IN-SITU U-VALUE MEASUREMENTS AND THERMOGRAPHIC SURVEY: SPACE COMMUNITY CENTRE EDINBURGH CRAIGMILLAR

IN-SITU U-VALUE MEASUREMENTS AND THERMOGRAPHIC SURVEY: SPACE COMMUNITY CENTRE EDINBURGH CRAIGMILLAR Technical Report No. E4632_JS_2012 10 - October - 2012 Prepared for: CastleRock Edinvar Housing Association Edinburgh 1 Hay Avenue Edinburgh, EH16 4RW For the attention of: Chris Thompson Prepared by: Jon Stinson. Brief: To conduct side-by-side comparative study. Using in-situ U-value measurements and IR images to compare a thermal enhanced office space with a similar office with no thermal enhancement. [Forming part of a Technology Strategy Board funded project] Revised by: Julio Bors Williamson, SEC Consultant 42 Colinton Road, Edinburgh, EH10 5BT T: (0131) 455-5235 F: (0131) 455-5121 e: j.broswilliamson@napier.ac.uk

Contents Acknowledgement...2 1 Introduction...3 2 Methodology...5 3 Results...8 4 Review of results... 19 5 Conclusion... 21 U-value and IR thermography report 1

Acknowledgement The author would like to thank the funding body; Technology Strategy Board (TSB); the staff at SPACE community centre; and the staff within the two offices for facilitating this work. U-value and IR thermography report 2

1 Introduction This report forms part of the Technology Strategy Board (TSB) funded Invest in Innovative Refurbishment project awarded to Castlerock Edinvar Housing Association (CREHA). The (SEC) carried out a series of building performance tests upon the community centre called SPACE, an early 20 th century school converted into rentable offices, arts and community spaces. The work undertaken forms part of a wider report and submission to TSB for further funding to proceed with phase 2 of the innovative refurbishment plan. Work by the SEC was done to inform CREHA and the design teams appointed to instruct and project manage the fabric refurbishment work. In-situ U-value measurements were carried out on the building fabric. Equipment to measure thermal transmittance was installed within two rooms of the SPACE building One room to represented the existing fabric a second room underwent thermal upgrading. Infrared thermography survey was conducted localised to the same locations as the U-value probes. The monitoring and fabric evaluation study focused on four areas within each room: Window South wall - area below window South wall - top of wall North wall - top of wall U-value measurement equipment was installed in eight locations across two rooms on 24 th September 2012. Loggers were retrieved prematurely 2 weeks later in line with phase 1 TSB funding schedule, also to allow the results to populate the dynamic computer modeling and simulation scenarios undertaken by Green Energy Partners (GEP). The orientation and forecasted boundary temperature differential for September would result in an appropriate period of monitoring for this project to be 6 weeks. After the interim logger download to inform this report the data loggers have been redeployed into the building to allow for further measurements. The aim of the thermographic survey is to visually display areas of heat loss through the building fabric. Thermograms of the control room established a baseline representing the construction of the rest of the external envelope. Comparative thermograms of the refurbished room will assist in identifying areas where the thermal upgrading may require improvements and learn from areas that haven t been insulated properly or where thermal bridging may be occurring. U-value and IR thermography report 3

The following objectives have been identified; Establish thermal transmittance value (U-value) for existing and refurbished fabric elements Assess the continuity of insulation across the building facades Identify, where possible, air leakage paths resulting in excessive heat loss. The infrared survey was undertaken on the 2nd of October 2012, between 00:00 hrs 02:00 hrs, by Jon Stinson. 1.1 Survey description To take infrared images of the building, certain environmental conditions have to be met. To create a sufficient temperature gradient across the building structure, the survey is conducted in low ambient temperatures at night, and the building heating system is left on for a 24 hour period prior to the survey to maximise the temperature differential between internal and external environment. The survey is conducted with a minimum temperature internal to external temperature difference of 10 o C. This also ensures that the camera is detecting heat from within the building, and not latent heating remaining within the building structure from the day-time/solar gains. 1.2 Environmental conditions The environmental conditions were recorded before and after the survey. The measurement location of these parameters was selected to be representative of the mean. At the start of the survey, the internal temperature of the control room was 20.3 C. The refurbished room experienced an internal temperature of 22.9 C. External temperature of 3.3 C and a wind speed of between 1.3 m/s -1. At the end of the survey, the conditions had remained relatively stable, with an internal temperature similar to the above and an external temperature of 3.1 C and a wind speed of between 1.2 m/s -1. This provided a mean internal to external temperature differential of approximately 18.4 C. The survey was conducted in low wind conditions, which is within the permitted range which provides a more equal pressure and temperature differences over the facades of the building, increasing the comparability of images. The sky conditions were approximately 10% overcast throughout the survey. All building surfaces were dry and remained so for the duration of the survey. The relatively uniform night sky presented good survey conditions. U-value and IR thermography report 4

