada.net/thermally-improving-curtain-wall-systems/
Copyright ©2021 Construction Canada unless otherwise noted.