December 28, 2018
by Antonio Renzullo and Domenic Chiovitti
Over the last 60 years, aluminum and glass curtain wall systems have been incorporated within a vast number of modern building envelope designs. The broad use and acceptance of these wall systems by the design community is indicative of their success. The systems’ flexibility and ability to offer a range of building façade design possibilities make them popular with architects and owners. Although curtain wall system designs have proven themselves over the course of time with advancements in several areas, experience has shown these systems are not exempt from problems.
In these authors’ experience, commonly encountered curtain wall performance issues include water infiltration, excess air leakage, and condensation. In northern climates, condensation and ice formation may occur on wall surfaces during the coldest periods of the year. Surprisingly, for winter conditions in cold climates, condensation-resistance performance for standard curtain walls is acceptable for typical occupancy conditions. However, under extreme cold or higher levels of interior humidification, the limitations of many curtain wall systems become apparent in the form of condensation or frosting. In some cases, occupancy factors, unconventional building geometry, design details, and the design of heating systems and interior finishes may also contribute to the reduction of condensation resistance during cold periods.
In this case study, a building in the Montréal area was experiencing condensation on interior curtain wall surfaces, specifically at the exterior corners formed by the primary façades (Figure 1). In many of these instances, interior drywall finishes were damaged under conditions of extreme cold (from the wetting of surfaces), and frequent repairs were required (Figure 2). In an attempt to mitigate the reoccurrence of condensation during the winter months and to minimize damages and occupant complaints, the client requested an investigation of the problem areas in the building.
The property under study is a 12-storey commercial office building constructed in 1989. The exterior building envelope consisted primarily of a standard aluminum and glass curtain wall assembly. The corners formed at the junction of the building’s primary façades were particular in design and comprised areas of sloped glazing. The building envelope design also included precast concrete wall panel inserts distributed over the four façades. HVAC systems consisted of a zoned variable air volume (VAV) system for general air/cooling requirements. Additionally, electric baseboard heaters installed along the building perimeter provided heating. The heaters were recessed below the vision units of the curtain wall assembly.
During the initial phases of the investigation of condensation issues at the building corners, the following methodology was utilized.
Collection of information
Meetings were organized with building management, building operations personnel, HVAC contractors responsible for environmental systems operation, and tenants of office spaces to determine:
Survey of conditions and data collection
During periods of intense cold, site visits were made to assess curtain wall performance and to determine the severity of condensation on interior surfaces. Specifically, the following was undertaken at selected floor levels of the building:
Thermal scans and point surface temperature measurements confirmed the greatest condensation potential occurred on the curtain wall interior surfaces at the building’s exterior corners.
Interior surface temperature measurements at those corners obtained during periods of extreme cold (–18 C [–0.4 F]) revealed temperatures as low as –0.1 C (31.8 F) for areas with minimal or no perimeter heating. In areas where heating systems were in operation, temperatures as low as 2.4 C (36.3 F) were measured. In comparison, for similar surfaces within typical areas of curtain wall, temperatures were measured to be approximately 4 C (7.2 F) warmer. These measurements were taken on the surfaces of vertical mullions, at the junction of vertical and horizontal mullions (nearest sealed glass unit). The inside glass surface temperatures were consistently measured to be higher than the coldest aluminum surface temperatures.
The following specific issues were identified during the investigation as being problematic and likely to promote condensation on interior curtain wall surfaces at the building corners. Surprisingly, many of these issues are unrelated to the design of the curtain wall system itself.
The laws governing heat flow dictate the exterior corners of any building present an inherent thermal weakness due to geometry. During winter, an exterior corner mullion assembly creates a condition in which more exterior surface area is exposed for cooling in comparison to the surface area available from interior surfaces. Consequently, an exterior corner mullion assembly is generally subject to a thermal penalty in comparison to a typical wall mullion. As a result, slightly colder interior surface temperatures are expected. The opposite is generally true for interior corners where more surface is exposed—as a result, comparatively warmer temperatures are expected.
Interior building structural obstructions
The placement of the building’s structural columns in close proximity to the exterior corners of the building envelope was found to impede air circulation in those areas. Poor air circulation due to interference by the building structure was also found to be a contributing factor in reducing the energy available to heat interior wall surfaces at the building corners. Thus, building structural obstructions were a significant contributing factor to the low surface temperatures recorded (Figure 4).
