Climate change and the indoor environment: impacts and adaptation CIBSE TM36: 2005 The Chartered Institution of Building Services Engineers 222 Balham High Road, London SW12 9BS

The rights of publication or translation are reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means without the prior permission of the Institution. ��February 2005 The Chartered Institution of Building Services Engineers London Registered charity number 278104 ISBN 1 903287 50 2 This document is based on the best knowledge available at the time of publication. However no responsibility of any kind for any injury, death, loss, damage or delay however caused resulting from the use of these recommendations can be accepted by the Chartered Institution of Building Services Engineers, the authors or others involved in its publication. In adopting these recommendations for use each adopter by doing so agrees to accept full responsibility for any personal injury, death, loss, damage or delay arising out of or in connection with their use by or on behalf of such adopter irrespective of the cause or reason therefore and agrees to defend, indemnify and hold harmless the Chartered Institution of Building Services Engineers, the authors and others involved in their publication from any and all liability arising out of or in connection with such use as aforesaid and irrespective of any negligence on the part of those indemnified. Layout and typesetting by CIBSE Publications Printed in Great Britain by Page Bros. (Norwich) Ltd., Norwich, Norfolk NR6 6SA Cover illustration: Winter Garden at Canary Wharfe East (artist���s impression). Reproduced by courtesy of Cesar Pelli & Associates/rendering by dBox. Printed on 100% recycled paper comprising at least 80% post-consumer waste

Foreword New buildings are designed to last for a significant period and although many will be substantially refurbished during their life spans, their envelopes will remain significantly unchanged. This fact imposes severe limitations on how the building can be modified to take account of changing climatic conditions. Historically, this was of little importance because changes in the climate were insignificantly small but the widely acknowledged onset of global warming has changed this scenario. The climatic information currently published by CIBSE is historical and does not take into account our changing climate. Using historical data when significant increases in temperature are forecast due to global warming is clearly not sensible and the Institution was therefore keen to sponsor a research project which indicates the expected increased temperatures, the effect they will have on current buildings, the way in which these effects can be mitigated and advice on how buildings may be better designed and serviced in the future. The research also was seen as providing a guide to the information which the Institution should provide to designers to enable them to take account of the changing conditions. Although this may appear to be straightforward and logical it will require the Institution to base its climatic data on forecasts not historical facts. Because there is considerable uncertainty about the severity of global warming CIBSE will probably publish future weather information using all the UKCIP02 climate change scenarios, but ���morphed��� in a manner similar to that pioneered in this research project. Choice of the appropriate scenario will then be a matter of judgement by the design team. The Project Steering Group was multi-disciplinary and this led to vigorous and well informed debate. However, there was total unanimity in deciding that one single report serving all in the design team should be published, emphasising the fact that good environmental design could only be achieved through a team approach. Hopefully, architects, engineers, planners, surveyors, developers, and all those concerned with buildings, will find this publication instructive and useful. Brian Moss Chairman, TM36 Project Steering Group Acknowledgements This work was co-funded by Arup and the Department of Trade and Industry through the Department of Trade and Industry Partners in Innovation Scheme. The CIBSE is grateful to the project steering group for contributing their valuable time to attend eight one-day workshop sessions held over the course of the project and numerous hours reading and commenting on reports. Their input greatly assisted the shape and direction of the work. The project team is also grateful to the following individuals: Joanne Koenig (Sustainable Homes), Richard Daniels (Department for Educations and Skills), Bill Dunster (Bill Dunster Architects ZEDfactory), David Dennis (Property and Design, London Borough of Newham), Stefan Waldhauser (Waldhauser Haustechnik AG) and, from Arup, Barry Austin, Duncan Campbell, Guy Channer, Alistair Guthrie, Alan Jefcoat, Ian Munro, Roger Olson, Tim Thornton, Chris Twinn and Ken Wiseman. The CIBSE acknowledges the endorsement of this publication by the Royal Institute of British Architects and the Royal Town Planning Institute. Crown copyright material is reproduced with the permission of the Controller of HMSO and the Queen���s Printer for Scotland under licence no. C02W0002935.

TM36 Project Steering Group Prof. Brian Moss (CIBSE) (chairman) Martin Best (Hadley Centre, Met Office) Dr Richenda Connell (UK Climate Impacts Programme) Dr Hywel Davies (CIBSE) Dr Chris Gordon (Hadley Centre, Met Office) Hilary Graves (Building Research Establishment) Prof. Vic Hanby (Institute of Energy and Sustainable Development, De Montfort University) George Henderson (WS Atkins on behalf of DTI) Dr Gary Hunt (Department of Civil and Environmental Engineering, Imperial College) Prof. Susan Roaf (Department of Architecture, Oxford Brookes University) Edward Williams (Hopkins Architects) Dr Clare Goodess (Climatic Research Unit, University of East Anglia) Principal authors Dr Jacob Hacker (Arup) Prof. Michael Holmes (Arup) Prof. Stephen Belcher (Arup and University of Reading) Dr Gavin Davies (Arup) Arup Project Team Project Director: Dr Gavin Davies Project Manager: Dr Jacob Hacker Thermal modelling: Prof. Michael Holmes, Dr Jacob Hacker and Michael Edwards Climate scenarios: Prof. Stephen Belcher, Dr Jacob Hacker and Daniel Powell Editor Ken Butcher CIBSE Editorial Manager Ken Butcher CIBSE Research Manager Dr Hywel Davies CIBSE Publishing Manager Jacqueline Balian Note from the publisher This publication is primarily intended to provide guidance to those responsible for the design, installation, commissioning, operation and maintenance of building services. It is not intended to be exhaustive or definitive and it will be necessary for users of the guidance given to exercise their own professional judgement when deciding whether to abide by or depart from it.

