Authors: Rosanne Walker and Sara Pavía, Department of Civil, Structural and Environmental Engineering, University of Dublin Trinity College.
Introduction
The building sector is one of the key consumers of energy in Europe using approximately 450Mtoe per annum of which space heating accounts for about 70 per cent of the total energy use [1]. Retrofitting thermal insulation to existing buildings can reduce unnecessary heat loss minimising energy consumption.
Historic buildings (built before 1944) form a large proportion of the building stock of most European cities. They are often of architectural and historic interest and it is essential that any thermal upgrading does not undermine their special character. In most cases, a balance can be found between protecting the building's heritage value and energy saving interventions that lessen adverse environmental impact, reduce building energy costs and improve occupant comfort ensuring the long-term viability of historic buildings.
The architectural and historic significance of the exterior facades of historic buildings precludes the use of external insulation for most structures. Internal insulation is considered a more viable alternative although it can still be invasive by introducing new materials, replacing historic linings, disturbing internal features such as joinery and distorting the original room proportions.
A further concern with internal insulation is its physical compatibility with traditional and historic construction. The application of thermal insulation on the interior of a wall can result in the accumulation of moisture within the wall and potential interstitial condensation, frost damage, timber decay and mould growth.
This research investigates the in-situ thermal and moisture behaviour of a range of internal insulations applied to historic solid brick walls in a case study. It considers improvements in thermal performance and addresses durability concerns relating to moisture accumulation following the application of insulation.
1.1 Thermal performance of buildings
The building industry commonly uses thermal transmittance (U-values- W/m
2K) to measure heat transfer through walls. Thermal transmittance of a wall is commonly measured in-situ using heat flux sensors. In-situ measurement is important, as there is significant discrepancy between in-situ-measured and calculated/modelled wall U-values.
Inaccurate wall U-values can result in misguided assessments of energy saving options, over specification of insulation requirements, lower than expected improvements in thermal performance and incorrect estimation of energy savings.
The amount of heat loss through walls is frequently speculated and figures ranging between 10 and 45 per cent are commonly quoted. There is no average value, and actual heat loss depends on numerous variables including wall surface area, age, composition, condition and construction technology.
However, the application of insulation to walls and the resulting reduction in thermal conductivity should significantly contribute to an overall improvement in the energy performance of a building. The magnitude will depend on several variables such as building type, climatic conditions and performance of insulation materials [2].
Thermal insulation materials retard the rate of heat flow by conduction, convection and radiation [2]. Insulation reduces heat loss through a wall by reducing the thermal transmittance (U-value) with further thermal benefits including warmer surface temperature and reduced air permeability through the wall.
1.2 Effect of thermal insulation on wall moisture
Changing the balance between heat, air and moisture movement in a wall can affect the building’s integrity [3]. The application of internal insulation on the interior of a wall can result in moisture accumulation within the wall and potential interstitial condensation, frost damage, timber decay and mould growth.
Moisture accumulation in the insulated wall can occur on account of two reasons: the reduced permeability of the insulation impeding the wall drying towards the interior and the lowering of the wall temperature by the insulation increasing the likelihood of water vapour condensation leading to moisture accumulation. The Sustainable Traditional Buildings Alliance’s (STBA) reports knowledge gaps on the hygrothermal performance of insulated walls [4].
Water is widely regarded as the most prevalent cause of decay in historic buildings. Modern buildings are typically constructed using hard impervious materials that prevent moisture ingress. Traditional and historic buildings, however, function under a different premise, as their breathable fabrics allow moisture to dry from the wall (both internally and externally) preventing moisture accumulation over time. Heritage constructions are consequently more vulnerable to moisture loads than modern buildings [5].
The risk of moisture accumulation in a particular wall is dependent on multiple factors including weather conditions (exposure to driving rain, sunlight, wind) and orientation. The properties of the wall itself such as thickness, condition and material properties (pore structure, hydrophillicity/hydrophobicity and others). The behaviour of occupants conducting water generating tasks such as drying clothes and showering, can also influence moisture accumulation in walls.
