Building Research Establishment Report

Improving the habitability of

large panel system dwellings

J P Cornish, BSc, PhD, CEng, AFIMA, MCIBSE

G Henderson, BSc, MSC, CEng, MlEE
Christine E Uglow, BSC, MSC
R K Stephen, BSC
J R Southern
C H Sanders, BSC, MSC, DIC

Building Research Establishment

Garston
Watford
WD2 7JR
Price lists for all avail&le
BRE publications can be
obtained from:
Publications Sales
Building Research Establishment
Garston, Watford, WDZ 7JR
Tel: Garston (0923) 664444

This publication is one of a series being prepared as part of the Building Research
Establishment's programme of investigation to assist local authorities and their consultants in
appraisal, maintenance and repair of large panel system dwellings.

Other BRE publications on large panel system dwellings are:

Building Research Establishment. The structure of Ronan Point and other Taylor Woodrow-
Anglian buildings. BRE Report. Garston, BRE, 1985.

Building Research Establishment. The structural adequacy and durability of large panel system
dwellings. BRE Report. Garston, BRE, 1987. (Includes BRE Information Paper IP8/87.)

Building Research Establishment. The structural adequacy and durability of large panel system
dwellings: summary of the report. BRE Information Paper IP8/87. Garston, BRE, 1987.

Edwards M J. Weatherproof joints in large panel systems: 1 Identification and typical defects.
BRE Information Paper 1P8/86. Garston, BRE, 1986.

Edwards M . IWeatherproof joints in large panel systems: 2 Remedial measures. BRE

.
Information Paper IP9/86. Garston, BRE, 1986.

Edwards M J. Weatherproof joints in large panel systems: 3 Investigation and diagnosis of

failures. BRE Information Paper IP10/86. Garston, BRE, 1986.

Edwards M J. Weatherproof joints in large panel systems: 4 Flat roofs, balconies and deck
accessways. BRE Information Paper IP15/86. Garston, BRE, 1986.

Harrison H W, Hunt J H and Thomson J. Overcladding external walls of large panel system
dwellings. BRE Report. Garston, BRE, 1986.

Hotchkiss A R and Edwards M J. Bison large panel system dwellings: constructional details. BRE
Report. Garston, BRE, 1988.

Hotchkiss A R and Edwards M J. Reema large panel system dwellings: constructional details.
BRE Report. Garston, BRE, 1987.

Morris W A and Read R E H. Appraisal of passive fire precautions in large panel system blocks
of flats and maisonettes. BRE Information Paper IP18/86. Garston, BRE, 1986.

Reeves B R. Large panel system dwellings: preliminary information on ownership and condition.
BRE Report. Garston, BRE, 1986.

BR 154
ISBN 0 85125 404 7

@ Crown copyright 1989

First published 1989

Applications to reproduce extracts

from the text o f this publication
should be made to the Publications
Officer at the Building Research Establishment
Contents

Page

Introduction 1

Chapter 1 The habitability of LPS dwellings 2

Chapter 2 Heating systems for LPS dwellings 8

Chapter 3 Ventilation in high-rise LPS flats 12

Chapter 4 Air leakage measurements in LPS flats 16

Chapter 5 Improving thermal insulation of LPS dwellings 23

Chapter 6 Two case studies in LPS dwellings 33

Conclusions 50

References 51
Introduction

This report is concerned with the habitability (internal

living conditions) of large panel system (LPS)
dwellings. It is based on a number of investigations
carried out by the Building Research Establishment
(BRE) in which measurements were taken in a large
number of LPS dwellings, and on other BRE research
aimed at understanding the physical principles
underlying many of the problems found in these and
other dwellings. The aim is t o present the results of
these investigations in a coherent way, as a basis for
decisions on remedial measures which are both
effective and economic.

Complaints t o local authorities about the internal

living conditions in dwellings are typically concerned
with dampness, with mould growth on internal
surfaces, furnishings and clothes, and with the costs
and difficulties of providing adequate heating. These
are complex problems whose solution requires an
understanding of the technical, social and economic
causes underlying them. Their solution will require
consideration of a number of different factors: the
provision of adequate ventilation, the insulation of
the external envelope, the costs and adequacy of
heating to acceptable levels, and the elimination of
cold bridges. Other main sources of complaint are
concerned with noise and sound, security and the
external environment of estates, but these are beyond
the scope of the present report.

The effectiveness and costs of remedies to habitability

problems cannot be considered in isolation since they
are likely to be affected by work which has t o be
undertaken to improve weathertightness, fire
performance or structural adequacy. For example,
major repairs and weatherproofing work to the
external face of an LPS building may provide an
opportunity t o improve the insulation, which may not
otherwise be economically viable. Similarly, the
replacement of windows provides an opportunity to
improve thermal performance and weather sealing.
Improving the habitability of LPS dwellings should be
considered as part of the overall plan of remedial
measures. BRE publications have already been issued
on many of these aspects (p ii).

Each chapter of this report deals with a different

aspect. Chapter 1 provides an overall framework for
dealing with problems of dampness, mould and
heating. It identifies the factors contributing to their
solulion, and the way in which they interact.
Following chapters deal with each factor in turn -
heating systems, ventilation and thermal insulation -
and describe the results of BRE investigations. The
final chapter describes two case studies in LPS
dwellings.
Chapter 1
The habitability of LPS dwellings

Complaints to local authorities about habitability The SDD asked local authorities to classify the
or internal living conditions in large panel system severity of condensation problems into the following
(LPS) dwellings are typically concerned with three categories:
dampness, mould growth and inadequate heating.
This report is concerned with providing an Grade I (Least severe of the three grades)
understanding of these problems, and thus defining Minimal mould growth on ceilings, and on walls
a framework within which investigations of specific behind stationary items such as beds and furniture
problems can be carried out. It provides a rational
basis for making decisions about effective and Low or zero moisture meter readings on wall
economic remedial measures, and ones which are surfaces and at ceiling angles
likely to be acceptable to the occupants. Conditions are inconvenient but tolerable
This chapter provides a background to the report.
Grade 2
It identifies the causes of these problems and
outlines the factors which have to be taken into Persistent patches of mould growth on walls and
account when searching for cost-effective solutions. ceilings
The rest of the report gives a detailed description
Traces of mildew on soft furnishings and clothing
of these factors, the present condition of LPS
buildings; and what remedial measures are Medium moisture meter readings on wall surfaces
available. It is assumed that building owners, when and at ceiling angles
considering remedial measures for problems of
Redecoration is necessary at the end of each
habitability, will at the same time be reviewing
heating season
other aspects of performance, such as structural
stability and weathertightness. Remedial measures Prevailing conditions are causing discomfort
here will often lead to opportunities to improve
habitability: eg if overcladding has to be Grade 3 (Most severe of the three grades)
undertaken to overcome rain penetration problems, Extensive mould growth on walls and ceilings,
it is cost-effective to undertake external insulation
wallpaper falling off and paintwork peeling
at the same time, thus making the dwellings easier
to heat. Soft furnishings and clothing damp and coated
with mildew
High moisture meter readings within plaster/
Dampness and mould growth plasterboard on walls and at ceiling angles
The scale of the problem
Most surveys of dampness complaints which have Redecoration is necessary both during and at the
been undertaken do not identify LPS dwellings as a end of each heating season
particular group. However the Scottish Prevailing conditions constitute a health hazard
Development Department's (SDD) Housing Plan and are causing distress
Returns do identify LPS dwellings, and
information on condensation in these dwellings can The data indicate that the assessment of the severity
be obtained. These Returns from the Scottish local of the problems in affected LPS dwellings is very
authorities provide a guide of the proportion of a similar to that found in the total stock of housing
local authority's stock that is suffering from affected by condensation. Of all affected LPS
condensation: they also identify the severity of properties 55% are classified as being of grade 1, 37%
dampness or mould growth. as grade 2, and 7% as grade 3 - this compares with
54%, 36% and 10%. respectively, for all affected
The data returned to the SDD, based on dwellings.
inspections by 12 local authorities and an analysis
of tenants' complaints from a further 21, indicate There is evidence to indicate that the condensation
that on average an LPS dwelling is nearly 40% problem throughout the rest of the United Kingdom is
more likely to suffer from condensation than is any as severe as that in Scotland. On this assumption it is
other type of local authority dwelling. Whereas it is therefore likely that some 39 000 LPS dwellings are
estimated that 27% of the local authority housing suffering from condensation and that of these nearly
stock in Scotland suffers to various extents from 2750 are in such a condition that they may constitute
condensation, the figure is 37% for dwellings of a health hazard and are causing distress to the
LPS construction. occupants.
Investigating dampness water. Temperatures within dwellings are suitable for
Complaints which are made to local authorities about mould growth and in addition there will be adequate
the habitability of LPS dwellings are likely to be nutrition and oxygen: the principal factor controlling
concerned with dampness and mould growth or with mould growth in dwellings is therefore moisture. It is
the expense of beating. If the complaint is one of possible that mould fungi can obtain sufficient
dampness, then the local authority needs to identify moisture without liquid water being present. Studies
whether the cause is rain penetration or condensation. have indicated that moulds may develop on surfaces
If the problem is one of rain penetration, then the when relative humidities at the surface are about
solutions may involve replacing windows, 85%. As a general rule of thumb, if the ambient
overcladding walls or over-roofing. In these cases, as relative humidity within an LPS dwelling remains
in most remedial work, it is likely to be cost-effective above 70% for a reasonable period, conditions are
to improve the insulation standard of the building conducive for mould development.
fabric at the same time. This will reduce the heat
losses from the dwelling, improve the internal One of the criteria adopted for ensuring that LPS
environment and reduce the risk of surface dwellings are habitable is that mould growth will not
condensation and mould growth. develop on internal surfaces. Whereas many occupiers
will tolerate quite high levels of dampness, especially
If the complaint to the local authority is primarily if intermittent, fewer will tolerate mould growths on
concerned with mould growth, the basic cause for the walls, furnishings or clothing. Mould growth presents
development of mould needs to be determined and problems because of its appearance, the associated
corrected. In any event the local authority needs to musty smell, and fears for an adverse effect on
ensure the safe removal of mould growths. BRE has health, particularly that of children. Thus there are
undertaken tests to assess the likely performance of strong incentives for at least adopting a cosmetic
anti-fungal paints and anti-condensation paints which approach or removing or masking the mould growth
include a fungicide. A list of products which have itself even if the primary cause of the condensation
been tested in BRE's 'high condensation' facility and cannot be resolved.
have remained free of mould for a long period is
given in Table 1.
Factors affecting condensation
Table 1 Effective fungicidal paints
In order to determine what remedial measures might
Trade name Manufacturer
be most cost-effective it is useful to understand how
condensation occurs. Condensation is a physical
Emulsion paints phenomenon which occurs whenever the temperature
Biocheck Mould Growth Consultants Ltd of air is reduced to a critical temperature (the
Clear Mould The Paint Factory Ltd dewpoint) or the moisture content of the air is
Fungi-Chek Bio-Kil Chemicals Ltd
Fungi-Shield Glixstone Ltd
increased until saturation actually occurs. The
Mandaval Mandoval Coatings Ltd relationship between moisture content in the air, air
Masotex/Aquagloss/Aquasheen Joseph Mason plc temperature and relative humidity is illustrated in
Steracoat Signpost Paints Figure 1. In order to reduce the risk of surface
Sleracryl Donald Macpherson & Co Ltd condensation and mould growth, temperatures within.
Steridex Liquid Plastics Ltd
an LPS dwelling should be increased (horizontal
Antisondensstion paints arrow) or the moisture content of the internal air
Aquarid Anglia Ins & Roofing Supplies reduced (vertical arrow). In practical terms the
Ltd occurrence of condensation is therefore dependent
Steracote AntiLCondensation upon four factors, namely:
Compound Signpost Paints
Therma-Shield Glixtone Ltd heating,
thermal insulation,
ventilation, and
If the mould has developed as a result of persistent moisture generation.
condensation or high relative humidities, the local
authority needs to determine the most cost-effective Heating
remedy. In the past there has been some reluctance to Surveys of fuel consumption in local authority
spend large amounts of money on remedies which dwellings have indicated a very large spread in the
may not be totally successful, and the BRE studies energy usage in dwellings which have nominally the
concerning the habitability of LPS dwellings have same heat loss. Figure 2 shows the distribution of
been endeavouring to assess the most effective way electricity usage in a scheme of Scottish local
forward. authority housing. Even allowing for the extremes of
the distribution there is a spread of 4:l between low
and high fuel users. In order to be successful,
Factors affecting mould growth remedial measures need to be successful for most of
The basic requirements for the development of mould the fuel usage and in particular for fuel users at the
growths are a suitable temperature, food, oxygen and lower end of the spectrum. The problem is
 Graphs of relative humldlty
100% 90% 80% 70% 60% 50%

1 5 10 15 20
Air temperature j°C/

Figure 1 Relationship between air temperature, moisture content

and relative humidity

x lo3 k w h
Figure 2 Distribution of fuel usage in scheme of electric warm air houses

compounded in LPS dwellings since, for structural circumstances background heating under landlord
safety reasons, gas may not be used. control might prove useful, but any such proposal
would require discussions between occupant and
As indicated previously, the aim of remedial measures landlord.
should be to provide internal temperatures that avoid
mould growth at an energy cost which is affordable Thermal insulation
by the occupant. Once the amount of heating that is A number of LPS dwellings are insulated to a
required to avoid mould growth is determined, an reasonable standard compared with the building
assessment needs to be made of the fuel cost and of regulations in force at the time they were built. The
its affordability to the occupant. The Building U-value for walls for some LPS dwellings is about I .O
Research Establishment Domestic Energy Model W/m2K (this compares with a requirement of 1.7
(BREDEM)' provides such a means of assessment. W/m2K in the building regulations enforced at the
time). However, from complaints received by local
As well as the cost for fuel, consideration should also authorities it appears that, because of the type of
be given to the methods of payment. In certain heating used, a number of LPS dwellings are not
 Heat input 4 kW
Oufslde ,empersture O°C
Outside RH 9 0 %

0
0 1 2 3 4 5 6
Ventilaf8on rate lair changes per hour]

Figure 3 Relationship between insulation standards, ventilation and

relative humidity

heated to a standard which will avoid condensation In general, improving thermal insulation will not only
and mould growth. Consideration should therefore be reduce the relative humidity within a dwelling, there
given to improving the insulation standards of the may also be some fuel saving. Figure 4 plots the
walls. This is particularly true if other remedial works annual fuel consumption against the fabric
are being done, such as weather-screening which transmittance for about 1500 local authority
makes the installation of insulation very cost-effective. dwellings. For the 25% of users using the lowest
amount of fuel Oowest 25%), improving the
Figure 3 indicates the effect of insulation on the insulation of the dwellings is unlikely to result in a
relative humidity within dwellings for different fuel saving but will improve internal conditions. For
ventilation rates. Improving insulation implies that the users using more fuel than that used by the lowest
dwelling will be warmer and consequently the relative 25070, there is also likely to be a saving in fuel.
humidity will be reduced. The figure indicates that
with a ventilation rate of about 1 air change per hour Methods of insulating LPS dwellings are by either
(ach), a flat built to the regulatory requirements at the internal or external insulation and a decision has to be
time most LPS dwellings were built would have a made on the effectiveness and cost of different
relative humidity of about 70% - the value at which systems. It must be borne in mind that, in general,
mould growth is likely to develop on internal surfaces. external insulation is up to five times more expensive
Insulating the flat to a standard required by present than internal insulation. However, if other works are
building regulations would reduce the relative undertaken at the same time, external insulation may
humidity to about 40% - well clear of the risk of be cost-effective since the additional cost may be only
mould growth. the cost of the insulating material.