Contents Summary 1 Introduction 2 The climate scenarios 2.1 UKCIP02 scenarios 2.2 Use of the UKCIP02 scenarios for environmental design 2.3 UKCIP02 climate changes for UK sites 3 Performance indicators 3.1 ���Overheating��� criteria 3.2 Energy usage 4 What does the future look like? 4.1 Temperatures in future climate 4.2 Space heating 4.3 Risk of summertime overheating 4.4 Comfort cooling 4.5 Performance of air conditioning systems 5 Case studies: detailed assessment of existing building types 5.1 Introduction 5.2 Dwellings 5.3 Offices 5.4 Schools 5.5 Other locations: Manchester and Edinburgh 5.6 Other emissions scenarios 5.7 Ventilation control for the advanced naturally ventilated buildings 6 Adaptation strategies 6.1 Dwellings 6.2 Offices 6.3 Schools 7 Conclusions 7.1 Passive measures 7.2 Mechanical cooling 7.3 Final conclusions References Annex: data sheets for case studies D1 19th century house D2���D4 New-build house D5 1960s flat D6 New-build flat O1 Naturally ventilated 1960s office O2 Modern mixed-mode office O3 Mechanically ventilated high thermal mass office O4 Advanced naturally ventilated office O5 Fully air conditioned office S1 1960s school S2 Advanced naturally ventilated school 1 3 4 4 6 6 6 6 10 10 10 11 13 14 14 15 15 16 17 17 17 17 18 19 20 23 24 25 26 27 28 28 29 30 32 36 38 40 42 44 46 48 50 52

Plain text is unavailable for this page.

1 Summary There is compelling scientific evidence that our climate is changing, and it is probable that average temperatures will increase by several degrees over the coming century. These increases in temperature are expected to have a major impact on the indoor environment of buildings in the UK. Key questions are: ��� To what extent will climate change increase the occurrence of summertime thermal discomfort and ���overheating���? ��� To what extent will passive measures be able to improve summertime thermal comfort and amelio- rate the increased propensity for overheating? ��� How effective will different approaches to comfort cooling be under the changing climate? ��� What are the energy use implications of the various strategies? This publication addresses these questions through dynamic thermal computer modelling of 13 case study buildings, chosen to provide a cross section of UK building types, including dwellings, schools and offices, and illustrate a range of different approaches to comfort cooling provision. The starting point is the set of UKCIP02 Climate Change Scenarios for the United Kingdom(1), which provides the best currently available scientific projections for UK climate over the coming century. These scenarios indicate a warming climate through the century. For example, maximum temperatures in mid-summer in London are projected to increase by between 3.6 and 6.9 K above the 1961���1990 average by the latter part of the century. To assess the performance of both mechanical and passive design, detailed climatic information is required. For the present day, CIBSE has produced a set of ���weather years���, derived from Met Office data, with climatic variables recorded at hourly intervals for three sites(2): London, Manchester and Edinburgh (hereafter referred to as the ���CIBSE/Met Office weather years���). For each site there are two recommended years: a Test Reference Year (TRY) which represents average conditions and is typically used for energy use predictions, and a Design Summer Year (DSY) which is an actual year with a ���near-extreme��� summer and is typically used for overheating risk assess- ment and cooling system sizing. The CIBSE/Met Office weather years are drawn from the period 1976���1995 and so may be thought of as representing the UK climate of the ���1980s���. In order to provide similar data for the future, the UKCIP02 predictions for changes to mean climate are here combined with the CIBSE/Met Office weather years to produce synthetic TRY and DSY weather years for the 2020s, 2050s and 2080s timeslices using a technique known as ���morphing���. The ���morphed��� future weather years have the mean properties of the monthly climate of the UKCIP02 scenarios but the hour-to-hour weather variability of the CIBSE/Met Office weather years. This variability might change in the future, in addition to the changes in mean climate, and so the new weather years provide only a first order assessment. However, this assessment is one that gives valuable insight into the likely impacts of climate change on the indoor environ- ment. The quantitative dynamic thermal modelling has focussed on London under the present-day CIBSE DSY (1989) and its synthetic future counterparts under the UKCIP02 Medium-High climate change scenario. The results of this analysis suggest that: ��� In some buildings it is not possible to meet the comfort targets used in this study using only ventilation cooling by opening windows (even for the ���present-day��� 1980s climate of the CIBSE DSY) because of the high external temperatures in summer. However, advanced passive cooling measures were found to enable the targets to be met in a number of cases. ��� For dwellings, the results suggest that buildings with very good control of solar shading, ventila- tion and internal heat gains can meet targets until the 2050s. Further benefits were found to be provided by the use of high thermal mass con- struction elements. In living areas, use of high mass construction enabled the performance targets to be met into the 2080s and had the particularly desirable effect of reducing peak space tempera- tures. However, some problems were found in bedrooms by the 2080s. A simulation of a medium thermal mass house with air conditioning installed indicated significant increases in carbon emissions due to the use of the cooling system. ��� For offices, a range of buildings have been considered. The HVAC systems for the offices include both passive and mechanical ventilation, passive and mechanically assisted thermal mass cooling, low energy mechanical cooling and full air conditioning. The modelling suggests it would be difficult as the climate warms to meet the thermal performance targets considered here using passive measures alone. A mixed mode approach, in which mechanical systems are available at times of peak cooling need may be the most practical way to achieve the performance targets considered here. In mixed-mode buildings, the use of energy for HVAC services will be, as now, largely determined by the standard of selection, design, maintenance and management of systems. Climate change and the indoor environment: impacts and adaptation