Lack of perimeter wall heating at corners
Perimeter heating was notably absent at the corners of the building. It is speculated the obstruction presented by the building structure may also have contributed to a decision not to install heaters directly at the building corners. In some instances, the temperature setpoint for the perimeter heating system was adjusted to a lower setting by the occupant and the existing heaters were found to not be in operation. In some cases, equipment failure was suspected and heating capacity was further reduced. Interior air temperatures at corner offices were measured as low as 18 C (64.4 F) due to these issues. The lack of perimeter heating was identified as one of the most important contributing factors in low interior curtain wall surface temperatures (Figure 5).
Erroneous HVAC RH measurements
Measurements of interior air conditions obtained at several floor levels were compared with values reported by the control system from the floor-level air returns. Lower than actual RH level readings were reported by the HVAC control system. In one instance, approximately 10 per cent RH error was noted to exist between a reading reported by a humidity sensor located in a return air duct and measurements made with a wet/dry bulb psychrometer within occupied spaces. As a result, a condition of over-humidification was noted to exist in certain areas.
As part of the normal building humidification process, humidification levels were initially found to be maintained at 30 per cent RH during periods of occupancy, regardless of exterior temperature levels.
Ventilation and heating system schedules
As an energy conservation measure, ventilation and perimeter heating systems were found to be in limited operation during unoccupied periods or not in operation in areas of certain floor levels due to vacancies. These practices were also found to be maintained under conditions of intense cold.
As an initial mitigation strategy, several measures were proposed to the building management to minimize the formation of condensation at the building corners in extreme cold. These changes were applied to areas of a test floor. The following modifications, repairs, or adjustments were employed due to ease of implementation and relatively low cost:
Due to high costs, a series of modifications proposed for the perimeter ventilation system to improve airflow around structural columns (through the installation of linear diffusers) were not implemented. The option was, however, retained as a last-resort measure if initial attempts proved insufficient to significantly improve condensation resistance at building corners.
Following the implementation of these initial changes on a test floor, the corner office spaces on that level were reviewed regularly during periods of cold weather to verify their effectiveness. During these site visits, thermal scans and surface temperature measurements of curtain wall interior surfaces were recorded.
An increase in interior curtain wall surface temperatures was observed. In some instances, localized condensation was still noted and some accumulation had occurred on sills of some of the offices of the test floor. In these cases, the perimeter heaters were marginally operational due to a low thermostat setting (set by the occupant of the office space).
Temperature measurements indicated the effect of installing supplemental heaters at building corners produced an increase in interior curtain wall surface temperatures of approximately 2.5 C (4.5 F) near the sills of vision areas (with an exterior temperature of approximately –18 C [–0.4 F]). Significantly higher temperatures were observed near the head of vision areas. Although initial efforts produced positive results with a marked reduction in surface condensation, estimates of projected interior surface temperatures for colder exterior temperatures still indicated additional improvements were required.
For interior air conditioned to 22 C (71.6 F) and 25 per cent RH (reduced from 30 per cent RH), surface temperatures significantly in excess of the dewpoint temperature (approximately 1 C [33.8 F]) would be required during periods of extreme cold to provide sufficient margin against the formation of condensation on interior surfaces.
Prior to revisiting considerations for the modification of perimeter ventilation systems for additional temperature gains (or other exotic and equally expensive alternatives), some exploratory dismantling was performed within the cavity wall finish below sill level (within a corner space on an unoccupied floor) to examine the spandrel area of the curtain wall assembly.
Subsequent observations indicated a layer of batt insulation measuring approximately 152 mm (6 in.) in thickness had been installed within the cavity space between the curtain wall insulated pan and the interior drywall finish. A similar condition was also confirmed to exist within the suspended ceiling space at the junction of the curtain wall and overhead slab (Figure 6).
The installation of batt insulation within the cavity space of the interior finishes was subsequently analyzed to determine its impact on curtain wall condensation resistance.