25 -
-
2
",
z- 2 0 -
c
.+
0

15-
0
-
z
m

-
1 10
m
3
-
0
0 n 0
Lowest 25%
-
c
z 5 -
0 Lower, 10%

0 10 I
100
I
200 300
I
400
I
Fabric rransmlrrancs IZAU W/K/

Figure 4 'Useful' energy supplied to 1500 Scottish local authority houses

Again, when deciding which method of insulation In order to remove moisture from dwellings two types
should be used, consideration should be given to how of ventilation rate are required: high rates of
the dwellings are likely to be used. With certain types ventilation in kitchens and bathrooms where most of
of occupancy, internal insulation may be the preferred the moisture is produced, and background ventilation
system since this provides a thermal responsive throughout the rest of the dwelling.
structure.
Figure 5 indicates the ventilation rates required in
Whether external insulation or internal insulation is order to prevent condensation occurring on single-
chosen, decisions will have to be made about the glazed kitchen windows. During cooking the
extent of the areas to be insulated. For example, ventilation rate required is of the order of 25 ach and
insulation is unlikely to be cost-effective if the whole the only reasonable means of achieving this is by
exterior of a block of flats which has many projecting mechanical extract fans.
fins and balconies is insulated. However, in the same
block it may he cost-effective to insulate plane areas There are advantages if mechanical extract fans are
such as gables or the walls adjacent to stairwells, controlled by humidistats. There are also advantages
which are likely to be unobstructed. if background ventilators are controllable by the
tenant, otherwise they are likely to be blocked off and
Similarly, decisions on where to insulate need also to not serve the purpose for which they were installed.
be based on the relative heat losses from building
elements, eg a large proportion of the heat loss may Individual measures are not likely to be introduced
be through single glazing in which case consideration for individual dwellings; consequently serious
should be given to replacing some glazed areas with consideration is needed about ventilation remedies
insulated panels. which are applicable to high-rise blocks of flats. This
is because ventilation rates can vary by a factor of 10
Ventilation between ground-floor and top-floor flats.
Ventilation is a critical factor in the thermal
performance of LPS dwellings. Figure 3 indicates that Moisture generation
at both very low and very high ventilation rates the The production of moisture within the dwelling is
risk of mould growth is high (ie relative humidity completely outside the designer's control and is
above 70%). At low ventilation rates the risk of dependent on the living habits of the occupants. Table
mould growth is caused by the non-removal of the 2 indicates sources and rates of moisture generation
water vapour, and at high ventilation rates it is caused within a dwelling. The main source of moisture is the
by the reduction in temperature within the dwelling. drying of clothes indoors and the obvious places
where this might occur are kitchens and bathrooms.
When considering remedial measures, local authorities For this reason consideration should be given to
need to determine whether the ventilation rates within providing means of removing this moisture at source.
the dwellings are high or low. In general an acceptable
ventilation rate is between 0.5 and 1.5 ach. One Advice should be provided to the occupant on means
means of determining an assessment of the ventilation of reducing moisture generation and of reducing the
rate is by pressurisation testing, which is quick and risk of condensation. A useful document is the DOE
relatively easy. leaflet 'Keep your home free from mould and damp'2.

10 - Gar

0.8 -
-
.L
m
rn
-
2 - Ele~fll~lt"

0.6 -
c
e
"=
n
em
-,
-
04
= H e a l lnpuf 1 kW
..
m
0
I

Figure 5 Ventilation rale required to prevent condensation on

0 5 10 15 20 25 30 kitchen windows (the arrows indicate typical values
venf8laton rate (a,, changes per hour) when the cooling fuel is gas or eleclricity)
Table 2 Typical moisture generation rates in a five- Conclusions
person dwelling
In order to be certain that LPS dwellings are
Amounts habitable, the aim is to ensure that mould growth
Sources of moisture Oitres/day) does not occur and the dwellings are heated to a
reasonable standard at a cost which is affordable by
Maim sources the occupant. In general the following points should
5 people asleep for 8 hours
5 people active for 16 hours
be considered when determining remedial measures.
Cooking
Bathing, dish-washing Heating
Sub-total 7.2 a The system must be capable of providing a
reasonable temperature at a reasonable cost
Additional sources
Washing clothes 0.5 a BREDEM' is a means of assessing heating costs
Drying clothes 5.0
Paraffin heater 1.7
Thermal insulation
Sub-total 7.2
Thermal insulation reduces the energy requirement
Combined total 14.4
Decisions on the means of insulating dwellings
(internal versus external insulation) depend upon
other works which might be undertaken
Energy and capital costs a Areas to insulate depend on relative heat losses
Any remedial measure must be effective at a cost from elements and the cost and ease of insulating
which is affordable by the occupant. Figure 6 them
indicates the cost of remedial measures in a typical
LPS dwelling with a heat input of 2 kW, moisture Ventilation
generation of 7 litres per day and a ventilation rate of
1 ach. In this typical flat the relative humidity would
a Pressurisation tests should be considered to obtain
an indication of the ventilation within dwellings
be about 80% and there would be a considerable risk
of condensation and mould growth. In order to a Provision of ventilation should permit the rapid
remove the risk of mould growth, the relative removal of moisture from kitchens and bathrooms
humidity should be reduced to 70% either by
Background ventilation should be provided to
insulation at a cost of £3000-£4000 or by the
living areas
occupant spending more on additional heating (about
£60 - f 150 per annum depending upon fuel and Remedial measures should be acceptable in high-
tariff). rise as well as in low-rise dwellings

Graphs of
relaf~vehumidify
5 - Outhtde tempersture = 3OC
Outride vapour pressure = 7 m b
Moisture generation = 7 litreslday
Venttiatton rate = 1 air change per hour

4 -

z:3 -
-
3
m
c
.
a
I
2 -
E 3000-E 4000

0 50 100 150 200 250 300

Fabric tiansmlttance (WIK]

Figure 6 The cost of remedial measures needed to reduce relative humidity, eithe. by insulation
or by heating, in a typical LPS flat
Chapter 2
Heating systems for LPS dwellings

Introduction
It has long been recognised that a heating system must expenditure is not an overriding constraint, ie typical
be adequately sized, ie it must be capable of achieving conditions in UK housing. This would suggest living-
design temperatures indoors when the outdoor room temperatures of 2 l 0 C during heated periods,
temperature drops to an appropriate design figure for allowing temperatures to drift downwards overnight
the location. This does not, however, by itself as is customary for the great majority of UK
guarantee that the system will be considered adequate dwellings. The rest of the dwelling might be assumed
by the occupants of the dwelling. It must also be to be about 2-3°C cooler than the living-room.
capable of providing heating at a cost which they Higher standards might be appropriate in special
consider reasonable. If it cannot, it is likely to be circumstances, such as housing for the elderly or
under used or simply ignored and some other means infirm.
of heating used in its place. Many local authorities
have properties which are otherwise sound but are There is then the question of how to calculate annual
unpopular because they are hard to heat, and the heating costs. The Building Research Establishment
problem is at least partly due to inadequate heating Domestic Energy Model (BREDEM)' has been
systems. extensively tested against results obtained from
monitoring occupied buildings. It is the technical basis
Large panel system (LPS) dwellings are, in this for a number of commercially available computer
respect, no different from many other types of programs and for the British Standards Institution's
dwelling, often being better insulated than some older code of practice on energy efficiency in housing3 and
properties. That is not to underestimate the its accompanying designer's manual' which sets out a
significance of problems with heating systems, step-by-step procedure for its use. Whilst the results
however, since many of the LPS blocks were built at obtained from a BREDEM calculation are unlikely to
a time when it was assumed that electricity prices match actual consumptions in individual cases very
would not prove to be too high to be afforded for precisely, they are considered to be a reliable
space heating, even at on-peak rates. In many cases, indication of relative heating costs of dwellings. Local
structural safety considerations rule out individual gas experience may be useful in indicating how to
supplies in multi-storey blocks, so the range of interpret results in terms of what tenants perceive and,
possibilities for replacement of systems is restricted. at worst, such calculations should be capable of
This chapter considers how the adequacy of heating examining design options by comparison with other
systems might be assessed more realistically than by similar cases in the locality.
simple sizing considerations, and discusses the
practical options for improving heating systems. The steps to be followed in order to assess a heating
system are:
1 Calculate heat loss and check sizing
Assessing adequacy 2 Calculate annual energy requirements, using
A proper assessment of a heating system must take BREDEM
into account how much it costs to run. This obviously 3 Determine a 'reasonable' level of fuel expenditure
requires assumptions to be made about indoor for the likely tenant group and compare with the
temperatures and occupancy levels and it may rightly calculated energy cost
be noted that these are notoriously variable, as
demonstrated by field trial results. This variability If a favourable comparison is not obtained,
should not be allowed to obscure the overall improvements will have to be sought in either the
objective. What is needed is an assessment of how heating system or the level of thermal insulation of
much it costs to heat a dwelling to a reasonable the dwelling. In cases where both aspects have been
standard for a full occupancy schedule, since it is for improved to their economic limit, the only option may
those who want to heat to such a standard that the be to manage the estate in a way that minimises the
problem is of greatest relevance. effects of the problem.

The question of what constitutes a reasonable

temperature is a difficult one. Many local authority Some examples of running cost
dwellings were designed originally to Parker Morris calculations
(or revised Parker Morris) standards of heating which
may now be considered to be too low, especially in Table 3 shows running costs calculated using
bedrooms. It is perhaps more reasonable to base BREDEM for a two-bedroomed flat with different
standards on what is practised in homes in which fuel levels of heat loss. The result shows that space heating
Table 3 Running costs calculaled for typical flats

The example below is a two-bedroomed flat with 25 m 2 of external wall and a total floor area of 61 ma. Total energy costs
are calculated assuming that exprndilurr on starldir~gcharges and non-space-heating uses ir £17O/year.

Unimproved.
U-value of
r o d and As Icfl, but As left, but also
walls = 1. with roof double glared and
single glazed U-value = 0.35 walls insulsted
~ ~-

Modern electric Space heating

storage heating, conaumpliorl (kWh/y) 5200 4300 3300
whole tlaf Total energy cost (L/wk) £5.75 £5.30 E4.90
heated mrlrning Averagc temperature
and evening in Zunr Ze("C) 18.5 18.7 19.0

On-peak electricity, Spacc heating

only living-room hratcd consumption (kWh/y) 2900 2450 2000
morning and evening Total energy cost (£/wk) fh.110 f5.60 £5.20
Average temperature
in Zonc 2' ("C) 14.5 15.2 15.9

I f the fiat is on the lop flour, hrat l u ~ a e sthrrlugh the roof may bc very important. The following values are for whole-tlat
heating as above but for a flat on the top floor.
~ -

Unimproved.
U-value of
rouf and As lufl, but As lefl, but also
walls = 1, with roof double glared and
single glazed (1-value = 0.35 walls insulated
--

Modern eleclric Space heating

storagc hcating, consumption (kWh/y) 9400 6500 46W
whole flaf heated Total energy cost (f/wk) L7.80 £6.40 £5.50
morning and evening Avcragc temperature
in Zone 2* ("C) 17.8 18.1 18.6

* Zone 2 is the remainder of the dwelling when the living-room (Zone I) is excluded.

energy consumption can be considerably reduced by value for money obtained from an inadequate heating
insulating external walls and by installing double system.
glazing. For a top-floor flat, roof insulation is also
very important in this respect and is normally able to A further comparison with gas heating (not shown in
be installed at low cost. Other geometric effects, such Table 3) showed that for all the levels of insulation
as a design with two or three outside walls, may also considered, gas heating produced lower fuel running
contribute to higher heat losses. costs. However, any such savings were more than
offset by the extra standing charges and the cost of
The costs are calculated for modern electric storage maintaining gas heating equipment. This is generally
heating and are shown to be reasonable whilst true for a dwelling with a total specific heat loss of
maintaining a good standard of heating, especially in less than 200 W/K, which includes a large proportion
the better-insulated cases. In calculating total energy of purpose-built flats.
costs, fl70/year has been allowed for water heating,
lighting, appliances, cooking and standing charges. In considering heat loss, it has been assumed that
This figure exceeds space heating costs substantially adjacent flats have been heated to a similar standard.
for the better-insulated cases with electric storage If a tenant is unfortunate enough to be surrounded by
heating. unheated flats, his own heat losses may be increased
substantially, perhaps doubling in extreme cases, and
For comparison, a second case is calculated in which fuel costs increased accordingly.
only the living-room is heated using on-peak
electricity, a situation which is encountered all too
frequently in practice. This shows both higher energy Landlord control of heating systems
costs and much lower standards of beating (as In blocks of flats, there is an inherent dilemma for
indicated by the average temperature in Zone 2) than the landlord regarding the control of heating. A good
for the case with storage heating. Such low standard of heating which is good for the property
temperatures are likely to lead to problems with and, arguably, good for the tenant may only result if
condensation and mould growth and indicate the poor the landlord exercises a degree of control over the
heat supplied to the dwelling. This also ensures that fabric of the dwelling in relation to its floor area. A
'heat robbing' from adjacent flats does not take value of 1.5 W/K per square metre of floor area can
place. On the other hand, many tenants prefer be considered to be well insulated and suitable for this
individual control of heat output and, with it, form of heating, whilst figures in excess of 2.5
individual control of costs. The local authority policy indicate that careful consideration should be given to
on this issue will determine whether 'heat with rent' further insulation.
or individual metering is favoured. In either case,
running costs are important to the success of the In 1978 the Joint Working Party on Heating and
solution adopted. Heat with rent breaks the link Energy Conservation in Public Sector Housing
between individual consumption and payment and so produced a note (DEN 3)5 on remedial work for
leads to high usage, and hence high rent charges if the existing electrically heated dwellings which contains
system is expensive to run. Equally, with individual much useful material. A later publication by the
metering, high running costs lead to problems since Electricity Council, The DEN 3 book6, interprets the
widespread underheating is likely to occur if the original publication in terms of currently available
tenants see the heating system as giving bad value for products. Overall, the DEN 3 approach continues to
money. be a sound basis for improving electric heating
systems.
Some authorities have experimented with a system in
which only background heating is supplied with rent
and the control of comfort is left to the tenant. This Some examples of heating system
system has attractions since it prevents gross
underheating but still leaves an important element of
improvements
control with the tenant. The problem is that such a Individually metered electric storage systems
system can add considerably to technical complexity The city of Newcastle upon Tyne has taken a very
and hence to capital costs, particularly if it is based active role in improving the energy efficiency of its
on a communal heat supply, and if it requires heat housing stock. The Westgate Hill flats are a well
metering. publicised example of such improvements. External
walls were insulated with dry lining, secondary double
glazing fitted in the living-rooms and electric storage
Practical considerations radiators fitted in every room. Water heating circuits
were fitted with Economy 7 controllers and each flat
Natural-gas supply to individual flats is ruled out in was changed to Economy 7 tariff.
many blocks of flats for structural safety reasons. The
choice of heating system is therefore constrained in The level of fabric insulation achieved was equivalent
practice to either a communal heat supply or a system to the 1.5 W/mZK standard referred to above for the
using off-peak electricity. Communal supply may well two-bedroomed flats and rather better for the one-
provide the lowest cost per unit of heat delivered but bedroomed flats. After the improvements, electricity
is likely to involve a very high capital cost, especially consumption rose by about 6% while expenditure fell
when retrofitting an existing block. Therefore its use by an average of £83/year. A smaller group of flats
is likely to be restricted to special circumstances such which had the heating system improved but no
as the proximity of an existing heat supply with spare additional insulation, showed an increase in
capacity. consumption of 18% and a reduction in expenditure
In many cases, electric heating will be the only viable of £55/year. Overall, there was a clear benefit to the
tenants in reduced fuel expenditure but only a modest
option. The public perception of electric heating as
increase in energy used and, by implication, the
expensive to run is largely based on obsolete systems
standard of heating. Individual results showed a
installed in badly insulated dwellings. Modern storage
significant minority of tenants using very little off-
radiators, correctly sized and operating on current
peak electricity so it was clear that underheating was
off-peak tariffs, offer a much lower running cost than
not eliminated.
their older counterparts and are incomparably better
than systems which rely entirely on on-peak
electricity. It is important, however, that they are Heat with rent
Glasgow District Council has undertaken
installed in dwellings that are well insulated since the
improvements to a number of blocks of flats by
rapid cooling down of a badly insulated dwelling will
put high demands on the storage capacity of the units installing modern storage radiators and upgrading
and lead to excessive needs for additional on-peak thermal insulation. Data were obtained on two Reema
blocks, one of which was modified and operated on a
heating during the evening period.
heat-with-rent basis and one which was not. The
The best basis for deciding how well insulated a U-value of the walls even before improvement was
dwelling should be to make it suitable for electric relatively good and in the modified block was about
1.1 W/mZK. Direct comparisons between the blocks
storage heating is to calculate the running costs, as
described above. As a first indicator, however, it may could be misleading since the insulation levels differed
as well as the method of payment. It was clear,
be instructive to look at the specific heat loss of the
however, that temperatures were markedly higher (2 Cyclo control
o r 3°C) in the heat-with-rent group and underheating Cyclo control is the name given to a system that uses
was eliminated in that block, except for two flats the mains wiring to send signals for controlling
where the tenants were allowed to opt out of the heating equipment from a single control point located
scheme. This was achieved at a cost that most tenants nearby. Typically it is used to control the charging of
could afford so the scheme must be judged to be very storage radiators in groups of 100 or more flats. Each
successful. The only difficulty is that such high levels flat may be allocated its own control code and receive
of insulation will not be easy to achieve in all other a signal to match output to its heat loss as well as
cases: if all flats were so well insulated, heating taking into account the prevailing outdoor
problems would not he a big issue. In many cases, the temperature. This system has been installed in about
cost of achieving such a high standard of insulation 20 000 dwellings over the past 10 years, mostly in the
would he excessive and the resulting compromise London area. In all cases the system has been
would push up running costs and hence rent charges. operated on a heat-with-rent basis, and tenant control
Experience from older district heating schemes shows limited to the dampers on storage radiators. Cyclo
that high charges for heat with rent are very control is a well proven technology that enables good
unpopular with tenants. Overall, the conclusion must control of storage radiator systems and deserves wider
be that heat with rent is successful in eliminating application than at present.
underheating and can be popular with tenants if the
running costs are low.
Conclusions
Partial heat with rent
The Central Technical Unit of the GLC developed a 1 The running cost of a heating system is a key
system called 'Owner Based Resident Adjustable factor in determining whether it will be considered
Control' (OBRAC) which allows control over satisfactory by tenants and whether it will be fully
background heating by the authority but control of utilised. Good thermal insulation and a system
comfort level by the resident. A standing charge, paid which can produce useful heat at low unit cost are
with rent, covers the cost of the background heating both needed t o achieve that purpose.
while the remainder of the heat is charged directly to
the tenant based on readings from a heat meter 2 Running costs for typical occupancy patterns may
installed in each flat. Several such systems were be calculated using BREDEM'. Such calculations
installed, in each case operating from a gas-fired should be undertaken when heating systems are
communal boiler room. designed or renovated.

