Valor U Paredes de Piedra
Valor U Paredes de Piedra
Valor U Paredes de Piedra
Conservation Group
Technical Paper 10
U‐values and traditional buildings
In situ measurements and their
comparisons to calculated values
Dr. Paul Baker (Glasgow Caledonian University)
January 2011
Historic Scotland Technical Paper 10
The views expressed in the research report(s), presented in this
Historic Scotland Technical Paper, are those of the researchers,
and do not necessarily represent those of Historic Scotland.
This paper is available as web publication
on the Historic Scotland website:
www.historic‐scotland.gov.uk/technicalpapers
This paper should be quoted as
‘Historic Scotland Technical Paper 10’.
Acknowledgements
We like to thank Crichton Development Company Ltd, The Great Stewart of Scotland’s
Dumfries House Trust, Lister Housing Association, The National Trust for Scotland, Reidvale
Housing Association, and all the private building owners for the opportunities to carry out
measurements in their properties. We also like to thank the building occupiers who shared
their homes for two weeks with measurement sensors and data loggers.
We also like to thank Dr. Caroline Rye for her help in developing this research further.
Definitions
U‐value (or thermal transmittance co‐efficient) is a measure of how much heat will pass
through one square metre of a structure when the temperature on either side of the
structure differs by 1 degree Celsius. The lower the U‐value, the better is the thermal
performance of a structure. The U‐value is expressed in W/m2K.
Page I
Historic Scotland Technical Paper 10
U‐values and traditional buildings
In situ measurements and their comparison to calculated values
Contents
Introduction by Historic Scotland …………………………………………………………………………… III
Research report by Dr. Paul Baker, Glasgow Caledonian University ……………………………. VIII
Executive summary
Historic Scotland Technical Paper 10 provides the results of a thermal performance study of
traditional building elements. The study focused on U‐values as an indicator of thermal
performance, and involved their in situ measurements and the subsequent comparison with
U‐values calculated with software programs, and with often used ‘default’ U‐values. The
study was carried out from 2007 to 2010 by Dr. Paul Baker, Glasgow Caledonian University.
U‐values are normally calculated with computer programs developed with present‐day non‐
traditional construction in mind. Historic Scotland felt that the suitability of such programs
when used to assess traditional buildings needed investigation, and therefore appointed
GCU to carry out this study, the results of which will help construction professionals and
assessors of energy building performance to make better informed and more balanced
decisions when assessing and improving the energy performance of traditional buildings.
For the study, 67 in situ U‐value measurements were carried out, mostly of uninsulated solid
walls but, for comparison, some cavity walls, and building elements retrofitted with
insulation, were also measured. The non‐invasive measurements were generally taken of
building elements with their internal and external finishes intact.
The study then compared the U‐values measured in situ with their calculated equivalents.
A particular focus of the comparison was the impact of the lime‐and‐stone core of a
traditional solid stonewall.
The study found that software programs for U‐value calculations tend to overestimate
U‐values of traditional building elements: traditional building elements tend to perform
better thermally than would be expected from the U‐value calculations. The study suggests
that the in situ measurement of U‐values is a useful tool which can aid in the assessment of
the thermal performance of traditional building elements.
The study recommends further research on the thermal properties of traditional building
materials and construction components; improvement to U‐value calculations; and a
standardised methodology for in situ measurements of U‐values.
Page II
Historic Scotland Technical Paper 10
Introduction
Historic Scotland Technical Paper 10 provides the results of a thermal performance study of
traditionally constructed building elements. The study focused on U‐values as an indicator
of thermal performance, and involved their in situ measurements and the subsequent
comparison with U‐values calculated with software programs, and with often used ‘default’
U‐values. The study was carried out from 2007 to 2010 by Dr. Paul Baker, Centre for
Research on Indoor Climate & Health, Glasgow Caledonian University (GCU).
U‐values are generally used to describe the thermal performance of building elements, and
also form part of the base data used to assess the energy performance of whole buildings.
U‐values are normally calculated with computer programs developed with present‐day non‐
traditional construction in mind. Historic Scotland felt that the suitability of such programs
when used to assess traditional buildings needed investigation, and therefore appointed
GCU to carry out this study, the results of which will help construction professionals, and in
particular assessors of energy building performance, to make better informed and more
balanced decisions when assessing and improving the energy performance of traditional
buildings. The study also provides recommendations on where further research is required.
Some of the measurement results in this paper have already been presented in 2008 in
Historic Scotland Technical Paper 2. 1 Similar research has now also been carried out by
Dr. Paul Baker, on behalf of English Heritage, assessing brick walls, 2 and by Dr. Caroline Rye,
on behalf of The Society for the Protection of Ancient Buildings (SPAB), measuring a variety
of wall construction types. 3 Historic Scotland has also published research on the thermal
performance of windows. 4
U‐values as heat flow indicators
Protection from weather is a fundamental function of any building, and protection from
wind and cold is of particular importance in a climate like Scotland’s. Heat is lost (and
occasionally gained) through the building envelope, and heat flow (also referred to as heat
transfer or thermal transmittance) occurs, to different degrees, in any structure. Heat flow
can be measured, and subsequently expressed, as U‐value (or thermal transmittance
co‐efficient) being the heat flow through one square metre of a structure when the
temperature on either side of the structure differs by one degree Celsius. Therefore, the
U‐value is dependent on the thermal conductivities of the building materials and their
respective thicknesses.
U‐values are commonly used to describe the thermal performance of building elements, and
subsequently the overall energy performance of a building. Generally, U‐values are
calculated with readily available software programs rather than measured in situ. Such
calculations are normally carried out in order to show compliance with building standards
requirements 5 for new buildings, and for conversions of existing buildings, prior to
construction or conversion.
Page III
Historic Scotland Technical Paper 10
Page IV
Historic Scotland Technical Paper 10
In comparison, non‐traditional cavity wall construction is built with two distinctly separated
leaves (often made from brick and/or blockwork bedded in cement mortar). The cavity
between the masonry leaves is either left unfilled (and generally ventilated to the outside),
or filled with insulation. This form of construction can easily be thought of as a set of
separate vertical layers: outer masonry leaf, cavity (with/without insulation), and inner
masonry leaf (see Figure 1b). In this form of construction, often only one of the two leaves is
load‐bearing.
Figure 1a Figure 1b
Figure 1. Comparison of (a) traditional solid masonry wall construction and (b) non‐
traditional cavity wall construction
Key findings and recommendations
The GCU research provides the following key findings:
• The in situ measurement of U‐values is a useful tool which can aid in the assessment of
the thermal performance of traditional building elements, particularly where calculation
methods may suffer from deficiencies resulting from lack of knowledge of the actual
build‐ups used, and of the thermal properties of traditional materials.
• Indicative U‐values for 600 mm thick traditional stonewalls are as follows:
o Uninsulated walls finished with ‘plaster on laths’: 1.1 ±0.2 W/m2K
o Uninsulated walls drylined with plasterboard: 0.9 ±0.2 W/m2K
Page V
Historic Scotland Technical Paper 10
• Generally, an increased wall thickness, and building materials of higher thermal
resistance, results in a lower U‐value. However, careful consideration needs to be given
to establish the actual build‐up of the building element as defective areas, building
irregularities, ventilated cavities etc. can have a significant impact on the heat flux, at
least locally.
• Walls with internal finishes which incorporate an (unventilated) air‐filled cavity, such as
‘plaster on laths’, drylining or timber lining, have lower U‐values than walls of the same
thickness finished with ‘plaster on the hard’.
• Internal drylining and insulating of solid stonewalls can improve their thermal
performance significantly. 10 However, careful detailed and correct installation is
essential, and issues of vapour transfer need to be taken into account.
• Software programs for U‐value calculations tend to overestimate U‐values of traditional
building elements compared with the results from the in situ measurements. Traditional
building elements tend to perform thermally better than would be expected from the
U‐value calculations.
The GCU research provides the following key recommendations:
• Further research should be carried out to establish a better understanding of the
thermal properties of traditional building materials and construction components.
• Baseline databases of U‐value calculation programs should be extended to include more
data on traditional building materials, and allow for easier, and more user‐friendly,
modelling of traditional construction techniques, such as solid stonewalls.
• A standardised methodology for in situ measurements of U‐values should be established
to ensure that future measurement results are comparable.
1
Baker, P., 2008. In situ measurements in traditional buildings: preliminary results. (Historic Scotland
Technical Paper 2) Edinburgh: Historic Scotland.
Available at http://www.historic‐scotland.gov.uk/technicalpapers
2
The publication of Dr. Baker’s research report by English Heritage is expected in spring 2011.
3
Rye, C. (2010) The SPAB U‐value Report. London: The Society for the Protection of Ancient Buildings.
Available at www.spab.org.uk
4
Historic Scotland has also tested the thermal performance of windows, and window glazing, and the results
have been published as Historic Scotland Technical Papers 1 and 9:
Baker, P., 2008. Thermal performance of traditional windows. (Historic Scotland Technical Paper 1) Edinburgh:
Historic Scotland. Available at www.historic‐scotland.gov.uk/technicalpapers
Heath, N., Baker, P. & Menzies, G. (2010) Slim‐profile double glazing: thermal performance and embodied
energy. (Historic Scotland Technical Paper 9) Edinburgh: Historic Scotland.
Available at www.historic‐scotland.gov.uk/technicalpapers
5
Guidance on U‐value requirements for new buildings, and for conversions of existing buildings, is available
from the Building Standards Division of the Scottish Government, and from Historic Scotland:
Page VI
Historic Scotland Technical Paper 10
Scottish Government, 2010. Building standards technical handbooks. Edinburgh: Scottish Government.
Available at http://www.scotland.gov.uk/topics/built‐environment/building/building‐standards
Historic Scotland, 2007. Conversion of traditional buildings: application of the Scottish building standards.
(Guide for Practitioners 6) Edinburgh: Historic Scotland.
Available at http://www.historic‐scotland.gov.uk/conversionoftraditionalbuildings1and2.pdf
6
BuildDesk software by BuildDesk Ltd., Pencoed, Bridgend, CF35 6NY; more information about this program is
available at http://www.builddesk.co.uk
7
BRE U‐value Calculator software by Building Research Establishment Ltd., Bucknalls Lane, Garston, Watford,
WD25 9XX; more information about this program is available at http://projects.bre.co.uk/uvalues
8
Anderson, B., 2006. Thermal properties of building structures. In: The Chartered Institution of Building
Services Engineers, 2006. CIBSE guide A: environmental design. 7th ed. London: CIBSE. Ch.3.
9
Energy Saving Trust, 2004. Scotland: assessing U‐values of existing housing. (Energy Efficiency Best Practice in
Housing: CE84) London: Energy Saving Trust.
10
Historic Scotland is currently also researching options of internally retrofitted insulation which do not
require the removal of existing wall finishes, such as ‘plaster on laths’.
Page VII
Historic Scotland Technical Paper 10
Research report
U‐values and traditional buildings
In situ measurements and their
comparison to calculated values
A research report by
Dr. Paul Baker
Glasgow Caledonian University
January 2011
Page VIII
U‐values and traditional buildings
In situ measurements and their comparison to calculated values
Dr. Paul Baker
Centre for Research on Indoor Climate & Health
School of the Built & Natural Environment
Glasgow Caledonian University
January 2011
Page 1
Contents
1. Introduction 3
2. Building descriptions 4
2.1 Buildings 4
2.2 Building elements 6
3. Monitoring and analysis procedures 9
3.1 Principles 9
3.2 Procedures 9
4. U‐value calculations 10
4.1 Assumptions for the U‐value calculations 10
5. Results and discussion 13
5.1 Uncertainties 13
5.2 Graphic presentation of measurement and calculation results 15
5.3 Discussion of in situ measurement results 20
5.4 Laboratory measurement on a Locharbriggs sandstone wall 23
5.5 Comparison of results from in‐situ measurements and calculated U‐values 24
5.6 Comparison to often quoted used U‐values 27
6. Conclusions 30
I thank Dr. Caroline Rye for her help in developing this research further, Carsten Hermann of Historic
Scotland for fruitful technical discussions and his rigorous editing of this report, and Duncan
McCormick and David Nicol for their assistance in carrying out the in situ measurements of the
cavity walls as part of their undergraduate dissertations at GCU.