2 ��� For schools, the high internal heat gain from classroom occupants together with the high fresh air ventilation rates required to maintain good air quality means that as the external air temperature increases, it becomes increasingly difficult to achieve comfort standards through use of passive systems alone. The results suggest that, as for offices, a move to a mixed mode approach may be the most practical way to achieve the thermal performance targets considered here. ��� In all the case studies, the warmer climate con- ditions point to the need to limit summertime heat gains to spaces as far as possible as the first and most energy efficient measure to reduce the need for mechanical comfort cooling. This means employing solar shading, reducing the density or power output of lights, machines and possibly the density of occupants, and providing the ability to reduce ventilation to minimum levels during hot periods of the day. For buildings with exposed thermal mass it also means enabling the spaces to be purged with cool air at night and during periods of cooler weather to maximise the capacity for passive heat absorption by the building fabric. In Manchester and Edinburgh the overheating risk has been found to be lower. Broadly, overheating in Manchester during the 2050s and 2080s is similar to London during the 1980s and 2020s respectively. Overheating in Edinburgh during the 2050s and 2080s is similar to Manchester in the 1980s and 2020s respectively. A key concern with respect to the indoor environment of buildings is the potential for significant increases in build- ing energy consumption due to the use of mechanical comfort cooling systems. This increased use of mechanical systems will hamper efforts to reduce greenhouse gas emissions and limit climate change. It is estimated that buildings account for approximately 45% of total energy consumption in the UK(3) and 41% across the European Community(4). There is, therefore, considerable potential to reduce emissions through good practice in building design and methods of use, e.g. by up to 50% for new buildings and following major refurbishment(2). The investigation was based upon a set of performance criteria and because relative performance was the main interest it was not essential to set absolute targets. It is probable that future Building Regulations will require engineers and architects to demonstrate the need for mechanical cooling and air conditioning. In that case it will be essential that the industry has: ��� appropriate overheating risk criteria ��� a standardised calculation method so that all designers can obtain the same predictions. ��� standardised climatic data ��� standardised methodology for performance predic- tion. These issues are outside the scope of this study, but are likely to be addressed in the future by the CIBSE and others. References 1 Hulme M, Jenkins G J, Lu X, Turnpenny J R, Mitchell T D, Jones R G, Lowe J, Murphy J M, Hassell D, Boorman P, McDonald R and Hill S Climate Change Scenarios for the United Kingdom: The UKCIP02 Scientific Report (Norwich: Tyndall Centre for Climate Change Research, School of Environmental Sciences, University of East Anglia) (2002).(available from http://www.ukcip.org.uk/resources/publications) 2 Weather, solar and illuminance data CIBSE Guide J (London: Chartered Institution of Building Services Engineers) (2002) 3 Energy efficiency in buildings CIBSE Guide F (London: Chartered Institution of Building Services Engineers) (2004) 4 Directive 2002/91/EC of the European Parliament and of the Council of 16 December 2002 on the energy performance of buildings Official J. of the European Communities L1/65 (4.1.2003)

3 1 Introduction This publication addresses the issue of how climate change in the UK over the 21st century may affect summertime thermal comfort in buildings and the energy use of associated heating, ventilation and air conditioning (HVAC) systems. There is compelling scientific evidence that our climate is changing and it is probable that average annual tempera- tures will increase by several degrees during this century. Changes in climate will impact upon the energy used for heating and cooling in buildings, may cause overheating in naturally ventilated buildings and affect the ability of low energy cooling systems to provide comfortable conditions. Many buildings, particularly dwellings, are designed to last for several decades and longer consid- eration of climate change issues is therefore necessary now to ensure the longevity of the building stock. Not to do so will result in a generation of buildings that are likely to become obsolete within their useful lifetime, or require costly and difficult retrofits. Designing for the anticipated future climate is therefore very much a current issue. Until now, however, there has been little information available regarding the magnitude of these effects. While mechanical air conditioning is an obvious tech- nological solution to adapt to the warming climate, this route is undesirable for two reasons. First, inclusion, retrofitting and maintenance of air conditioning in many buildings is likely to be beyond the bounds of economic viability. This is particularly important in the domestic sector in which the very young, old or physically infirm are likely to suffer greatest harm from thermal discomfort and heat stress. Secondly, and perhaps more fundamen- tally, use of air conditioning has the potential to increase significantly the energy burden of, and consequently the greenhouse gas emissions from, a building, thereby exacerbating the problem for which the adaptation is needed. It is estimated that buildings account for approximately 45% of total energy consumption in the UK(1) and 41% across the European Community(2). There is, therefore, considerable potential to reduce emissions through good practice in building design and methods of