3 The design and replacement of heating systems in

The OBRAC system has some attractions, combining
LPS dwellings does not present any problems
the best features of heat with rent and individual
unique t o that form of construction. However,
charging, but is inherently expensive. Not only is it
many LPS flats are in high-rise blocks where
necessary to provide a boiler room and heat
individual gas supplies are ruled out by structural
distribution system but also a wet central heating
system for each dwelling. The heat metering is also a safety considerations. In such cases off-peak
electric heating is likely to be the only practical
considerable extra expense, especially if it is combined
option. The DEN 3 guidelines5' are a good basis
with a remote management and control system to
for the design of replacement electric heating
record the readings and control the background heat.
systems.
The capital costs of OBRAC appear too high to make
4 The method of payment for heating will be
it viable for widespread application. There is no
determined by individual local authority policy.
reason, however, why the principle of partial heat
There are examples of successful schemes using
with rent cannot be implemented more simply if it is
heat with rent and similarly for individual
felt to be advantageous. For example, the authority
could make a charge with rent which was then used to metering. It may be possible to devise a scheme
which combines the best features of both methods
pay for the first (say) 3000 k w h of electricity used at
off-peak rates, directly t o the fuel supplier. Such an by charging for background heating only with rent.
approach would ensure that no tenant had any
financial pressure to switch off his storage radiators
and therefore background heating should be assured.
The remainder of the tenant's bill would be paid
directly, as metered. Such an approach would require
neither any extra equipment nor excessive
administrative effort. Alternatively, an unmetered
supply could be used for one or two radiators via a
separate circuit, but this would require additional
wiring and be open to tampering by tenants.
Chapter 3
Ventilation in high-rise LPS flats

Introduction
An adequate supply of fresh air in dwellings is changes per hour, and either one, two (adjacent or
essential for breathing, for controlling indoor parallel) or three exposed walls.
pollutants to acceptable levels and for ensuring the
safe and efficient operation of combustion appliances. In addition t o the analytic studies, some field
Good ventilation is also a key factor in preventing measurements of air infiltration rates were made in a
condensation and mould, a significant problem in flat of Reema construction (Chapter 6), using the
many dwellings in the UK, including those of large technique of tracer gas decay. These measurements
panel system (LPS) construction. However, if provided useful confirmation of the trends identified
ventilation rates are too high, excessive fuel bills and in the theoretical studies.
occupant discomfort can result. In the provision of
ventilation, the aim is to strike a balance between the
need for fuel economy and the need to maintain the Natural ventilation
health and well being of the dwelling's occupants. Natural ventilation is driven by two mechanisms: the
wind effect, resulting from air moving around and
BRE's programme of research into ways of improving over the structure, and the stack effect, resulting from
the habitability of LPS dwellings has included studies, a difference in temperature, and hence density,
both in the field and using theoretical models, between the indoor and outdoor air. These two
designed to identify the nature of the ventilation mechanisms create pressure differences across
problem and to develop appropriate practical openings in the dwelling fabric, leading t o the
iolutions. The first stage of the ventilation exchange of fresh outdoor air with stale indoor air.
programme was a survey of air leakage rates under These openings are of two distinct types: firstly, those
artificially induced pressure differences, in a sample which are provided specifically for the purpose of
of 87 LPS flats of different types of construction. ventilation, such as openable windows, airbricks and
This work is described in detail in Chapter 4. One of window-head ventilators, and secondly, adventitious
the important findings to emerge was that the flats gaps in the building fabric, such as gaps around
tested were relatively airtight compared with low-rise service outlets and around the openable and fixed
dwellings of traditional construction. However, it does parts of windows and doors.
not follow automatically that the ventilation rates
found in these flats under normal weather conditions The movement of air through the adventitious gaps is
will necessarily be low, because high-rise flats generally referred to as infiltration; it ensures a
experience considerably more severe exposure to the background level of fresh air entry, even when all
wind than low-rise dwellings in typical suburban purpose-provided openings are closed. The rate of air
environments. The significance of the air leakage infiltration is governed by the size, nature and
measurements can only be determined by a systematic position of the gaps and the pressures acting on them.
evaluation of all the additional factors which The purpose of carrying out the air leakage tests
influence ventilation, including exposure, weather and reported in Chapter 4 was to gather information on
occupants' behaviour. the effective area and position of adventitious
openings. This information is one of the key inputs to
The purpose of the second stage of the ventilation the mathematical model used to predict natural
studies was to determine likely natural ventilation ventilation rates.
rates in high-rise LPS flats, during the heating season,
taking all of the above factors into account.
Ventilation rates were studied as a function of wind Variations in ventilation rates
speed, wind direction and internal-external
temperature difference, using a mathematical model Most high-rise LPS blocks are at least 10 storeys high
of ventilation developed at BRE7. Assessments were and some are as high as 30 storeys, which means that
also made of the effects on ventilation rates of: the flats on the top floor can be well over 100 metres
height of the flat above ground level, the number of from the ground. Wind speed increases with
exposed walls, and various measures to improve increasing height from the ground, and for flats more
ventilation, including airbricks and extract fans. These than 50 m from the ground, wind speeds can be more
studies were based on typical LPS flats which had than double those experienced by low-rise dwellings
ventilation characteristics corresponding to the under the same weather conditions. This means that
average of the flats tested in the field studies. They the pressures exerted by the wind on flats at these
were assumed to have a volume of 150 m3, an air heights are at least four times greater than those on
leakage rate at 50 P a pressure difference of 8 air low-rise dwellings. Even on the lower levels of a high-
Figure 7 Air flow patterns around a tall building (A and B denole areas of high wind speed)

rise block, the wind environment can be hostile, as ventilation rate with changing weather conditions and
shown in Figure 7O. For these reasons, ventilation with changes in the built form, position and
rates in high-rise flats tend to be dominated by the orientation of the flat. Under given weather
effects of wind alone, rather than the combined conditions, there is likely to be considerable variation
effects of stack and wind which dominate in low-rise between ventilation rates in flats in the same block.
housing. However it is not practical to design ventilation for
individual flats, and the approach recommended to
Examination of the effects of wind speed on likely owners of high-rise LPS flats must be a compromise.
ventilation rate showed that even with a comparatively The underlying requirement is that the ventilation
high degree of airtightness, the ventilation rate in a provision should enable an adequate minimum level
flat on the top floor of a 14-storey block could of ventilation to be maintained at all times.
increase by a factor of 10. As the wind speed
increased from 'calm' (Beaufort Scale 0) to a 'fresh In many UK dwellings, air infiltration alone is relied
breeze' (Beaufort Scale S), the ventilation rate varied upon to provide ventilation for the major part of the
from less than 0.5 air changes per hour to over 4 air heating season, despite the lack of control by
changes per hour. occupants. For high-rise flats this is impracticable,
because of the sensiiivity to weather conditions and,
The effects of wind direction were also found to be in the case of LPS flats, the relative airtightness. A
significant, particularly for flats with only one or two high degree of airtightness is in fact almost essential
exposed walls. In a flat with only one exposed wall, for high-rise flats, since it helps to lessen the effects
the ventilation rate fell to very low levels when the of extremes of high winds and cold weather.
exposed wall was in the lee of the wind. For a flat
with two adjacent exposed walls and conditions of a The best way of providing ventilation during the
'fresh breeze', it was found that the ventilation rate heating season, especially in high-rise flats, is to
could vary by a factor of more than 4, depending on complement background air infiltration with finely
the direction of the wind in relation to the exposed controllable, purpose-provided ventilation, designed
walls. For the flat with two parallel exposed walls, the so that openings do not pose a security threat and do
variation was even larger. not cause draughts. Fine control is essential if
occupants are to use the ventilators to meet their
Variations in temperature difference between inside varying ventilation requirements, especially during
and outside were also investigated, to establish at windy weather. Unless ventilators are designed to
what point the stack effect becomes as important as avoid uncomfortable draughts, occupants will
wind effects. Even though temperature difference was probably close or even block the ventilators
varied between 3°C and 16°C. it was found that in all permanently. The role of the occupant is crucial in
flats but those on the first few storeys, this large ensuring the effectiveness of controllable ventilation.
variation in temperature difference had a relatively
small effect on ventilation rates. This confirmed the However, in the majority of LPS flats, openable
earlier indications that wind is the dominant influence windows and fixed ventilators are the only means of
in determining ventilation rates in most flats in high- providing additional ventilation. Windows are usually
rise blocks. very large and have poorly designed hardware,
resulting in high ventilation rates even when the
window is open to only the first stop. Whilst this is
Approach to ventilation essential for summertime cooling, it means that
The most important factor to emerge from the openable windows are not an appropriate means of
ventilation studies was the wide variation in providing ventilation during the heating season. Fixed
airbricks are frequently blocked by occupants, in an The types of ventilator available include:
attempt to prevent draughts and reduce heat losses. a Adjustable 'wind-driven' fans, set into walls or
Clearly there is room for improvement in the glazing
specification of purpose-provided ventilation.
Adjustable window-head ventilators
a Adjustable 'through-the-wall' ventilators
Practical solutions
It has already been stated that a high degree of For kitchens and bathrooms, further ventilation
airtightness is to be advised in high-rise flats, so that provision should be made to ensure the removal of
the effects of severe weather are lessened. This is of moisture at source. This can most effectively be
course subject to the proviso that additional provision achieved by the use of mechanical extract fans, sized
for ventilation is made, otherwise there will be many to provide an air flow rate equivalent to at least 3 air
days during the heating season when ventilation rates changes per hour when running. The fan can be
are inadequate. As a general rule for high-rise flats, controlled manually, by time switch or by humidistat
an air leakage rate, at 50 Pa pressure difference, of control. The latter is the preferred option, although it
more than about 10 air changes per hour may be can result in fans running unnecessarily during warm,
considered to be excessive, and consideration should humid summertime conditions. For the other two
then be given to the installation of routine modes of control, the fans should run during periods
draughtproofing measures. of moisture generation and for a short period
afterwards. The fan should not he positioned too
The difficulty for the building owner is in knowing closely to the other ventilation openings, otherwise
whether or not particular LPS flats are airtight. An 'short circuiting' can easily occur.
inspection of the major leakage paths, in particular
windows, will give an indication of the likely scope Some degree of education of the occupants will
for reducing leakage - if good-quality windows are usually he required to emphasise the importance of
in place and there are no obvious cracks or openings occupant control over ventilation, and to advise on
in the fabric, the BRE experience of LPS flats how ventilators should be used to meet varying
suggests that the flat is likely to he reasonably requirements. Evidence from a number of field
airtight. Where there are obvious cracks, or windows studies, including DOE'S Better Insulated Homes site
are poorly fitting, measures should be taken to at AbertridwrlU,shows that occupants WIN make good
improve airtightness. In cases where it is difficult to use of controllable ventilation, provided that it meets
assess likely airtightness, or where some confirmation the requirements set out above. Where appropriate,
of the visual assessment is needed, the ideal solution is particularly when new heating systems have been
to conduct an air leakage test, as described in installed, user education should be set in the context
reference 9, on a sample of dwellings. of the whole package of refurbishment measures.

Bearing in mind the variations in ventilation rate

within a block of flats, there is little merit in Conclusions
introducing permanently open ventilators, such as A relatively high degree of airtightness is an
airbricks, since their effect is simply to increase the advantage for high-rise flats, since it helps to limit the
overall leakage rate of the flat, with the likely effects of severe exposure to the wind. If flats are not
consequence of uncomfortable draughts and excessive airtight, the variations in natural ventilation rate,
ventilation during certain weather conditions and their which result from extremes of wind and cold weather,
consequent blocking-up by the occupants. are unacceptably large.

On the basis of BRE's research to date in LPS flats, it BRE's recommendations on ventilation provision in
is recommended that additional ventilation provision high-rise LPS flats, based on experience to date, are
with the following characteristics should be installed as follows.
in all rooms of LPS flats:
a Ventilators should provide a minimum of 400 mmz 1 Owners of high-rise LPS flats should assess the
free opening area (eg 20 mm by 20 mm) per m 2of degree of airtightness of a sample of flats,
room floor area preferably by means of a fan-pressurisation test as
described in Chapter 4, or by visual inspection of
Ventilators should permit fine control by occupants the more significant air leakage paths (windows,
The controls should be easily accessible doors, major gaps in the fabric).
a Ventilators should be positioned to avoid draughts 2 If the fan-pressurisation test shows an air leakage
Ventilators should provide a fine 'trickle' of air, rate at 50 Pa pressure difference in excess of 10 air
when in the nominally closed position changes per hour, or the visual inspection identifies
large gaps around windows or doors or in the
Ventilators must not jeopardise security
 fabric, install routine draughtproofing measures to
improve airtightness.

3 In all habitable rooms in all flats, install additional

controllable ventilation provision with the areas of
opening and required characteristics as listed in the
previous section.

In kitchens and bathrooms, install extract fans

sized to provide at least 3 air changes per hour
when running.

4 Educate occupants on how to use the controllable

ventilation to achieve comfortable conditions.
Chapter 4
Air leakage measurements in LPS flats

Introduction
Ventilation by natural infiltration, ie the entry and the ventilation of the dwelling, greatly. For example,
exit of air through a myriad of small adventitious LPS construction uses only concrete floors/ceilings.
openings in the structure, is an important factor in Concrete is, for our purposes, impermeable to air so
controlling the indoor environment in all dwellings. It there is not the scope for air leakage which a
takes place in addition to natural ventilation through suspended timber floor/ceiling would provide.
intentional openings, such as airbricks, trickle vents, Furthermore, a junction between an LPS floor/ceiling
etc. and a wall panel is likely t o be very different from the
equivalent junction in a conventional house.
Total natural ventilation rates can be assessed in two
ways: directly, by tracer gas techniques, or indirectly,
by measuring air leakage rates under artificially Fan-pressurisation technique
induced pressure differences using the fan- A full description of the fan-pressurisation technique
pressurisation technique. Natural ventilation rates, used t o determine the air leakage characteristics of
measured directly by tracer gas methods, are dwellings is given in reference 9. The equipment used
profoundly affected by prevailing weather conditions. for the measurements was developed at BRE and
Such measurements must therefore be made over a comprises two principal components: a fan to blow
period of 2 weeks or more in order to cover a air into (pressurisation or positive pressure) or suck
representative range of weather conditions. The air out of (depressurisation or negative pressure) the
pressurisation method measures not ventilation rate dwelling, and an air-flow measuring device (Figure
but the air leakage characteristics of a building 8(a)). The complete unit has been calibrated for air
envelope. The measurements are not greatly affected flow rate in the BRE pressurisation calibration
by weather conditions and are quick and easy to carry chamber. When in use, the unit is connected by means
out, and it is possible to identify and quantify specific of a short flexible duct t o an expanding door panel
air leakage paths (through which air infiltration and which is sealed into a doorway in the envelope to be
ventilation take place). Air leakage data are useful in tested. A micromanometer measures the pressure
their own right, and they can also be used as one of difference which the fan generates across the envelope
several inputs to a mathematical model t o predict (Figure 8(b)). All windows and external doors are kept
likely ventilation rates under natural weather closed, and internal doors held open, when
conditions7. measurements are being made.

As part of the BRE's programme of research into the The air leakage characteristics of the dwelling are
habitability of large panel system (LPS) dwellings, a measured by noting the air flow rate and envelope
large number of pressurisation tests in LPS flats have pressure difference at each of a range of fan speeds.
been carried out. This chapter aims to present and Results are obtained for both positive and negative
interpret the air leakage data obtained from tests in 87 pressure by turning the pressurisation unit round.
flats in Glasgow and London, covering the following These two sets of results are corrected for
six common LPS systems: Taylor Woodrow-Anglian, indoor/outdoor temperature difference and then
Harley Haddow, Bison, Skarne, Reema and Tracoba. plotted graphically t o ensure that there are no
The implications of these data and estimates of likely irregularities and that the relationships between air
ventilation rates obtained from a mathematical model flow rate and pressure difference follow smooth
are reported in Chapter 3, which includes a discussion curves. Each curve may be fitted to a simple power-
on how to achieve an adequate level of ventilation in law function of the form:
LPS flats.

