I S S U E 2/2020  VOLUME 37  NUMBER 2

Journal of architectural technology published by Hoffmann Architects, Inc., specialists in the rehabilitation of building exteriors. Performance Considerations for Historic Window Replacement and Repair

Rachel C. Palisin, PE, LEED AP BD+C and Richard W. Off, AIA

F rom the stained glass windows Determining a path for historic win- that welcome variegated light into dow treatment that balances aesthet- cathedrals to the steel-framed geo- ics and historic integrity with contem- metric designs of Art Deco and the porary performance standards can be bold leaded glass of the Prairie style, challenging, particularly if the windows windows have served to define the are architecturally significant. character of buildings and spaces since Through thoughtfully designed and ex- the advent of glazed fenestration. Not ecuted repairs, some historic windows only do windows estab- can realize performance gains that ex- lish the aesthetics and tend their lifespan and improve indoor ambiance of a structure, comfort. However, where deteriora- they form an integral part tion is advanced, or where there is a of the building enclosure. More than ornament, compelling need to modernize the as- windows serve a vital role sembly to meet current performance in protecting the building standards, replacement may be war- from the elements and in ranted. In such cases, careful consider- modulating the transfer of ation of materials and window design heat, moisture, and light is critical to respecting historic char- from exterior to interior. acter while meeting project require- ments. Decisions about wood frames As windows age, their versus metal or composite, insulating components become glazing versus single-pane, true divided subject to the ravages of lights versus applied muntins, and time and weather: sealant historical versus modern anchorage, crumbles, wood decays, Meeting performance demands while preserving building among other considerations, require metal corrodes, glass character presents challenges for historic window projects. expert evaluation of the available op- deflects. If not properly tions. Testing, both in the laboratory maintained, historic win- and in the field, is a valuable tool to dows are in danger of deteriorating verify performance and adjust the final to the point where they are no longer design to meet the unique demands salvageable. Even for those windows of the building and situation. that have been carefully protected over the years, performance demands By applying the principles of window may raise considerations for replace- design with a sensitivity to the treat- ment with materials that provide ment of historic properties, build- improved efficiency and durability. ing owners and project teams can

Rachel C. Palisin, PE, LEED AP BD+C, Senior Engineer with Hoffmann Architects, oversees structural, thermal, and building envelope analysis to develop customized historic window solutions. Richard W. Off, AIA, Project Architect, has specialized expertise in the restoration and adaptive reuse of historic and landmark structures. JOURNAL