The installation of batt insulation within the wall cavities of interior finishes is inadvisable, but is not an uncommon practice. Certain designers and builders sometimes feel compelled to stuff all wall cavities with insulation. However, experience has shown the placement of batt insulation within an interior wall cavity requires careful analysis and may significantly impact the condensation resistance of the wall assembly.
In regards to curtain wall installations, experience has also shown the placement of batt insulation within the wall cavity of interior finishing will affect condensation performance. The batt insulation isolates the curtain wall from the warming effects of the interior heated air. Consequently, a reduction of interior curtain wall surface temperatures within the spandrel area occurs. Given the thermally conductive nature of metallic curtain wall materials, the cooling of adjacent curtain wall components would be expected. Interstitial condensation may also occur within the batt insulation or on hidden interior curtain wall surfaces within the spandrel area if the dewpoint temperature of interior air is reached.
With interior air at 21 C (70 F) and 30 per cent RH, a computerized condensation analysis of a simplified representation of the curtain wall construction indicated the formation of interstitial condensation within the batt insulation at a rate of 1 g/m2/day at exterior temperatures of –17 C (1.4 F). The analysis further indicated the rate of formation of condensation increased to 8.5 g/m2/day as the exterior temperature approached –23 C (–9.4 F).
When the analysis was repeated for an interior air condition of 21 C and 25 per cent RH, interstitial condensation was reduced within the batt insulation to a rate of 0.7 g/m2/day at exterior temperatures of –17 C. The rate of formation of condensation also decreased to 1.9 g/m2/day as the exterior temperature was decreased to –23 C.
The analysis also indicated the presence of batt insulation within the wall cavity would also produce a temperature drop of approximately 16.5 to 19 C (29.7 to 34.2 F) within the assembly, resulting in a corresponding backpan metal surface temperature ranging from 2.4 to –0.2 C (36.3 to 31.6 F) for an exterior temperature varying from –17 to –23 C, respectively (Figure 7).
In comparison, a similar analysis of the wall assembly undertaken with identical temperature conditions and with the batt insulation removed from the wall cavity indicated a temperature drop of approximately 3.3 to 3.8 C (5.9 to 6.8 F) across the wall cavity. As a result, the backpan metal surface temperature would range from 15.7 to 14.9 C (60.2 to 58.8 F) for an exterior temperature of –17 C and –23 C, respectively (Figure 7).
Based on these comparisons, it became clear the presence of batt insulation within a wall cavity significantly reduced surface temperatures of curtain wall components and increased the potential for condensation and ice formation.
The computations also demonstrated the benefits of having reduced humidification levels during periods of extreme cold. The simulations indicate for a reduction of five per cent RH, interstitial condensation would only occur for exterior temperatures at or below approximately –22 C (–7.6 F).
The localized cooling effect produced within the spandrel sections of the curtain wall likely contributed to the cooling of apparent surfaces and the formation of condensation. Large, thermally conductive pathways provided by the mullions at the building corners likely provided a heat flow path to the colder, opaque sections of the curtain wall assembly, much like a heat sink. This phenomenon further decreased the temperature of exposed surfaces.
The computer simulations also suggested a reduced risk of condensation for the flat areas of the curtain wall assembly and condensation was not visibly apparent during site verifications. It was speculated the reduced levels of condensate production in the typical wall areas was actually stored and frozen within the insulation during periods of intense cold and these areas would subsequently thaw and dry out during warmer periods.
At the building corners, the compounding of condensation-promoting factors was sufficient to produce conditions that saturated the batt insulation. Once sufficient moisture had accumulated, moisture migration along surfaces occurred to eventually produce an accumulation on sills.
This time lag between the visible identification of condensation and the conditions that actually produced it may have contributed to the misdiagnosis of the condensation issue in the past.
To improve condensation resistance at the building’s exterior corners, it was clear the batt insulation within the interior wall cavities would have to be removed. A building corner was selected on the test floor for the removal of the wall cavity insulation and additional testing. To minimize the effects of heat sinking to levels above and below the test floor, the insulation from the wall cavities on those levels was also removed.