Ventilation paths in LPS buildings where: Q = air volume flow rate

The construction methods and materials used in LPS Ap = envelope pressure difference
buildings are very different from those used in
conventional dwellings, and one might expect the K = a constant whose value depends on
ventilation characteristics t o be different also. envelope size, envelope leakiness and
Although some of the variety of cracks through which the system of units used
air infiltrates and exfiltrates may have similar shapes n = an exponent which is independent of the
and dimensions, eg cracks between opening and fixed system of units used
window frames, the number and distribution of other
types of cracks in an LPS dwelling are likely t o The parameters K and n are useful in that they allow
influence its air leakage characteristics, and therefore the data to be expressed in a simple form, hut they do
 (a1 Unit showing air-flow measuring device
View on intake honeycomb
Side view (cut-away) [Honeycomb cut away to show
Direction of air flow anemometer head)

Intake Honeycomb Vane Duct Fan Fan Fan rotor

flow anemometer motor rotor safety grille
straightener head

Ib Schematic general arrangement of pressurisation unit installed in a building and shown depressurising

Key

1 Fan
2 Air-flow-measuring anemometer head
3 Honeycomb flow straightener
4 Mounting board
5 Building envelope
6 Pressure difference micromanometer
7 Pressure difference tube
4 8 Flexible duct

Flgure 8 BRE pressurisation fan unit

not represent any simple physical characteristics of the The contribution made by specific air leakage paths to
air leakage paths. Normally, the value of the exponent the total leakage can be assessed by using the
n is found to lie between 0.5 and 0.7. pressurisation technique with reductive sealing. The
envelope is first tested as found. Leakage paths of a
In the presentation of the results it is usual to specific type (eg gaps around opening window frames,
conform with Swedish practicell and t o quote the airbricks, cracks at wall/ceiling joints, or holes where
mean air leakage rate taken from the air leakage pipes or cables pass through the envelope) are then
curves at + 50 P a and - 50 P a pressure difference. sealed (with masking tape, modelling clay and
Division of this figure by the volume of the envelope Polythene) and the envelope tested again. The
tested gives a figure in units of air changes per hour reduction in air leakage due to the sealing is then
which can be used to compare the air leakage determined by difference. This process can be
performance of different dwellings. Note that this is repeated until all the principal air leakage paths have
not a ventilation air change rate - the pressure been sealed, the remaining air leakage being termed
differences which drive natural ventilation are much 'background'. In this way the relative importance of
less than 50 P a and they act both positively and different air leakage paths is measured.
negatively on different parts of the building at the
same time. In semi-detached, terraced and flat-type dwellings,
some of the measured leakage may not be to or from Results
outside but to or from adjoining dwellings. This can The results of the air leakage measurements made in
be overcome by carrying out additional tests in which all 87 flats are presented in the form of a stacked bar
each adjoining dwelling in turn is pressurised (or chart in Figure 9. The overall mean air leakage rate at
depressurised, as appropriate) to the same level as that 50 Pa pressure difference is 7.3 air changes per hour
being tested such that there is no pressure difference (ach), but there are differences in mean air leakage
across the party wall or floor (or compartment wall or rates between system types. The mean air leakage
floor). The cross-leakage rates to the adjacent rates at 50 Pa applied pressure difference by system
dwellings may then be established by subtraction of type are: Taylor Woodrow-Anglian (TWA), 6.1 ach;
the retest results from the original simple test results. Harley Haddow, 8.9 ach (excluding the flat with
In practice this is not easy. poorly fitting windows); Bison, 7.5 ach; Skarne, 9.4
ach; Reema, 6.3 ach; Tracoba, 3.8 ach.

Air leakage rates in LPS flats These results may be compared with a sample of 284
Flats tested low-rise houses of traditional construction - mostly
Air leakage measurements were carried out in 87 LPS semi-detached and terraced houses. The distribution
flats of six different systems. The breakdown of the of air leakage rates for the sample of low-rise housing
sample is: Taylor Woodrow-Anglian, 12 flats; Harley is given in Figure 10(a) and that for the LPS flats in
Haddow, 6; Bison, 20; Skarne, 21; Reema, 18; Figure 10(b). The horizontal axes of both bar charts
Tracoba, 10. The sample included 30 one-bedroomed are to the same scale to illustrate the relative
flats, 50 two-bedroomed flats and 7 three-bedroomed airtightness of the LPS flats.
flats. Most of these were true flats occupying only one
floor, but in nine flats the front door was on the floor The mean air leakage rate for the low-rise dwellings
below the flat itself. Another nine dwellings were split was 14.7 ach, ie about twice the corresponding figure
between two floors and are thus more correctly called for the LPS flats. Furthermore, the mean air leakage
maisonettes, but for convenience are referred to here rates for each individual system were all below that
-- I l d l S .
a3
"-2- - - esuecially the Tracoba,
for the low-rise dwellings
Reema.. Tavlor
. Woodrow-Analian - and Bison svstems.
Many of the flats tested had had some modification, This suggests that low air leakage rates are a feature
such as replacement windows or the fitting of extract of LPS systems generally, and are not peculiar to
fans in kitchens and bathrooms. Only four had been particular system types.
totallv refurbished. these being- Tavlor
. Woodrow-
Anglian flats in London. One flat which was
narticularlv leakv was a Harlev Haddow two-
r
Reductive sealing measurements
bedroomed flat in which the windows were very Tests in a Harley Haddow flat
poorly fitting. Measurements, discussed later, showed Description of the flat tested
that the flat would have been much more airtight if The first air leakage measurements made by BRE in
the windows had been well fitting. an LPS flat were in a Harley Haddow block in

t%? Reem.: maan 6.3 ach I

~ t s o n mean - 7 5 ach
@ Skarne mean = 9 4 ach

1 Harley Haddow m e a n - 00 7 ach

Tracoba mean - 3 8 ach
l w A mean = 6 1 ach

Total number in sample 87 I

Figure 9 Distribution of air leakage rates in LPS flats, by systemtype

18
 la1 Traditional low-rise dwellings
40, 7

Ibl LPS flats

30
28 n - 87 flats
26

Figure 10 Distribution of air leakage rates in traditional and LPS

dwellings

Glasgow. This two-bedroomed flat was somewhat 5 Large holes in plasterboard, service duct boxing
unusual in that it was on two floor levels, but other and meter box sealed
features were typical of LPS flats generally, including 6 Principal cracks sealed
corridor access to the front door, a small private
balcony and deck access to a bedroom on the upper Note: (a) The holes in the plasterboard lining, on the
floor. external wall of the main bedroom, were
the result of an earlier structural
The windows were wood-framed with very poorly examination of the block.
fitting centre-pivot-type opening lights, daylight being
visible round many of them. Two windows had direct (b) Principal cracks were those at window
passages to the outside as much as 10 mm wide. No frame/wall joints, external wall/ceiliug
attempt had been made to draughtproof any of the joints, and poor wood joints on the bend
windows. of the stairs over the enclosed corridor. In
all cases there were visible gaps, not just
Test method hair-line cracks.
The pressurisation fan was set up in the front
doorway and a series of measurements were made Test results
with reductive sealing as follows: The results are presented in the form of a bar chart in
1 Flat as found Figure 11. This shows that the results for the negative
and positive pressure tests are similar, though not
2 All openable window frames sealed
identical. This is quite normal and is thought to occur
3 Two ventilation grilles sealed for two reasons: firstly, the physical size of some air
4 External doors sealed (not front door - occupied leakage paths may vary due to small movements of
by equipment) some building components (eg openable windows)
 F] Venfilaflon
grilles
grilles

w4
External doors
Exlernsl doors
Large
holes
Large holes 2.8%
10%

looH
x
Principal
- cracks
h

Background
33.2%

0 Negaflve Dre6SU.e

Figure 11 Distribution of air leakage in an LPS flat @y reductive sealing, at

50 Pa pressure difference)

under different pressure conditions; secondly, the Pa pressure difference. This would bring the flat well
aerodynamic characteristics of a complex flow path inside the range of air leakage rates measured in other
depend on which way the air is flowing through it. In Harley Iiaddow flats.
view of this, and the accuracy of the technique, there
is good agreement between the two sets of results. Other reductive sealing fesfs
Windows
The badly fitting windows accounted for some 30% Reductive sealing of the openable windows was
of the total air leakage of the flat, the ventilation carried out in another nine flats of four system types.
grilles for a further 10% and the two external doors The mean reduction in air leakage was 19%, with a
accounted for about 8%. Surprisingly, the large holes maximum of 45% and minimum of 6%. The mean air
in the plasterboard, service duct boxing and meter box leakage rate for these nine flats was reduced from 8.0
did not contribute significantly to the air leakage. The ach to 6.48 ach.
principal cracks accounted for 17% of the overall
leakage leaving the background leakage at about 30%. In practice, the fitting of good-quality replacement
windows with integral draughtproofing would not
Discussion necessarily achieve similar reductions in air leakage to
With the exception of the gaps in the woodwork on those given above for two reasons. Firstly,
the stairs and the large holes, as previously defined, draughtproofing does not usually make a perfect seal
all of the leakage paths sealed were on, or connected between fixed and moving frames. Secondly, the seal
with, the external walls of the flat. The results show between the new window frame and the opening in
that these leakage paths accounted for about 70% of the wall may be less airtight than what was achieved
the total air leakage. The importance of this is with the original window. Air leakage from both of
emphasised by the fact that only 28% of the surface these sources has been observed, with the aid of
area of the envelope was external, the remaining 72% smoke tubes, in refurbished flats.
being party or compartment walls, floors and ceilings.
The leakage paths connected with the external Trickle ventilators
surfaces of the envelope were clearly much more In three flats, measurements were made of the
significant than the paths between the flat and the rest contribution made by trickle ventilators to the overall
of the block. air leakage. In all three flats there was one trickle vent
per room, except for the living-room which had two.
It is interesting to note that all the air leakage paths The vents were built into the frames of replacement
sealed for these measurements could easily be sealed windows.
using readily available DIY materials. If this were
done the air leakage could, in principle, be reduced to In a refurbished flat the eight trickle vents provided
the background level, ie 6.3 air changes per hour at 50 21% of the overall leakage. In the other two flats the
,- ,-
Flat No 39 Flu No 40

Whole flat Whole flat

d 6 0 Flat No 35 i52 Flat No 36

Whole flat Zero Whole flat

; 6.5

;
i
Whole flat
Flat No 31

- ;T
Whole flat
Flat No 32

Flat No 27 Flat No 28

Whole flat Whole flat

'Not measured based on average

values excluding flats 28/24

Figure 12 Average air leakage rates at 50 Pa pressure difference expressed

as air changes per hour: indoar/autdoor figures derived from
whole-flat and party wall/floor results

six vents provided 31 % and 34% of the overall air measurements was carried out in eight adjacent flats
leakage. Clearly, trickle ventilators can be very using the method of balanced pressures across
important in the provision of permeable area for compartment walls and floors, as described
natural ventilation in these flats, but they do not previously. The block effectively consisted of three
necessarily provide a proportionate increase in natural columns. A central column contained the lifts, stairs
ventilation rate (see Chapter 3). and some services, whilst the outer two columns, one
each side of the central column, contained the flats,
two flats per column per floor. The columns were
Cross-leakage between LPS dwellings linked by enclosed hallways on every floor, and each
Measurements in a Taylor Woodrow-Anglian block flat (except those on the top floor) had adjacent flats
When a pressurisation test is carried out in a flat, above, below and to one side.
some of the measured air leakage is not to the outside
but to the adjacent flats above, below and to the side. More than 40 pressurisation tests were carried out,
Under natural conditions there is usually very little air including some repeat measurements, and the results
movement between flats, and that air which is were analysed to give a complete picture of air
exchanged is not fresh outside air. Therefore, if leakage in the top half of one column (Figure 12). NO
allowance is not made for the magnitude of cross- significant cross-leakage was found between adjacent
leakage, the natural (fresh air) ventilation rate flats on the same floor, ie across compartment walls,
predicted using the results from the pressurisation test but significant leakage was found to flats above and
may be overestimated. below, ie across compartment floors. The latter
amounted to about 10% of the 'whole flat' air
An unoccupied Taylor Woodrow-Anglian block in leakage rate for each floor, except between flat
east London provided a unique opportunity to numbers 24 and 28 where a hole was found in the
investigate cross-leakage between LPS flats. A floor around service pipe penetrations.
comprehensive programme of air leakage
If the pressurisation results are to be used to predict
natural ventilation rates, the cross-leakage should be
subtracted from the 'whole flat' air leakage to give an
'effective' air leakage rate (ie to the outside). The
effective air leakage rates illustrated in Figure 12 are
about 20% less than those for the whole flat.

Other cross-leakage measurements

Cross-leakage measurements were also made, on an
ad-hoc basis, in four other LPS systems, but only one
or two compartment walls or floors were tested in
each flat type so it would be wrong to draw firm
conclusions from the data. However, the results are
interesting: air leakage across compartment walls,
expressed as a percentage of whole-flat leakage,
ranged from zero (within experimental error) to
19.5%. while that across compartment floors ranged
from 4.3% to 25%. The average values were 8.3%
and 11.5% respectively for a sample of seven
compartment walls and eight compartment floors.

Clearly, these results show that cross-leakage can be

significant in flats and has important implications for
the interpretation of air leakage measurements in
multi-dwelling buildings.

Conclusions
1 The air leakage measurements reported here show
that LPS flats are generally much more airtight
than typical low-rise dwellings, the average air
leakage rate for the LPS flats being about half that
for typical low-rise dwellings. However this does
not necessarily imply that the flats are
underventilated.

2 In one flat, about two-thirds of the overall air

leakage was connected with the external walls and
about half of that was due to very badly fitting
windows.

3 In a sample of nine flats, air leakage around

openable windows was found to represent between
6% and 45% of the overall air leakage, the mean
value being 19%.

4 The fitting of trickle ventilators can provide a

substantial increase in air leakage, but this could be
offset by the improved airtightness of properly
installed, well draughtproofed, replacement
windows.

5 Cross-leakage between adjacent flats was found to

be significant in many, though not all, flats. This is
important when air leakage results are used in the
prediction of natural ventilation rates,
Chapter 5
Improving thermal insulation of LPS dwellings

Introduction
Increased thermal insulation can provide benefits insulation needed to achieve U-values of 0.6 or 0.45
nationally, locally and individually. The main national (starting from 1.0 and 2.0) and 1.2 (starting from
benefits are saved energy and extended life of the 2.0).
stock of houses. Potential local benefits include
improved local environment, lower long-term The options for thermally upgrading the walls of LPS
maintenance and factoring costs, reduced complaints buildings are to fix insulation materials to the inside
from tenants, and extended life of the housing stock. or outside surfaces. The insulants themselves are
The individual tenants may have benefits of improved generally low in cost, but fixing and protecting the
internal and external environment, reduced health insulants makes the whole process relatively expensive.
problems and reduced fuel costs. The owner of large Internal and external systems are therefore rarely
panel system (LPS) dwellings should aim to choose applied for their insulating value alone. The reasons
the most advantageous option at the best time. for insulating internally or externally are usudly
Options relating to the need to improve insulation are complex and likely to include reducing problems such
no action, increase the heating input, partially as condensation, rain penetration and cold bridging,
insulate, fully insulate, transfer ownership or or improving shabby appearance as well as saving
demolish. The correct time to insulate may be dictated energy.
by such factors on the positive side as grants, a cash
vote, work which can be allied to thermal insulation, The thermal insulation of walls, particularly on the
the weight of complaints, high number of empty outside surface, is a relatively new technique although
dwellings, and high maintenance load. On the other the constituent materials have been used in buildings
side of the equation are capital cost, effort spent in for many years. There have been the inevitable
design supervision and disputes with contractors, and technical problems that arise from the introduction of
the possibility of increased maintenance because of these new methods and, in the case of external
the insulation option. Note that maintenance may be insulation, the development of a whole new industry.
increased or decreased by insulation measures. In general, however, the problems have been no
greater than those experienced with traditional walls
If the decision is taken to insulate there is no and in only a very few instances could the problems
regulatory requirement to achieve a particular level of be classed as failures. There is still a great deal of
insulation when renovating LPS buildings, except change in both internal and external insulation, with
when change of use is involved, but it is preferable to new materials and methods being developed, new
aim for at least the minimum standard. Most LPS systems being introduced and systems being
panels have a U-value of around 1.0 W/m2K. The withdrawn.
walls adjacent to common accessways may be as low
as 2.0. Suggested target U-values for upgading are
0.6 W/mZK for compliance with the current standard, External insulation systems
0.45 W/m2K for the next likely standard and There are five principal function's which external
additionally 1.2 W/m2K for small localised cold insulation may be required to perform either singly or
bridges as outlined in The Building Standards in combination:
(Scotland) regulation^'^. Table 4 illustrates the
1 To upgrade the thermal performance of external
approximate thickness of either internal or external
walls
Table 4 Approximate thickness of insulation needed to upgrade walls

Thickness of insulation (mm)

Original U-value: 1.0 W/m% Original U-value: 2.0 W/m2K

Target U-value ON/mZK): Target U-value ON/m2K):
Insulation
materials 0.6 0.45 0.6 0.45 1.2

Expanded polystyrene
or fibre batt 24 43 41 M) 12
Polyisocyanurate
or polyurethane 17 30 29 43 8
Fibre mat or
cellular glass 28 51 49 72 14
2 To prevent water penetration into the building equivalents of methods used successfully for over 60
3 To prevent pedestrian injury from falling aggregate years.
or spalled concrete
Example 2
4 To improve visual appearance
Aluminium or steel rail sections are fixed to the
5 To retard deterioration of the existing surface existing wall by screws or bolts and support cladding
sheets of colour-coated steel or aluminium or
There are three basic components in an external alternatively external-grade non-metal boards. Of
insulation system: the insulant, the cladding and the these the metal cladding systems are the most
fixing. The following sections describe some different important. A large number of companies manufacture
systems classified by the method of fixing. and market proprietary systems of both steel and
aluminium. The properties of the coated materials are
Indirect mechanically fixed well documented at the time of manufacture, usually
Vertical and/or horizontal battens or rails are fixed to in accordance with British Standard BS 5427 'Code of
the wall. Insulation can be placed between them and practice for performance and loading criteria for
they support various claddings. Alternatively, profiled sheeting in building'I3. It is necessary in
insulated sandwiched panels can be fixed to rails. specifying the material itself only to describe it and to
make reference to the standard and to the selected
Advantages thickness of material. Metal claddings have now been
The methods are variations of traditional used on LPS dwellings for several years and there is
techniques. wide experience of their use on commercial and
industrial buildings. Performance has varied from
A wide range of claddings can be used, such as
excellent to very poor, with the main problems being
metal or other sheet materials, tiles, rendered metal
poor sealing against rain penetration, corrosion,
lathing or timber siding.
especially at edges, and early deterioration of the
Battens or rails can span from one sound fixing to protective coating.
another with only stabilising supports between
them. A wide range of non-metallic sheets and boards exists
and many of these are capable of forming the basis
Cavities can be formed behind the cladding for
for suitable exteinal cladding systems. Most of these
ventilation or drainage.
boards are similar to colour-coated asbestos cement
The application of most of the systems is not boards. Experience with asbestos cement has been
restricted by weather conditions. mixed, with found problems similar to those with
metal sheeting. One group of multi-storey buildings
Replacement of damaged sheet or tile claddings is
was successfully overclad with an insulated asbestos
possible without removing the whole system.
cement panel system in 1976. Some of the non-
High thermal insulation value can be achieved. asbestos boards developed in the past few years have
cracked in use and modified fixing methods are now
It is not necessary to provide movement joints to
used.
coincide with joints in the original wall.