Page 2
1. Introduction
This report summarises the results of a thermal performance study of traditionally
constructed building elements. The study focused on U‐values as an indicator of thermal
performance, and involved the in situ measurements of U‐values and their subsequent
comparison with calculated U‐values 1 . The study was carried out from 2007 to 2010 by
Dr. Paul Baker, Centre for Research on Indoor Climate & Health, Glasgow Caledonian
University (GCU).
The main objective of the study was to assess the actual thermal performance of traditional
building elements in order to provide guidance for energy performance assessments. For
the study, 70 in situ U‐value measurements of walls, roofs and a floor were carried out at
15 properties over three heating seasons between November 2007 and March 2010. 2
The sample of properties used for the measurements mostly represents Scottish traditional
construction techniques (e.g. solid wall construction). However, for comparison, some
measurements of buildings with non‐traditional construction, namely cavity walls, have also
been included in this study.
Additional measurements were also made on a traditionally constructed sandstone wall in
an environmental chamber at GCU, with and without a drylining.
This report also presents the measurement and analysis procedures used to determine the
in situ U‐values. Measurements were made of the heat flow directly through the building
element using heat flux sensors mounted on internal surfaces, and of room and outdoor
temperatures. Most measurements were taken in occupied properties, and therefore
included external and internal wall finishes (in some occasions also including cavities behind
such finishes).
The U‐values, measured for the study, were subsequently compared with calculated
U‐values using two standard software programs, BuildDesk and BRE U‐value Calculator, to
assess the applicability of such programs when used for assessing traditional building
construction. Both programs are commonly used as assessment tools for new buildings and
conversions of existing building, to ensure compliance with the U‐value requirements of
building standards.
1
Where the expressions ‘calculated U‐value’ and ‘U‐value calculation’ are used in this report, they refer to
calculating U‐values ‐with software programs‐ using standardised assumptions for material characteristics.
Results from actual in situ heat flux measurements are normally not used in such calculations. These
calculations are, generally, carried out in accordance with the calculation methods set out in British Standard
BS EN ISO 6946:1997 and the BRE publication Convention for U‐value calculations (Anderson, 2006a). The
analysis of the in situ heat flux measurements for this study also require some ‘calculation’ to convert the
measured heat flux results into U‐values. However, the term ‘calculation’ has been avoided in this context in
order to not confuse such conversions with the U‐value calculation carried out with the standard software
programs.
2
Some of these measurement results have already been published in Historic Scotland Technical Paper 2
(Baker, 2008).
Page 3
Furthermore, the measured U‐values were also compared with often used default U‐values
and with the requirements of current building standards for new buildings, and for
conversions of existing buildings.
This report firstly describes the measured buildings and building elements. It then outlines
the procedures for in situ heat flux measurements and their analysis and conversion into
U‐values. These procedures are described in more detail in Appendix A. The report
continues with listing the assumptions used for the U‐value calculations carried out with
software programs. The report closes with a discussion of the results of the in situ
measurements, and a comparison with calculated U‐value, and to the U‐value requirements
of the building standards. Specific details of the buildings, building elements, measurement
locations, and measurement results are presented in the form of datasheets in Appendix B.
2. Building descriptions
In situ U‐value measurements were made of different building elements in a variety of
buildings. The main focus of the measurements was on walls constructed with a range of
materials and techniques. Some
measurements were also made of roofs and
floors. The building elements were
measured including existing surface
finishes.
2.1 Buildings
15 properties throughout Scotland were
visited for measurement (Figure 1). 3 The
majority of building measured for this study
were constructed pre‐1919 with traditional
construction techniques.
A brief description of the buildings used for
measurements is given in Table 1.
More detailed descriptions, together with
measurement results, are presented in
form of Building Datasheet in Appendix B.
A sample of such a building datasheet is
shown in Figure 2. Figure 1. Location map
3
Some of the properties actually consisted of more than one building. The Building Datasheets in Appendix B
give details on how many different buildings were measured using the same measurement ID number.
Page 4
Table 1. Brief description of the buildings used in the study
Page 5
Figure 2. Example of building datasheet with description and measurements results
2.2 Building elements
The building elements measured for this study were external walls, roofs (ceilings to roof
spaces and ceiling coombs) and floors (in contact with soil). The main focus of the
measurements was on the walls, and the results for the other building elements are
reported in this study without much further analysis and discussion.
Walls
The majority of the walls used for measurements were built pre‐1919 and were therefore of
solid wall construction, the traditional form of building construction in Scotland, typically
constructed from stone or brick, bedded in lime mortar.
The original walls of Logie Schoolhouse (ID5) were of earth/mud construction. However,
some of the walls measured were of mixed construction, e.g. some of the older repairs to
the mud walls of Logie Schoolhouse were carried out using bricks and stone.
Of the solid walls measured, most were made from stone (of differing stone types) bedded
in mortar (generally lime mortar). Such stonewalls are not homogenous in their build‐up but
consist of inner and outer stone leaves with the centre of the wall being packed with
smaller stones and mortar (Figure 3).
Page 6
Figure 3: Solid wall construction. The
example is of a test wall built by Historic
Scotland’s masons for GCU from
Locharbriggs sandstone (see Laboratory
measurement in section 5). The outer
leaf (right) has an ashlar finish, whilst
the inner leaf is a rubble construction. It
has been estimated that the proportion
of mortar in the wall is about 30% by
volume.
For comparison with solid wall construction, four more recent buildings with cavity wall
constructions, typical of post‐1919 construction techniques, have also been included in the
study. Two of these buildings were semi‐detached 1930s houses (ID14), and the other two
were detached houses constructed in the 1970s (ID15). The cavity walls were constructed
from brick or concrete blocks, and one of each house pairs had cavity wall insulation
installed subsequently.
Measurements of the walls (and also of the roofs) were taken in situ, i.e. of existing building
elements. The measurements, therefore, included any existing surface finishes, externally
and internally. Sometimes the internal finishes could also include cavities.
Generally, four types of internal wall finishes were included in the measurements:
• plastered on the hard
• plaster on laths
• drylining
• timber lining
Except for the plaster on the hard finish, all finishes were fixed to battens / studs (generally
made from timber) which were then fixed back to the wall faces. The depth of the
battens / studs could vary. Typical sizes are 25 mm for plaster on laths, and 50 mm for
drylining.
Page 7
Where battens / studs were used as fixing medium for the wall finish, a cavity was generally
formed between the battens / studs. This cavity was uninsulated unless otherwise stated
and therefore air‐filled.
A few of the measured buildings also had external finishes, such as harling / render which
were included in the measurements.
Figure 4 shows the distribution of measurement samples by construction types and surface
finishes.
Figure 4. Distribution of measurement samples by construction type and surface finish
Roofs
U‐value measurements of some ceilings (to uninsulated roof spaces) and some ceiling
coombs were made where the opportunity arose. The roofs and ceilings were generally
constructed with timber joists. All roofs were slated, and finished internally with plaster on
laths, plasterboard or timber lining (similar as described above for walls). The roofs and
ceilings were generally uninsulated unless otherwise stated.
Floors
The U‐value of a basement floor (in the Georgian Tenement, ID01) was also measured to
validate the performance of an insulation product used to improve the thermal
Page 8
performance in basement flats in the building as part of a refurbishment project
(Changeworks, 2008). The flats had had their original solid floors removed in the 1970s as
part of a previous refurbishment and replaced with concrete laid over aggregate. The in situ
U‐value of the floor was measured before and after the addition of a sample of insulation.
3. Monitoring and analysis procedures
3.1 Principles
The monitoring and analysis procedures have been developed during the first phase of the
project (Baker, 2008) and similar projects with other organisations. The procedures are
based on the principles of prEN 12494:1996 which are summarised below.
The U‐value, or thermal transmittance, of a building element is defined in
BS EN ISO 7345:1987 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.”
In the laboratory suitable steady state conditions can be achieved to determine the U‐value
of a building element for standardised boundary conditions. However, during in situ
measurements, the boundary conditions (temperature, wind velocity and solar radiation)
change with time. It is therefore recommended that the surface‐to‐surface thermal
resistance of the element is obtained by measuring the heat flow rate through the element
and the surface temperatures on both sides of the element for a sufficiently long period of
time to give a good estimate of the steady state from the mean values of the heat flow rate
and temperatures. The U‐value can then be calculated by applying standardised surface
heat transfer coefficients. This averaging approach is valid if the following conditions apply:
• the thermal properties of the materials in the element are constant over the range
of temperature fluctuations;
• the change in the internal energy of the element is negligible if compared to the
amount of heat going through the element.
An alternative is to use a dynamic method to account for the fluctuations in the heat flow
and temperature in the recorded data.
It is assumed that the element is sufficiently homogeneous or made of sufficiently
homogeneous layers to use a heat flow meter.
3.2 Procedures
The test and analysis procedures are summarised as follows and explained in greater detail
in Appendix A.
Page 9
Actual measurements, recorded using one or more data loggers, were made over a period
of at least two weeks of the heat flow through the internal surface of each wall, and of the
internal and external temperatures. The measurement period was found to give a stable
average U‐value (Baker, 2008) which takes into account the thermal inertia of the wall.
Sensor locations were chosen to avoid probable thermal bridge locations near to windows
and corners, with the heat flow sensor ideally located about half‐way between window and
corner, and floor and ceiling. Where possible a North‐facing or sheltered elevation was
selected to reduce the influence of solar radiation on the wall. If possible, both external air
and surface temperatures were measured.
4. U‐value calculations
The U‐values of building elements are estimated as part of any new‐build construction
project by using software programs to show compliance with the U‐value requirements of
building standards prior to start of construction. In this study the objective was to assess the
suitability of such software to estimate the U‐values of traditional construction build‐ups.
Such calculations may be used as part of a thermal performance assessment of a traditional
building, for example, to aid in the choice of refurbishment options.
BuildDesk U 3.4 4 and BRE U‐value Calculator 5 software programs were used to calculate the
U‐values of the building elements measured for this study. The software programs calculate
U‐values in accordance with the Convention for U‐value calculations (Anderson 2006a)
published by the Building Research Establishment (BRE).
4.1 Assumptions for the U‐value calculations
The main assumptions made in order to model the build‐ups used for the calculations,
allowing for the restrictions in the program’s database, are outlined below.
Masonry
The BuildDesk database provides only two options for stone types (sandstone and granite)
which were used for the calculations. The software has the ability to calculate the effect of
mortar joints in brick and block constructions, in accordance with Anderson (2006a), using
the joint thickness and brick or block dimensions.
4
BuildDesk software by BuildDesk Ltd., Pencoed, Bridgend, CF35 6NY; more information about this program is
available online at http://www.builddesk.co.uk
5
BRE U‐value Calculator software by Building Research Establishment Ltd., Bucknalls Lane, Garston, Watford,
WD25 9XX; more information about this software is available online at http://projects.bre.co.uk/uvalues
Page 10
A rubble wall is somewhat different, since the wall is not a uniform construction with
regular mortar joints. If the proportions of the constituents of the wall (stone, mortar and
voids) are known or assumed from prior knowledge, the wall may be simply modelled as a
multi‐layer build‐up. For example, Figure 5 represents a rubble wall with 60% stone and
40% mortar which can be modelled as two layers representing the correct proportions of
the materials.
However, it is likely that the proportions of stone, mortar and voids are unknown, without
intervention, which is the case for the buildings in this study. The sandstone test wall
(Figure 3) described in Baker et al. (2007) was inspected and found to consist of about
70% sandstone and 30% mortar.
For calculation purposes a lower stone/mortar ratio of 60/40 was considered to be realistic
with an upper limit of 100% stone assumed for the worst case. Figure 6 shows the effect of
various assumed stone/mortar ratios on the calculated U‐value of a sandstone wall of an
overall thickness of 600 mm including a 25 mm plaster on the hard finish. The difference
between the U‐value calculated assuming 40% mortar and that assuming only sandstone
is 30%.
Earth / mud walls
At Logie Schoolhouse (ID05) solid walls from earth / mud were measured.