use, e.g. by up to 50% for new buildings and following major refurbish- ment(2). This publication aims to address these issues by providing guidance on measures to ensure summertime thermal comfort in UK buildings without incurring excessive energy use. There are a number of key questions: ��� To what extent will climate change increase the occurrence of summertime thermal discomfort and ���overheating���? ��� To what extent will passive measures be able to improve summertime thermal comfort and amelio- rate the increased propensity for overheating? ��� How effective will different approaches to comfort cooling be under the changing climate? ��� What are the energy use implications of the various strategies? These questions are addressed here by quantitative assess- ment of the effect of climate change on building and HVAC system performance, measured by the frequency of overheating, energy consumption and carbon emissions. The risks posed by climate change to these performance measures are assessed in two ways. First, properties of the future climate are examined to provide an initial, qualita- tive, assessment. Secondly, dynamic thermal modelling is used to make quantitative assessments of case study buildings drawn from three generic building types: dwellings, offices, and schools. The case study buildings are chosen to illustrate the response of different HVAC strategies, including manually operated natural ventila- tion, full mechanical air conditioning, and passive and low energy methods. An important aspect of the study was to analyse the performance of design features that successfully cool buildings without mechanical means, e.g. the control of solar radiation and ventilation, or the use of thermal storage. A number of approaches currently used in the UK and other parts of the world were applied to the case study buildings and tested under present and future climates. Novel techniques such as embodying phase change materials within the building fabric were not considered, because the objective was to examine what can be done with existing technology. Similarly it was assumed that there will not be significant changes in modes of building use, including internal heat gains and occupation patterns, over the time periods considered. Following this introduction, the structure of the docu- ment is as follows: ��� Section 2: describes the climate change scenarios used and the method used to produce design weather years for projected future climates. ��� Section 3: discusses design targets for thermal performance and energy use. ��� Section 4: describes some of the general impli- cations of the climate changes on the performance of different types of building based on the charac- teristics of the future weather years. ��� Section 5: forms the core of the document and presents the results of the dynamic thermal modelling of the case study buildings. Climate change and the indoor environment: impacts and adaptation

4 Climate change and the indoor environment: impacts and adaptation ��� Section 6: considers further strategies and remedial options for those buildings where limitations to performance have been identified. ��� Section 7: presents the conclusions. ��� Annex: contains the data sheets for the case study buildings and the results of thermal modelling. 2 The climate scenarios 2.1 UKCIP02 scenarios In 1998 the United Kingdom Climate Impacts Programme (UKCIP) released the first set of comprehensive climate change scenarios for the United Kingdom. This was done in recognition of the need to make quantitative assess- ments of the possible impacts of climate change. These scenarios were subsequently updated in 2002 as the ���UKCIP02��� scenarios(3). (A further update is expected in 2007/8.) The principal changes in the 2002 scenarios are that (a) they make use of the more recent Met Office global climate model (HadCM3) and (b) they contain information from a regional model (HadRM3) embedded within the global climate model with a resolution of 50 km. The scenarios are being widely used to assess the possible impacts of climate change on the UK (see www.ukcip.org). It is likely that the scenarios will be further refined and developed in the future, but at present they represent the best available information on the likely course of climate change in the UK over the 21st century. Some types of climate scenario are excluded, e.g. sudden or gradual cooling of the northern hemisphere due to changes in the Gulf Stream. However, these types of climate change are considered to be of very low probability within the next 100 years and lie outside the range of scenarios presently being considered in climate change impacts adaptation and planning. The following is a brief outline of how the scenarios were produced full details are available in Hulme et al.(3) 2.1.1 Emissions scenarios The basis for the UKCIP02 climate scenarios is a set of four ���storylines��� for greenhouse gas emissions, which are taken from the Intergovernmental Panel on Climate Change (IPCC) SRES emissions scenarios. Each storyline represents a possible future, as described in Table 2.1, ranging from one relatively intensive in fossil fuel use and greenhouse gases emissions, to one in which sustainability is given high priority on a global level and fossil fuel use decreases. Figure 2.1 shows the predicted changes in atmospheric carbon dioxide over the coming century under each of the scenarios. These changes in atmospheric composition are computed independently of the climate models. They form the input ���forcing��� to the climate models, which then aim to calculate the resulting future climates . Note that even under the Low Emissions (Global Sustainability) scenario, atmospheric carbon dioxide continues to increase until around the middle of the century due to the projected timescale to phase out fossil fuel use. In the scenarios it is therefore anticipated that there will be an appreciable level of climate change over the course of the century even if substantial efforts are made now to reduce greenhouse gas emissions. 