develop window rehabilitation solu- Energy and Thermal Performance is required depends on the location tions that respect the original building and the use of the space, as there may The thermal performance of a win- fabric while providing lasting, reliable be scenarios in which a high SHGC is dow is typically defined by U-factor, performance. preferred to allow passive heating, or the measure of how much heat is lost a high VLT may be desired to maxi- or gained in an assembly by radiation What Are the Principles? mize daylighting to reduce usage of and conduction. The numerical value, electrical lighting and power. Although Many of the performance require- expressed in decimals, is the transfer there are some jurisdictions that do ments and standards that should be rate of heat divided by the difference not have specific energy codes, most considered when approaching window in temperature and is the weighted building codes have minimum require- replacement projects are governed by average of U-factors for the center of ments tied to occupancy and climate. code. For residential projects, includ- the glass, edge of glass, and window ing single-family homes and duplexes frame, as defined by the National Structural Considerations of not more than three stories, the Fenestration Rating Council (NFRC). International Residential Code (IRC) Chapter 16 of the IBC defines struc- The lower the U-factor, the better the tural requirements for windows, applies; for all other commercial and energy performance. (R-value, used to residential projects, the International primarily regarding lateral loads due to measure the performance of opaque Building Code (IBC) and International wind and other sources. To establish wall materials, is the opposite.) For sin- Energy Conservation Code (IECC) are a uniform design pressure, commonly gle-pane historic windows, which have the prevailing model codes. While given in pounds per square foot, the high U-factors, condensation may be replacement windows must comply project’s structural consultant uses a an issue, particularly in environments with the performance standards out- wind speed mandated by applicable with tightly controlled mechanical sys- lined in these codes, in many jurisdic- buidling code, with adjustment fac- tems that limit natural ventilation. tions, historic and landmark buildings tors that account for building height, are exempt, so long as such windows Other common ways to evaluate topography, exposure, and occupancy. are replaced in kind, matching his- energy performance with windows Windows and their anchorage must toric conditions. However, the IBC still are Solar Heat Gain Coefficient (SHGC) be designed to withstand these pres- requires that safety glazing be installed and Visible Light Transmittance (VLT). sures, along with any other loads to in potentially hazardous locations, such SHGC measures the amount of solar which they may be subject. The code as windows at enclosed fire stairs. In radiation that passes through win- also requires that window glazing itself general, performance requirements dow glazing, and VLT measures the have appropriate wind and/or other may be organized into three main cat- amount of visible sunlight that passes applicable lateral load resistance, and egories: energy/thermal, structural, and through the entire window assembly. this is typically established via deflec- envelope considerations. Whether a high or low SHGC or VLT tion requirements, which indicate that glass cannot bend beyond a given maximum to avoid breakage. Certain regions, especially those susceptible to very high winds, such as hurricane-prone coastal areas, also have windborne debris resistance requirements. During storm events, various forms of unsecured debris may become windborne and can po- tentially impact windows, break glass, create even more debris, and poten- tially subject the interior of buildings to wind loads that could destabilize the structure. Building height, configu- ration, and proximity to the coast are Samples of original hardware (top left) facilitate restoration and, where necessary, replacement. all critical factors when determining Existing historic hardware may be removed and restored (bottom left), then re-installed (right). wind loads on buildings.

2 VOLUME 37  NUMBER 2

Building Envelope Integrity require a specific window class for any Envelope requirements for windows particular situation, but it does require are essentially concerned with keep- an AAMA 101 certification, which ing air and water out of the building. demonstrates compliance with mini- The two main metrics for evaluating mum performance criteria. building enclosure performance are Evaluating Window Performance air leakage and water penetration Energy/thermal, structural and enve- testing, both of which test windows in lope standards can all be evaluated in pressurized chambers to determine some capacity through testing. First, a the extent of air and water that pass series of laboratory tests establishes from one side of a sealed window to the performance baseline, through another. Although the IECC outlines both manufacturer/fabricator qualifica- typical envelope requirements for win- tion testing and also performance mock- dows, such as maximum requirements ups, or window units constructed for air leakage, ratings set forth by the solely for the purpose of laboratory American Architectural Manufacturers testing, to ensure their ability to meet Association (AAMA) are considered For this landmark military building, window specified standards and requirements. security was a top replacement consideration. the industry standard for new win- Depending upon project specifica- dows. (Note that, as of January 2020, tions, samples and visual mock-ups they may also be subject to in-situ/ AAMA and IGMA have unified to might also be incorporated into the field testing to further evaluate their create the Fenestration and Glazing submittal process to better evaluate Industry Alliance, or FGIA.) ability to meet performance require- window appearance, including review ments, and the extent to which the According to AAMA, following a series of materials, finishes, profiles, and sight complete fenestration assembly can of performance tests, new windows lines, before installation. This is espe- meet specified criteria, not just inde- can be designated one of four perfor- cially critical with historic replacement pendent window units. Typically, such mance classes, each of which is appro- projects, to verify that replacement on-site evaluation includes additional priate to a different degree of usage windows match original or landmark air and water testing, with some ac- conditions. and type of setting: counting for field conditions. • AW, generally used for high-rise If one or more laboratory tests fail, and mid-rise buildings, or if appearances are not satisfactory, How Are the Principles Applied? • C W, often used for mid and low- then mock-ups may be modified as Industry standards and code require- rise buildings, required to achieve acceptable results. ments are well established for new • LC, typically used for multi-family Then, such modifications are incorpo- windows; however, when it comes to dwellings, and rated into the final design and manu- applying these performance criteria to • R, used for one and two-family factured windows. Once windows historic window replacement and re- have been fully installed and sealed dwellings. pair projects, it can be more challeng- into their respective facade openings, Each window class requires a different ing, especially for projects with special combination and quantity of tests, with considerations. decreasing performance requirements Replacement Projects as the classes shift downward from AW to R, but all of which consider One example that illustrates these minimum structural loading, resis- complexities is a window replacement tance to air and water infiltration, and project conducted at a turn-of-the security, or resistance to forced entry. century National Historic Landmark These performance classes can apply armory building in a large metropoli- to windows of any material, although tan area. The project included com- the highest classes are commonly prehensive replacement of all original, achieved with steel- and aluminum- Visual mock-ups are useful in historic deteriorated windows, but, since the framed windows. The code does not window projects for verifying aesthetics. building serves as an active military