The East Anglia Earth Buildings Group suggests a thermal conductivity of 0.6 to 0.8 W/mK
for this material. 6
Plaster on laths
Plaster on laths was modelled as 25 mm layer of lime plaster with two widths of air cavity:
7 mm or 25 mm.
Timber linings
Timber linings were assumed to be 12.5 mm thick and modelled with two widths of air
cavity behind, 7 mm and 25 mm, assuming that studs were used to fit the linings to the
surface of the stonewall.
Multi‐foil insulations
At the Colonies Flat (ID4), a multi‐foil insulation had been used in an earlier roof
refurbishment. The thermal resistance, or R‐value, of the layer was assumed to be
1.9 m2K/W as stated in a typical manufacturer’s datasheet which is equivalent to a thermal
conductivity of 0.034 W/mK. 7
6
East Anglia Earth Buildings Group. Tech spec: clay lump wattle and daub. Available online at
www.eartha.org.uk/Eartha‐TechSpec.pdf
7
Kontrol R1.9 multi‐foil insulation by Kontrol Building Insulation; the manufacturer’s information sheet is
available online at www.kontrol‐insulation.com/uploads/resources/Kontrol_R19_4pp.pdf
Page 11
Figure 5. Traditional solid wall construction: the left figure (a) shows a schematic diagram of
a rubble wall construction with 40% mortar; and the right figure (b) shows its
representation as two layers for modelling.
Figure 6. Influence on calculated U‐value of the assumed proportion of lime mortar in a
600 mm sandstone wall with an internal finish of 25 mm lime plaster on the hard
Page 12
Insulating board
At the Georgian tenement (ID1), the existing concrete floor was measured with an insulting
board applied to its top face. The board was Spacetherm C board which is a 36 mm thick
laminated composite board consisting of a 9 mm insulating Spacetherm Blanket, a fibrous
matting impregnated with silica aerogel, and 21 mm MDF, a particle board made from
wood. The thermal resistance of the board was assumed as 0.13 W/mK. 8
Thatch
At Beaton’s Croft House (ID6), the measured roof was finished externally with thatch.
CIBSE Guide A, in chapter 3 (Anderson 2006b), gives a value of 0.09 W/mK for the thermal
conductivity of water reed thatch.
5. Results and discussion
The results of the in situ U‐value measurements and the comparable calculated values are
summarised in Table 2 and, in more details, in the table on each Building Datasheets in
Appendix B.
The tables state the locations of the in situ measurements; the overall thickness and build‐
up of the building element; the results of in situ U‐value measurements; and also the
comparable calculated U‐values which is expressed as a range of values where there is an
uncertainty regarding the build‐up, particularly for the ratio of mortar to stone in a
traditional solid stonewall constructions (see Masonry in section 4).
The tables on the Building Datasheets provide more details about the build‐up of the
measured building element, and also the measured inside and outdoor temperatures.
Some of the results have increased uncertainties which are described below, followed by
the discussion of the measured in situ results, and by comparisons with the calculated
U‐values, and with often quoted default U‐value. The findings of the study are then
summarised in section 6, Conclusions.
5.1 Uncertainties
Two reasons are noted below which resulted in increased uncertainties of some of the
in situ measurements: too small a differential between the inside and outdoor
temperatures; and unknown factors in the build‐up of a building element.
8
Spacetherm board by A. Proctor Group Ltd.; product information is available online at
http://www.spacetherm.com/pdf/apg5738%20spacetherm%20v5.pdf
http://www.spacetherm.com/pdf/Lister%20Housing.pdf
http://www.proctorgroup.com/index.asp?lm=523
Page 13
Uncertainty due to small temperature differential
The following results have increased uncertainty due to the small temperature difference
between the room and exterior:
• Georgian Tenement (ID01), uninsulated floor slab
(However, for the measurement with insulation, a daily heating cycle was applied which
gave sufficient variation in temperatures to produce a significant result.)
• Victorian Tenement (ID02), stairwell walls
(The temperatures in the close / stairwell were similar to the room temperatures.)
• Logie Schoolhouse (ID05)
(The building had been recently refurbished and was awaiting occupation by the new
tenant, and was therefore not fully heated.)
Uncertainty due to unknown factors in build‐up of building element
The results for McCowan House (ID13) are also uncertain since an existing ventilation
system in the walls has some influence on the U‐value estimates, particularly on the top
floor of the building. It is thought that warm air from rooms may enter the ventilation
system through grills in the internal walls and circulate behind the internal leaf of the wall,
before exiting via external grills. External air may also enter the system via the external grills
(Figure 7). There is also some evidence that wind speed and direction, measured at a local
Figure 7. McCowan House (ID13) with
several external ventilation grilles
visible at high level either side of the
windows.
Page 14
weather station, may influence this behaviour. Using the analysis method described in
Section 3 and Appendix 1 does not produce a stable value, for example if the cumulative
average U‐value is plotted daily for measurement location ID13.6 (heat flux sensor on
North‐facing 2nd floor wall) the U‐value does not approach an asymptote after a reasonable
test duration, ≥14 days (Figure 8).
Figure 8. Cumulative average U‐value over a 36‐day period for McCowan House,
measurement ID13.6 (North elevation, 2nd floor). Compare this graph to Figure A8 in
Appendix A.
5.2 Graphic presentation of measurement and calculation results
The results of the in situ wall measurements are presented below in Table 2. The results for
the wall measurements are also presented graphically in Figures 9 to 12 sorted by internal
wall finishes. The results of the measured roofs are presented in Figure 13.
The calculated U‐values are included in these figures, for comparison, with upper and lower
estimates, for the stonewalls, representing stone/mortar ratios of 100/0 and 60/40
respectively.
Page 15
Table 2. Summary of results of in situ U‐value measurements and calculated values
Page 16
Table 2. continued…
Page 17
Figure 9. Results for walls finished with plaster on the hard
Figure 10. Results for walls finished with plaster on laths
Page 18
Figure 11. Results for walls finished with drylining
Figure 12. Results for walls finished with timber lining
Page 19
Figure 13. Results for ceilings (to uninsulated roof spaces) and to coombs
5.3 Discussion of in situ measurement results
The in situ measurements for this study show a wide range of U‐value results. Given the
small size of samples, the results should be treated cautiously. However, in the following
general observation apply to the results.
Thermal impact of wall thickness
Stonewalls of same build‐up, but different thicknesses, were measured at the Georgian
tenement (ID1). The measurements of the 600 mm thick walls resulted in U‐values ranging
from 0.8 to 1.6 W/m2K; whereas the measurement of a 300 mm thick wall (in a wall press)
resulted in a U‐value of 2.3 W/m2K. This indicates that increasing wall thickness improves
the thermal resistance, and results in a lower U‐value.
The same conclusion can be drawn from Figure 12 showing the measurement results for
stonewalls finished with timber lining. The measurements of 300 mm thick stonewalls
resulted in U‐values ranging from 1.1 to 1.5 W/m2K (IDs 3.4, 12.10 and 12.14); whereas the
measurements of 760 and 1200 mm thick stonewalls resulted in U‐values ranging between
0.6 and 0.9 W/m2K (IDs 10.1, 6.1 and 6.2). This also demonstrates the beneficial impact of
an increased wall thickness on the thermal performance of the building element.
Page 20
Thermal impact of wall finishes with air‐filled cavities
At the Georgian tenement (ID1), measurements were carried out of solid stonewalls of
similar build‐up and thickness (600 mm) but with different internal wall finishes. The
measurement of the wall finished with plaster on the hard resulted in a U‐value of
1.6 W/m2K; the measurement of the wall finished with plaster on laths was 1.4 W/m2K; and
the measurements of two walls finished with drylining (plasterboard) were 0.8 and
1.0 W/m2K. This demonstrates the insulating effect of an air cavity behind a plaster on laths
and a drylined wall finish.
Similarly the measurement results for the Garden Bothy of the Dumfries House Estate
(ID12) show a range of values depending on internal surface finish. Here 600 mm thick solid
stonewalls finished with plaster on laths, in nine locations, and drylining in one
measurement location resulted in U‐values ranging between 0.9 and 1.3 W/m2K; whereas
the measurement of the same wall without any finishes, i.e. with a bare wall face, resulted
in a U‐value of 2.4 W/m2K. In two locations, a thinner wall section (300 mm) was measured
which included a timber‐lined finish: this resulted in U‐values of 1.1 and 1.2 W/m2K which is
similar to the results for walls finished with plaster on laths with greater thickness
(600 mm), possibly due to the higher thermal resistance of timber compared to lime plaster.
Other walls finished internally with timber lining also produced results in the lower range of
U‐values found in the study: Balmacara Estate office (ID10) with 600 to 750 mm thick solid
stonewall 0.9 W/m2K; Beaton’s Croft House (ID6) with 1200 mm thick stonewalls
0.6‐0.8 W/m2K; and Logie School House (ID5) with 600 mm mud walls an average value of
0.6 W/m2K. In these cases either the benefits of greater wall thickness (ID6 and ID10) or
lower material thermal conductivity (ID5) combine with those of the timber lining.
Indicative U‐values for stonewalls finished with plaster on laths or drylining
Given the range of wall materials, thicknesses and finishes measured, the two categories of
wall which showed some general consistency were 600 mm thick traditional stonewalls
which were (a) finished with plaster on laths (excluding the McCowan House results, ID13),
and (b) drylined with plasterboard (without insulation). These two categories resulted in
U‐values as follows.
• Walls finished with plaster on laths: 1.1 ±0.2 W/m2K
• Walls finished with drylining: 0.9 ±0.2 W/m2K
Thermal impact of cavity wall insulation
Cavity walls with and without insulation were measured at the pairs of 1930s semi‐
detached houses (ID14), and of 1970s detached houses (ID15). In both pairs, one building
still had the cavity unfilled, whereas in the other building the cavity had been retrofitted
with blown‐in cavity fill insulation. For ID14, the measurements of the retrofitted wall
resulted in a U‐value of 0.3 W/m2K (compared to 1.3 W/m2K when uninsulated); and for
Page 21
ID15, the U‐value of the retrofitted wall was 0.6 W/m2K (compared to 1.1 W/m2K). This is
equivalent to U‐value reductions by 76% and 45% respectively.
Thermal impact of internal drylining and insulating of walls
A solid stonewall drylined internally with insulation was measured at the Balmacara Square
steading conversion (ID11). The walls were 600 mm thick, and had, when the conversion
was being carried out, been drylined with plasterboard and insulation fitted between the
studs. The measurement of the insulated wall resulted in a U‐value of 0.3 W/m2K.
Unfortunately, there was no opportunity to measure the same wall build‐up either in its
bare state or without insulation. But the result could be compared to the above‐stated
indicative U‐values of 0.9 and 1.1 W/m2K (for walls with plaster on laths and drylining
respectively).
Thermal impact of walling material
At Logie Schoolhouse (ID5), measurements were taken of 600 mm thick solid walls
constructed from earth/mud. (However, only one measurement, ID5.1, was actually of a
pure earth/mud construction; in all other measurement locations, the external wall face
had been repaired with either brick or stone.) The wall measurements, which included
internal timber‐lined finishes, resulted in U‐values ranging between 0.4 and 0.8 W/m2K. This
is significantly better than solid walls of similar thickness but constructed from stone
(e.g. Georgian tenement (ID1) with U‐values between 0.8 to 1.6 W/m2K; Victorian tenement
(ID2) with U‐values of 0.9 and 1.0 W/m2K; or the Garden Bothy of the Dumfries House
Estate (ID12) with U‐values between 0.9 and 1.3 W/m2K). However, none of these
comparative values were of wall with a timber‐lined wall finish, but all were of wall finished
internally either with drylining or plaster on laths. Nonetheless, all three types of internal
finishes incorporate air‐filled cavities, although of potentially different thicknesses.
The very low U‐value results for Logie Schoolhouse indicate the higher thermal resistance of
a (at least partial) earth/mud wall compared to walls constructed from other walling
materials. This obviously relates to the higher thermal resistance of 0.6 to 0.8 W/mK for
earth/mud (see Earth/mud walls in section 4) compared to 1.8 W/mK for sandstone
(Anderson (2006a), table 3.47).
Results of roof measurements
The measurement results of roofs, i.e. ceilings to uninsulated attic spaces and ceiling
coombs, show a range of U‐values depending on the construction type and build‐up of the
building element.