1.3 Building description The building originally constructed in 1930 s for use as a local primary school. The building underwent a change of use from a primary school to a mixed use office and community centre, during this time some refurbishment work was undertaken in early 2000 s with the addition of internal wall insulation at various point across the front facade. The exact make-up of the existing building elements is not known, architect assumptions have been used and comments have been made based on drawings used in this report. The report will focus upon two rooms, hereby referred to as a control room and a test room. Both rooms occupy the same general location within the building, frontal orientation and room dimensions, at time of this report both room were being occupied by the same tenant. The rooms are located beside one another on the first floor, East wing and facing approximately South-West. Each room measures 7.9m by 6.2m with one main entrance and four 1.2m by 1.9m timber sash and case windows. The test room is occupied by 10 workstations accommodating office staff for the average 35hr working week. The control room was occupied less often and used primarily for staff meetings and training space. To compensate for the difference in occupancy the control room was artificially heat to a similar level as that experienced during in the test room. Temperature loggers were deployed in both rooms to validate the heat profile in the rooms. For the purpose of this report the control room is representative for the construction type and elemental build-up for the rest of the existing structure. The test room has undergone thermal upgrading to select portions of 2 walls and each of the four windows. Detailed breakdown of existing and additional material is explained in Section 3. 2 Methodology The testing methodology for calculating the in situ U-value measurements was undertaken in the wellestablished format described in Baker (2008, 2011), Rye (2011) and Currie (2012). The U-value of a building element or component is defined in BS EN ISO 7345:1996 as the heat flow rate in the steady state divided by the area and the temperature difference between the surroundings on each side of a system. (BSI, 1996, p.5) U-value and IR thermography report 5

2.1 In-situ U-values The in-situ U-value measurements were taken with a Hukseflux HFP01 thermopile-based heat flux transducers (Figure 1) of 80 mm diameter and 5 mm thickness. These were attached to each building element being tested throughout the initial period of monitoring (for 2 weeks duration). The elemental U-values were determined by recording the heat flow through the element together with internal surface and external air or surface temperature. This was done by logging differential voltage from the heat flux transducers and temperature from calibrated K-type thermocouples (Figure 2). Grant Squirrel data loggers with 24bit A/D conversion resolution (Figure 3) were used to log data from the heat flux transducers and the thermocouples. Where required, Tinytag temperature and relative humidity loggers (Figure 3) were used for locations where cabling could not be run back to the main data logger. Figure 1 : Heat flux transducer (heat flow mat) mounted to wall face Figure 2: K-type thermocouple mounted to face of heat flux transducer Figure 3 : Grant Squirrel data logger to log data from heat flux transducers and thermocouples Figure 4: Stand-alone Tinytag View2 data logger measuring air temperature and relative humidity U-value and IR thermography report 6