1.3 Methods to accommodate moisture when internally insulating solid walls
There is no agreement on the most appropriate method to accommodate water when internally insulating solid walls. Broadly, two approaches are used to avoid moisture accumulation when insulating historic structures: vapour tight system and capillary active/vapour open systems. The vapour tight systems use a vapour retarder on the interior side of the insulation to prevent moisture from entering the wall.
This, however, prevents drying towards the interior and does not allow moisture buffering of room humidity. Furthermore, perforations can result in substantial underperformance. Capillary active insulation works on the premise that wall moisture can be transported towards the interior by capillarity and the wall allowed to dry.
The moisture behaviour of vapour tight and capillary active insulation systems reported by different authors in traditional structures is inconsistent probably due to the wide variety of insulation materials, walls, sources of moisture ingress and individual circumstances.
There is a wide range of internal insulation currently available. In recent years, there has been increased interest in vapour permeable insulations in place of conventional vapour tight systems. Sustainable insulation materials with lower embodied energy and reduced environmental emissions are also increasing in popularity and a large number of innovative insulations are constantly entering the market.
A suitable internal insulation system should be tailored to individual specific building requirements taking into account the multiple variables that influence moisture accumulation. As noted Nielsen et al. (2012), the same refurbishment method might have different outcomes when applied to different buildings [6].
Methods
2.1 Description of case study
In-situ testing of the thermal and moisture behaviour of the insulation was undertaken at the Adjutant General’s Building in the Royal Hospital Kilmainham, Dublin (Fig 1 and main pic). The building was designed by the internationally renowned architect Francis Johnson (1760-1829) and constructed in 1805. It consists of approximately 770mm solid brick walls (400mm beneath the windows).
[caption id="attachment_27722" align="alignright" width="300"]
Adjutant General’s Building in the Royal Hospital, Kilmainham, Dublin[/caption]
The exterior is a roughcast lime render that was re-rendered in c.2005. The interior plaster was removed approximately 30 years ago and the building was treated for timber decay. Additional detail on the case study can be found in a further publication, 'Thermal performance of a selection of insulation materials suitable for historic buildings' [7].
2.2 Insulation materials
A selection of seven insulations are investigated and compared to a traditional lime plaster. These include thermal paint; aerogel, cork lime; hemp lime; calcium silicate board, timber fibre board and polyisocyanurate (PIR) board. The composition and properties of the insulation investigated can be found in detail in a further publication, ‘Thermal and hygric properties of insulation materials suitable for historic fabrics’ [8].
A traditional lime plaster was used as a control by which to compare the performance of the other insulation materials. Lime plaster is not considered an insulation material although it has good thermal properties [9].
The insulation was applied in separate sections to the walls of three rooms on the north and west elevations of the first floor. Seven insulation and two lime plaster (control) sections were installed in the building over a four-month period as set out in table 1. The walls were levelled using lime plaster (table 1) before the application of the insulation.