2.1.2 The global climate model Predictions for global temperature change in UKCIP02 were obtained in the following way. First, the global climate model was run for the period from 1860 (a nominal pre-industrial starting point) until 1990 using observed changes in greenhouse gases and other natural forcings of climate change such as volcanoes. The data for the thirty-year period 1960���1990 were averaged to form the ���baseline��� climate. Next, the global climate model was run forward until 2100 for each of the four emissions scenarios. Values of global average temperature in the runs are shown in Figure 2.2. Finally, these data were averaged over three 30-year timeslices: the 2020s, 2050s and 2080s Table 2.1 Characteristics of the UKCIP emissions scenarios (from tables A.2 and A.3 of the UKCIP02 report(3)) UKCIP02 climate change IPCC SRES UKCIP socio-economic Description scenario emissions storyline scenario title Low Emissions B1 Global Sustainability Clean and efficient technologies reduction in material use global solutions to economic, social and environmental sustainability improved equity population peaks mid- century Medium-Low Emissions B2 Local Stewardship Local solutions to sustainability continuously increasing population Medium-High Emissions A2 National Enterprise Self-reliance preservation of local identities continuously increasing population economic growth on regional scales High Emissions A1F1 World Markets Very rapid economic growth population peaks mid- century social, cultural and economic convergence among regions market mechanisms dominate. 1960 1980 2000 2020 2040 2080 2060 A1F1 A2 B2 B1 2100 1000 900 800 700 600 500 400 300 200 1000 900 800 700 600 500 400 300 200 Carbon dioxide concentration / ppm Figure 2.1 Global carbon dioxide increases (reproduced from UKCIP02 Scientific Report(1) Crown copyright)

The climate scenarios 5 corresponding to the periods 2011���2040, 2041���2070 and 2071���2100, respectively. The four emissions scenarios and three timeslices in UKCIP02 make a total of twelve climate examples to consider. Dealing with the complete set of scenarios is therefore a considerable undertaking. However, the scenarios are mathematically linked and the differences between them are proportional. The proportionality is given by a ���climate scaling factor��� (CSF), which is defined as the ratio of the global average temperature change in a scenario relative to that in the Medium-High 2080s scenario (the CSF is called the ���pattern scaling factor��� in UKCIP02). The scenarios are listed in Table 2.2 in order of increasing average global temperature change and CSF. The CSF values in this table may be used to relate the climate changes under a given scenario to those in the Medium-High Emissions 2080s scenario which has a CSF of 1.0. A graphical comparison of CSFs is shown in Figure 2.3. It can be seen that in the 2020s the level of climate change in the four emissions scenarios is similar, which is because the levels of CO2 in the atmosphere are similar at this time (Figure 2.1). By the 2050s timeslice, however, the four scenarios are starting to diverge, with differences being quite appreciable by the 2080s timeslice. For example in the 2080s the climate scaling factor associated with the High Emissions scenario is around twice that of the Low Emissions scenario. Figure 2.3 also indicates that the climate scaling factor of different scenarios is similar at different timeslices. For example, the level or warming in the Low Emissions scenarios 2080s is similar to that in the Medium-High scenario 2050s, and that in the Medium- Low scenario 2080s similar to that in the High scenario 2050s. 2.1.3 The regional climate model The size of the computational grid boxes for the global climate model (HadCM3) is approximately 300 km over the UK. This spatial resolution is too coarse to resolve the geographical variations due to factors such as topography and coastline morphology. To produce such information, a ���regional climate model��� (HadRM3), covering only the UK and part of northern Europe was used. The regional model takes boundary conditions from the global climate model and the size of the computational grid boxes was approximately 50 km. Running the regional climate model is computationally intensive, requiring several months of ���super-computer��� power. For this reason only a limited number of model runs were made. All the regional detail in the UKCIP02 scenarios is based on regional model runs for the Medium-High Emissions scenario 2080s and the baseline 1961���1990 climate. Results for the present day climate were then subtracted from the 2080s results, giving the change in the climate parameters across the UK on a 50 km grid. The philosophy adopted in UKCIP02 is that the geographical variations in climate changes across the UK are the same for all scenarios but vary in magnitude in direct propor- tion to the global average temperature change. To obtain regional climate changes for the other scenarios and timeslices, the changes for the Medium-High Emissions scenarios are simply multiplied by the CSF values given in Table 2.2. This method is called ���pattern scaling���. The resulting UKCIP02 climate scenarios contain monthly averaged values of climate variables recorded on the 50 km computational grid. The variables available are: ��� temperature (daily average, maximum and mini- mum dry-bulb) ��� total precipitation ��� snowfall rate ��� 10 m wind speed ��� relative and specific humidity ��� total cloud in the longwave radiation band ��� net surface long and shortwave radiation ��� total downward shortwave radiation ��� soil moisture content ��� mean sea-level pressure ��� surface latent heat flux. 1850 1900 1950 Observations 2000 2050 2100 6 4 2 0 -2 T emperature change / K A2 A1F1 B2 B1 Figure 2.2 Predictions of annual average temperature in the UKCIP02 global climate model runs (Crown copyright) Table 2.2 UKCIP02 scenarios ranked by magnitude of global average temperature change and the derived climate scaling factor (CSF) Average global Climate Emissions scenario Timeslice temp. change scaling relative to factor (CSF) 1960���1990 0.79 0.24 Low 2020s 0.88 0.27 Medium-Low 2020s 0.88 0.27 Medium-High 2020s 0.94 0.29 High 2020s 1.4 0.43 Low 2050s 1.6 0.50 Medium-Low 2050s 1.9 0.57 Medium-High 2050s 2.0 0.61 Low 2080s 2.2 0.68 High 2050s 2.3 0.71 Medium-Low 2080s 3.3 1.0 Medium-High 2080s 3.9 1.18 High 2080s 1��2 1��0 0��8 0��6 0��4 0��2 0 High Medium-high Medium-low Low Emissions scenario 2020s 2050s 2080s Figure 2.3 The range of climate scaling factors in the UKCIP02 scenarios