Error analysis of in-situ measured U-values The certainty of the measured in-situ U-values are influenced by sensor related errors. The sensitivity of the senor or probe will impact on each recorded value during the period of monitoring. Calculating the uncertainty related to temperature probes and HFM allows for an error range to be found which provides a ± value indicating the level of uncertainty derived from the individual temperature and heat flux measurements. An error analysis for the results from each case study has been calculated by using the established error analysis methodology (Baker 2011). The results are presented as part of the analysis of U-value results chapter 4. 2.2 Thermography Infrared thermography was used to locate areas of significant temperature differences and inhomogeneous construction elements. The IR camera also assisted with ensuring that the sensors were not placed on cold bridges, by identifying the placement of metal studs which assisted with the optimal placement of the heat flow mat. Table 1 provides equipment and calibration details for those items of equipment used during the survey. Table 1: Equipment and calibration information Equipment Serial No. Date of calibration expiration Calibration certification no. FLIR thermal imaging camera B335 with 320 x 240 pixel resolution and 25 o lens 48803639 11/03/12 n/a Anemo Rotating Cup Anemometer N/A N/A 55007 Testo 110 Thermometer with Probe. 33922213/805 11/03/12 UK04606 U-value and IR thermography report 7

3 Results 3.1 Window The windows in the South facing facade tested in the study are single glazed timber sash and case window (Figure 5) these are representative of all the windows along the listed front facade. During the refurbishment the windows in the test room were thermal enhanced with the installation of toughened low-e double secondary glazing unit in an openable upvc frame with trickle vent added to the four windows. The secondary glazing unit called Thermal Shield developed by Adam Dudley Architects of Edinburgh, is installed internally with approx. 100mm gap between glazing units and the existing window. Additionally a novel radiant heat barrier window blind installed between glazing units (Figure 6). Figure 5:Internal view of window in control room, with HFM attached Figure 6: Internal view of window in test room, Thermal Shield secondary glazing installed and opened on vertical hinge Interim results from the U-value in-situ measurements have shown that the addition of Thermal Shield secondary glazing system improved the U-value of the glazing from 5.2 to 0.6 W/m²K results in Table 1. U-value of 5.2 W/m²K is typical for single glazing, whilst the measured U-value of 0.6 W/m²K is typical for the secondary glazing system that has been previously measured in past SEC historic building refurbishment performance evaluation projects. Therefore, the 2 week interim results would suggest that the addition of the radiant heat barrier blind has not influenced the U-value. However, four U-value and IR thermography report 8

temperature probes embedded either side of the blind has identified a noticeable temperature gradient, see Section 3.1.1. Table 1: Comparative measured U-value results for glazing Building element Control room (Existing fabric) U-value (W/m²K) Enhanced room (Refurbished fabric) U- value (W/m²K) Glazing 5.2 0.6 The significant improvement in U-value is visually represented in the thermogram displayed in Figure 9. Four windows highlighted in the left of Figure 9 have been treated with Thermal Shield in the test room, the four windows highlighted on the right are of the control room, and these windows are representative of the windows in the rest of the front facade. The darker hue across the four treated show less heat loss compared to their counterparts. The brighter spot above the darker windows is a security perimeter light clearly seen in Figure 10 and not a source of heat loss. Figure 7: Thermogram, front of SPACE building. Two rooms highlighted by black outline. Test room on left control room on right. U-value and IR thermography report 9

Figure 8: Front of SPACE building. Showing four windows of the test room (Highlighted by black outline) 3.1.1 Heat barrier blind The results from the six temperature probes embedded at four points within the space between each the two glazing units are displayed in Figure 9, showing a temperature profile for two days at various points within and on either side of the window. Two temperature probes were installed on the outside of Thermal Shield facing toward the space between the double glazing and the heat barrier blind recording air temperature at two points, 100mm from the sill and 100mm from the header. The same was replicated on the inside of the single glazed pane, recording the air temperature at two points between the single glazing and heat barrier blind. The results suggest that the blind was acting in a manner that could be attributed to a heat barrier. However, the interim results suggest that Thermal Shield is doing most of the heat resisting show temperatures both sides of Thermal Shield. It was anticipated that a microclimate would form between both sets of glazing as a result of the introduction of the heat barrier blind. An advantage may have been the onset of convective heat currents operating within the microclimate. The results appear to suggest that the heat captured between Thermal Shield and the blind is rising as the heat increases inside the area between Thermal Shield and the heat barrier blind. However, it is difficult to ascertain if the trapped heat is being reintroduced into the room via the trickle vent at the top of Thermal Shield. U-value and IR thermography report 10