| Material |
Coat/thickness |
Composition |
| Plaster to level all walls |
Scud coat c.5-6mm |
2.5:1:0.63 sand:NHL5:water |
| Scratch coat 25-75mm to make walls plumb |
2.3:1:0.67 sand:NHL3.5:water |
| Lime Plaster (control)(LP) |
Floating coat c.12mm |
3:1 :0.60sand:NHL3.5:water |
| Skim coat 3mm |
1:1:0.5 sand:NHL2:water |
| Painton lime plaster (P) |
Floating coat c.12mm |
|
| Skim coat 3mm |
As above |
| 3 coatsof paint |
Emulsion with ceramic additives |
| Aerogel
(AG) |
19.5mm aerogel and plasterboard |
As per manufacture’s spec using mechanical fixings |
| Gypsum skim coat 3mm |
As per manufacture’s spec |
| Lime and hemp(HL) |
2*20mm layers |
Hemp:NHL2:water 1:2.9:3.5 (by weight) |
| Lime and Cork(CL) |
2*20mm layers |
Cork/lime: water 2.15:1 (by weight) |
| Calcium silicate board
(CSB) |
30mm |
As per manufacture’s spec using adhesive to affix board |
| proprietary skim coat c.6-7mm |
Basecoat with mesh imbedded (4-5mm) and finish coat (2mm) |
| Timber fibre board(TB) |
40mm |
As per manufacture’s spec using mushroom fixings |
| proprietary skim coat c.6-7mm |
Basecoat with mesh imbedded (4-5mm) and finish coat (2mm) |
| Thin PIR with foil |
37.5mm |
As per manufacture’s spec using mechanical fixings |
| Gypsum skim coat 3mm |
As per manufacture’s spec |
Table 1. Details of lime plaster and insulation materials. NHL –Natural Hydraulic Lime.
2.3 Thermal transmittance (U-value measurement)
The thermal transmittance of the walls with and without insulation materials was measured in-situ in accordance with BS ISO 9869-1:2014, using the Hukseflux TRSYS01 measurement system and Loggernet software for in-situ measurement complying with ISO 9869 and ASTM C1155. For each measurement, two heat flux pads were affixed with two adjoining internal surface temperature thermocouples (Fig. 3 and 4) and a further two external surface temperature thermocouples mounted directly opposite on the exterior of the wall.
The U-value readings were taken sequentially for approximately three weeks over a six-month period. The interior of the building was heated and the temperature difference between the interior and exterior was maintained at an average of c.9°C. Thermal imaging of the wall was undertaken prior to the application of the sensors to select a representative section of wall and avoid any thermal bridges or wall anomalies (Fig 5).
Pictures, right: Fig 3. Thermal and moisture monitoring set-up; Fig 4. Insulated walls with monitoring set and Fig 5. Thermal image of a section of wall between the two windows showing homogeneous wall temperature and no anomalies.
The thermal conductivity and thermal transmittance (U-value) were calculated using the average method (BS ISO 9869-1:2014) according to equation 1. An estimate is computed after each measurement and converges towards a final value.
W/m
2K Equation 1
U
U-value (thermal transmittance) (W/m
2K)
T
si internal surface temperature(K)
T
se external surface temperature (K)
Q
heat flux (W/m
2)
R
int internal surface resistance (standard 0.13m
2K/W)
R
ext external surface resistance (standard 0.04m
2K/W)
The error in the U-value is calculated as per Baker (2011) [10], who incorporates the error in all the sensors/probes and in the standard deviation of the results.
2.4 Gravimetric analysis
Holes were drilled in the walls and the dust collected at a depth of 100-130mm. It was not possible to drill several holes owing to the historic importance of the building and only one sample for each insulation was collected. The wet weight and dry weight (following oven drying at 105°C for 24 hours) was used to calculate the moisture content in accordance with equation 2.
MC = ((wet weight – dry weight)/dry weight) x 100 per cent Equation 2
2.5 Monitoring of wall with temperature and RH with probes
The wall temperature and RH were monitored using Lascar EL-USB 2+ temperature and humidity probes inserted into holes drilled in the wall and sealed with tape. The hole depth was c.130mm, extending through the insulation (c.40mm) and levelling plaster (c.70mm) to the interior surface of the brick wall.
3 Results
3.1 U-values of the brick wall and the brick wall with insulation
The, in-situ-measured, U-value of the c.840mm brick wall (with the external render and the c.70mm internal plaster) is 1.32W/m
2K. This is in the range of former experimental values [3, 11, 12] however higher than expected given the wall thickness (the thick wall should have lowered U-values).