6 Climate change and the indoor environment: impacts and adaptation 2.1.4 Uncertainties in the UKCIP02 scenarios The climate projections in the UKCIP02 scenarios are subject to a number of uncertainties beyond the uncertain- ties in the emissions scenarios. A discussion of these uncertainties is given by Jenkins and Lowe(5). An important point to recognise is that the UKCIP02 scenarios were based on just one climate modelling framework, that of the Hadley Centre. Other climate models in other countries would yield somewhat different rates and patterns of climate change for the UK. However, the Hadley Centre model is one of the best validated models in the world and the UKCIP02 scenarios are the climate change scenarios approved for use by the Department of Environment, Food and Rural Affairs (DEFRA). 2.2 Use of the UKCIP02 scenarios for environmental design Not all of the variables contained in the UKCIP02 scenarios correspond directly to those needed for environmental design, but relevant parameters may be derived. More fundamentally, while the scenarios contain values for changes in monthly averaged values of climate variables, environmental design and HVAC system sizing need information regarding extremes and hour-to-hour variability. This type of information is typically not directly available from climate models. This is a common problem in climate change impacts assessment known as ���temporal downscaling���. An additional problem, ���spatial downscaling���, is that while the UKCIP02 scenarios data are at relatively high resolution, the grid box containing the location of the building may not be truly repre- sentative of local microclimate effects such as unresolved topography, local land use and urban heat island effects. The weather data were collected at airports and so have a local microclimate characteristic of an urban area. For example, London (Heathrow) has a maximum ���heat island��� of about 5 K, which compares with a maximum heat island in central London of about 6 K. To address the spatial and temporal downscaling problems use is made here of the temporal and spatial information contained in the CIBSE/Met Office weather years for London, Manchester and Edinburgh(4). All these weather years have been combined with the UKCIP02 scenarios for monthly climate changes for the three sites, thereby producing synthetic future weather years. The future weather years contain the diurnal variations and vari- ability of the present day, and the microclimate of an urban area, but the average climatic properties (e.g. daily average temperature, solar irradiance, wind speed etc.) of the UKCIP02 scenarios. This method is referred to here as ���morphing��� as it involves shifting and stretching the present-day weather time series to produce new weather time series with the required monthly climate statistics. The full details of the method used here are described in Belcher et al.(6) 2.3 UKCIP02 climate changes for London, Manchester and Edinburgh The CIBSE/Met Office weather years span 1976���1995, and may be considered here to constitute a ���1980s��� timeslice. This is the baseline climate onto which the UKCIP02 changes are applied*. Figure 2.4 shows the baseline monthly mean values of some important climate variables in the 1980s for London, Manchester and Edinburgh. As expected, London is slightly warmer than Manchester, which is in turn slightly warmer than Edinburgh. The three locations received comparable solar irradiance. Figure 2.5 (see page 8) shows the changes to the monthly mean values of a number of key variables taken from UKCIP02 for the 2080s Medium-High Emissions scenario. The changes for other scenarios may be obtained by multiplying these changes by the CSF values, shown in Figure 2.3. The greatest changes are in temperature, particularly in summer and in London. There are also appreciable increases in solar irradiance in summer (principally due to reduced cloud cover). Air moisture content increases in winter and decreases in late summer and autumn, but relative humidity is reduced in all seasons due to the increase in temperature, decreasing quite sharply in summer. Average wind speeds show smaller magnitude changes, typically less than 5%, increasing in winter and decreasing in summer. 3 Performance indicators In order to assess the impacts of climate change discussed in section 1, performance indicators are defined, based on: ��� the level of summertime thermal performance ��� associated changes in energy consumption and carbon emissions. These two aspects are discussed below. 3.1 Summertime thermal performance Summertime thermal performance is usually measured against a criterion expressed in terms of a benchmark temperature that should not be exceeded for a designated number of hours or percentage of the year. The bench- mark temperature is usually related to a temperature at which occupants begin to feel thermal discomfort, although may be related to other factors, such as produc- tivity or health. When the benchmark temperature is exceeded, the building is said to have ���overheated��� and if this occurs for more than the designated amount of time, the building is said to suffer from ���overheating���. Consequently the design target is called an ���overheating criterion���. In the UK, there is no universally agreed overheating criterion for buildings with the exception of schools(7), to which standard Building Regulations Approved Document L2 now refers(8). Other countries, e.g. Germany(9), have fixed standards for overheating in offices. In the UK, thermal performance targets for offices and many other buildings types are decided upon on a project-by-project basis, through discussion between the design team, the client, and the other building stakeholders. *The CIBSE/Met Office data, which cover the period 1976���1995 and are used here as the base period, are about 0.3 K higher than the UKCIP02 base period, which is 1961���1990. Hence the ���morphing��� here leads to a slight exaggeration of the climate change.