Figure 9: Temperature profile at six points across the window with Thermal Shield and Heat barrier Blind The results from the internal temperature probes shows a noticeable increase in temperature as the temperature difference either side of the blind increases at 16:00-17:00hrs on the two example days in Figure 9. A pattern has emerged from the peak temperature profile that would suggest that the marginal increase in room temperature may have been affected by the heat trapped on the inside of the heat barrier blind and reradiated back through the glazing and into the room. However, room user involvement with the heat system can be ruled out at this stage. It is worth commenting at this interim stage that the large amount of heat build-up between the blind and thermal Shield may be the result of heat struggling to pass back through the double glazing and into the room. Therefore the advantages of Thermal Shield as a barrier to internal heat escape may be negatively influencing the inherent properties of the blind as an alternative heat escape barrier. 3.2 Top of South wall Heat flow mats were applied to the area above the suspended ceiling and between each of the windows (see Figure 12 and 14). The elemental build-up of the existing structure consists 500mm precast concrete lintel, 75mm unventilated air gap, 50mm mineral wool insulation, 12.5mm plasterboard on 50mm metal framing (see Figure 10). During the refurbishment stage 75mm hemp insulation added to air gap between concrete lintel and mineral wool (see Figure 11). U-value and IR thermography report 11

100mm hemp insulation added atop joists 100mm hemp insulation added between joists 500mm precast concrete 75mm hemp insulation added into void between concrete and mineral wool Suspended ceiling tiles Existing 50mm mineral wool insulation Figure 10: Architect s detail of wall header section listing elemental construction. Materials added during refurbishment highlighted in black outline Table 2 lists the results from the in-situ U-value measurements. 50mm of mineral wool insulation added during a pervious upgrade programme resulted in the U-value now representative of the existing structure to be measured at 0.43 W/m²K. The addition of hemp to the void between the lintel and existing insulation has improved the U-value of the wall to 0.24 W/m²K. Extra insulation in this area coupled with lapped insulation to the joints above will reduce air leakage pathways and improve the junction s vulnerability to thermal bridging. Further work into the specificity of airtightness and thermal bridging is scheduled to be explored during further research in this area. Table 2: Comparative measured U-value results for wall above ceiling tiles Building element Control room (Existing fabric) U-value (W/m²K) Enhanced room (Refurbished fabric) U- value (W/m²K) Wall: above ceiling tiles between windows 0.43 0.24 Thermogram in Figure 11 and Figure 13 show the temperature difference between the ceiling tiling and insulated walls for the control and test room respectively. Figure 11 indicates that the area above the suspended ceiling is colder and heat is escaping. In comparison, Figure 13 shows the temperature of the area above the ceiling tiles is similar to that on the insulated wall, showing that the added insulation in to wall header and joists have reduced heat loss. U-value and IR thermography report 12

Figure 11:Thermogram of wall in control room showing temperature difference between ceiling tile and wall Figure 12: Photograph of image in Figure 11 Figure 13: Thermogram of wall in test room showing temperature of ceiling tile Figure 14: Photograph of image in Figure 13 The brighter area at the top of the blinds (Figure 11) in the control room is representative of the highly reflective metal material and therefore does not represent actual heat loss through that area. Compared to the same area in Figure 13 which shows colder areas to the top of the window is common for timber surround and the opened trickle vent. U-value and IR thermography report 13

3.3 Wall: Below window The area below the windows on the South facing façade is built-up of 25mm roughcast external render, 230mm solid brickwork, 50mm mineral wool insulation, 170mm unventilated air gap, 12.5mm plasterboard internal lining. The refurbishment work included 170mm of bead insulation blown into the gap between mineral wool and plasterboard (see Figure 15). 230mm solid brickwork Existing 50mm mineral wool insulation between metal studwork 170mm blown insulation added to existing framed out void behind plasterboard Figure 15: Architect s detail of wall below window section listing elemental construction. Materials added during refurbishment highlighted in black outline Results from in-situ measured U-value have shown that the addition of blown bead insulation has improved the thermal transmittance of that area of wall from 0.36 W/m²K to 0.12 W/m²K (see Table 3). Table 3: Comparative measured U-value results for wall below window Building element Control room (Existing fabric) U-value (W/m²K) Enhanced room (Refurbished fabric) U- value (W/m²K) Wall: below window 0.36 0.12 The results from the infrared thermography show that the area below the window in the control room is colder than that witnessed in the refurbished room (see Figure 16, 18, 20). The images confirm the finding in Table 3. The three thermograms have been scaled to within the same temperature range to allow for visual comparison. The darker coloured gradient spread across the surface of the wall below U-value and IR thermography report 14