Disadvantages Lightweight coating

A low-density cement-bound rendering is trowelled or
They are generally high-cost systems because sprayed onto the wall and then protected by polymer
several operations are needed. or cementitious coatings.
The decorative coatings on sheet claddings may
require recoating several times during the lifetime Advantages
of the building. Usually the lowest-cost system.
Several of the claddings are vulnerable to impact Uneven surfaces can easily be coated.
damage at ground-floor level.
The thickness can be altered to work around and
Most systems are medium to heavy weight, 20-40 under fixtures.
kg/m2.
The material maintains its insulation value in a fire
Example I and protects the substrate.
Treated timber battens are fixed with corrosion- Patching is relatively easy.
resistant screws, insulation with rigid expanded
plastics is pushed between the battens, or mineral Disadvantages
fibre quilts bonded to selvedged building paper are
stapled to the battens. External cladding is of Experience has shown that there can be problems
rendered stainless or galvanised metal lathing, timber with delamination.
siding, tiles or sheet materials. These are modern
 Insulation value is about one-third that of most There have been some problems due to both faulty
other insulants for a given thickness. application and coating in adverse weather conditions.
All panel joints have to be brought through the The appearance can degenerate within a few years in
material. damp areas because of organic growths on the
surface.
Concrete surfaces have to be specially prepared
with bonding agents, or covered with metal lathing Direct mechanically fixed
to achieve adequate bond. The cladding and insulation are fixed directly to the
wall by pins or screws which are located in holes
Example drilled through both the cladding and the insulation.
The coating is lightweight aggregates such as
expanded polystyrene or perlite in a cement matrix at Advantages
thicknesses of 25 7 5 mm. Final protection is by
cement render or polymer-bound coatings. The The systems are semi-traditional in concept and
material has been used on low-rise buildings in the little training should be needed.
United Kingdom since 1978 and on high-rise since The methods of application are relatively simple
1983 with mixed results, the main problem being and there are few components to check.
delamination either of the protective coat or within
the lightweight render. One system has been used These systems contain the widest choice of
successfully for over 7 years whereas other systems insulants of any of the external insulation types of
have been withdrawn from the market following overcladding.
serious failures.
Disadvantages
Adhesive fixed There is a restricted range of finishes, mainly
Rigid slabs of insulation are adhered to the wall. A variations of different renders.
protective coating, usually reinforced with glass fibre,
is then applied followed by a thin decorative coat. The systems are amongst the heaviest, up to about
Supplementary mechanical fixings are sometimes 50 kg/mz, which is a disadvantage where there are
needed. limitations on loading.

Advantages Example
Composite boards of insulation material bonded to a
Very little load is added to the building, breather membrane and metal lathing are held against
approximately 12 kg/mz. the wall while holes are drilled through to the
A high insulation value is easily obtained. substrate to take fixing pins which are hammered in.
The insulation can be expanded plastics, semi-rigid
No movement joints are needed with some systems. mineral fibre slabs or cellular glass. Finishing is with a
traditional sand:cement rendering. This has been the
Disadvantages most widely used external insulation type in the
Very restricted range of finishes. United Kingdom and experience dates back to 1976
mainly on low-rise dwellings with examples on LPS
Not suitable for application during poor weather. dwellings dating from 1979. There has been long
High level of skill needed for satisfactory experience of rendered metal lathing in this country
application. and there is some confidence in the fact that it has
lasted for over 60 years on timber-framed dwellings.
Surface finish will need renewing several times The traditional rendered finishes can last for 30 years
during the lifetime of a dwelling. or more without deteriorating in any performance
The original wall surface must be firm and fairly respect except dirtying and staining. Problems with
even. the system have mainly resulted from poor
workmanship such as inadequate fixing or bad
Example rendering practices which have resulted in cracking.
Adhesives based on sand:cement mixes modified by
the addition of polymer emulsions are used as strips
or dabs to fix insulation boards to the wall. The Internal insulation systems
boards are covered with a thin layer of cementsand A great deal of information on plasterboarding and
modified by polymer emulsion and reinforced with also some information on insulated dry-linings is
included glass fibre or layers of woven glass fabric. contained in trade literature. This section utilises some
The final decorative coating is usually a trowel- or standard information but also incorporates data from
roller-applied filled co-polymer. These systems have site experience and user surveys.
been used on the Continent since the mid-1960s and
include examples on multi-storey buildings. In the In upgrading work the internal insulation may be used
United Kingdom use has mainly been on low-rise for improving the thermal properties of the wall or as
buildings since the late 1970s. part of a package of repair-and-renewal or, as was
25
found in a local authority survey, mainly to reduce Adhesive fixed
condensation problems. Boards of insulation bonded to plasterboard are
adhered to the wall. Adhesive is either a thin layer
The components of an internal insulation system are over all the surface or dabs or strips of thicker
some or all of the following: adhesive. The plasterboard is then finished by
Fixing traditional plasterboarding techniques. Mechanical
Water barrier fixings may be used to give supplementary support.
Insulation
Cavity This is a quick and low-cost method and the adhesive
Vapour control layer adds little to the thickness. Therefore the intrusion
Lining into the room is less than with other methods. Where
there is cold bridging along ceilings and intersecting
Three main methods of supporting the insulation are walls, the insulation can easily be extended onto their
used. The insulation can be adhered to the face of the surfaces.
wall, or loosely supported between battens fixed to
the wall, or supported away from the face of the wall The main disadvantages are that services can only be
in a similar manner to an internal partition. The main accommodated on the surface of the lining or in
categories therefore are as follows: grooves in the insulant or walling behind the system.
The method is unsuitable for use on damp walls.
Adhesive fixed
Supplementary fixings are usually needed to prevent
Mechanically fixed
the boards from bowing or to support those with
Separated lining
thermoplastic insulants during fire. It is not easy to
support medium- to heavy-weight fittings on the
The above classifications are used in the following
lining.
text to cut down the amount of repetition in
describing the systems and their properties but, as
Mechanically fixed
with external insulation, there are inevitable overlaps
Three basic methods are described below, all of which
of methods with some systems using partially adhesive
use the background as the basic support.
and partially mechanical fixing support and with
plasterboard/insulation laminate being used for both
I Plasterboard/insulation laminates: direct fixed
adhesive and mechanically fixed systems.
Plasterboardiinsulation laminates are held against the
background and fixed with self-drillingiself-tapping
The type of system chosen is mainly dependent on the
screws or metal pins in predrilled holes into the
condition and materials of the background. Factors
which influence this choice are whether the concrete.
background is weak or strong, rough or smooth, wet
With skill a finish suitable for plasterboard jointing
or dry. Another factor which may influence choice is
techniques can be produced. Otherwise the
the skill available and whereas some methods are
plasterboard is best finished with a skim coat of
suitable for use by joiners, others require a plasterer.
LPS walls are smooth and strong but may be dry or plaster.
wet.
The method is simple, quick and uses the minimum of
different techniques. Any of the plasterboard/
The reason for any damp-staining must be established
insulation laminates can be used although those with
and steps taken to remove the source of damp. This
some resilience are preferred because a slight amount
may include repairing plumbing work, repairing
of alignment can be achieved by adjusting the
joints1', flashings, etc, or taking measures to reduce
pressure on the fixing. Metal fixings may act as local
condensation problems.
cold bridges when fixed directly into concrete. This
The walls should also be inspected for evidence of method is not suitable for damp walls.
cold bridging, particularly at the intersection with
flanking walls and intermediate floors. These 2 Plasterboard/insulation laminates on metal firring
problems may be increased after a wall has been Special channel-shaped strips of thin zinc-coated steel
internally insulated. are stuck to the wall with a plasterlike material, then
a plasterboard insulation laminate is screwed to the
channels with self-drilling/self-tapping screws.
Where it is decided to remove an existing inner lining
before internal insulation, a special check should be
made to ensurc that sound transmission and fire The method is suitable for any background which can
stopping between different rooms and dwellings are be plastered. A slight amount of initial dampness can
not adversely affected. be tolerated but it is unsuitable for permanently damp
walls.
The requirements for wiring, electrical outlets,
Care has to be taken not to overtighten the screws
plumbing and fitments must be determined and
with compressible insulants as the area of contact
related to the type of insulation method planned.
between firring and insulant is relatively small and the internal insulation. An approximate order of cost
plasterboard can be distorted. would be that external insulation is two to five times
the cost of internal insulation.
3 Plasterboard or laminates on battens
Battens are fixed to the wall and either insulation is 2 Wall construction
placed between the battens then covered with a No insurmountable problems have been found which
vapour control layer and plasterboard, or a preclude the choice of either internal or external
plasterboard insulation laminate is fixed to the systems for LPS walling.
battens. The battens may be separated from direct
contact with the wall, and a cavity ventilated to the 3 Repair and upgrading
outside may-be provided between the insulation and The ideal time for specifying internal or external
the walling. insulation is alongside repair or rehabilitation work.
The nature of the work will often dictate the most
This method is suitable for walls which are dry or advantageous method.
intermittently damp. Permanently wet backgrounds
should not, however, be covered. On damp walls it is Where internal wall surfaces have to be disrupted by
advisable to provide an isolating layer of building electrical and plumbing work, the extra cost of
paper together with spacers to hold the main areas of providing insulation becomes small. Similarly when
the battens away from the wall. In these conditions replacing kitchen fitments, repairing damaged plaster
corrosion-resistant fixings are essential. The use of a surfaces or even redecorating walls, internal insulation
plasterboard/insulation laminate even when additional may be profitably included in the work.
to insulation between the battens has the advantages
that a sandwich vapour control layer within the External insulation can be combined with the repair
laminate is protected and the effect of cold bridging or redecoration of surfaces that have deteriorated or
by the battens is considerably reduced. Electrical and be applied as a cure for rain penetration.
other services can easily be fitted behind the battens.
The battens provide a firm anchorage for fixings and Window replacement may cause damage to both the
fitments. Disadvantages with this and the firring internal and the external wall surfaces. This provides
supported method (number 2 above) are that they are an ideal opportunity to modify the window
relatively expensive, disruptive and reduce the floor construction or location so that it can be linked to
area considerably. either internal or external insulation systems.

Separated linings 4 Wall thickening

One method is to build an extra masonry inner leaf. Internal insulation reduces the room size and in small
More often a lightweight lining is fixed to the floor dwellings this may be the most important reason for
and ceiling, with added stability being provided by not using the method.
flanking walls and possibly centre packers for bracing
to the original wall. External insulation may thicken the wall sufficiently
to obstruct common accessways.
As the lining and main supports are separated from
the wall by a cavity, the method can be considered for 5 Fittings on walls
damp walls, but only where provision is made to Internal radiators, water pipes, electrical outlets and
remove water from the cavity between the lining and switches, skirting boards, facings, window sills,
the wall. These methods have the major difficulty that cupboards, and bathroom and kitchen fittings may
they are expensive and reduce the usable internal area have to be removed and refitted during work on
by the largest amount compared with other methods. internal insulation.

When insulation is external, rainwater pipes, overflow

Comparison of insulation position pipes, gas supply pipes, balanced flues, wall
Several factors govern the choice of where to place ventilators, window and door surrounds, canopies,
the insulation, ie externally or internauy, and these television cables, lightning conductors and telephone
are discussed below. One factor may, in a particular wires may have to be removed, replaced or f'med on
instance, be of paramount importance, particularly if extended brackets.
a main objective of insulation is to cure condensation
or to improve the appearance of the wall as well as to 6 Disturbance and access
improve thermal performance. All the factors should During the fitting of internal insulation the occupants
be initially assessed on a simple comparative basis to are inconvenienced and temporary rehousing is often
enable decisions t o be made and to identify factors needed.
which may require further detailed investigation.
With external insulation, garden plants on or near the
wall may be destroyed. Scaffolding or a similar
1 Capital
External insulation is usually more expensive than
method of access is required and fitting usually 13 Weather during application
requires noisy impact drilling. Internal insulation can he applied whatever the
weather conditions.
7 Heating system
Any existing heating system will probably be oversized External insulation is very dependent on the weather
and less efficient after insulation and should conditions during application. There may be long
preferably be replaced by a lower output system. delays caused by frost, rain and wind.

The added insulation will cover only between one- 14 Appearance change
third and one-sixth of the surface area of a room and There will be little or no change in appearance after
so may not significantly change the thermal response internal insulation.
of the room.
External walls of shabby or dingy appearance can he
8 Thermal insulation markedly improved by external insulation, but stains
The improvement in the level of insulation mainly and organic growths similar to those present on the
depends on the thickness of insulation material. In existing wall are likely to be deposited eventually on
general, greater thicknesses of insulation can be used the new surface. Attractive wall surfaces may he
with external insulation than with internal insulation, rendered bland by external insulation, and planning
although there can be features of the wall which permission will be required in some instances where
restrict the insulation thickness with either type of the appearance will he changed.
system.
15 Maintenance
9 Cold bridging Internal maintenance may be considerably reduced
Externally applied insulation is usually more effective where the insulation reduces problems of
in overcoming the effects of cold hridges than is condensation, although differential staining may
internally applied insulation. It should not be increase if the system contains thermal bridges.
assumed, however, that all cold bridges are eliminated
because external insulation is specified. A special External insulation systems finished with polymeric
section on cold bridges is included later in this materials are likely to need more frequent
chapter. maintenance than concrete in good condition.

10 Surface condensation 16 Building height

Both internal and external insulation will lead to a Building height places no restrictions on internal
reduction in surface condensation, but it is unlikely insulation except the difficulty of moving materials
that either will cure serious condensation problems via lifts or stairways.
unless heating is increased and the amount of water
vapour in the air is reduced. Effects of rain and wind increase with the height of
the building and impose more stringent conditions on
11 Interstitial condensation the performance and difficulty of application of
Condensation within the thickness of the wall can external insulation. The cost of providing access
theoretically occur with some internal and some platforms, and more frequent maintenance, may be
external insulation systems. decisive factors in rejecting external systems altogether
or only considering specific types.
Ensuring the installed integrity of vapour control
layers can be a problem with internal insulation. 17 Wall surface complexity
Interstitial condensation can generally be avoided in The easiest wall surfaces to cover with internal or
external insulation by using permeable claddings or external insulation are large flat areas with few
ventilating behind impermeable claddings. openings, returns or fittings.