3 JOURNAL

facility, the client required that the re- Low-E coatings can pro- placement windows meet contempo- duce a colored tint, which rary performance standards, including may not be a desirable ap- energy code and anti-terrorism blast pearance for historic build- resistance. ings, and, therefore, coating The windows to be replaced were selection should be care- fully considered. As with old-growth softwood, and although SHGC and VLT ratings, the historically appropriate replace- there can be situations ments could not achieve the same where not using Low-E thermal performance as metal win- coatings may be desirable,

dows with thermal breaks, the new ABS GROUP PHOTO: to increase the potential mahogany hardwood frames offered Shock tube apparatus for blast resistance window testing. for passive heating. better performance than non-thermal- ly-broken metal windows. The original Even more challenging was given the position of the windows glazing was single pane, so to achieve the need to meet anti-terrorism blast- within the masonry walls, as well as an acceptable U-factor, Insulating resistant standards. All the replace- the anticipated design pressures, ep- Glazing Units (IGU) were used, or two ment windows and glazing needed to oxy anchorage was not employed be- lites (panes) of glass separated by an be able to resist a shock wave from cause it could not achieve manufactur- air cavity. To accommodate the IGUs, explosions or other ballistic forces at er-specified minimum edge distances, which are larger and heavier than the the exterior of the building. To achieve or the smallest allowable distance original single-pane units, the wood the requisite blast resistance, the IGUs between anchors and the outer edge frame needed to increase in depth incorporated layers of poly-vinyl-butyl of the masonry. Instead, closely spaced and weight. Fortunately, the replace- (PVB) plastic lamination, the same mechanical anchors were ultimately ment window frames were able to material used in car windshields. In this specified, which are both suitable to achieve the same overall exterior ap- project, the lamination was applied to the masonry substrate and able to pearance, and accommodate required the inside surface of each glass pane achieve structural requirements. operating hardware, in a manner that of the IGU, creating what is known as Another example of the unique chal- was acceptable to the State Historic a double-laminated IGU. For single- lenges of windows within historic Preservation Office. laminated assemblies, lamination can buildings is a project where a series be applied to either the interior or the To further increase energy efficiency, a of storefronts were installed at a large exterior lite. Double lamination maxi- Low-E (emissivity) coating was incor- commercial building. In this case, the mized the glazing’s resistance to lateral porated into the IGU assembly. These windows were not replacements, but forces, as demonstrated by subject- coatings limit the amount of ultraviolet were new metal windows and doors ing window performance mockups to light that can pass through the glass, installed in original loading deck en- laboratory-based shock tube testing, reducing the greenhouse effect on the trances throughout the ground floor where the designed windows were of a former warehouse, as part of building interior. However, it should evaluated for their ability to meet be noted that some high-performing an adaptive reuse project. Since the specified blast loads. loading docks were originally open In addition to the glazing without solid doors, and the building assembly, the anchorage is a landmark, the design team opted also required special atten- for large expanses of storefront glass tion, as any lateral forces with as few intermediate supports to which the IGUs are as possible to preserve the original subject will be transferred building profile. However, the building through the frames to the is located in a low-lying coastal area, surrounding masonry walls. so flood resistance and the potential A series of epoxy adhesive for glazing deflection when subject to anchors were originally flood loading were both critical con- considered for use due siderations. Therefore, it was necessary PHOTO: ENGINEERING EXPRESS, TEST REPORT ENGINEERING EXPRESS, PHOTO: Flood debris testing of a proposed window assembly. to their high strength, but to understand the stresses imposed (continued on page 6)