The in-situ-measured U-values of the brick wall with insulation are included in table 2. As it can be seen from this table, the insulation considerably reduces the U-value of the wall between 35 and 61 per cent (except for the thermal paint which shows a slight increase in U-value). The percentage improvement measured depends on the U-value of the original wall therefore, it is important that this is accurately established when considering thermal upgrades.
The aerogel and PIR provide the greatest thermal advantage, improving U-values by 61 per cent and 59 per cent respectively. The other insulations also considerably reduce U-values by 54.5 per cent (timber board), 45 per cent (cork lime), 36.9 per cent (hemp lime) and 34.1 per cent (calcium silicate board). The increase in U-value by the thermal paint is not statistically significant, therefore the paint has no measurable effect on the wall U-value.
The existing building regulations specify a U-value requirement for a wall of 0.21W/Km
2 and 0.35W/Km
2 (which can be relaxed to 0.60W/Km
2) for new and existing buildings respectively. Only three of the retrofitted walls achieve this target complying with the 0.60 W/m
2K value: timber board, PIR and aerogel. The minimal thickness of the insulation required to preserve historic structures is responsible for these non-conforming U-values.
A more relaxed, slightly higher U-value requirement could be applied to historic structures as a compromise to balance thermal performance requirements with heritage conservation. This would lessen the risk of moisture accumulation in the wall that can undermine the long-term durability of historic and traditional materials.
|
U- value |
U-value improvement compared to lime plaster only (%) |
Conditions ISO 9869-1:2014 |
Error in U-value measurement (%) |
| Test>72h |
24hr R variation % |
1st 2/3 R = last 2/3 R |
| LP |
1.321 |
0 |
Yes |
2.20 |
3.20% |
3.92 |
| P |
1.349 |
-2.12 |
Yes |
1.89 |
5.90%* |
5.00 |
| AG |
0.514 |
61.09 |
Yes |
0.45 |
1.08% |
7.97 |
| CL |
0.727 |
44.97 |
Yes |
0.90 |
1.90% |
4.69 |
| HL |
0.834 |
36.87 |
yes |
0.36 |
3.33% |
3.37 |
| CSB |
0.870 |
34.14 |
yes |
0.16 |
12%* |
4.30 |
| TB |
0.601 |
54.50 |
yes |
0.21 |
2.39% |
6.67 |
| PIR |
0.540 |
59.12 |
yes |
0.98 |
3.80% |
10.77 |
|
|
|
|
|
|
|
|
|
|
Table 2. U-values of the RHK wall with insulation measured in-situ. Error incorporates standard deviation in moving weekly averages and errors in equipment as set out in Baker (2011) [10]. Three conditions are stipulated in section 7.1 of ISO 9869-1:2014 to ensure testing is long enough to obtain accurate U-value readings. * This value is above the five per cent value set out in the standard.
3.2 Moisture content in the walls
The moisture content in the walls was measured 20 months after the application of the insulation. No internal heating was used for six months prior to this measurement and consequently the wall temperature is similar in all wall sections (table 3). Moisture was present in the wall on account of residual construction moisture (levelling lime plaster and the wet-applied insulation) and typical sources of ingress such as humidity.
The moisture content, measured by gravimetric analysis, is set out in table 3. Additionally, the temperature and RH at a wall depth of 130 mm (levelling plaster- insulation interface) were measured using a probe.
The authors previously demonstrated that a brick moisture level of c.1.8 per cent (measured gravimetrically) equated to an environmental RH of c. 96 per cent [13]. Even though the brick moisture content may vary depending on the brick density, this finding shows that water vapour saturated environments coexist with low moisture content in the brick wall. This is consistent with the results in table 3 that show low brick moisture contents (0.92-3.59 per cent) and yet relatively high relatively humidity (63.5 – 93 per cent).
|
% moisture content |
RH % (probe) |
Temp °C |
| LP |
1.13 |
67.5 |
15 |
| P |
1.36 |
67.5 |
15 |
| AG |
1.92 |
69 |
15 |
| CL |
0.95 |
65 |
15 |
| HL |
1.46 |
66 |
15 |
| CSB |
0.92 |
63.5 |
15.5 |
| TB |
- |
- |
- |
| PIR |
3.59 |
93 |
14.5 |
Table 3. Moisture content by gravimetric analysis, RH and temperature in the wall at 20months with no heating applied.