The results in the refurbished loft in the Colony Flat (ID4) show a significant difference
between the two measurement locations: 0.4 and 1.1 W/m2K. (A thermal imaging survey is
recommended to assess the integrity of the insulation in the coomb of the loft.)
Page 22
The results of the measured roof at the Balmacara Estate office (ID10) show the
improvement of retrofitted insulation can make: the uninsulated ceiling coomb was
measured with a U‐value of 1.2 W/m2K; whereas the U‐value for the insulated ceiling to the
attic space was 0.8 W/m2K. However, the attic space, as a thermal buffer zone, will also
have contributed to the improved U‐value.
The measurement of the recently insulated roof at the Balmacara Square steading
conversion (ID11) resulted in an even better U‐value of 0.3 W/m2K (compared to ID10).
Results of floor measurements
Insulating the concrete floor in the basement of the Georgian tenement (ID1) with an
insulating board (see Insulating board in section 4) resulted in a U‐value of 0.6 W/m2K
compared to 3.5 W/m2K for the uninsulated bare concrete floor. This is equivalent to
U‐value reductions by 83%.
5.4 Laboratory measurement on a Locharbriggs sandstone wall
In addition to the measurements within buildings, two situ U‐value measurements were
made on a Locharbriggs sandstone wall constructed within an environmental test chamber
at GCU. The wall thickness was 550 mm and has an ashlar exterior, and a rubble interior,
face (Figure 14). A heat flux sensor was mounted in the centre of the interior wall face.
Figure 14. Locharbriggs sandstone wall in test chamber. The left photo shows the internal
rubble face; the right photo the external ashlar face.
Page 23
Temperatures of 23°C on the warm side and 8°C on the cold side of the wall were
maintained. The U‐value was determined from 10 days data which are sufficient under
steady conditions.
Following the test on the solid wall, timber studs were fixed to the sides of the wall and a
sheet of plasterboard added. The cavity formed was sealed off. A second heat flux sensor
was mounted on the plasterboard. The U‐value of the wall was re‐measured with the
plasterboard finish.
The results of the wall measurements were as follows:
• bare 550 mm thick sandstone wall 1.4 ±0.1 W/m2K
• same wall finished with drylining 1.1 ±0.1 W/m2K
Whilst the result with the drylining is in general agreement with the in situ results for the
drylined walls, the U‐value for the unfinished wall is significantly lower than the value for
the Garden Bothy’s (ID12) unfinished wall of 2.4 W/m2K. This may be due to the damp
condition of the Garden Bothy walls. The heat and moisture transport simulation software
program WUFI v2.0 (Künzel et al., 1997) gives dry and wet thermal conductivity values for
nine German sandstones: the average values are 1.9 W/mK for dry material and 3.0 W/mK
for wet sandstone. If the Garden Bothy’s walls would have been saturated it would have
result in a calculated U‐value of about 2.4 W/m2K for 70% stone / 30% mortar compared to
1.5 W/m2K for a dry wall.
5.5 Comparison of results from in‐situ measurements and calculated U‐values
In the following, the U‐values measured in situ are compared to their calculated equivalents
produced by using standard U‐value calculation software programs. First the comparison of
wall measurements is discussed, then the roof measurements, and finally the results of the
floor measurements.
Measured and calculated U‐values of walls
Figure 15 summarises the comparison between measured and calculated U‐values for the
walls (including both solid and cavity walls), and shows the number of measured U‐values
(including the uncertainty) which are (a) lower than the calculated value or range,
(b) measurements within the calculated range, and (c) measurements higher than the
calculated range.
Of the total number of wall measurements (57), 25 (44%) are lower than the calculated
U‐value range, 24 (42%) are within the calculated range, and 8 (14%) are higher than
calculated. By category, the wall plastered on the hard and with plaster on laths finishes
show a high proportion of results in agreement with the calculated values; whereas for
drylined and timber‐lined walls, the highest proportion of results is lower than calculated.
Page 24
Figure 15. Comparison between measured and calculated results by category and total
number of measurements. A range of calculated U‐values was assumed for stonewalls
reflecting the possible range of stone/mortar ratios. The percentages refer to the values
within each category.
Considering solid stonewall constructions only, if only the upper calculated value is taken, as
may be the case if the solid stone part of a wall build‐up is modelled solely as “sandstone”
without considering the proportion of mortar in the wall, then only two of the
measurements (5%) are within range of the predicted U‐value compared with 18 (45%)
considering the range of possible calculated results for different stone/mortar ratios
(Figure 16).
The comparisons between calculated and measured U‐values show the uncertainty in
modelling the build‐ups of traditional building elements.
Figure 16 indicates that if the build‐up for solid stonewall is simply calculated without
considering the proportion of mortar in the construction, this will result in an
overestimation of the U‐value. Some improvement in matching measured and calculated
values is achieved if the solid stone portion of the wall is modelled as a mixture of stone and
mortar. However, this approach is unsatisfactory for more than 50% of the in situ
measurements.
Page 25
Figure 16. Comparison between measured and calculated results by category and total
number of measurements for solid stonewall constructions only. ”Upper limit” categories
represent the calculated U‐values for stonewalls modelled as internal finish and stone
without allowance for mortar. ”Range” categories represent the possible range of
calculated values for stone/mortar ratios between 60/40 to 100/0.
In order to better model the wall build‐ups, more information about the construction
details are required, for example as follows:
• Thickness of layers
• Status of cavities (such as insulated/uninsulated and ventilated/unventilated)
• Ratio and types of stone and mortar (and voids) in solid stonewalls
• Thermal properties of materials used in traditional construction (such as local stone
types, historic brick, earth/mud walls, traditional plasters and mortars)
It is recommended that further research is carried out to establish more accurate U‐values
for the above listed materials / construction components to allow more accurate thermal
assessment of traditional building elements.
Page 26
Measured and calculated U‐values of roofs
The comparison of measured and calculated U‐values present a mixed picture. Ceilings and
coombs (Figure 16) showed a high proportion of calculated U‐values indicating that the
build‐up should perform better than the in situ result. Ceilings and coombs were perhaps
the most difficult build‐ups to model, since it was not easy to identify the actual
construction details on site, particularly for combed ceilings, where only the internal lining
and the outer roof material were observable.
On the other hand, the in situ U‐values of the thatched roof at Beaton’s Croft House (ID6)
are significantly higher than the estimate by the U‐value calculation: 1.1 to 1.5 W/m2K
compared to 0.3 to 0.4 W/m2K. Again the actual construction details are unknown, for
example, the space behind the internal timber lining may be ventilated which would
increase the U‐value. However, because of the small samples size (one roof only) the results
should not be taken as indicative. It is recommended that further measurements are carried
out in other thatched buildings.
Measured and calculated U‐values of floors
The measured and calculated U‐values of floors are in very good agreement. U‐values were
measured as 3.5 W/m2K uninsulated and 0.6 W/m2K insulated, compared to calculated
U‐values of 3.3 W/m2K and 0.5 W/m2K respectively. However, this should not be taken as
indicative as the sample size was small (one floor only), and the floor was of non‐traditional
construction (concrete slab), for which accurate U‐value calculations could be expected.
5.6 Comparison to often quoted used U‐values
In the absence of in situ U‐value measurements of traditionally constructed elements,
U‐values are often estimated or calculated using less appropriate assumptions and
simplifications.
In the following the U‐values of the in situ measurements for this study are compared to
often used default U‐values. For this comparison U‐values have been used from the
following two publications: CIBSE Guide A: Environmental Design (Anderson, 2006a,
chapter 3); and Scotland: Assessing U‐values for Existing Housing (Energy Saving Trust,
2004).
The measured values of retrofitted insulation are also compared to current U‐value
requirements by building standards (Scottish Government, 2010).
Page 27
Comparison to default U‐values for walls
The Energy Saving Trust (EST) publication suggests the use of the following default U‐values
(Energy Saving Trust, 2004, table 2):
• 1.7 W/m2K for traditional sandstone (or granite) dwelling with solid walls: stone
thickness typically 600 mm with internal lath and plaster finish
(for the pre‐1919 period)
• 1.7 W/m2K for cavity walls involving brick and block with external render
(for 1919‐1975)
• 0.3 W/m2K for brick/block cavity walls with insulation (for 2003‐present)
The CIBSE Guide suggests the use of U‐values as follows (Anderson, 2006a, tables 3.49 and
3.50):
• 1.38 W/m2K for a 600 mm stonewall with a 50 mm airspace and finished with
25 mm dense plaster on laths
• 2.09 W/m2K for a 220 mm solid brick wall with 13 mm dense plaster
• 1.41 W/m2K for a 220 mm solid brick wall with 50 mm airspace/battens and
12.5 mm plasterboard
• 1.44 for a brick/brick cavity wall with 105 mm brick, 50 mm airspace, 105 mm brick,
and 13 mm dense plaster
The in situ measurements of walls for this study indicate, as detailed below, that existing
buildings can perform thermally better than the above default U‐values suggested by EST
and CIBSE.
• For the case of 600 mm thick solid stonewalls with plaster on laths, the results of
this study indicate a typically performance of 1.1 ±0.2 W/m2K compared to above
default U‐values of 1.38 and 1.7 W/m2K.
• For the case of uninsulated cavity walls, the measurements at the 1930s semi‐
detached house (ID14), and the 1970s detached house (ID15), resulted in U‐values
of 1.3 and 1.1 W/m2K respectively, compared to above default U‐values of 1.44 and
1.7 W/m2K. For the case of cavity walls retrofitted with insulation, the U‐value for
ID14 (retrofitted after 2003) was 0.3 W/m2K which matches the default U‐value
suggested by EST.
However, the measurement results for this study have also shown that U‐values can vary
significantly from building to building, and also within one building itself. The transfer of the
U‐value results from this study to other building elements should only be made with
caution. Where the thermal performance of a building element is expected to be better
than the default U‐value for that element, in situ measurements should be considered to
confirm the actual U‐value of the construction. Such in situ measurements should generally
be carried out in a number of locations to minimise the risk of measuring in a location with
an untypical wall build‐up (e.g. a throughstone in a stonewall). A thermographic survey of
Page 28
the internal surface is recommended in order to identify and avoid thermal bridges and
anomalies in the construction.
Comparison to default U‐values for roofs
The CIBSE Guide suggests 2.30 W/m2K for cold pitched roofs with horizontal ceiling,
constructed with 12.5 mm plasterboard, no insulation, roof space, and tiling; and
0.42 W/m2K for the same roof build‐up where insulated with 100 mm mineral wool above
the plasterboard. The only example in the study comparable to this build‐up is the
measurement at the Balmacara Estate office (ID10.3), which has a timber‐lined ceiling to a
cold pitched roof, with a value of 0.8 W/m2K. It is assumed that the roof is insulated,
however, the depth of insulation is unknown.
The majority of the other roofs studied were coombs, for which no default value is quoted.
Similarly there is no default value for a thatched roof for Beaton’s Croft House (ID6).
Comparison of thermal upgrading to current building standards requirements
The building standards (Scottish Government, 2010) require under section 6.2 that building
elements of new buildings, and of conversions of existing buildings, achieve certain
maximum U‐values. These U‐values, applicable as of October 2010, are outlined in Table 3
below. However, building standards recognise that achieving these U‐values can be difficult
in traditional buildings, and higher U‐values can be acceptable. 9
Table 3. U‐value requirements of 2010 buildings standards (U‐values are expressed as
W/m2K. The U‐values for conversions apply to conversion of previously heated buildings.)
Building Domestic Non‐domestic
element New building Conversion New building Conversion
Wall 0.25 0.30 0.27 0.30
Roof 0.18 0.25 0.20 0.25
Floor 0.20 0.25 0.22 0.25
The measurement results of the solid stonewalls (600 mm) of the Balmacara Square
steading conversion (ID11) demonstrate that retrofitting such walls internally with insulated
drylining can achieve the current requirements. The measured U‐value here was 0.3 W/m2K
as required by building standards.
Similarly, the upgrading of one of the 1930s semi‐detached houses (ID14) with cavity fill
insulation resulted in a U‐value of 0.3 W/m2K.