4 VOLUME 37  NUMBER 2

Window Performance Testing

(continued from page 3) While contractors or manufacturers may refer to a win- dow as having been “tested” in accordance with AAMA standards, building owners should note that the code only mandates qualification testing, which is performed on a manufacturer’s sample unit to demonstrate that the typical unit produced in similar configurations will meet the required thresholds. This type of testing is what allows a window to be designated as a certain AAMA perfor- mance class, such as “LC” or “AW.” In the field:Spray rack water test of a rehabilitated historic window. Code-compliant qualification and labeling of a window system is a general certification and does not guarantee that a specific window unit will meet performance standards in a given installation. Although not required by building code, best practice standards for project- specific performance testing of window assemblies may be included in a window replacement project. Testing confirms the design intent and level of perfor- mance and can provide quality assurance during construc- tion. There are two types of performance testing com- monly specified for window projects: off-site testing that In the lab: Wind and water test preparation for a window PMU. occurs before construction, and on-site field testing that occurs during construction. While similar in the perfor- or E 331), air infiltration (ASTM E 283), and structural mance metrics they are capturing, the methodologies and performance (ASTM E 330). Additional tests may be setups differ. specified depending on project requirements, such as Off-Site Testing life-cycle testing or lateral force (earthquake resistance). Tests may be repeated after each other to determine if Also referred to as a performance mock-up (PMU), off-site air infiltration, for example, increases after the window is testing is usually carried out at an architectural testing loaded with the maximum structural pressure. The PMU laboratory. The parameters of a PMU are set by the proj- generally represents testing the “best-case” scenario for ect specifications to any installation, since it is a controlled and easily accessible follow the proce- setup in a lab. dures of AAMA 101: North American Field Testing Fenestration Standard/ During construction, on-site tests provide quality assur- Specification; the ance. The general procedure for construction testing design professional should follow AAMA 502: Voluntary Specification for Field may elect to test a Testing of Newly Installed Fenestration. Testing consists of single typical unit or visual assessment of seals, alignment, and window opera- multiple configura- tion, as well as air and water infiltration tests according tions. A general test to the standards of ASTM E 783 and E 1105, respectively. program includes The project designer must determine the number and multiple tests for frequency of field tests, as well as passing thresholds for water penetration the specific metrics; it is not sufficient to only reference Water test of newly installed window. (ASTM E 547 and/ “test according to AAMA 502” in the specifications.