The brick walls retrofitted with CSB and CL show the lowest moisture contents which is attributed to the ability of these insulations to wick moisture and minimise its accumulation. Good moisture transfer is similarly responsible for the low wall moisture content for the HL insulation (the gravimetric result is higher than anticipated probably on account of a single experimental reading). The lime plaster has a higher moisture content than the lime-based insulation (CL and CSB) due to its lower moisture permeability.
Both the aerogel and thermal paint impede moisture transfer and consequently show higher levels of moisture in the wall. The PIR has a very high moisture content – virtually saturation. This is attributed to its impermeable nature preventing moisture from dissipating. This high RH value is above the critical moisture content acceptable in the wall and could result in long-term durability issues. Moisture values in the TB are not included in the results as an external system was mistaken applied.
3.3 Moisture fluctuations over time in insulated walls in response to changing environmental conditions
The temperature and RH probes, inserted into the wall at a depth of 130 mm (levelling plaster- insulation interface) were monitored for one month (March 2015), 12 months after the application of the insulation. During this month, the mean external temperature was 5.8°C and rainfall 57.5mm [14]. The average room temperature during monitoring was 18°C.
The internal wall temperatures measured using the probes are 16.3, 16.5, 15.3, 15.3, 16.1, 16.2, and 13.9°C for the LP, P, AG, CL, HL, CSB, PIR respectively. The RH measured in the wall depends on moisture conditions but also on wall temperature as RH increases with decreasing temperature. Therefore, insulation with low thermal conductivity should reduce the wall temperature to a greater extent resulting in higher RH.
This effect is apparent in the aforementioned temperatures whereby the LP and paint, with the lowest insulating values, have the highest wall temperature (16.3 and 16.5°C) and the PIR and aerogel which are better insulators with lower U-vales have low wall temperatures (13.9 and 15.3°C).
[caption id="attachment_27725" align="alignright" width="300"]
Fig 6. Relative humidity in the wall, over one-month period, 12 months after application of insulation[/caption]
Figure 6 shows the response of the walls to fluctuations in environmental moisture conditions over a month. The PIR has unacceptably high moisture levels which appear relatively stable and unaltered by changing environmental conditions: while the room RH fluctuated between c.45 and c.65 per cent, the PIR wall humidity remained high between 87 and 91 per cent. This is due to its impermeable nature preventing moisture from dissipating and the low internal wall temperature.
The RH levels in the lime based wall sections (LP, CL and HL) slightly undulate over time which is likely a reflection of fluctuating environmental conditions although changes are too small to accurately ascertain this. The lime plaster has a relatively lower average RH than the lime-based insulations (56.5 vs c.63 per cent at 12 months). This is attributed to the higher wall temperature of the lime plaster wall lowering RH.
The RH levels recorded in the wall with the paint and AG appear stable and are not influenced by fluctuating environmental conditions which can be attributed to the low vapour permeability of these materials.
The CSB shows the largest fluctuations in moisture content (from 55 to 65 per cent) which mirror changes in RH room conditions with a slight time lag. This can be explained by the high moisture transfer properties of the CSB and illustrates its ability to buffer internal environmental conditions.
The high moisture content of the CSB at 12 months compared with its value lower than the other insulations at 20 months is due to the presence of residual construction moisture in the wall at 12 months as this was the last insulation applied which had fully dried by 20 months.