9
U‐value requirements by building standards for conversions of building should be discussed with the
building standards officer at the local councils. Historic Scotland has published guidance to this regard:
Historic Scotland (2007): Conversion of traditional buildings: application of the Scottish building standards.
(Guide for Practitioners 6) Edinburgh: Historic Scotland.
Page 29
The measurement results of the recently insulated roof at the Balmacara Square (ID11)
resulted in U‐values of 0.3 W/m2K which does not achieve the U‐value requirement by
building standards but does constitute a significant improvement to the existing prior to
conversion, and can therefore be acceptable.9
However, careful detailing in refurbishment projects is recommended to minimise thermal
bridging, avoid the associated risk of interstitial condensation, and take into account issues
for vapour transfer in traditional building materials.
6. Conclusions
The in situ U‐values of 57 walls and 9 ceilings / coombs have been carried out covering part
of the range of traditional Scottish constructions and internal finishes, and also including
examples of non‐traditional cavity wall constructions. Six measurements were carried out in
pre‐1919 buildings retrofitted with insulation.
Given the sample size it is not possible to differentiate between different masonry
materials.
However, the following conclusions can be drawn from the study.
• Increased wall thickness improves the thermal resistance, and results in a lower
U‐value.
• Walls with internal finishes which incorporate an air‐filled cavity, such as plaster on
laths, drylining or timber lining, have lower U‐values than walls of the same
thickness finished with plaster on the hard. This demonstrates the insulating effect
of such an air cavity, especially where the air is stagnant or moving slowly.
(However, cavities, especially when ventilated to the outside, can sometimes also
have a detrimental effect on thermal performance.)
• Thus far, indicative U‐values for 600 mm thick traditional solid stonewalls are as
follows:
o Uninsulated walls finished with plaster on laths: 1.1 ±0.2 W/m2K
o Uninsulated walls drylined with plasterboard: 0.9 ±0.2 W/m2K
• The walling material impacts on the thermal performance. The study has shown
that earth/mud walls perform thermally better than stonewalls of the same
thickness.
• Internal drylining and insulating of solid stonewalls can improve the thermal
performance of the wall significantly. (However, careful detailing in refurbishment
projects is recommended to minimise thermal bridging, avoid the associated risk of
interstitial condensation, and take into account issues of vapour transfer in
traditional building materials.)
Page 30
• The thermal benefits of retrofitting insulation relies heavily on the correct
installation of the insulating systems. The measurements at the Colony Flat (ID4)
have shown that U‐values can vary significantly, presumably due to inappropriately
installed insulation. (A thermal imaging survey can be used to locate / verify
differences in thermal performance.)
• Generally U‐value calculations tend to overestimate the U‐values compared with
the results from the in situ measurements. This is particularly the case if no account
of the proportion of mortar in a solid stonewall construction is considered. Better
agreement was achieved, in some cases, if the wall was considered as a
stone/mortar mix, although the actual construction details may remain elusive.
• Considering the actual build‐up of a traditional stonewall can have a significant
impact on the walls thermal performance. Often such walls are considered to be
fully made of stones with regular mortar joints. However in reality, such walls are
not a uniform construction but have a centre filled with small stones and a larger
proportion of mortar. The overall mix of stone to mortar might be up to 60/40%.
The difference between the U‐value calculated assuming 40% mortar, and that
allowing for 0% mortar (i.e. 100% stone), is 30%.
• U‐value calculations for the non‐traditional cavity wall construction with better
defined build‐ups gave closer agreement to the in situ measurement results.
• U‐value calculations for ceilings and coombs showed particularly poor agreement
with the measured U‐values.
• Improvements to the baseline data for U‐value calculations for traditional
constructions are recommended including the following matters:
o Thickness of layers
o Status of cavities (such as insulated/uninsulated and ventilated/unventilated.
o Ratio and types of stone and mortar (and voids) in solid stonewalls
o Thermal properties of materials used in traditional construction (such as local
stone types, historic brick, earth/mud walls, traditional plasters and mortars)
• U‐value calculation software programs only provide limited baseline data for
traditional building materials for their calculations (e.g. the BuildDesk database
provides only two options for stone types, sandstone and granite). Although the
modelling of mortar joints in masonry is included in such software programs, the
modelling of a traditional solid stonewall (i.e. with a centre packed with small
stones and mortar) is not allowed for. This can, however, be modelled as a mulit‐
layer build‐up.
• It is recommended that producers of U‐value calculation programs extend their
baseline database to include more traditional building materials, and allow for
easier, and more user‐friendly, modelling of traditional construction techniques,
such as solid stonewalls.
• It is recommended that further research is carried out to establish a better
understanding of the thermal properties of traditional building materials and
Page 31
construction components (e.g. different stone types, different mortar types, plaster
on laths, vented/unvented air‐filled cavities behind internal finishes).
• It is recommended to establish a standardised methodology for in situ
measurements of U‐value to ensure that future measurement results are
comparable.
The in situ measurement of U‐values is useful tool which can aid in the assessment of the
thermal performance of traditional building elements, particularly where calculation
methods may suffer from deficiencies resulting from lack of knowledge of the actual build‐
ups used, and of the thermal properties of traditional materials.
Page 32
Appendix A Monitoring and analysis methodology
This appendix describes the procedures used in this study for the monitoring and analysis of
the thermal heat flow through building elements. The procedures have been developed
during the first phase of this project (Baker, 2008) and projects for other organisations. This
appendix first describes the monitoring equipment used for the measurements, and its set‐
up; and then the analysis of the collected data including two methods of error analyses.
A1. Monitoring
For this study, the heat flow (thermal conductivity) through a building element, and the
associated indoor and outdoor temperatures were measured. Ideally the temperature
measurements would be of both air and surface temperatures, with surface temperature
being the preferred option where only one such measurement would be possible.
Campbell Scientific CR1000 data loggers equipped with heat flux and temperature sensors
were generally used, however some external measurements were made with Tinytag Plus 2
loggers in locations were access was restricted.
Hukseflux HFP01 heat flux sensors were used to measure heat flows through the selected
walls (Figure A1). The sensors are 80 mm in diameter and 5 mm thick. The sensors were
mounted by firstly applying a layer of double sided adhesive tape to the back of the sensor.
Secondly, low tack masking tape was applied to the wall. Finally, the heat flux sensor was
applied firmly to the masked area. This arrangement was generally satisfactory for two or
more weeks monitoring on painted surfaces. Wallpapered surfaces were not generally used
in case of damage. Sensor locations were chosen to avoid probable thermal bridge locations
near to windows and corners, with the sensor ideally located about half‐way between
window and corner, and floor and ceiling (Figure A2). Where possible a North‐facing or
sheltered elevation was selected to reduce the influence of solar radiation on the wall.
Figure A1. Heat flux sensor
Page 33
Figure A2. Typical heat flux sensor and room temperature measurement locations
To measure room air temperature, stainless steel‐sheathed thermistors, Campbell Scientific
type 107 temperature probes, were used internally mounted within a simple radiation
shield in order to minimise the influence of solar radiation and other heat sources
(Figure A3).
The surface temperature of the face of each heat flux sensor was measured using type‐T
thermocouples taped onto the surface of the heat flux sensor (Figure A4).
Figure A3. Room air temperature shield Figure A4. Type‐T thermocouple mounted
on surface of heat flux sensor
To measure the outdoor air temperature, stainless steel‐sheathed thermistors, Campbell
Scientific 107, were used externally placed in a radiation shield, either mounted onto the
Page 34
exterior wall surface by screw‐fixing the bracket of the shield to a mortar joint (Figure A5),
or by tying the bracket to a rainwater downpipe. External temperatures were also measured
using separate data loggers (Gemini Tinytag Plus 2 with Tinytag Standard Thermistor
Probes) which could be mounted outdoors (Figure A6), as it had been found that, during the
first phase of the project, it was not always possible, or practical, to run an external sensor
cable back into the building, particularly through sash windows without leaving the window
slightly open to accommodate the cable. In contrast modern windows fitted with a gasket
seal can be closed onto a cable.
Figure A5. Campbell Scientific 107 Tempera‐ Figure A6. Tinytag Standard Thermistor
ture Probe in a radiation shield screw‐fixed Probe in a radiation shield tied to a
into a mortar joint of the external wall face. rainwater downpipe. The probe is
The probe is connected to a data logger connected to a Tinytag Plus 2 data logger
inside the building. hanging from the shield bracket.
External wall surface temperatures were generally measured using type‐T thermocouples.
Crimp‐on terminals were used to secure surface temperature sensors to mortar joints, by
drilling and plugging joints (Figure A7). External surface temperature measurements were
Crimp-on terminal Thermistor
Heatshrink sleeving
Figure A7: Method of mounting external surface temperature sensor to mortar joint
not made where access was a problem, or fixing in the manner described would damage the
external finish.
Page 35
Room sensors and external sensors were logged at 5 second intervals and averaged over
10 minutes using the Campbell Scientific logger. Tinytag loggers were set to record at 1 or 2
minute intervals, depending on the expected duration of the monitoring period.
A2. Data Analysis
Given that the monitoring conditions are non‐steady state, it is considered necessary to
monitor for about two weeks or, preferably longer, in order to collect sufficient data to
estimate in situ U‐values. The period should be sufficient to take into account the thermal
capacity / inertia of the wall. Figure A8 shows the effect of increasing the length of the
monitoring period on the estimate of the U‐value using a simple averaging procedure as
described below. A period of at least a week is required before the U‐value estimate
stabilises to within ±5% of the final value determined from about 27 days data.
Figure A8: The effect of increasing the monitoring period
For example, the U‐value may be estimated by a simple averaging procedure as follows:
i =t
∑Q i
U= i =t
0
i =t
W/m2K Eqn. A1
∑ Ti − ∑ Te
0
i
0
i
where U is the average U‐value after time t, Q, Ti and Te are, respectively, the heat flux,
room temperature and external temperature collected at time intervals of i.
Page 36
There are drawbacks to using internal and external air temperatures in terms of the
uncertainties introduced. In the case of internal air temperature stratification may occur,
therefore the measured temperature may not be representative for the location of the heat
flux meter. Whilst the external air temperature measurements may be representative for
the building, there may be exposure of the external surface to solar radiation, and radiative
exchange with its surroundings will occur. Therefore an alternative to using air
temperatures to calculate U‐values using Equation A1, is to use the surface temperature
difference across the wall to determine its thermal resistance and add the standard internal
and external surface resistances, respectively rint = 0.13m2K/W and rext=0.04 m2K/W, as
follows:
1
U t = i =t W/m2K Eqn. A2
∑0 Tsii − Tsei
i =t
+ rint + rext
∑ Qi
0
where Tsi and Tse are respectively the internal and external surface temperatures.
In some cases it is not possible to measure the external surface temperature. Therefore the
difference between the internal surface temperature and the external air temperature can
be used as follows:
1
U t = i =t W/m2K Eqn. A3
∑0 Tsii − Tei
i =t
+ rint
∑ Qi
0
A small correction is applied for the thermal resistance of the heat flux sensor
(<6.25×103 m2K/W).
A3. Error analysis
The uncertainty of the U‐value estimate is derived from the individual measurement
uncertainties and the standard deviation (s.d.) of the average value.
For the averaging method, the calculated U‐value contains all the information available;
therefore the uncertainty of this value cannot be easily determined.
Error analysis by averaging method
One approach is to calculate moving averages for, say, weekly periods, i.e. the first period is
the average over day 1 to day 7; the second period day 2 to day 8; etc. The standard
Page 37
deviation (s.d.) of these N averages can then be calculated, which will give some indication
of the uncertainty of the estimated U‐value. This approach is justified since a week is the
minimum period which may be expected to give a result.