Performance indicators 7 For the present study, thermal performance benchmark temperatures have been chosen for the case study building types. The temperature thresholds are based on thermal comfort models, which are discussed in section 3.1.1 below. It should be noted, however, that the temperature thresholds are only intended to be illustrative and are not advocated here as being universally appropriate. In each case, two temperature thresholds have been defined: a lower temperature threshold, which is taken to indicate when occupants will start to feel ���warm���, and a higher threshold temperature, which is taken to indicate when occupants will start to feel ���hot���. Using two tempera- ture benchmarks is helpful, as it is likely that occurrences of either intense periods of hot conditions or more prolonged periods of warm conditions can have an equally detrimental impact on building users. In section 5, the percentage of occupied hours that the two threshold temperatures are exceeded are displayed graphically so that an assessment can be made of the degree to which the building is predicted to overheat. However, to define a fixed measure of ���overheating���, an exceedance of more than 1% of occupied hours in a year over the higher temperature benchmark has been adopted to indicate a failure of the building to control overheating risk. The benchmark temperatures for each of the buildings are given in Table 3.1. They are discussed further below. 3.1.1 Adaptive and deterministic thermal comfort models The majority of research on thermal comfort in buildings has taken one of two approaches to the specification of comfort conditions: (a) deterministic methods (e.g. Fanger(10)), which relate given space conditions, e.g. in terms of tempera- ture, humidity and air speed, and given clothing and activity levels, to the likely level of occupant comfort (b) adaptive methods (e.g. Brager and de Dear(11)), that are empirically based on the outcomes of occu- pancy surveys, and aim to capture the variation in comfort expectations with different climates. Typically, the level of thermal discomfort in both types of model is expressed as the ���percentage of persons dis- satisfied��� (PPD). Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 30 25 20 15 10 5 0 Daily minimum temperature / �� C Jan Feb Mar Apr May Jun (b) (a) Jul Aug Sep Oct Nov Dec 30 25 20 15 10 5 0 Daily maximum temperature / �� C Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 250 200 150 100 50 0 Average solar shortwave irradiance / (W/m 2 ) Jan Feb Mar Apr May Jun (d) (c) Jul Aug Sep Oct Nov Dec 30 25 20 15 10 5 0 Daily average temperature / �� C London Manchester Edinburgh London Manchester Edinburgh London Manchester Edinburgh London Manchester Edinburgh Figure 2.4 Average ���baseline��� climate of the CIBSE/Met Office ���1980s��� period (a) daily maximum temperature, (b) daily minimum temperature, (c) daily average temperature, (d) solar shortwave irradiance Table 3.1 Benchmark temperatures and overheating criteria Building type ���Warm��� threshold ���Hot��� threshold Overheating criterion temperature / ��C temperature / ��C Dwellings: ��� living areas 25 ��C 28 ��C 1% occupied hours over 28 ��C ��� bedrooms 21 ��C 25 ��C 1% occupied hours over 25 ��C Offices 25 ��C 28 ��C 1% occupied hours over 28 ��C Schools 25 ��C 28 ��C 1% occupied hours over 28 ��C

8 Climate change and the indoor environment: impacts and adaptation The adaptive approach offers a way to relate acceptable space conditions to those found outside. Figure 3.1 shows the ASHRAE Standard 55.2004(12) adaptive comfort model relationship between comfort threshold temperatures and the average monthly external temperature. Comfort thresholds for both warm and cold discomfort are shown for comfort levels of 80% and 90% PPD. The shaded band of temperatures between the two curves for each case gives the comfort temperature band. The internal temperatures on the y-axis should be taken to be ���operative��� temper- ature, which is the average of the air temperature and the room surface temperatures, since this provides a better indication of comfort than the air temperature alone, as the radiative heating or cooling from surfaces is taken into account. The outside temperature is the ambient dry bulb temperature. Note that no account of humidity is made in the model, as adaptive studies have not established a clear link between comfort experience and humidity level. The Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 7 6 5 4 3 2 1 0 T min change / K Jan Feb Mar Apr May Jun (b) (a) Jul Aug Sep Oct Nov Dec 7 6 5 4 3 2 1 0 T max change / K Relative humidity change / % Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 -2 ���4 ���6 ���8 ���10 ���12 ���14 Jan Feb Mar Apr May Jun (d) (f) (c) Jul Aug Sep Oct Nov Dec 35 30 25 20 15 10 5 0 Solar irradiance change / (W/m 2 ) London Manchester Edinburgh London Manchester Edinburgh Moisture content change / % (e) Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 25 20 15 10 5 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 8 6 4 2 0 ���2 ���4 ���6 Wind speed change / % London Manchester Edinburgh London Manchester Edinburgh London Manchester Edinburgh London Manchester Edinburgh Figure 2.5 Changes in monthly average climate variables for 2080s predicted under UKCIP02 Medium-High scenario changes for the other emissions scenarios and timeslices can be obtained by multiplying by the climate scaling factors given in Figure 2.3 (a) maximum temperature, (b) minimum temperature, (c) solar irradiance, (d) relative humidity, (e) wind speed, (f) moisture content 0 40 35 30 25 20 8 out of 10 satisfied 9 out of 10 satisfied 15 10 5 Indoor comfort temperature, T / �� C Mean monthly outdoor air temperature / ��C 34 32 30 28 26 24 22 20 18 16 14 Figure 3.1 Adaptive comfort model (after ASHRAE(12))