the window in the control room (Figure 15) indicates areas of heat loss. This becomes more evident towards the bottom of the plastboard. Figure 16: Thermogram of wall in control room showing temperature of area below window Figure 17:Photograph representing Figure 16 Figure 18: Thermogram of wall in test room showing temperature of area below and left of window Figure 19: Photograph representing Figure 18 Retrofitting isolated areas of a wall presents issues with maintaining continuity of the thermal envelope. Figure 18 shows a dark area representing a cold area of heat loss left of the area beneath the window. Although this area has not been thermal enhanced it does show a clear contrast to area that has been treated. U-value and IR thermography report 15

Figure 20: Thermogram of wall in test room showing temperature of area below window Figure 21: Photograph representing Figure 21 Darker patches highlighting areas of heat loss in Figure 20 are limited to sections of the plasterboard where metal studwork has been detached. It is typical for IR images to highlight these areas therefore they are not of concern. This observation differs from Figure 16 as the wall area beneath the window in the test room is of a similar hue across the thermogram and is within the higher end of the temperature scale. U-value and IR thermography report 16

3.4 Top of North wall A small portion of the North facing wall in the main offices is a non-list cavity wall with the upper portion being an external wall. This upper external wall consists of 25mm roughcast external render, 90mm solid brickwork, 110mm cavity, 110mm brick, 50mm unventilated air gap, 50mm metal stud framing finished internally with 12.5mm plasterboard. During refurbishment 110mm blown bead insulation was pumped into the cavity between the brick leaf (see Figure 21). Further to this hemp insulation was installed to the cavity behind the plasterboard and to the overhead joists to form a continuous insulating layer in an attempt to mitigate cold bridging. 100mm hemp insulation added atop joists 100mm hemp insulation added between joists 75mm hemp insulation added wall above suspended ceiling 90mm brick outer leaf 110mm blown bead insulation added to cavity 110mm brick inner leaf Figure 22: Architect s detail of wall header and roof connection on North facing wall section listing elemental construction. Materials added during refurbishment highlighted in black outline With the addition of blown bead insulation the U-value of the wall in the test room is lower to that of the North wall U-value in the control ( see Table 4). However the measured U-value is not as low as that predicted for a thermal upgrade of this type and application. A number of variables can be identified to explain this result. Firstly, it is important to be cautious with results retrieved before the minimum period of monitoring has past as explained in Section 4. Therefore this reported value maybe misrepresented at this stage and by a very large uncertainty factor. Secondly, some or most of the blown bead insulation may have disproportionally settled and/or shifted after install resulting in less insulation within the cavity. Alternatively, the results may be an indication that the air gap between the internal lining and internal brick leaf is experiencing the infiltration of colder external air ventilating the cavity space. This U-value and IR thermography report 17

observation becomes more evident when considering the results witnessed in the thermograms of the North wall in both rooms. Table 4: Comparative measured U-value results for North wall Building element Control room (Existing fabric) U-value (W/m²K) Enhanced room (Refurbished fabric) U- value (W/m²K) Wall: North facing below ceiling tile 1.11 0.53 The thermograph of the control room North wall shows heat loss through air leakage paths along the perimeter beneath the suspended ceiling on the North side. The darker areas in Figure 23 are inconsistent with the general surface temperature along the upper portion of the wall. Observations made in the test room show similar signs of ventilation heat loss along the area beneath the suspended ceiling. Although to a lesser degree, Figure 25 shows that heat is being lost in the same area which may be an indication of an issue with continuity when placing and lapping of insulation. Figure 23: Thermogram of wall in control room showing temperature of gradient across North wall Figure 24: Photograph representing Figure 23 Figure 25: Thermogram of wall in Test room showing temperature gradient across North wall Figure 26: Photograph representing Figure 25 U-value and IR thermography report 18