12 Wall condition after insulation 18 Added load

After the application of internal insulation the Internal insulation adds so little load that it is of no
original wall will experience slightly increased ranges significance.
of temperature and moisture content. There is no
evidence that this will create problems with walls in External insulation weighs between 10 and 50 kg/m2.
good condition. The stability of the structure should he checked before
fitting external insulation.
External insulation will protect the original wall from
changes in temperature and moisture content. There 19 Regulations and standards
will normally be a reduction in moisture content of After improvement by internal or external insulation,
the wall, so slightly improving the insulation value of the building must not be of a lower standard for any
the treatment. of the requirements than before the improvements.
Fire performance may be affected by both internal 25 Access to structure after insulation
and external insulation. Internal insulation may affect The structural concrete elements may have to be
space standards and sound transmission. Thermal inspected for signs of deterioration at intervals during
insulation will always be improved of course. their projected life. Such inspection can disrupt
Guidance is given in reference 15. finishes or vapour control layers. In these cases it is
important that access to the structure is considered
20 Planning permission when choosing internal or external insulation.
No planning permission will be needed for internal Guidance on the appraisal of LPS is given in
insulation. reference 16.

The change in appearance brought about by external 26 Total area insulated

insulation may restrict or prohibit the use of some The wall area to be insulated internally is usually less
finishes. than that externally because of the smaller internal
area at corners and reduced area where intersecting
21 Site expertise walls and floors meet the internal wall surface. In a
Most internal insulation is fixed by joiners or corner flat, the internal wall area is about 12% less
plasterers and no special skills are usually required. than the external wall area. The accessible internal
area may be further reduced by built-in furniture and
External insulation usually requires specially trained bathroom fittings. Both the above factors reduce cost
operatives. at the expense of greater energy loss compared with
external insulation.
22 Site supervision
Methods of applying internal insulation are regarded The total insulated area may also include reveals and
as traditional and within the understanding of most soffits which can further alter the balance between
clerks of works. internal and external areas depending on whether the
windows are fixed towards the outside or inside
External insulation systems range from the semi- surface of the wall.
traditional activities of fixing battens and rendering
on metal lathing to specialised techniques. With the
specialised systems, supervision by normal site Cold bridging
personnel is difficult because they do not know the External insulation systems can help overcome
relative importance of operations, nor what checks to problems of cold bridging, which occur when
make, unless specific detailed instructions are given. materials of high thermal conductivity span across the
Even semi-traditional methods can give problems insulating layer. The additional energy lost through a
when it is assumed that operatives and supervisors will cold bridge may be a small proportion of the overall
have had experience of the particular skills required input. Often the main benefit of treating cold bridges
for that system. is a reduction of condensation.
23 Impact and vandalism It is sometimes assumed that all floor/wall
The plaster or plasterboard used over most internal intersections and internal wall/external wall
insulation provides a traditional finish and will stand intersections are cold bridges. This may not be so with
up to normal use. Areas susceptible to vandalism or some LPS dwellings where careful detailing reduces
higher than normal impact will be obvious during the the potential cold bridges considerably (Figure 13).
inspection made before deciding on a system. Nevertheless, the inclusion of a narrow strip of
insulation on a drawing does noi ensure that it is in
External insulation systems vary in impact resistance, place in the building.
but can be modified by thicker protective coats or
extra reinforcement to withstand impacts expected Some information on cold bridges can be obtained by
with normal usage. All external insulants are generally thermographic surveys of buildings, but the
more susceptible to destructive vandalism and impact interpretation of the data needs to be thorough. The
damage than the traditional materials they cover. The important factor is the expected temperature of the
ground-floor walls of multi-storey buildings are cold bridge relative to the air temperature and
usually treated differently from the other storeys. moisture content of the air.
24 Rain penetration Valid information on the cold bridges can be obtained
Walls showing signs of persistent rain penetration are from a detailed examination of the dwellings and
unsuitable for internal insulation until the defect is from an analysis of past complaints from a particular
cured. estate or block of dwellings even where there is to be
a major change in occupancy, heating levels or
External insulation can overcome rain penetration ventilation. It may be that the main problems
problems. associated with cold bridging occur in greater numbers
 extreme cases no insulation may be possible, but every
effort should be made to provide some insulation to
the reveals in situations where there have been
condensation problems in these areas.

The top of a wall, especially in multi-storey buildings,

is often terminated by a parapet. If the parapet is
low, external insulation can be taken up and over the
parapet and linked with roof insulation (Figure IS), or
in the case of internal insulation, ceiling insulation
can be installed. High parapets present a similar case
to balconies, which are described in some detail
below.

Insulation at back
of parapet linked
n

Dividing wall Cornel

Figure 13 Cold bridging at LPS panel intersections reduced by

linking insulation Figure 15 External insulation at the roof/wall junction

in certain flats or in specific locations in flats. This Where the walkway of a balcony forms the roof of a
knowledge will help in determining whether selective habitable room, or where the ceiling over a balcony
or comprehensive or no special treatment will be forms the floor of a habitable room, these areas
needed to deal with cold bridging. The locations dealt should always be insulated where practicable and
with below are where typical cold bridges may be linked as far as possible to the wall insulation (Figure
found. 16). Balconies and accessways which do not enclose
habitable rooms but are connected to floor slabs
Solid ground floors are sometimes continued through provide a thermal bridge to the inside. The decision of
to the outside and this can lead to condensation both whether or not to insulate these areas can be a
on the floor and on wall elements in contact with the difficult one as the cost of providing insulation to
floor. Possible remedies are overfloor insulation both the top and bottom surfaces of a balcony is very
protected with chipboard, perimeter insulation below high, whereas the benefits may not be very great.
floor level or external insulation from damp-proof
course level to below ground level (Figure 14).

Figure 14 Cold bridge will be present even after external

insulation unless the wall below the damp-proof course
is insulated
el ing ~ n s ua #on is
The boundaries surrounding windows and doors often a ternative t o floor
contain cold bridges but the opportunity for
insulation may be restricted, for example because the Figure 16 Insulation far balconies
window frames are of relatively narrow section. In and accessways
 External

! Outside air O°C

Internal
insulation L)Q
10.3O :\,ti
~ a n y

' , , , I \

Inside air 16.5OC 1

,L,xio 41 Outside air 0 ° C

Figure 17 Comparison of internal and external wall insulation. Effect of insulation on surface temperatures where
floor slab forms cold bridge and is extended as a balcony

The likely effects of external and internal insulation priority. Where there is rain penetration or
and the internal surface temperature at a typical deteriorating external surfaces, recladding may be
balcony are shown in Figure 17. In this instance the required and estimates can be obtained for this which
internal insulation would not be effective in reducing can then be subtracted from the cost of external
the incidence of condensation on the floor or ceiling insulation. In one large contract the estimated cost of
near to the outside wall, but external insulation would sealing joints and redecorating the external surface
reduce the risk considerably. was greater than the estimate for an insulated
overcladding. Similarly the cost of repairs to plaster
and redecoration may be offset against internal
Cost considerations insulation.
Cost-effectiveness calculations for insulation take
Another method of quantifying benefits is to use a
account of only the cash benefits for reduced fuel
composite method of analysis. A simple assessment
usage. In reality the situation is much more complex
can be made for rented property where tenant
and other benefits may arise after insulation. Factors
dissatisfaction results in a larger than normal turnover
vary considerably from example to example and
in occupants. Each re-let costs a considerable amount
depend on assumptions that are relevant only to the
in lost revenue, increased repair costs and
time when the analysis is undertaken.
administrative time during the period the property is
empty. Vandalism is often increased or security costs
The following notes give some indication of other
have to be increased to keep out vandals. Each
factors which may be more significant than energy
landlord will have different cost items and values but
saving on its own.
the simple example below illustrates the principle:
The total capital cost of a measure is easy to Empty dwellings on
determine from estimates. The insulation part of a substandard estate 12%
system typically accounts for less than 10% of the
Empty dwellings on
total, and in a realistic assessment a decision has to be
average estate 3%
taken on what proportion of the overall cost should
be used in calculations. Average period dwelling
empty 6 weeks per dwelling
Most insulation activities are associated with repairs
Cost of closing, repairing,
and cures for problems such as rain penetration, £210 per dwelling
redecoration and re-letting
.
condensation., Door wall surfaces and shabbv
appearance, with energy saving possibly being a lower Weekly revenue f 30 per dwelling
Extra lost revenue improvement in the worst areas of cold bridging.
for substandard (12-3) x 52 x £30 Typical results are illustrated in Figure 17.
houses on 100-house estate = £14 040 per year
The following insulation studies were made:
Extra cost of repairs
and re-letting for Calculations of expected energy usage at theoretical
substandard houses (12-3) x 52 ? 6 x £210 design temperatures
on 100-house estate = £16 380 per year
Thickness of insulation required to give target
U-values with typical wall/roof and floor
Therefore the extra cost on the difficult-to-let estate constructions
compared with an average estate is £14 040 +
£16 380, or £304.2 per house per year. This benefit Change in calculated energy loss with different
which may be gained from upgrading including insulation strategies
insulation is likely to be greater than the value of the Effect of room positions on flat-to-flat heat
energy saving and can be a direct gain to the landlord,
transfer
whereas energy savings would be a gain to the tenant.
Nevertheless, in a realistic cost-benefit analysis these Assessment of the options for surface-applied
could be added together. insulation to the whole or parts of the structure,
and practical considerations including comparisons
of preferred methods of fixing and protecting
A study of 1950s 11-storey deck-access against impact, fire and weathering where
flats appropriate

This section gives a summary of an insulation Comparison of internal and external insulation,
assessment which was part of a study undertaken by a especially of walls fronting onto balconies
multi-disciplinary team drawn from the Building Methods of improving roof insulation
Research Advisory Service assisted by specialists in
condensation, thermal insulation, material science, The main findings of the investigation into the
microbiology and infra-red thermography. insulation part of the study were that selective
external insulation of the gable walls and the walls
An inspection of unoccupied dwellings showed little adjacent to the stair wells would be the most
evidence of severe condensation problems and no profitable, together with the replacement of some
widespread mould growth. The external walls glazing with insulated opaque panelling, the insulation
contained large areas of glazing which would result in of the underside of the floors of some rooms, and
a considerable heat loss. The Tenants' Association improved roof insulation. Other insulation measures
was contacted and from discussions with them the would be complex and expensive without reducing
main cause for complaint appeared to be excessive heat loss by a significant amount. However, these
heating bills rather than condensation problems. latter measures could be considered as part of a
Occupied flats chosen from a list provided by the staged package if a demonstration of the simpler
Tenants' Association were inspected. It was found insulation measures together with suggested heating,
that condensation and mould growth were more ventilation and maintenance improvements showed
prominent in rooms where the floor or ceiling slab that the complete insulation package was necessary.
was exposed on the outside because of balconies, etc.

An analysis of 'Section 99' Court Orders showed that

the complaints of dampness and mould growth were
highest in top-floor dwellings and those adjacent to
the stair wells.

Examination of the drawings identified several areas

of potential cold bridging, especially: where the floor
slab extended to the outside as balconies; at the
junction between the roof and the wall; in the area
around the refuse chute; at the wall and wall/floor
junctions adjacent to the stair wells; at the corners of
the building; and beneath some first-floor flats where
the underside of the floor was exposed.

Finite-element calculations were undertaken to

determine expected temperatures for the existing
construction and for insulated constructions with
internal and external insulation. These showed that
internal insulation would bring about little
 Chapter 6
Two case studies in LPS dwellings

Introduction
The Building Research Establishment's Scottish Many of the tenants had lived in the flats since they
Laboratory at East Kilbride has been investigating the were built; most were families with children. Few
risk of condensation and mould growth in dwellings complained of condensation problems; there were
built with the various large panel systems (LPS). This however cases of rain penetration.
chapter describes two monitoring exercises carried out
in Reema flats in Glasgow and in Bison Wall Frame Modified flats
houses at RAF Machrihanish over the winter of This was a 20-storey Reema block built along with
1985 -86. two others in the late 1960s on the north edge of
Glasgow. There were four two-bedroomed and two
Previous studies of non-LPS flats at Stirling and one-bedroomed flats on each floor. The three blocks
Inverclyde and houses at Harrow over four winters on the site formed a heat-with-rent scheme. Electric
produced a reasonably complete picture of the storage heaters running on the White Meter tariff
conditions that occur in traditional local authority (similar to Economy 7) installed in every room were
housing. The main objective of the present work was metered independently and Glasgow District Council
to determine whether remedies identified as successful were charged by the SSEB. The tenants paid a
in traditional housing would be as effective in LPS standard charge for heating (£16 or £20 per month for
dwellings. Throughout the analysis of the results, a one-bedroomed or two-bedroomed flat respectively)
comparisons were made between conditions in the along with their rent and were charged as usual by the
LPS blocks and conditions in the insulated flats at SSEB for the remaining electricity for lighting,
Stirling, where the U-values of the walls were very cooking, etc. It was possible for tenants to opt out of
similar to those of the LPS flats. It was expected that the scheme, but most chose to participate.
differences in conditions found would relate to other
features peculiar to LPS dwellings. To combat problems of rain penetration and upgrade
the insulation standard of the walls, all three blocks
were covered with profiled metal cladding
Reema flats, Glasgow incorporating 25 mm of insulation; this reduced the
The buildings studied U-value from 1.1 to 0.6 W/mZK. The resulting fabric
It was originally hoped to study two large panel transmittance was 46 W/K for the one-bedroomed
blocks with and without severe condensation flats and 57 W/K for the two-bedroomed flats: both
problems. However, discussions with Glasgow District unusually low values. The existing single glazing was
Council Department officials suggested that they have retained.
few major condensation problems in these structures.
It was, therefore, decided to compare conditions in As in the other block, most of the tenants were
two Reema blocks, one unmodified and one which resident for many years. The population structure was
had been externally insulated and was part of a 'heat rather different, however, with many older single
with rent' scheme. people, especially in the one-bedroomed flats.

Unmodified flats Monitoring

This was a 12-storey Reema block built in the late Internal air temperatures and humidities
1960s on a mixed high- and low-rise estate about 4 After approval for the work had been given by the
miles south of the centre of Glasgow. There were six Council, letters were circulated to all tenants in each
two-bedroomed flats on each floor, with off-peak block explaining the purpose of the work and asking
electric warm-air heating serving all rooms. for their co-operation. Ten flats with interested and
Bathrooms were internal with ventilation extract via helpful tenants were then selected in each block to
central ducts to the roof. provide a reasonable spread of locations within the
blocks. Thermohygrographs were installed in the
The standard Reema construction includes 25 mm of living-room, kitchen and one bedroom of each flat
insulation in the external walls giving a U-value of starting in early January 1986. These continuously
approximately 1.1 W/m2K. The resultant average recorded the air temperature and relative humidity in
fabric transmittance (the total theoretical heat loss) of the rooms on charts which were changed weekly for 3
76 W/K is low despite the fact that the windows are weeks. A further 10 flats were then selected in each
single glazed. Values of 110 W/K were typical of the block, and they were monitored for another 3-week
insulated flats studied at Stirling (200 - 250 W/K for period.
the unmodified flats).
Hourly readings of the outside weather conditions at the anemograph at East Kilbride, 5 miles away. While
Glasgow Airport over the whole period were obtained this did not represent the air flows around the tower
from the Meteorological Office. blocks, which may be complex, it did act as an index
of wind in the Glasgow area.
The thermohygrograph charts were digitised using a
graphics tablet and microcomputer, and hourly values Air leakage tests
of temperature and humidity were stored. These were A different technique was used in the occupied flats,
used to calculate the mean conditions for each flat where the tracer gas cannot be easily used. A large
over the three weekly periods. fan, fitted into a replacement front door, was used to
pressurise or depressurise the flat. Measurement of the
Surface conditions pressure difference across the fan as the flow was
The AGEMA thermovision camera produces an image varied allowed a leakage value to be calculated at a
in which variations in brightness are proportional to standard pressure difference. A thermovision survey
the surface temperature of the object in view. The of the flat while it was depressurised showed areas
images can be recorded on a modified video-recorder where cold air leakage reduced internal surface
and analysed with a dedicated microcomputer to temperatures and highlighted the leakage paths.
provide quantitative temperature information. If a
building has been well heated for some time (at least a With the help of Strathclyde University, who had a
day) a thermovision survey of the internal surfaces contract from BRE to carry out pressurisation tests,
will show the location of any areas which are less well three flats in each block were tested.
insulated than the remaining wall. These areas are
known as thermal bridges. Surveys taken from the Social survey
outside are rarely satisfactory, because the thermal At the conclusion of the exercise all the tenants were
images are distorted by the external weather interviewed to establish the household composition,
conditions, particularly wind and rain. the use of the heating and other appliances and any
problems they had with the flat.
A detailed thermovision survey of one empty flat in
the unmodified block, and general surveys of a Analysis of data
number of occupied flats in both blocks, were made The data collected in the two Reema blocks in
to assess the extent and severity of thermal bridging. Glasgow were analysed and compared with the data
from the insulated flats at Stirling.
As temperatures at internal surfaces, especially near
thermal bridges, are lower than the air temperature at Mean temperature and vapour pressures, weighted by
the centre of the room, the risk of mould growth can room volume, were calculated for each flat. The
be much higher in localised areas. To identify these difference between the average whole-house and
areas, sensors were installed to record surface average outside temperatures in the three groups of
temperatures and the adjacent air temperature and flats is plotted against the weekly energy consumption
humidity: the locations were identified using the during monitoring in Figure 18. Energy use in two of
thermovision camera. The data were recorded on the flats in the modified block (o in Figure 18(a)) was
small data loggers and interrogated with a very much less than that in the others. The occupants
microcomputer for further analysis. of these flats had opted out of the heat-with-rent
scheme and rarely heated their flats. As these are not
Energy consumption representative of the system, they have in general been
All electricity meters were read on each visit and eliminated from the subsequent analysis and are not
weekly energy consumptions were calculated. Final included in the calculated regression line. It is,
meter readings were taken in mid-April to provide however, interesting to note the relatively warm
data on the longer-term fuel usage in the dwellings. conditions in these flats. The data from the flat
represented by the outlying point (0) in Figure 18(b)
Drawings supplied by the Department of Architecture have also been eliminated. This flat, in the
and Related Services were used to calculate the fabric unmodified block, was immediately over a very well
transmittance or total theoretical heat loss from each heated communal laundry room; temperatures were
flat. therefore unrealistically high.