Conclusion
This paper investigates the thermal and moisture performance of a range of insulation materials for use in historic buildings. The aerogel and PIR showed the best thermal performance reducing wall U-values by 61 and 59 per cent respectively. The thinness of the aerogel (only c.20mm) makes it particularly appropriate in historic buildings. However, the aerogel resulted in an increase in moisture levels in the wall and should be used cautiously in buildings where moisture accumulation is unlikely.
The PIR resulted in unacceptably high moisture levels in the wall which may undermine the integrity of the wall and consequently is not appropriate for use in this situation.
The other insulations also considerably reduce U-values by 54.5 per cent (timber board), 45 per cent (cork lime), 36.9 per cent (hemp lime) and 34.1 per cent (calcium silicate board). These insulations do not have a significant adverse affect on wall moisture (timber board not measured) on account of their breathable nature and provide a good compromise between energy saving and hygrothermal risk.
The results in this paper are specific to the building investigated (its composition, location, exposure, weather conditions) and the insulation materials used. A building should be carefully investigated and assessed when considering internal insulation options. Additional information about this research in other papers by the authors is included in reference list numbers 7, 8 and 13.
Acknowledgements
The authors wish to thank the Irish Research Council and the Office of Public Works for funding this research. The authors also thank the OPW, in particular John Cahill, for collaborating in the project and also Ben Fay, George Whelan and Colin McAlloram for the installation of the internal insulation in the Adjutant General’s Building. Laboratory work and site support was provided by the Department of Civil Engineering, Trinity College Dublin, and the authors thank chief technician Dr Kevin Ryan, Dr Michael Grimes and Eoin Dunne for their help.
References
[1] BPIE, Europe’s buildings under the microscope -A country-by-country review of the energy performance of buildings, Buildings Performance Institute Europe (BPIE), 2011.
[2] M.S. Al-Homoud, Performance characteristics and practical applications of common building thermal insulation materials, Building and Environment 40 (2005) 353–366.
[3] S. Rhee-Duverne , P. Baker, Research into the Thermal Performance Of Traditional Brick Walls, English Heritage Report, London, 2013.
[4] M. Neil, C. Rye, Responsible Retrofit of Traditional Buildings- A report on existing research and guidance with recommendations, Sustainable Traditional Buildings Alliance (STBA), 2012
[5] H.M. Künzel, A.H. Holm, Moisture Control and Problem Analysis of Heritage Constructions, IBP, Fraunhofer Institute, Stuttgart, 2009.
[6] A. Nielsen, E.B. Møller, T.V. Rasmussen, E.J. Hansen, Use of sensitivity analysis to evaluate hygrothermal conditions in solid brick walls with interior insulation, Proceedings of the 5th International Building Physics Conference (IBPC) 2012.
[7] R. Walker and S. Pavía, Thermal performance of a selection of insulation materials suitable for historic buildings, 94 (2015) 155–165.
[8] R. Walker and S. Pavía, Thermal and hygric properties of insulation materials suitable for historic fabrics. Proceedings of the 3rd international COINVEDI Construction and Building Research Conference, Madrid, Spain 2015.
[9] M. Stefanidou, M. Assael, K. Antoniadis, G. Matziaroglou , Thermal Conductivity of Building Materials Employed in the Preservation of Traditional Structures, International Journal of Thermophysics, 31 (2010) 844-851.
[10] P. Baker, U‐values and traditional buildings- In situ measurements and their comparisons to calculated values, Technical Paper 10, Historic Scotland Report, Scotland, 2011.
[11] C. Rye, C. Scott, The SPAB Research Report 1 – U-value Report, SPAB, London, Published 2010 and revised November 2012.
[12] Chartered Institution of Building Services Engineers. Guide A: Environmental design (7th edition). CIBSE, London; 2006.
[13] R. Walker, S. Pavía, M. Dalton, Measurement of moisture content in solid brick walls using timber dowel. Materials and Structures (2015) 1-13.
[14]
http://www.met.ie/climate/monthly-data.asp checked January 5, 2015.