With the results gathered to date, the evidence would suggest that some infiltration ventilation effects are being witnessed behind the internal lining. Actions available to address that issue would include the insertion of additional hemp insulation into the framed out cavity behind the plasterboard, to a point below the suspended ceiling. However, this would prove challenging to identify the most cost effective depth and a mechanism by which to secure the hemp insulation without removing the internal lining. Alternatively, further post installation survey work can be undertake to identify if continuity of insulation has been compromised during the installation process along with the aid of airtightness testing and more exploratory infrared thermography. Further work into identifying the nature of this and similar issues is scheduled to be explored during further research in this area. 4 Review of results An error analysis has been carried out on the measured U-values results for both rooms. It is recognised that an uncertainty level of ±10% is common for this testing methodology; this is to account for diurnal cycles and measurement tolerance of the equipment. It has been identified that the results were calculated prematurely to satisfy various reasons. For work carried out on Southern façades, insufficient time was allowed to account for heat stored during prolonged period of solar heat again to the face of the building. In this respect the level of uncertainty for this round of results is between to ±50 and ±65%. These results correlate with previous SEC findings (Currie 2012), which demonstrated the relationship between the temperature differential and level of uncertainty attributed to the results collected on a South façade. Figures 27 and 28 have been generated to portray the level of uncertainty attributed to these sets of results. U-value and IR thermography report 19

In-situ measured U-value (W/m²K) In-situ measured U-value (W/m²K) Technical Report No. E4632_JS_2012 10 9 8 7 6 5 4 3 2 1 0 Existing Refurbished Glazing Building element Figure 27: Measured U-value results and error bars representing level of calculated uncertainty for glazing measured during first two weeks of monitorng 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Existing Refurbished Existing Refurbished Existing Refurbished Wall: above ceiling tiles between windows Wall: below window Building element Wall: North facing below ceiling tile Figure 28: Measured U-value results and error bars representing level of calculated uncertainty for walls measured during first two weeks of monitorng U-value and IR thermography report 20

5 Conclusion Through the use of side-by-side comparative data analysis, two rooms in the SPACE building at Craigmillar Edinburgh were monitored and surveyed for existing and enhanced levels of fabric thermal performance. The interim results from the first 2 weeks of in-situ U-value measurements have identified that the range of insulation and fabric upgrade work introduced into a test room has improved that room s ability to further resist the flow of heat lost through the fabric compared to the neighbouring room of similar dimensions and heating profile. Infrared thermography was undertaken to allow for the visual identification of heat loss. The external thermograms has shown that the Thermal Shield system has reduced the amount of heat loss through and around the existing windows as compared to the surrounding windows representing the primary window system type used in the front of the building. Internal images of the South wall have shown that the areas where additional insulation was added, the surface temperature of that element had become more uniformed show less heat loss through that portion of wall. The IR survey has highlighted a possible issue with the refurbishment strategy applied to the North wall. Therograms of the area beneath the suspended ceiling on the North side has shown signs of ventilation heat loss. Comparison between thermograms of each room would suggest that the ventilation heat loss has been mitigated to some extent by the addition of the insulation, however some still persists. The colder patches identified by the IR survey are an indication of heat escaping into a colder boundary, whether this be the cold attic space or the external environment is unclear. However, it would suggest that the insulation to the inside of the top of North wall and ceiling joists may have been compromised. This document is due for revision upon the completion of the full period of monitoring required for insitu U-value measurements. The longer required period of monitoring will allow of a measured U-value result that will be within the desired ±10 uncertainty level. Prepared by: Checked by: Jon Stinson Energy & Buildings Research Assistant Julio Bros Williamson - B Arch. (Hons), MSc Energy & Buildings Consultant U-value and IR thermography report 21