Natural ventilation measurements The data collected from the three blocks are
A series of measurements of ventilation rates was summarised in Table 5.
made in an empty flat in the unmodified block, using
an infra-red gas analyser to monitor the concentration Energy use
of nitrous oxide. After release and uniform mixing To put the flats in this study into the context of
throughout the flat, the concentration of tracer gzs Scottish local authority housing as a whole, annual
decays exponentially; the gradient from a plot of the energy consumptions were estimated from the 3
logarithm of concentration against time gives the months of meter readings using published degree-day
ventilation rate. Wind speed data were obtained from values for the period, as shown in Table 6.
Table 5 Summary of data from monitored flats

Large panel, Glasgow, 1986-87

Stirling,
Modified Unmodified insulated

Mean SD Mesn SD Mean SD

U-value of walls (W/m?() 0.6 - 1.1 - 0.5 -

Fabric transmittance (W/K) 56 6 76 8 110 17
Energy use (kWh/week) 426 91 330 125 214 87
Number of adults/flat 1.6 0.8 1.8 0.4 2.1 1.2
Number of childredflat 0.4 1.O 1.1 1.3 1.2 1.1

20 -
;; 1
(a) Modlfisd LPSflatr

0; + 0.008E
I
Table 6 Estimation of annual energy consumption

Degree-days for west of Scotland

1 February
March-

2L
0 0 60 120 180 240 300 360 420 480 540 600
April
May
June
242
167
85
304
164
101
Weekly energy con.umpt<on ( E : k w h )
-
" July
August
64
67
71
104
t (b) Unmadnfned LPSflats
0 20, - September 107 112
October 178 182
November 307 386
December 359 320
Total 2615 2918

1 To estimate the annual coilsumption from the I5 weeks

consumption during 9/1/86 -24/4/86:
Sum of degree-days from January to April 1986 = 1478
-
-
rn
a
weekly energy cansumprion (E: kwh)
Proportion of 20-year mean annual total = 1478/2615 = 0.57
Number of days from January to April = 120
Multiply weekly consumption by (52 x 120)/(365 x 0.57) =
( c ) Stirling. inrulat~dflata
20, I 29.99

2 To estimate the annual consumption from the 12 weeks

consumption during 30/1/86-24/4/86:
12 Sum of degree-days from February to April I986 = 1067
10 Proportion of U)-year mean annual total = 1067/2615 = 0.41
Number of days from February to April = 89
6 Multiply weekly consumption by (52 x 89)/(365 x 0.41) = 30.93
4 n

0 60 120 180 240 300 360 420 480 540 600

Weekly energy consumption IE kwh)

Figure 18 Difference between inside and outside temperature

plotted against energy consumption in the three groups
of flats

Histograms of the estimated values are shown in glazing; double glazing would reduce this to about
Figure 19. The mean annual consumption for each 30%.
block is plotted against fabric transmittance in Figure
20 together with the corresponding value from the Energy use in the unmodified block was very similar
insulated Stirling flats. The lines are derived from a to that in the Stirling flats where mould growth was
survey of 1500 local authority houses in central widespread; however, the low fabric transmittance
Scotland. The small external wall areas implied very would lead us to expect better conditions. In
low fabric transmittances in the Reema blocks, comparison, energy use in the modified block with the
especially the modified block, compared with the heat-with-rent scheme was significantly higher, near
Stirling flats, which had similar U-values. Between 40 the top 10% of users in the larger survey. This,
and 50% of the heat loss was through the single together with the very high standard of insulation,
 ~ o d i f i e dLPS flats temperatures in the two modified flats where there
was little energy use show that the external insulation
was keeping the whole block warm.

Internal conditions
The average temperatures and vapour pressures from
each room in both blocks and in the flats at Stirliug
are plotted in Figure 21 together with the outside
conditions. The curved lines of relative humidity
indicate the risk of mould growth; values over 70%
0 2 4 6 8 10 12 14 16
are critical.
Estimated annual energy cansumpt#on (MWh)
It can be seen from the data in Figure 21 that the
weather was unusually cold and dry during the
monitoring period, significantly different from the
unmodified LPS flate
5 Stirling experiment when milder, humid conditions
I
4 C M e a n = 8.0 i 3 . 1
1 (more typical of the winter conditions) prevailed. To
make the data collected in Glasgow more readily
comparable with the Stirling data and more generally
applicable, they have been standardised on the outside
condition at Stirling, as follows.

Using the hourly values, regression equations between

inside and outside temperatures and vapour pressures
were calculated for each room in each block. The
resulting equations are shown in Table 7. The
gradients were then used to standardise the Glasgow
Figure 19 Histograms of estimated energy consumption data on Stirling using:

should lead to unusually good conditions in the flats. = T, +B x (T, - T,)

Figure 19 also shows that the range of consumptions
was much less in the modified block. where: T, = standardised value
T. = original value
The relative slopes of the regression lines between B = gradient of relevant regression equation
temperature excess and energy use shown in Figure T, = outside temperature at Stirling
18, demonstrate the low total heat loss from the T,= outside temperature at Glasgow
modified Reema flats in Glasgow compared with the
traditional flats in Stirling. Although they are not The standardised values are plotted as before in
included in the analysis, the relatively high Figure 22.

upper 10%

25 - A Modlfled LPS
O Unrnodifled LPS
0 Sf,rl,ng. ,nsuls,ed

20 -

-
Ez 15-
Average

"

-
n
z
5

c
-
: 10- Lower258

Well lnrulafed Poorly ensolated

0 -
0 50 100 150 200 250 300 350 400
Fabr~cfransm#tiance(WlKj

Figure 20 Energy consumption in the three groups of flats compared with large-scale survey
of Scottish local authority houses
 Temperature (OC)

Figure 21 Average temperatures and vapour pressures from each group of flats, with graphs
of relative humidity (70)

Table 7 Regression equations between inside and outside hourly data

Modified block Unmodified block

Living-room T = 18.5 + 0.14 x OUT R = 0.14 Living-room T = 16.9 - 0.09 x OUT R = 0.09
P = 5.4 + 0.55 x OUP R = 0.34 P = 7.3 + 0.24 x OUP R = 0.19

Bedroom T = 17.4 + 0.02 x OUT R = 0.02 Bedroom T = 14.8 - 0.02 x OUT R = 0.02
P = 5.5 + 0.36 x OUP R = 0.32 P = 5.9 + 0.46 x OUP R = 0.41

Kitchen T = 18.2 - 0.02 x OUT R = 0.02 Kitchen T = 18.7 - 0.06 x OUT R = 0.05
P = 7.0 + 0.22 x OUP R = 0.14 P = 6.9 + 0.41 x OUP R = 0.24

Abbreviorions: T, inside air temperature ("C); P, inside vapour pressure (mb); OUT, outside air temperature ("C); OUP, outside vapour
pressure (mb); R, correlation coefficient.

Modlfted Unrnod~ffed Sflrilng

L
Bedroom B
Kgtchen @ @ K
Oufslde 0 0 0
100% 90% 80% 70% 60% 50%
12-

11-

10-

-
n
-E 9-

0 2 4 6 8 10 12 14 16 18 20 22 24
Temperature /OCI

Figure 22 Average temperatures and vapour pressures from each group of flats, standardised
on Stirling, with graphs of relative humidity (%)
(a) Modified LPS flat

Unmodified LPS flat

e n r e 23 Trpical thcrmogrums horn each block

It can be seen that temperatures in the modified block In comparison temperatures were er in the
were so high that there was little danger of mould unmodif~edblock than in the mol 'd block,
growth; the bedrooms in particular were little colder especially in the bedrooms, which re not in general
than the other rooms, showing the benefit of heated; however the temperatures .e higher than
insulation and full central heating in keeping the those observed at Stirling. Vapou essures were high
whole dwelling warmer. Vapour pressures were in the kitchens; this was probably :to the larger
unusually high in the living-rooms, perhaps reflecting family sizes in this block.
low ventilation rates,
Surface conditions Modified LPS flats
Except for air leakage through joints (see below), the
thermovision survey did not indicate any serious
80 r
problem areas. Examples of cold areas at wall/floor 0 A#r RH over 70%
0 Surlace RH over 80%
junctions from each block are shown in Figure 23.
The temperature in the bedroom of the particular flat
in the modified block was lower than that in the flat
in the unmodified block, leading to lower surface
temperatures overall. However, Figure 23 shows that
the coldest point in the corner was 1 to 1 . 5 degrees
colder than the overall wall temperature in the
modified flat, whereas in the unmodified flat the
difference was 3 degrees. The external insulation had 0
Flat 1 Far 2 Flat 3 Flat 4 Fiat 5
therefore smoothed out the temperature variations
over the wall, reducing the intensity of the potential
thermal bridge at this point. Unmodified LPS flats
I
-
The risk of mould growth on surfaces can be 0 A#, RH over 70%
estimated from the proportion of time that the 0 Surface RH over 80%
relative humidity was over 7 0 % in the room
(equivalent to 8 5 % at the wall surface). The surface
temperatures and humidities, measured with the
portable data loggers, at various high-risk locations
-
were averaged over each flat and the risk of mould
growth calculated as shown in Figure 24. It can be
- -
seen that, while there was little or no danger in the
modified block, at some locations in the unmodified
block readings would suggest that mould growth was
possible. It should be emphasised, however, that the
areas chosen were the worst possible cases and that
Flat 2 Flat 3
-
Flat 4

there was negligible risk of widespread problems in Figure 24 Times of mould growth risk based on relative humidity
any of these flats. (RH)

Ventilation and air leakage ventilation rate and wind speed. At wind speeds
The ventilation rates measured with nitrous oxide in a typical of urban conditions, ventilation rates were
sixth-floor empty flat in the unmodified block are near or below 0.5 air changes per hour (ach).
plotted against wind speed in Figure 25 with an
indication of wind direction. It can be seen that, An internal thermovision survey of the east-facing
except for those occasions when the wind was from wall during an easterly wind showed that there was
the east, there was a good correlation between considerable air leakage through the joint at floor

1.6

1.4 -
Unmodified LPS flat

-
-
2 1.2-
-
c

m
% 1.0-

-
m
"
z

G 0.8 -
-
Z
a-
0.6 -
P
-
.-
m

;0.4 - N

02 -

0
Figure 25 Ventilation rates measured by tracer gas (the arrows 1 2 3 4 5 6 7 8 9
indicate wind direction) w , " d speed ( E ~ ~s i ~~ b n d(v;
e ) m/s)
(a) Floor chilled by ax leakwe at joint

(b) C h w in floor tmwrature on deoressuriaation

Fie- 26 Thermograma in unmcdified block of LPS flats

level: this cooled the floor boards next to the wall building, which is generally sheltered from driving
(Figure 26(a)). This phenomenon was investigated rain, it should not be a major route for water
further by depressurising the flat with the air-leakage penetration; a similar gap in a flat on the other side
test rig while the thermovision camera was of the block wuld be more serious. No other major
continuously recording. By subtracting the image leakage paths through the panel joints were observed
recorded after depressurisation from that recorded in this flat.
before, the temperature differences caused by the air
leakage can be clearly seen (Figure 26@)). As this gap The leakage rates at a pressure difference of 50 Pa
between the panels was on the eastern side of the measured with the pressurisation rig in three flats
Figure 27

from each block are shown in Figure 27, together with

a histogram of results from traditional housing in
England. Also shown are the results from four flats in
below theResults
Stirling. mean for traditional
all six largehousing,

were relatively airtight.

panel flats
confirming
were well

the indication from the tracer gas tests that these flats

In Figure 28 the difference between internal and

external vapour pressure is plotted against the number
of occupants for the three groups of flats; this is
indicative of the moisture load. Both blocks of flats in
Glasgow showed significant increases of vapour
pressure with occupancy, supporting the impression
that they were poorly ventilated compared with the
Stirling flats, where there was no such relationship.
=
E