5 JOURNAL

(continued from page 4) on the glass and the maximum span it Repair Projects called cames, which are H-shapes for could achieve. While high performance standards middle section pieces and C-shapes To verify engineering calculations, flood often can be achieved in some capac- for edge pieces. All lead cames were resistance laboratory testing was per- ity in window replacement projects replaced and resealed to the adjacent formed, including flood debris testing, within historic buildings, it can be even stained glass with a linseed oil-based which evaluates the ability of windows more delicate and difficult an exer- waterproofing cement. Over time, as to withstand large debris carried by cise in window repair projects. One this cement deteriorates, windows fast-moving flood waters via impact example of what can potentially be are often temporarily repaired with loading, and also hydrostatic pressure accomplished is demonstrated in the elastomeric sealant, but when main- testing, which measures the pres- restoration of an early 20th century taining historic windows, it is essential sure of flood waters on glazing in a bronze-framed stained-glass window to use a sealant which is compatible controlled chamber. To achieve passing at a utility headquarters. The large or- with the original materials. Following results, it was again necessary to use namental window on the entry facade restoration, the window was rein- laminated glazing in the storefront showed numerous signs of deteriora- stalled, and although the performance itself, as well as within the glass fin-wall tion, and forensic in-situ water pen- and operability do not conform to system which was used to structurally etration testing, using an exterior spray contemporary standards, they are support the glass. The fin-wall system rack which mimics wind-driven rain, much improved. consists of a series of vertical glass fins, When evaluating alternative methods which run perpendicular to the glass for addressing deteriorated historic storefront, directly behind it at the windows or attempting to improve interior side, attached only by a very their performance, the use of protec- thin vertical line of structural silicone sealant. Glass fins were used in lieu of tive glazing or storm windows are often vertical metal mullions to enhance the considered, especially when replace- appearance of openness and mini- ment or full restoration is cost-pro- mize visual obstructions within the hibitive. Protective glazing is com- storefront system. Alternative and/or monly installed at the building exterior, supplemental means of flood resis- over existing stained-glass windows, to reduce air and water infiltration. tance measures were also consid- Documenting and reconstructing stained- ered, such as flood barriers or gates. glass mosaics is painstaking work. However, if such exterior glazing is This typically consists of metal posts not properly ventilated, hot air and and panels which would be securely confirmed that the window leaked condensation can become trapped in anchored to facades or in the ground and required extensive work. Given the air cavity between the interior and and requires a trained staff capable of that much of the decorative glass and exterior glazing and can further dete- rapidly deploying the system before an bronze frame components were in riorate the already distressed window. anticipated flood event. good condition, and that the win- Metal components can become cor- Other considerations in historic win- dow could be considered a priceless roded and displaced, glass can crack dow projects include general security artwork, the project team decided to and become dislodged, and additional concerns, as well as fire rating require- fully restore the window, rather than leaks may be generated. The use of ments. At a recent project at an urban replace it. The window was carefully engineered storm windows, on the library, determining whether wire disassembled, salvaged, and brought other hand, especially when installed mesh glass, security guards, or integrat- to a studio certified by the Stained at the interior side, can be an effective ed vandal guards were most appro- Glass Association of America. This means of improving both the energy/ priate was a primary design concern. allowed for the cleaning and archival thermal and envelope performance For lot line windows, stair towers, and documentation of each component, of historic windows. Nonetheless, de- fire-rated wall assemblies, steel-framed which would also facilitate replace- sign professionals and storm window fire-rated windows might be specified, ment should pieces become damaged manufacturers should be consulted to despite historical accuracy consid- during restoration. evaluate the best approach, as each erations, to comply with fire code Historic stained glass is typically held window repair situation presents its requirements. in place by thin lead components own unique problems. (continued on page 8)