Performance indicators 9 model takes no explicit account of air speed, but some of the data have been taken from buildings in which occu- pants have the ability to affect air movement, e.g. through use of desk and ceiling fans. Figure 3.2 shows the 80% PPD warm weather discomfort threshold temperatures, obtained using the model shown in Figure 3.1, for the London design summer years for the Medium-High Emissions scenario, discussed in section 2. The maximum threshold temperatures occur in July and are 27.7 ��C, 28.1 ��C, 28.6 ��C and 29.3 ��C for the 1980s, 2020s, 2050s and 2080s DSYs, respectively. Tightening the limits to 90% PPD reduces the acceptable condition by 1.0 ��C, for example in the 2080s, from 29.3 ��C to 28.3 ��C. The drawback with use of the adaptive model for perform- ance assessment is that results can be difficult to interpret, because the benchmarks change from month to month (or day-by-day if a running mean from the previous 30 days is used for the external temperature condition). The deter- ministic Fanger method has been used here to develop comfort temperature thresholds that include some allowance for occupant adaptability, in particular through adjusting dress levels. Curves for PPD against operative temperature obtained from the Fanger model for nom- inally fixed humidity and low air speed for different levels of dress are shown in Figure 3.3. The curves can be used to derive a single temperature threshold for a given level of dress (although if seasonal variations in dress are included the limit might again vary with external temperature). If the 90% satisfaction level is taken together with the lightest form of dress considered ��� light summer dress (e.g. likely to be acceptable in a office with a relaxed dress code) ��� then an upper threshold of just under 28 ��C is obtained. Taking the next ���lowest��� level of clothing (shirt and tie) the comfort threshold falls to around 25 ��C. This suggests two ���benchmarks���: an oper- ative temperature of 25 ��C when people might be assumed to notice their surroundings as warm and seek to adjust their clothing level, and an upper one of 28 ��C when they may well feel hot but are unlikely to adjust their clothing level further. It is interesting to note that the upper threshold provided by the deterministic Fanger model presented in Figure 3.3 is quite close that that suggested by the adaptive model. 3.1.2 Overheating criteria for offices For offices, a considerable amount of research has been carried out on thermal comfort and a number of different approaches have been taken to specify what is meant by overheating(13). CIBSE Applications Manual AM10: Natural ventilation in buildings(14) also discusses a number of criteria. With the exception of a Dutch standard(15) (which uses a method based upon Fanger���s predicted mean vote (PMV)(10)), the criteria usually reduce to a number of hours (or percentage of occupied hours) that certain tempera- tures are not to be exceeded. For the present study, the comfort threshold temperatures of 25 ��C and 28 ��C identified from the Fanger model have been used. These levels are also of interest as they were used for the comfort specification for the BRE Energy Efficient Office of the Future(16). The overheating criterion used in that specification was that temperatures should not exceed 25 ��C for more than 5% of the year and/or exceed 28 ��C for more than 1% of the year. As discussed above, here the simplified criterion of 1% of occupied hours per year over the higher temperature threshold will be taken as indicative of overheating. 3.1.3 Overheating criteria for schools In contrast to office buildings, clear standards for over- heating in school buildings are now referred to within the Building Regulations framework through the Department for Education and Skills��� Building Bulletin BB87(7). This document states that: For school buildings it is accepted practice to define overheating as occurring when the internal air temperature exceeds 28 ��C. An allowable degree of overheating in a school is that this may occur for up to 80 occupied hours in a year. The CIBSE Test Reference Years from CIBSE Guide J should be used as the basis of predicting the number of occupied hours when the temperature exceeds 28 ��C. The overheating threshold temperature of 28 ��C is the same as the ���hot��� temperature threshold identified from the adaptive Fanger model in section 3.1.1 and used for the office overheating criterion. The hours of occupancy of a typical school with a 6 week summer holiday is estimated to be around 1800 hours, so the criterion of 80 hours per year corresponds to just under 5% of occupied hours over the year. Rather than use this criterion the more stringent target of a 1% exceedance of 28 ��C has been adopted, for consis- tency with the office criterion. Note that the CIBSE Design Summer Years (DSYs) have been used for the assessment, not the Test Reference Year (TRY) as suggested by BB87. This also represents a more stringent test of overheating risk, since the peak summer temperatures in the DSYs are significantly warmer than the TRYs, which are 1989 Operative temperature / �� C 2020s 2050s 2080s 30 28 26 24 22 20 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 3.2 Threshold temperatures for 80% PPD from the ASHRAE adaptive comfort model shown in Figure 3.1 using changes for London 20 30 28 26 24 22 P ercentage persons dissatisfied Temperature / ��C 80 70 60 50 40 30 20 10 0 Light summer dress Two piece suit No jacket Heavy woollen suit 8 out of 10 satisfied 9 out of 10 satisfied Figure 3.3 Deterministic comfort model (after Fanger(10)) effect of clothing level on comfort temperatures