i
L
~~~~
Leakage rate sf 50 Pa apDlhed pressure difference (air changer per boor)

Pressurisation tests In flats compared with national sample of traditional housing

3
2
1
0
-1

!y
6
(=I ~ ~ c i f LPS

0.6 1.2
i ~ flats

(bJ unmodified LPS flats

D P = 1.26
R = 0.709
d

1.8 2.4

+ 0.515~
3.0 3.6 4.2
Number of occupants IN1
4.8 5.4 6.0

Discussion 2 2

The two Reema blocks in Glasgow contained very m1 1

;;;;
0
stable populations, predominantly families in the
unmodified block, more single people in the modified
ern -1
2
5 0 0.6 1.2 1.8 2.4 3.0 3.6 4.2 4.8 5.4 6.0
one. These may be typical of occupancies in this type
-
a Number of occupants (Nl
a
of housing, and the data on occupancy and
m
environmental conditions can therefore be usefully c
(EJ Stirling, insulated flats
generalised. 61

In comparison with the traditional flats at Stirling,

, ;, , A ,
temperatures were high and vapour pressures low in 4
both blocks leading to low relative humidities with 3
little or no risk of mould growth (Figure 21). 2
1 A
However, once corrections were made for the unusual 0 A
outside conditions during the Glasgow tests, it was -1
apparent that the vapour pressures were not -2
0 0.6 1.2 1.8 2.4 3.0 3.6 4.2 4.8 5.4 6.0
significantly different from those at Stirling; the Number of occupants iNJ

important factor was the high temperatures, especially Figure 28 Difference between internal and external vapour
in the modified block (Figure 22). pressure plotted against number of occupants in the
three groups of flats

41
 0 1 100
I
200
I
300
Fabric transmittance (WIK)

Figure 29 Fabric transmittance and heat input for the three groups of flats,with graphs of relative humidity (%)

AS can be seen in Figure 20, the Reema blocks were 14 kg/day for a large family drying washing indoors,
unusually well insulated in comparison both with the etc. Ventilation rates will normally lie between 0.5 and
national stock and with the Stirling flats. Energy 1.5 ach.
consumption in the unmodified block was similar to
that in Stirling, leading to a similar temperature In Figure 30(a) the hatched area is well below any
distribution, with warm conditions in the heated danger of mould growth (ie 70% relative humidity);
kitchens and living-rooms and significantly lower Figure 30(b) shows that conditions become marginal
temperatures in the unheated bedrooms. for high moisture productions; Figure 30(c) shows
that there is a risk of mould for a wide range of
The better insulation standards and higher energy use occupancies in the traditional flats at Stirling. The
(near the top 10% of users (Figure 20)), in the good standards of heating and insulation in the LPS
modified block led to warm conditions in all rooms, flats in Glasgow therefore permit a much wider range
especially the bedrooms. of moisture production than the flats at Stirling where
the risk of mould growth could impose constraints on
A calculation procedure has been developed as a the occupants.
prediction technique. Given the moisture production
and ventilation rate in the dwelling and the external
conditions, the relative humidity may be calculated as Bison Wall Frame houses,
a function of fabric transmittance and heat input, as Machrihanish
shown in Figure 29. Also plotted are the mean heat
During the autumn of 1985 BRE Scottish Laboratory
input derived from the 3-monthly means in Table 5
was approached by the Property Services Agency
and the fabric transmittances for the Glasgow and
(PSA) about problems of rain penetration,
Stirling flats also in Table 5. The benefit of the high
condensation and low temperatures in the married
energy consumption in the modified flats with the
quarters at RAE Machrihanish, near Cambeltown. As
heat-with-rent scheme is evident, as the predicted
these were of Bison Wall Frame construction and
relative humidity is as low as 30%.
PSA were keen to carry out a range of modifications,
they were included in the monitoring programme.
The benefits of good heating and insulation can be
seen if the calculation procedure is inverted and the
The estate of 109 two-, three- and four-bedroomed
relative humidity is calculated as a function of
semi-detached and terraced houses was built in 1964
ventilation and moisture generation for fixed heat
in a very exposed position in the south-westernmost
input and fabric transmittance. Figure 30 shows the
tip of Scotland, less than a mile from the Atlantic.
calculated values for the three groups of flats in
The wall slabs contained 25 mm of insulation giving a
Glasgow and Stirling. The hatched boxes shown in the
U-value of 1.1 W/mZK. All the roofs contained
figure indicate the normal range of occupancy in
75 mm of insulation. Windows were single glazed
housing. Moisture production may vary from
giving an overall fabric transmittance of 140 W/K for
5 kg/day for a single person living alone, up to
 (a) ~ o d i f i s dLPS f l a b
To reduce these problems the PSA embarked on a
programme of refurbishment of the whole estate. The
.
5. 20
rn
n
rn
external walls were to be insulated with 50 mm of
r expanded polystyrene covered with a two-coat render
r 15
P with dry dash on a stainless steel lath fixed through
"z
n the insulation with polypropylene pins. This would
e
.r
la bring the U-value down to 0.4 W/m2K. The exisiting
e
2 windows were to be replaced by double-glazed units.
P 5 New porches were to be constructed to give some
protection to front and rear doors.
0
"" - "
Ventllatlon rate (air changes per hour)
- Despite the exposed position, it was felt that
(b) Unmodified LPS flats ventilation rates were low in these houses. Passive
stack ventilators were therefore to be installed leading
-i: 20
m
from each bathroom to the ridge of the roof. It had
n
,
m been hoped to install these in the kitchen also, but
-
i 15 this proved impracticable. The double-glazed units
0
" were to include trickle ventilators.
n
g 10
A pilot scheme of 11 houses was completed, except
-
P
for the double glazing, in February 1986. Work was
g 5
to follow on the remainder of the estate.
0
n . d Monitoring
Ventilation rate (air changer per hour) Monitoring at Machrihanish fell into two phases:
before occupation and afer occupation.

Before occupation
During December 1985 thermovision surveys and
tracer gas measurements were carried out in four
unoccupied houses, two modified and two
unmodified.

A survey of the remaining houses on the estate was

carried out to assess the extent and distribution of
Venf,lation rate (air changes per hour1 dampness and other problems.
Figure 30 Relative humidity (Qo) as a function of ventilation and
After occupation
. . of flats (H =
moisture generation for the three groups
heat input: A = fabric transmittance). The hatched After the modified houses had been occupied, a
boxes indicate the normal range of occupancy in further period of monitoring took place to assess the
housing environmental conditions, fuel usage and, in
particular, the risk of condensation in both the
the centre terraced houses and 171 W/K for the semi- modified and the unmodified houses.
detached and end terraced houses. All houses
originally had electric night-storage heaters, running It had been hoped to cover all the 11 modified and
on the off-peak tariff. During the year before the nine unmodified houses. However, owing to late
study these had all been replaced with modern, more completions and difficulties of access, this was
efficient storage heaters running on the cheaper White reduced to five modified and seven unmodified; these
Meter tariff. were monitored for 3 weeks in March 1986.
Unfortunately three of the modified houses had been
The two-storey houses were occupied by RAE (and occupied only a week before; the tenants may not
some USAF) personnel many of whom were stationed have had time to settle in and use the house as they
at the base for periods of a year or less. There was would in the longer term.
thus a shifting population which had no long-term
commitment to the estate. All the occupants were Thermohygrographs were left in the living-room and
relatively young with many small children. two bedrooms in each house to record the
temperatures and relative humidities. The second
Problems of rain penetration were widespread on the bedroom was selected in preference to the kitchen
estate, some houses being so bad as to be unlettable. because the kitchens opened off the living-rooms and
There were also many complaints of condensation and affected the environmental conditions there. The
mould growth, unpleasantly cold temperatures and charts were analysed as for the Glasgow study.
excessive fuel bills. Weekly energy consumptions were obtained from
F l g w 31 Th -
around window frame
meter readings at the beginning and end of each week. Thermovision survey
The fabric transmittance was calculated for each A number of interesting features emerged from the
house from plans provided by the PSA. surveys of both the unmodified and modified houses.

Hourly readings of the outside weather conditions Unmodified houses

over the whole period &ere obtained from the Both houses showed cold areas under and beside
Meteorological Office on the RAF base about 1 mile the bedroom window frames. This was caused by
from the site. rain penetration between the wall panel and
wooden frames (Figure 31).
Analysis o f data
Before occupation The one house that was uninsulated showed the
Survey expected cold areas on external comers. There were
The 109 houses on the estate were classified as also signs that the first-floor slab was acting as a
follows: thermal bridge, cooling the adjacent external walls
Empty 14 in the bedrooms.
Under rehabilitation 11
No access 6 At some stage in the past the external walls of the
Annoying condensation on windows 46 other house had been dry-limed, leaving very cold
Slight damp penetration IS areas in external corners and cold spots where
Severe damp penetration 17 metal f ~ g had s been used. The bedroom floors
were very cold near the external walls showing that
All the tenants with a problem of damp penetration the internal insulation had not cured the thermal
also complained of condensation on the windows, bridge at that point Figure 32(a)).
suggesting that vapour pressures were high in all the
houses leading to the possibility that ventilation rates There were very cold bands about 100 mm wide
were low. around the ground floors of both houses (Figure
32@)). These were caused by the lower edge of the
Most occupants said that they used the storage heaters frame of the wall panel and were an obvious site
provided. However, 21 (over a quarter interviewed) for mould growth.
said they also used Calor gas heaters: as these
generate water vapour they add to the moisture Modified houses
probIems in the houses. Twenty-six (one-third) said Most of the problems outlined above had been
that their house was too cold in winter. This may be eliminated in the modified houses, where the
an underestimate as many of those interviewed had external cladding had in general led to warmer
not lived on the estate over the coldest part of a year. more uniform temperatures on the internal
(9) Cold flow next to r4---' - -"

Figme 32 Thennogram from unmodified LPS house at Ma~hrihanish

surfaces. However, two problem areas were not lead to major condensation problems if they were
resolved. Firstly, the external cladding stopped at not rectified.
the same level as the preexisting insulation,
exacerbating the thermal bridge at this point Ventilation measurements
(Figure 33(a)). Secondly, the detail'ig of the barge The whole-house ventilation rates measured in the two
board on the gable end of one of the houses led to unm&~ed houses are plotted against the wind speed
a gap in the insulation in the top corners of both from the adjacent RAF base in F i e 34. It can be
bedrooms (Figure 33@)). Both these areas could seen that rates were very low, especially in house 2
 base of external insulation

@I Thermal bri* on gable wall

Figure 33 Thermosram from modified LPS house st Machrihanish

which was relatively sheltered in the centre of the open stack caused unusual air flows around the
estate. house, inhibiting proper mixing of the tracer; further
analysis is needed to establish the situation. At any
The measurements taken in the modified house, with rate the ventilation rates measured were still low.
and without the bathroom passive stack open, are
shown in Figure 35. It can be seen that, while the After occupation
results with the stack open are reasonable, those with Energy use
it shut are confusing, showing an apparent decrease of To put the houses in this study into the context of
ventilation rate with wind speed. It is possible that the Scottish local authority housing as a whole, annual
 1.0
0 House 1 . edge of estate
A Hause 2 , centre o i estate

2 4 6 8 10 12 14
I
W8"d speed (rn13,
I
Figure 34 Whole-house ventilation rate plotted against wind speed 0 2 4 6 8 10 12 14 16
for unmodified LPS houses at Machrihanish W8"d speed ( Y ; m1.1

Figure 35 Whole-house ventilation rate plotted against wind speed

for modified LPS house at Machrihanish

energy consumptions were estimated from the 3 weeks houses, which have larger external wall areas, in the
of meter readings and plotted against fabric modified group. Fabric transmittances will be further
transmittance (Figure 36). The lines in this figure are reduced by the double-glazed units when they are
derived from a large-scale survey of 1500 local installed. It must be remembered that the main reason
authority houses in central Scotland. As can be seen for the external cladding was to reduce the incidence
the houses in both groups were well insulated (the of rain penetration; improved insulation was
fabric transmittances were low) and were well above secondary.
the average energy users.
There was no apparent reduction in fuel usage in the
The disappointingly small reduction in fabric insulated houses. As many of them had been occupied
transmittance due to the external insulation is partially for only short periods this may not reflect longer-term
explained by the greater number of end terraced energy use patterns.

A Modified LPS house. Machrihanish

O Unmodified LPS house. Machrihanish

0 Control flats. Sflrlcng

0 Insulated flats. Stirling

20

..c

10 Lower 25%

Well ~nrulated Poorly ~nsulafed

0
0 50 100 150 200 250 300 350 400
Fabric transmltfance (WIKI

Figure 36 Energy consumption in LPS houses and traditional flats compared with large-scale
survey of Scottish local authority houses
Table 8 Summary of data from monitoring at Machrihanish

Bison Wall Frame, Mnchrihnnish

Stirling,
Modified Unmodified insulsled
n = 5 n = 7 n = 15

Mean SD Mean SD Mean SD

U-value o f walls (W/mzK) 0.4 - 1.1 - 0.5 -

Fabric transmittance W / K ) 139 20 159 17 I10 17
Energy use (kWh/week) 471 141 473 170 214 87
Number of adults/house 2.0 0 2.0 0 2.1 1.2
Number of children/house 1.5 1 .O 1.3 1.1 l .Z 1.1

F8
~ ~

Outside
Modsfled
d

0
~
Unrnod~fled
~

0
~
Stlrllng
mB

0
L

Figure 37 Average temperatures and vapour pressures from each group o f houses at
Machrihanish compared with Stirling, with graphs of relative humidity (%)

Internal conditions than the insulated bedrooms at Stirling which were

The data collected from the two groups are not heated.
summarised together with data from the insulated
flats studied at Stirling in Table 8. Vapour pressures were significantly higher in the
modified group, giving less of a reduction in relative
The average temperatures and vapour pressures from humidity than might be expected from the
each room in both groups of houses and the outside temperatures. It appears that any reduction in
conditions are plotted on a psychrometric chart in infiltration caused by the cladding was not being
Figure 37 together with the data from the insulated offset by the effect of the passive ventilators in the
flats at Stirling. bathrooms. The trickle ventilators to be installed with
the double glazing should help in this respect.
It can be seen from Figure 37 that conditions in both
groups were well away from the risk area for mould Discussion
growth (ie 70% relative humidity). Temperatures were At Machrihanish, the houses were insulated to a
high in all the rooms in the modified group; the good, but not outstanding, standard, ventilation rates
bedrooms in particular were little colder than the were low and there were specific thermal bridging
other living-rooms, showing the benefit of insulation problems. They were occupied by a shifting
and full central heating. In comparison temperatures population of service personnel some of whom had
were lower in the unmodified group, especially in the moved in immediately before the monitoring period.
bedrooms. They were, however, considerably warmer It would therefore be unwise to attempt to generalise
the data on conditions within these houses to LPS
buildings as a whole. Useful information did emerge
from this work on the application of external
insulation systems in a very exposed site to overcome
rain penetration, and on the type of detailing needed
to overcome thermal bridging.

Conclusions
The following conclusions drawn from the two case
studies can be applied to the behaviour of LPS flats
in general.

Heating
The Reema flats in Glasgow were well insulated and,
especially in the modified block, well heated. There
was therefore little risk of condensation and mould
growth.

Thermal performance
There were few thermal bridges in the Glasgow flats;
the external cladding on the modified block had
reduced the severity of those that existed. The external
insulation on the Bison Wall Frame houses at
Machrihanish had successfully upgraded the insulation
of the walls and eliminated the severe rain penetration
problems. The thermal bridging problems that
resulted can be overcome by improved detailing.

Ventilation
Both the houses and the flats were poorly ventilated.
This could lead to problems in properties with a high
moisture load and substandard heating.
Conclusions
This report provides a framework for decisions on Remedial measures for ventilation should include
remedial work in large panel system (LPS) dwellings means of providing background ventilation in living
concerned with improving habitability and, areas and the provision of means of achieving large
particularly, the avoidance of condensation and rates of extraction in moisture-producing areas such
mould growth. It has identified heating, thermal as kitchens and bathrooms. When determining
insulation and ventilation as the principal factors remedial measures, surveys should be undertaken of
concerned and has shown the importance of providing the condition of windows and doors since these will
affordable heat. greatly affect ventilation rates. If windows are ill-
fitting it is highly likely that ventilation rates will be
Heating systems, whether existing or replacement, high and this has a cost implication for the occupant.
must he able to provide temperatures sufficient for Conversely, if windows are tightly sealed and no
both the comfort of the occupants and the avoidance background ventilation is provided, relative humidities
of mould growth at a cost which is affordable by the can he high and may result in mould growth. Whilst
occupants. In LPS flats, as in most high-rise in a lot of cases visual inspection will indicate whether
dwellings, bedrooms are likely to be located or not reasonable ventilation rates are being achieved
immediately above other bedrooms and therefore need (say 0.5 to 1.5 air changes per hour) a more accurate
to be heated, since fortuitous heat gains from other check can be made by pressurisation testing. Whilst
parts of the dwelling are likely to be small. This is this is not a direct method of measuring ventilation it
totally different from the conditions in two-storey nevertheless will give a good indication of likely
dwellings where upstairs bedrooms have gains from performance.
heated living-rooms below. The report indicates how
the BRE Domestic Energy Model (BREDEM) can be A large number of LPS dwellings are of high-rise
used to assess heating costs for different target construction. This can lead to problems of ensuring
temperatures and discusses the practical options for adequate ventilation in all dwellings in a particular
improving heating systems. block, eg ventilation provision which may be effective
in top-floor flats may not be effective in ground-floor
Thermal insulation can reduce energy requirements in flats because of prevailing wind conditions. Hence the
LPS dwellings, assist in improving internal conditions location of background ventilators may be dependent
which affect habitability and increase the likelihood on the design of the block of flats. Background
that occupants can afford to heat the dwellings ventilators may need to be sheltered from prevailing
adequately. The likely effect on heating costs can be wind conditions; extract fans should be of adequate
assessed from BREDEM. Heating costs are payable size so that they can, when required, work against
by the occupants. However, capital costs are usually external pressures.
met by the building owner who in most circumstances
will not be the occupant. Only external or internal
insulation is usually feasible in LPS dwellings. The
capital cost of external insulation is usually three to
four times higher than that of internal insulation.
However, it may be economically more attractive if
undertaken at the same time as overcladding which
may be introduced to overcome such problems as rain
penetration or deterioration of the facade.

When windows need to be replaced in flats it is very

worthwhile considering double glazing, since windows
represent proportionately much more of the fabric
heal loss in those properties.

Decisions on the areas of building fabric which need

to be thermally improved require an assessment to be
made of the relative fabric heat losses and the cost of
installation, which can be dominated by accessibility.
For example, it is likely to be cost-effective to replace
parts of large single-glazed windows with insulated
panels. Similarly, it may be attractive to insulate solid
walls adjacent to stair wells where access to the
walling is relatively easy and consequently the work
can be undertaken at reasonable cost.
References
1 Anderson B R, Clark A J, Baldwin R and 14 Edwards M J. Weatherproof joints in large panel
Milbank N 0. BREDEM - BRE Domestic systems: 2 Remedial measures. BRE Information
Energy Model: background, philosophy and Paper IP9/86. Garston, BRE, 1986.
description. BRE Report. Garston, BRE, 1985.
15 Harrison H W, Hunt J H and Tbomson I.
2 Department of the Environment. Keep your home Overcladding external walls of large panel system
free from mould and damp. London, DOE, 1986. dwellings. BRE Report. Garston, BRE, 1986.
(Leaflet available from local housing authorities.)
16 Building Research Establishment. The structural
3 British Standards Institution. Energy efficiency in adequacy and durability of large panel system
housing. Part 1. Code of practice for energy dwellings. BRE Report. Garston, BRE, 1987.
efficient refurbishment of housing. British
Standard BS 8211:Part 1:1988. London, BSI,
1988.

4 British Standards Institution. A designer's manual

for the energy efficient refurbishment of housing.
London, BSI, 1989.

5 Department of the Environment. Remedial work

for existing electrically heated dwellings. Domestic
Energy Note 3. London, DOE, 1978.

6 Electricity Council. The DEN 3 book. London,

Electricity Council, 1986.

7 Warren P R and Webb B C. The relationship

between tracer gas and pressurisation techniques
in dwellings. Proceedings of the First Air
Infiltration Centre Conference, October 1980.

8 Building Research Establishment. Wind

environment around tall buildings. BRE Digest
141. Garston, BRE, 1972.

9 Stephen R K. Determining the airtightness of

buildings by the fan-pressurisation method: BRE
recommended procedure. BRE Occasional Paper.
Garston, BRE, 1988.

10 Jones P J and O'Snllivan P E. The role of trickle

ventilators in domestic ventilation design.
Supplement to Proceedings of the Seventh Air
Infiltration and Ventilation Centre Conference,
September/October 1986.

11 Thermal insulation - determination of

airtightness of buildings. Issue 1. Swedish
Standard SS 02 15 5 1. BST
Byggstandardiseringen,
Standardiseringskommissionen i Sverige, 1980.

12 The Building Standards (Scotland) Regulations

1981. Statutory Instrument 1981 No 1596 (S 169)
Building and Buildings. London, HMSO, 1981.

13 British Standards Institution. Code of practice for

performance and loading criteria for profiled
sheeting in building. British Standard BS
5427:1976. London, BSI, 1976.