Each of the measured parameters (heat flux, and internal and external temperature) has an
associated uncertainty due to the sensor itself (ES) and the logging system (EL). These are
combined as follows:
+ EL
2 2
E S
Eqn. A4
In order to determine the error each measurement will have on the U‐value estimate, the
U‐value calculation is repeated with each measured parameter perturbed by its error in
turn. For example, the error on internal surface temperature (δTsi) measurement is applied
(Equation A5) to calculate Uerr_Tsi:
1
U err _ Tsi = i =t Equ.A5
∑0 [Tsii + δTsii − Tsei ]
i =t
+ rint + rext
∑ Qi 0
The overall uncertainty on the U‐value estimate, δU, is calculated as the root mean square
value (RMS) of the deviations of each error case from the base case (i.e. the value
determined from Eqn. A2 or Eqn. A3) and the standard deviation of U as follows:
δU = [(U − U ) + (U − U
errQ
2
errTi )2 + (U − U errTe )2 + (s.d .)2 ] Eqn.A6
where UerrQ, UerrTi and UerrTe are the U‐values calculated by applying the errors due to heat
flux, internal temperature and external temperature, respectively.
Table A1 gives an example of the error analysis.
Table A1. The estimation of the uncertainty of the U‐value of a wall in a heated building
with a temperature difference of 8.3K
Sensor Average Sensor U‐value W/m2K
Value Error
Base Case U 1.52
s.d. 0.02
2
Heat Flux 16.8W/m 5% U_err_Q 1.57
o
Internal Surface Temp. 16.4 C 0.5K U_err_Ts_int 1.45
o
External Surface Temp. 8.0 C 0.5K U_err_Ts_ext 1.59
Temperature Difference 8.3K
Overall uncertainty δU 0.11
8%
Page 38
Whilst the uncertainty of the U‐values estimates is generally about ±10%, the level of
uncertainty increases where the temperatures difference across the wall or building
element is small. An example is given below for a measurement in an unheated building
where the average surface temperature difference across the wall is less than 1K (Table A2).
Table A2. The estimation of the uncertainty of the U‐value of a solid stone wall in an
unheated building with a temperature difference of 0.9K
Sensor Average Value Sensor Error U‐value W/m2K
Base Case U 1.83
s.d. 0.58
2
Heat Flux 2.5W/m 5% U_err_Q 1.89
o
Internal Surface Temperature 2.6 C 0.5K U_err_Ts_int 1.34
o
External Surface Temperature 1.6 C 0.5K U_err_Ts_ext 2.87
Temperature Difference 0.9K
RMS error 1.70
93%
Whilst the U‐value of the wall in the unheated building appears acceptable (1.8W/m2K), the
result should be rejected since the uncertainty is ±1.7W/m2K (93%). The U‐value of the wall
in the heated building is 1.5±0.1W/m2K (8%) which is satisfactory.
Error analysis by dynamic method
An alternative to the averaging method is to use a dynamic analysis method which explicitly
takes into account the thermal capacity of the wall. Such a method may be more
appropriate if, for example, there are large diurnal swings in external conditions as may be
experienced during spring, or changes in the weather pattern during the test period. An
example of such software is the LORD program (Gutschker, 2004) which models the wall as
a network of conductances and capacitances, analogous to an electrical circuit. Figure A9
shows an example of a simple wall. The wall is modelled with four nodes: the boundary
conditions of the network at nodes 1 and 4 are the measured temperatures (at node 1 the
outside temperature Text and at node 4 the inside temperature Tint). The measured heat flux
is applied at the interior node 4. The nodes are connected by thermal conductances
(H 1‐2, etc.). Each node has a certain thermal capacity (C2, etc.). Storage of heat is only
possible at the nodes. The program calculates the best fit values for the conductances and
Figure A9. Example of a wall modelled as a network of conductances and capacitances
Page 39
thermal capacitances. The number of nodes used to model the wall depends on its thermal
mass. However the selection of the optimum number of nodes may require a process of
trial and error and can be somewhat dependent on the user’s experience of interpreting the
output of the program.
Page 40
Appendix B References
Anderson, B., 2006a. Conventions for U‐value calculations. (BR 443:2006) Bracknell,
Berkshire: BRE Press.
Anderson, B., 2006b. Thermal properties of building structures. In: The Chartered Institution
of Building Services Engineers, 2006. CIBSE guide A: environmental design. 7th ed. London:
CIBSE. Ch.3.
Baker, P. H., Sanders, C., Galbraith, G. H. & McLean. R. C., 2007. Investigation of wetting and
drying behaviour of replica historic wall constructions. In: Cassar, M. & Hawkings, C., ed.,
2007. Engineering historic futures: stakeholders dissemination and scientific research report.
London: Centre for Sustainable Heritage, University College London.
Available at http://www.ucl.ac.uk/sustainableheritage/ehf_report_web.pdf
Baker, P., 2008. In situ U‐value measurements in traditional buildings: preliminary results.
(Historic Scotland Technical Paper 2) Edinburgh: Historic Scotland.
Available at http://www.historic‐scotland.gov.uk/technicalpapers
BS EN ISO 7345:1987. Thermal insulation: physical quantities and definitions. London: BSI.
BS EN ISO 6946:1997. Building components and building elements: thermal resistance and
thermal transmittance: calculation methods. London: BSI.
Changeworks, 2008. Energy heritage: a guide to improving energy efficiency in traditional
and historic homes. Edinburgh: Changeworks.
Available at http://www.changeworks.org.uk/uploads/83096‐EnergyHeritage_online1.pdf
Energy Saving Trust, 2004. Scotland: assessing U‐values of existing housing. (Energy
Efficiency Best Practice in Housing: CE84) London: Energy Saving Trust.
Available at http://www.energysavingtrust.org.uk/business/content/download/180012/
441763/version/4/file/ce84.pdf
Gutschker, O., 2004. LORD: modelling and identification software for thermal system: user
manual. Cottbus, Germany: Brandenburg Technical University.
Künzel, H. M., Radon, J., & Schmidt, T., 1997. WUFI v2.0: simulation of heat and moisture
transport in building components. Holzkirchen, Germany: Fraunhofer Institute for Building
Physics.
prEN 12494:1996. Building components and elements: in situ measurement of the surface to
surface thermal resistance. London: BSI.
Scottish Government, 2010. Building Standards Technical Handbooks. Edinburgh: Scottish
Government.
Available at http://www.scotland.gov.uk/Topics/Built‐Environment/Building/Building‐
standards/publications/pubtech
Page 41
Appendix C Building Datasheets
1. Georgian tenement, Edinburgh
2. Victorian tenement, Glasgow
3. Victorian villa, Glasgow
4. Colonies flat, Edinburgh
5. Logie Schoolhouse, Angus
6. Beaton’s Croft House, Highlands
7. Stalker’s cottage, Highlands
8. Weens Garden Cottage, Scottish Borders
9. Castle Fraser Estate, Aberdeenshire
10. Balmacara Estate, Highlands
11. Balmacara Square, Highlands
12. Dumfries House Estate, East Ayrshire
13. McCowan house, Dumfries and Galloway
14. 1930s semi‐detached houses, East Renfrewshire
15. 1970s detached houses, Dumfries and Galloway
Page 42
1. Georgian tenement Location: Tollcross, Edinburgh Date: early 19th century
Building description
A Georgian tenement built in Tollcross, Edinburgh, in the early 19th century. The building is
category ‘B’ listed, in a conservation area and within a World Heritage Site.
Wall measurement
The external walls are solid stonewalls built with Craigleith stone, a local blond sandstone. The front
elevation is of ashlar, the rear elevation of squared rubble stonework. The basement wall face is
finished with cement render.
Five wall measurement were taken in flat on the basement, ground and first floors. Measurements
were taken on front and rear elevations with the measurement locations having varying wall
thicknesses and internal and external finishes.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
1.1 Basement flat, front 600 mm ashlar solid wall from blond no plaster 1.6 W/m2K 1.5 ‐ 2.2 W/m2K Dec 2007 11.1 ˚C 4.6 ˚C
elevation Craigleith sandstone on hard
1.2 Basement flat, 300 mm ashlar solid wall from blond no plaster 2.3 W/m2K 2.3 ‐ 3.0 W/m2K Jan 2008 18.5 ˚C 5.9 ˚C
front elevation, Craigleith sandstone on hard
below window
1.3 Basement flat, 600 mm cement solid wall from blond yes plaster‐ 1.0 W/m2K 1.1 ‐ 1.4 W/m2K Dec 2007 13.3 ˚C 4.5 ˚C
rear elevation render Craigleith sandstone board
1.4 Ground floor flat, 600 mm ashlar solid wall from blond yes plaster 1.4 W/m2K 1.2 ‐ 1.7 W/m2K Dec 2007 15.8 ˚C 4.6 ˚C
front elevation Craigleith sandstone on laths
1.5 First floor flat, 600 mm random solid wall from blond yes plaster‐ 0.8 W/m2K 1.2 ‐ 1.7 W/m2K Jan 2008 17.8 ˚C 4.6 ˚C
rear elevation rubble Craigleith sandstone board
1. Georgian tenement continued…
Floor measurement
The basement floor in one flat was tested. The floor was a solid concrete floor slab installed in the
1970s. The slab was cast onto a damp‐proof membrane on gravel hardcore.
To enable the measuring the floor, a core of 100mm diameter was drilled out of the slab to allow the
insertion of a heat flux sensor. The original core was then re‐inserted into the slab.
Two measurements were taken on the basement floor. The first measurement was taken on the
exposed concrete slab. For the second measurement, 30 mm thick Spacetherm‐F insulating board
added to the slab on the room‐side. The board is a composite material consisting of a 21 mm thick
layer of Aerogel insulation and a 9 mm thick layer of Fermacell particle board.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
1.6 Basement flat, 150 mm unfinished solid concrete floor no unfinished 3.5 W/m2K 3.3 W/m2K Dec 2007 ‐ 9.0˚C 9.8˚C
floor, without slab with no Jan 2008
insulation insulation
1.7 Basement flat, 180 mm unfinished solid concrete floor no unfinished 0.6 W/m2K 0.5 W/m2K Jan‐Feb 2008 14.2˚C 12.0˚C
floor, with slab with 30 mm
insulation insulating board on
room‐side
2. Victorian tenement Location: Dennistoun, Glasgow Date: 1880s
Building description
A Victorian tenement built in Dennistoun, Glasgow, in the 1880s. The four‐storey building comprises
of shops on the ground floor, and six tenements of a common close, i.e. stairwell.
Wall measurement
The external walls are solid stonewalls build from sandstone. Although the front elevation, not
measured, is of red sandstone ashlar, the measured rear elevation is built in coursed rubble
stonework from blond sandstone.
In addition to external walls, measured were also taken on the internal walls between close and flats.
These walls were solid brick walls finished with plaster on the close side.
The interiors of the flats were refurbished in the 1980s, and the wall finishes, at the time of
measurement, were plasterboarded dry‐lining with mineral wool insulation in‐between the studs.
Four wall measurements were taken: two on the rear elevations and two on the close walls The
measurements were taken on the first and second floors.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
2.1 First floor flat, 600 mm coursed solid wall from blond insulated plaster‐ 1.0 W/m2K 1.2 ‐ 1.5 W/m2K Apr‐May 15.6 ˚C 9.9 ˚C
external rear wall rubble sandstone board 2009
2.2 First floor flat, 200 mm plaster solid brick wall insulated plaster‐ 2.4 W/m2K 0.6 W/m2K Apr‐May 15.6 ˚C 14.5 ˚C
wall to stairwell board 2009
2.3 Second floor flat, 600 mm coursed solid wall from blond insulated plaster‐ 0.9 W/m2K 1.2 ‐ 1.5 W/m2K Apr‐May 20.6 ˚C 9.9 ˚C
external rear wall rubble sandstone board 2009
2.4 Second floor flat, 200 mm plaster solid brick wall insulated plaster‐ 1.7 W/m2K 0.6 W/m2K Apr‐May 20.6 ˚C 18.4 ˚C
wall to stairwell board 2009
3. Victorian villa Location: Cathcart, Glasgow Date: 1880s
Building description
A detached Victorian villa built in Cathcart, Glasgow, in the 1880s.
Wall measurement
The external walls are solid stonewalls build with blond sandstone. The front elevation is of ashlar, the
other elevations of squared rubble stonework.
Four wall measurements were taken in the northwest facing bedroom on the first floor. The
measurements were taken on the north and west walls, with one measurement taken in a wall press.
The interior wall finish was plaster on lath on studs with no insulation. The wall press was finished
with timber lining, presumably on studs with no insulation.