6 VOLUME 37  NUMBER 2

Historic Window Rehabilitation Lexington Armory (1906) United States Capitol (1829/1866) , New York Washington, District of Columbia For more than 40 years, Hoffmann Window Replacement Dome Window Restoration Architects has guided clients in finding the right solution for historic windows, Ryan Hall (1903), Hawkins Hall (1833) Scholastic Inc. Headquarters (1889) from nineteenth century stained glass Georgetown University New York, New York to midcentury steel. Our architects Washington, District of Columbia Landmark Window Rehabilitation and engineers evaluate existing Exterior Restoration, Window Repairs Old Town Hall (1905) conditions, including anchorage, Saint Fidelis Church (1894) Stamford, hardware, wall opening, frame, sealants, College Point, New York Exterior Restoration and Window coatings, and glazing, and determine Restoration of Steel-Framed Stained- Replacement the best approach to balance historic Glass Windows integrity with modern performance. Tower Court Complex (1915-30) Our window projects at historic Jenkins-Waggoner Laboratory (1936) Wellesley College structures include: CT Agricultural Experiment Station Wellesley, Massachusetts New Haven, Connecticut Building Envelope and Window Aquarion Water Company (1931) Wood-Framed Window Replacement Bridgeport, Connecticut Rehabilitation Restoration of Bronze-Framed Stained- Flagg Building, Corcoran School of Bellarmine Hall (1920) Glass Windows the Arts & Design (1897) Fairfield University The George Washington University Fairfield, Connecticut Countee Cullen Library (1941) Washington, District of Columbia Window, Facade, , and Terrace New York, New York Building Envelope and Window Restoration Building Envelope Rehabilitation and Rehabilitation Window Replacement Benton Museum of Art (1920) 15 West 12th Street (1959) University of Connecticut Saint Bridget’s Church (1887) New York, New York Storrs, Connecticut Jersey City, New Jersey Analysis of Cold-Rolled Steel Windows Building Enclosure and Window Exterior Restoration and Window Consultation , Replacement (1933), New York, New York Window Replacement The Washburn Shops (1868) Worcester Polytechnic Institute Worcester, Massachusetts Dormer Window, Slate Roof, and Tower Rehabilitation Arts and Industries Building (1879) Packer Memorial Church (1885), Lehigh Smithsonian Institution Former Southern New England Univ., Bethlehem, Penn., Masonry Restoration Washington, District of Columbia Telephone Headquarters (1938), New and Window Replacement/Rehabilitation. Window Restoration Haven, Conn., Steel Window Refurbishment.

7 JOURNAL

Hoffmann Architects, Inc. 2321 Whitney Avenue Hamden, CT 06518

ADDRESS SERVICE REQUESTED

(continued from page 6) Balancing Performance with goals while making necessary com- Historic Integrity promises. Unfortunately, there is no JOURNAL is a publication of Hoffmann perfect window that lasts forever, and Architects, Inc., specialists in the With historic window projects, it is im- ongoing inspection and maintenance rehabilitation of building exteriors. The portant to understand limitations and firm’s work focuses on existing struc- will always be required. In particular, trade-offs. While it may not be possi- tures, diagnosing and resolving prob- maintaining sealant joints and finishes ble to realize all aspects of the desired lems within roofs, facades, windows, waterproofing materials, structural sys- aesthetic and performance, each proj- is critical, as they are the first line of defense. tems, plazas/terraces, parking garages, ect is governed by different dominant and historic and landmark structures. criteria, allowing achievement of salient Historic restoration is dependent We also provide consulting services for upon the availability of materials, as new building construction, as well as well as the properties and charac- litigation and claim support. teristics of the existing assembly, the For address changes, free subscriptions, surrounding structure, and repair and or information on the topics discussed replacement components. Certain in this issue, contact our Marketing materials and forms may no longer be Department at 800-239-6665, produced. Existing framing and attach- [email protected], or: ment elements may not be able to 2321 Whitney Avenue support replacement glazing. Original Hamden, CT 06518 materials may not perform as nec- 203-239-6660 essary to meet code requirements. 1040 Avenue of the Americas, Ste. 14C These and a host of other consider- New York, NY 10018 ations mean that historic window re- 212-789-9915 pair or replacement is rarely straight- forward. Understanding how to priori- 2711 Richmond Highway, Ste. 333 Arlington, VA 22202 tize project requirements and evaluate 703-253-9800 available options, as well as how to test and adjust the design to improve www.hoffarch.com This bronze-framed Art Deco window performance and aesthetics, are key to Editor/Production: Alison Hoffmann was removed, restored, and re-installed. a successful project outcome.

JOURNAL offers AIA/CES Learning Units Earn Health, Safety, and Welfare (HSW) continuing education credit for reading the JOURNAL. To learn more, visit www.hoffarch.com.

8