To gain access to the wall cavity insulation at the building corners and while minimizing cost and disruption to tenants, the complete dismantling of interior wall finishes was not an option. In lieu of the selective dismantling of finishes, a series of 102-mm (4-in.) diameter holes was cut within the interior wall finishes of one exterior corner below the curtain wall assembly and behind the obstructing structural columns at three floor levels of one particular building corner (with the middle level being the test floor). From these openings, any batt insulation within arm’s reach was extracted (Figure 8). Following removal of the insulation, round plastic vents were installed within the circular openings as an attractive finishing and to promote ventilation of hidden curtain wall surfaces (Figure 9).
The insulation situated within the suspended ceiling space was also removed by partially dismantling sections of the drywall. Following removal of the insulation, the drywall within the ceiling space was reinstalled.
The modifications to the existing wall assembly were completed in preparation for a period of sustained cold exterior temperature, when additional performance evaluations were undertaken.
A series of site visits were made in order to collect additional surface temperature measurements following the modifications to the wall assembly at the test corner of the building and to monitor performance. In between the site visits, consistent and extreme cold exterior temperatures from –18 to –28 C (–0.4 to –18.4 F) were experienced. During site visits, exterior temperatures of –18 C and –19 C (–2.2 F) were measured.
The removal of insulation from interior wall cavities and the installation of vent caps increased interior surface temperatures by approximately 3 C (5.4 F). Temperature readings collected at the test corner over three consecutive levels also indicated the combined effect of the modifications over successive floor levels further increased surface temperatures. The coldest surface temperatures in these areas were measured between 4 and 6 C (39.2 to 42.8 F) at sill level, with an exterior temperature of approximately –18 C, an improvement of approximately 4 C (7.2 F).
During one of the site visits within the period of intense cold, condensation was noted at a floor level outside the test corner (for which wall cavity insulation had not been removed). No condensation or water accumulation was observed on the curtain wall surfaces within the corner offices of the three floors comprising the test corner.
Subsequent reviews of the test area by building operations staff during the weeks following the monitoring period (with periods of very cold weather) also did not produce any reports of condensation.
Significant increases in surface temperatures were obtained for problematic areas of this curtain wall system by implementing simple measures.
A summary of recommendations and remedial efforts undertaken to mitigate the formation of condensation included:
Following a monitoring period to verify performance, the implementation of simple and cost-effective modifications proved successful and the condensation of interior curtain wall surfaces was mitigated. Based on the success of the limited tests undertaken, a recommendation was presented to the client to proceed with full-scale implementation of the mitigation strategies at all affected areas of the building.
This case study demonstrates the resistance to condensation for curtain wall assemblies can be adversely affected by factors external to the assembly itself or by site conditions. Condensation reduction strategies were applied to resolve a particular issue with good results. In this case study, issues specific to a particular building design and construction that reduced the condensation resistance of the curtain wall assembly were identified. Slight modifications to the building environmental controls and interior wall construction proved successful in the mitigation of condensation in problematic areas of the building envelope.
The successful resolution of this case required a detailed investigation of specific building issues affecting curtain wall performance during periods of extreme cold. Many of the recommendations discussed can be implemented on other buildings. However, due to the variation in building designs and the strong likelihood of special conditions, it is recommended that a thorough investigation be conducted and remedial efforts be verified prior to large-scale implementation.
Antonio Renzullo is vice-president of Chiovitti Consultants with almost 15 years of experience in building envelope consulting. His specialization includes building envelope inspection, technical investigations, and analyses of building envelope deficiencies and assembly failures. Renzullo has also conducted numerous façade condition surveys and safety inspections. He has participated in field performance compliance testing of building envelope assemblies and has served as expert witness. His additional experience includes data acquisition, instrumentation, research and development, and equipment procurement and development. Renzullo can be reached via e-mail at firstname.lastname@example.org.
 Domenic Chiovitti is president of Chiovitti Consultants and specializes in building envelope consulting. Chiovitti has more than 25 years of hands-on experience and has been recognized by the court as an expert in envelope performance failures. His specializations include onsite quality control (QC) monitoring for new and renovated structures, review of building envelope design and specifications, and recommendations on materials, components, and systems.
Source URL: https://www.constructioncanada.net/condensation-on-curtain-wall-surfaces-an-investigation/
Copyright ©2022 Construction Canada unless otherwise noted.