Coomb measurement
One measurement was also taken on the ceiling coomb in the same bedroom. The roof is a timber
construction with slate covering. The interior coomb finish is plaster on lath. The roof is not insulated.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
3.1 Bedroom, north / front 600 mm ashlar solid wall from yes plaster 1.0 W/m2K 1.2 ‐ 1.5 W/m2K Nov‐Dec 18.2 ˚C 6.9 ˚C
elevation blond sandstone on laths 2007
3.2 Bedroom, west / gable 600 mm squared solid wall from yes plaster 1.1 W/m2K 1.2 ‐ 1.5 W/m2K Nov‐Dec 19.1 ˚C 4.4 ˚C
wall, measurement 1 rubble blond sandstone on laths 2008
3.3 Bedroom, west / gable 600 mm squared solid wall from yes plaster 1.1 W/m2K 1.2 ‐ 1.5 W/m2K Nov‐Dec 19.1 ˚C 4.4 ˚C
wall, measurement 2 rubble blond sandstone on laths 2008
3.4 Bedroom, wall press in 300 mm squared solid wall from yes plaster 1.5 W/m2K 1.5 ‐ 1.9 W/m2K Nov‐Dec 19.1 ˚C 4.4 ˚C
west / gable wall rubble blond sandstone on laths 2008
3.5 Bedroom, ceiling coomb unknown slate timber roof yes plaster 0.7 W/m2K 1.7 W/m2K Nov‐Dec 19.1 ˚C 4.4 ˚C
on laths 2008
4. Colonies flat Location: ‘Shaftesbury Park’ Colonies, Edinburgh Date: around 1900
Building description
A ‘Colonies’ house built, as part of the ‘Shaftesbury Park’ Colonies, Edinburgh, around 1900. The
building forms the end of a terrace of houses. As typical for ‘Edinburgh Colonies’, the house consists of
a ‘lower flat’ on the ground floor and an ‘upper flat’ on the first floor extending into the roof space.
The ‘Shaftesbury Park’ Colonies form part of a conservation area.
Wall measurement
The external walls are solid stonewalls build with blond sandstone in ashlar finish. One wall
measurement was taken on the first floor, i.e. in the ‘upper flat’, in the dining room which was
internally finished with plasterboarded dry‐lining, presumably with no insulation.
Coomb measurement
Two measurements were also taken on the ceiling coombs on the attic floor. The roof is a timber
construction with slate covering. The interior coomb finish is plaster on lath. The roof is insulated
in‐between the rafter.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
4.1 'Upper flat', first floor, 400 mm ashlar solid wall from blond yes plaster‐ 0.6 W/m2K 1.2 ‐ 1.5 W/m2K Mar 2009 16.0 ˚C 4.3 ˚C
dining room, wall sandstone board
4.2 'Upper flat', attic floor, unknown slate timber roof with no plaster‐ 0.4 W/m2K 0.4 W/m2K Mar 2009 14.1 ˚C 6.4 ˚C
ceiling coomb, insulations in‐between board
measurement 1 rafters
4.3 'Upper flat', attic floor, unknown slate timber roof with no plaster‐ 1.1 W/m2K 0.4 W/m2K Mar 2009 14.1 ˚C 6.4 ˚C
ceiling coomb, insulations in‐between board
measurement 2 rafters
5. Logie Schoolhouse Location: Logie, near Montrose, Angus Date: late 18th century, converted 2005‐2006
Building description
A rural schoolhouse built in Logie near Montrose, Angus, in the late 18th century. The building is
originally a mud construction, a traditional Scottish building technique utilizing earth with a clay
content and straw. The building was converted in 1929 for church purposes, and in 2005‐2006 for
residential use. The building is category ‘A’ listed.
Wall measurement
The external walls are solid walls of mud construction. Alterations and repairs were carried out to the
walls over time, replacing the outer face of the mud wall with brick or random rubble. Some areas of
original full width mud wall survive. The building has been re‐harled in 2006 using a clay render.
Six wall measurements were taken on the north and south walls measuring different a variety of wall
build‐ups. All measurements were taken below the level of the window cills where the walls were
finished with timber lining on studs with no insulation.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
5.1 North elevation, 600 mm clay solid mud wall without yes timber 0.6 W/m2K 0.7 ‐ 0.8 Mar‐May 9.9 ˚C 7.8 ˚C
measurement 1 harling repairs lining 2009
5.2 North elevation, 600 mm clay solid mud wall with yes timber 0.5 W/m2K 0.8 ‐ 1.0 Mar‐May 9.9 ˚C 7.8 ˚C
measurement 2 harling exterior stone repairs lining 2009
5.3 North elevation, 600 mm clay solid mud wall with yes timber 0.8 W/m2K 0.8 ‐ 1.0 Mar‐May 9.9 ˚C 7.8 ˚C
measurement 3 harling exterior stone repairs lining 2009
5.4 South elevation, 600 mm clay solid mud wall with yes timber 0.6 W/m2K 0.7 ‐ 0.8 Mar‐May 9.9 ˚C 7.8 ˚C
measurement 1 harling exterior brick repairs lining 2009
5.5 South elevation, 600 mm clay solid mud wall with yes timber 0.4 W/m2K 0.7 ‐ 0.8 Mar‐May 9.9 ˚C 7.8 ˚C
measurement 2 harling exterior brick repairs lining 2009
5.6 South elevation, 600 mm clay solid mud wall with yes timber 0.8 W/m2K 0.7 ‐ 0.8 Mar‐May 9.9 ˚C 7.8 ˚C
measurement 3 harling exterior brick repairs lining 2009
6. Beaton’s Croft House Location: Bornesketaig, Isle of Skye, Highlands Date: mid 19th century
Building description
A thatched croft house built in Bornesketaig on Isle of Skye in the mid 19th century. The building
follows the Scottish ‘blackhouse’ typology with relatively thick solid walls. The building is category ‘A’
listed, and owned by the National Trust for Scotland.
Wall measurement
The external walls are solid stonewall built with local rubble. Two measurements were taken: one
each in the bedroom and the living room. Both rooms are finished internally with timber lining on
studs with no insulation.
Ceiling and coomb measurement
Three measurement were also taken at the ceilings and coombs. The roof is thatch on a timber
construction. The ceilings to the loft space were timber joisted with no insulation. Ceiling and coombs
were finished internally with timber lining.
In the living room, both ceiling and coomb were measured, whereas in the bedroom, only the coomb
was measured.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
6.1 Bedroom, 1200 mm rubble solid wall made from yes timber 0.8 W/m2K 0.7 ‐ 1.1 W/m2K Jan‐Feb 16.6 ˚C 4.7 ˚C
short wall local stone lining 2009
6.2 Living room, 1200 mm rubble solid wall made from yes timber 0.6 W/m2K 0.7 ‐ 1.1 W/m2K Jan‐Feb 16.6 ˚C 4.9 ˚C
long wall local stone lining 2009
6.3 Bedroom, unknown thatch timber roof yes timber 1.2 W/m2K 0.3 W/m2K Jan‐Feb 16.6 ˚C 4.7 ˚C
ceiling coomb lining 2009
6.4 Living room, unknown thatch timber roof yes timber 1.5 W/m2K 0.3 W/m2K Jan‐Feb 16.6 ˚C 4.9 ˚C
ceiling coomb lining 2009
6.5 Living room, unknown thatch timber ceiling (to attic yes timber 1.1 W/m2K 0.4 W/m2K Jan‐Feb 16.6 ˚C 4.9 ˚C
ceiling space under thatched lining 2009
roof, i.e. warm roof)
7. Stalker’s cottage Location: Torridon, Highlands Date: mid 19th century with 1950s extension
Building description
A stalker’s cottage built in Torridon in the northwest Highlands in the mid 19th century. The building
was extended in the 1950s.
Wall measurement
The external walls of the original cottage are solid stonewalls built with Torridonian stone, a local
sandstone. The extension was built as a cavity wall construction with two leaves of concrete block; the
wall cavity is uninsulated. The exterior wall finish of both, cottage and extension, is harling. The
interior wall finish is plaster on lath on studs with no insulation.
Three measurements were taken: one in the original cottage, and two in the extension.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
7.1 Original cottage 650 mm harling solid wall from yes plaster 1.6 W/m2K 1.1 ‐ 1.5 W/m2K Feb 2009 18.5 ˚C 6.1 ˚C
Torridonian sandstone on laths
7.2 1950s extension, 250 mm harling Two leaves of 100mm yes plaster 1.5 W/m2K 1.3 W/m2K Feb 2009 19.8 ˚C 6.1 ˚C
measurement 1 concrete block with on laths
50mm uninsulated cavity
7.3 1950s extension, 250 mm harling Two leaves of 100mm yes plaster 1.1 W/m2K 1.3 W/m2K Feb 2009 19.8 ˚C 6.1 ˚C
measurement 2 concrete block with on laths
50mm uninsulated cavity
8. Weens Garden Cottage Location: Weens, Hawick, Scottish Borders Date: 1845 with 1950s extension
Building description
A cottage built in Weens near Hawick, Scottish Borders, in 1845. The west / back elevation of the
building forms part of a boundary wall of a walled garden. In the 1950s, a small rendered extension
was added to the west of gable end (south wall) of the building, also utilising the garden wall as back
wall (west wall).
Wall measurement
The external walls of the original cottage are solid stonewalls built with squared rubble of red
sandstone. The outer face of the wall (west wall) has been re‐fronted with brick.
The extensions is generally a brick cavity construction with no insulation. However, the measured back
wall (west wall) of the extension is, presumably, still the original garden wall, i.e. full width rubble
stonework, now finished in cement render.
Four measurements were taken: In the original cottage, two stonewalls were measured in addition to
the back wall (west wall). In the extension, only the back wall (west wall) was measured.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
8.1 Cottage, front / 400 mm squared solid wall from red no plaster 1.3 W/m2K 2.0 ‐ 2.7 W/m2K Mar‐Apr 15.6 ˚C 3.5 ˚C
east elevation rubble sandstone on hard 2008
8.2 Cottage, gable / 400 mm squared solid wall from red no plaster 1.1 W/m2K 2.0 ‐ 2.7 W/m2K Mar‐Apr 13.0 ˚C 5.2 ˚C
north elevation rubble sandstone on hard 2008
8.3 Cottage, west / 400 mm brick solid wall from red no plaster 1.1 W/m2K 1.2 W/m2K Mar‐Apr 16.5 ˚C 4.7 ˚C
back elevation sandstone fronted on hard 2008
with brick externally
8.4 Extension, west / 400 mm cement solid wall from red no plaster 1.5 W/m2K 1.4 W/m2K Mar‐Apr 14.4 ˚C 4.7 ˚C
back elevation, render sandstone on hard 2008
presumably part of
former garden wall
9. Castle Fraser Estate Location: Inverurie, Aberdeenshire Date: 17th century and mid 19th century
Building description
Within the Castle Fraser Estate in Inverurie, Aberdeenshire, the following three buildings were used
for measurements: the 17th century apartment wing of the actual castle, and the mid‐19th century
gardener’s bothy and stables (including its turret). The castle is category ‘A’ listed, and the other
estate buildings are included in this listing as part of the castle’s curtilage. The estate is now owned by
the National Trust for Scotland.
Wall measurement
The external walls of all the buildings were solid stonewalls built with Kemnay granite rubble. The
apartment wing is finished externally with lime harling.
Four wall measurements were taken: one in each building plus an additional measurement in the
turret of the stables. The interior wall finishes varied from plaster on lath (on studs) and plaster on the
hard to plasterboarded dry‐lining.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
9.1 Apartments, first floor, 600 mm harling solid wall made from yes plaster 0.8 W/m2K 1.2 ‐ 1.6 W/m2K Mar 2008 14.8 ˚C 5.0 ˚C
bedroom, east wall Kemnay granite on laths
9.2 Stables, ground floor, 350 mm rubble solid wall made from no plaster 1.8 W/m2K 2.2 ‐ 3.1 W/m2K Mar 2008 9.6 ˚C 5.0 ˚C
turret, north facing Kemnay granite on hard
9.3 Stables, ground floor, 600 mm rubble solid wall made from yes plaster‐ 0.9 W/m2K 1.2 ‐ 1.6 W/m2K Mar 2008 13.8 ˚C 5.0 ˚C
office, north facing Kemnay granite board
9.4 Gardener's Bothy, north 600 mm rubble solid wall made from yes plaster 0.9 W/m2K 1.2 ‐ 1.6 W/m2K Mar 2008 12.0 ˚C 5.0 ˚C
facing Kemnay granite on laths
10. Balmacara Estate Location: near Kyle of Lochalsh, Highlands Date: 1884‐1886
Building description
Within the Balmacara Estate, near Kyle of Lochalsh in the Scottish Highlands, measurements were
taken in an estate building erected in 1884‐1886. The estate is now owned by the National Trust for
Scotland, and the building measured is used as their estate office.
Wall measurement
The external walls are solid stonewalls from Torridonian stone, a local sandstone. The exterior wall
finish is squared rubble stonework. The internal wall finish is timber lining on studs with no insulation.
Only one wall measure was taken in an office room on the first floor.
Ceiling and coomb measurement
Two measurements were also taken on the ceiling and coomb in the same office. The roof is a timber
construction with slate covering. The ceiling to the loft space is timber joisted with insulation
in‐between the joists. The coomb is not insulated. The ceiling and coombs were finished internally
with timber lining.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
10.1 First floor office, 600 ‐ 750 squared solid wall from yes timber 0.9 W/m2K 1.0 ‐ 1.4 W/m2K Dec 2008 ‐ 18.6 ˚C 5.3 ˚C
wall mm rubble Torridonian sandstone lining Jan 2009
10.2 First floor office, unknown slate timber roof, not insulated yes timber 1.2 W/m2K 1.7 W/m2K Dec 2008 ‐ 18.6 ˚C 5.3 ˚C
ceiling coomb lining Jan 2009
10.3 First floor office, unknown slate timber joisted ceiling timber 0.8 W/m2K 0.6 W/m2K Dec 2008 ‐ 18.6 ˚C 5.3 ˚C
ceiling insulted in‐between lining Jan 2009
joists, i.e. cold roof
11. Balmacara Square Location: near Kyle of Lochalsh, Highlands Date: 19th / 20th century, converted 1999‐2000
Building description
The steading at Balmacara Square, built in the late 19th / early 20th century, is part of the Balmacara
Estate. The estate is located near Kyle of Lochalsh in the Scottish Highlands, and is now owned by the
National Trust for Scotland. The steading was converted in 1999‐2000 into two‐storey dwellings. The
steading is category ‘B’ listed.
Wall measurement
External walls are solid stonewalls built with local rubble sandstone. The exterior wall face is painted
in most places. Only one wall measurement was taken in a bedroom which was finished internally
with plasterboarded dry‐lining with insulation between studs.
Coomb measurement
One measurement was also taken on the ceiling coomb of the same bedroom. The roof was a slated
timber construction. The interior coomb finish is plasterboarded dry‐lining with insulation between
the studs.
ID Location Thickness External Construction type Studs / Internal finish U‐value U‐value Monitor Room External
finish air gap in‐situ calculated period temp. temp.
11.1 Bedroom, wall 600 mm rubble solid wall from Torri‐ no plasterboarded dry‐ 0.3 W/m2K 0.4 W/m2K Dec 2008 ‐ 19.3 ˚C 4.4 ˚C
donian sandstone lining with insulation Jan 2009
11.2 Bedroom, unknown slate timber roof insulated no plasterboard 0.3 W/m2K 0.4 W/m2K Dec 2008 ‐ 19.3 ˚C 4.4 ˚C
ceiling coomb in‐between rafters Jan 2010
Image copyright: The above image is an excerpt from an image by Richard Dorrell copyrighted under the Creative Commons Attribution‐Share Alike 2.0 Generic License.
The image, and the associated copyright details, can be accessed online at www.geograph.org.uk/photo/1590887
12. Dumfries House Estate Location: near Cumnock, East Ayrshire Date: 19th century
Building description
Within the Dumfries House Estate near Cumnock, East Ayrshire, measurements were taken in the
Garden Bothy, built in the 19th century. The Bothy is located at a walled garden. The garden wall also
forms the back wall of the Bothy. The building had been empty since the 1970s, and was in disrepair.
Wall measurement
The external walls are solid sandstone walls, except for the south wall, which is also part of the
boundary wall of the walled garden, which is a sandstone wall with out brick facing externally.
Walls were measured including their internal finishes, plaster on laths or uninsulated drylining with
plasterboard. Some measurements were also taken on bare walls without any finishes.
Temporary heating was provided during the measurement period.
ID Location Thickness External Construction Studs / Internal U‐value U‐value calculated Monitor Room External
finish type air gap finish in‐situ period temp. temp.
12.1 Kitchen, 600 mm rubble solid sandstone yes plaster on 1.3 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 11.0˚C 1.6˚C
east wall laths 2010
12.2 Kitchen, 600 mm rubble solid sandstone yes plaster on 0.9 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 11.0˚C 1.6˚C
north wall laths 2010
12.3 Kitchen, 600 mm brick solid sandstone yes plaster on 0.9 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 11.0˚C 1.6˚C
south wall and brick laths 2010
12.4 Living room, 600 mm brick solid sandstone no bare 2.4 W/m2K 1.6 ‐ 2.3 W/m2K Feb‐Mar 8.2˚C 1.6˚C
south wall and brick stone 2010
12.5 Living room, 600 mm rubble solid sandstone yes plaster‐ 1.3 W/m2K 1.2 ‐ 1.5 W/m2K Feb‐Mar 8.2˚C 1.6˚C
west wall board 2010
12.6 Living room, 600 mm rubble solid sandstone yes bare 2.4 W/m2K 1.6 ‐ 2.3 W/m2K Feb‐Mar 8.2˚C 1.6˚C
west wall stone 2010
12. Dumfries House Estate continued…
ID Location Thickness External Construction Studs / Internal U‐value U‐value calculated Monitor Room External
finish type air gap finish in‐situ period temp. temp.
12.7 West bedroom, 600 mm rubble solid sandstone yes plaster on 1.1 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 14.8˚C 1.6˚C
north wall laths 2010
12.8 West bedroom, 600 mm rubble solid sandstone yes plaster on 1.1 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 14.8˚C 1.6˚C
south wall and brick laths 2010
12.9 West bedroom, 600 mm rubble solid sandstone yes plaster on 1.1 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 14.8˚C 1.6˚C
west wall laths 2010
12.10 West bedroom, 300 mm rubble solid sandstone yes timber 1.2 W/m2K 1.5 ‐ 1.9 W/m2K Feb‐Mar 14.8˚C 1.6˚C
north wall lining 2010
12.11 East bedroom, 600 mm rubble solid sandstone yes plaster on 1.3 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 14.3˚C 1.6˚C
north wall laths 2010
12.12 East bedroom, 600 mm brick solid sandstone yes plaster on 1.3 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 14.3˚C 1.6˚C
south wall and brick laths 2010
12.13 East bedroom, 600 mm rubble solid sandstone yes plaster on 1.1 W/m2K 1.2 ‐ 1.6 W/m2K Feb‐Mar 14.3˚C 1.6˚C
east wall laths 2010
12.14 East bedroom, 300 mm rubble solid sandstone yes timber 1.1 W/m2K 1.5 ‐ 1.9 W/m2K Feb‐Mar 14.3˚C 1.6˚C
north wall lining 2010
13. McCowan House Location: Dumfries, Dumfries and Galloway Date: 1929‐1931
Building description
McCowan House was built in Dumfries in 1929‐1931 as nurses accommodation for the Crichton Royal
Hospital. It is now part of the Crichton Campus and academic institution. The building is connected to
the adjacent Rutherford House, and the building complex is category ‘B’ listed.
Wall measurement
The external walls were solid stonewalls built with Locharbriggs stone, a local red sandstone. Several
ventilation grilles were set into the walls at various locations and heights indicating that ventilation
ducts were running through the walls. No further information was available on the ventilation
network.
Six wall measurements were taken on different floors and walls of the building. The wall build‐up, and
interior finishes, at the various measurement locations were basically identical.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
13.1 Ground floor, 600 mm ashlar solid wall of Locharbriggs yes plaster 1.7 W/m2K 1.2 ‐ 1.7 W/m2K Nov‐Dec 21.0 ˚C 6.9 ˚C
south wall sandstone on laths 2007
13.2 Ground floor, 600 mm ashlar solid wall of Locharbriggs yes plaster 1.3 W/m2K 1.2 ‐ 1.7 W/m2K Nov‐Dec 22.2 ˚C 6.8 ˚C
north wall sandstone on laths 2008
13.3 First floor, 600 mm ashlar solid wall of Locharbriggs yes plaster 2.0 W/m2K 1.2 ‐ 1.7 W/m2K Nov‐Dec 20.7 ˚C 6.8 ˚C
south wall sandstone on laths 2009
13.4 First floor, 600 mm ashlar solid wall of Locharbriggs yes plaster 0.9 W/m2K 1.2 ‐ 1.7 W/m2K Nov‐Dec 21.7 ˚C 6.8 ˚C
north wall sandstone on laths 2010
13.5 Second floor, 600 mm ashlar solid wall of Locharbriggs yes plaster 1.5 W/m2K 1.2 ‐ 1.7 W/m2K Nov‐Dec 19.3 ˚C 7.2 ˚C
south wall sandstone on laths 2011
13.6 Second floor, 600 mm ashlar solid wall of Locharbriggs yes plaster 0.6 W/m2K 1.2 ‐ 1.7 W/m2K Nov‐Dec 20.3 ˚C 7.4 ˚C
north wall sandstone on laths 2012
14. 1930s semi‐detached houses Location: Giffnock, East Renfrewshire Date: 1930s
Building description
Two semi‐detached houses built in Giffnock, East Renfrewshire, in the 1930s.
Wall measurement
The external walls were cavity wall constructions with both leaves built with bricks. In house 1, the
cavity is not insulated; whereas in house 2, insulation had recently been filled into the cavity. The
exterior wall finish of both houses was cement render, and their interior wall finish, at the
measurement locations, was plaster on lath on studs with no insulation.
In total, two wall measurements were taken: one in each house. The measurements were taken in the
living rooms in similar locations.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
14.1 House 1, living room approx. harling two leaves of brick with yes plaster 1.3 W/m2K 1.1 ‐ 1.4 W/m2K Feb 2009 15.4 ˚C 3.0 ˚C
265 mm uninsulated cavity on laths
14.2 House 2, living room approx. harling two leaves of brick with yes plaster 0.3 W/m2K 0.3 W/m2K Feb 2009 17.3 ˚C 3.1 ˚C
265 mm insulated cavity on laths
15. 1970s detached houses Location: Dumfries, Dumfries and Galloway Date: late 1970s
Building description
Two detached houses built in Dumfries, Dumfries and Galloway, in the late 1970s.
Wall measurement
The external walls were cavity wall constructions with both leaves built with concrete blocks. In
house 1, the cavity is not insulated; whereas in house 2, insulation had recently been filled into the
cavity. The exterior wall finish of both houses was cement render, and their interior wall finish, at the
measurement locations, was plaster on the hard.
Two measurements were taken: one in each house at the gable end wall in similar locations. The room
finishes were plaster on the hard in both measurement locations.
ID Location Thickness External Construction type Studs / Internal U‐value U‐value Monitor Room External
finish air gap finish in‐situ calculated period temp. temp.
15.1 House 1, approx. cement two leaves of concrete no plaster 1.1 W/m2K 1.1 W/m2K Dec 2009 ‐ 13.7 ˚C 0.0 ˚C
gable end wall 265 mm render block with 65 mm on hard Jan 2010
uninsulated cavity
15.2 House 2, approx. cement two leaves of concrete no plaster 0.6 W/m2K 0.4 W/m2K Dec 2009 ‐ 15.6 ˚C 0.0 ˚C
gable end wall 265 mm render block with 65 mm on hard Jan 2010
insulated cavity
Published by
Historic Scotland
Conservation Group
January 2011
Historic Scotland, Longmore House, Salisbury Place, Edinburgh, EH9 1SH
Telephone: 0131 668 8600
Website: www.historic‐scotland.gov.uk
Email: hs.conservationgroup@scotland.gsi.gov.uk
© Crown copyright