Phil Symonds

Overheating in English dwellings: comparing modelled and monitored large-scale datasets

ABSTRACT Monitoring and modelling studies of the indoor environment indicate that there are often discrepancies between simulation results and measurements. The availability of large monitoring datasets of domestic buildings allows for more rigorous validation of the performance of building simulation models derived from limited building information, backed by statistical significance tests and goodness-of-fit metrics. These datasets also offer the opportunity to test modelling assumptions. This paper investigates the performance of domestic housing models using EnergyPlus software to predict maximum daily indoor temperatures over the summer of 2011. Monitored maximum daily indoor temperatures from the English Housing Survey’s (EHS) Energy Follow-Up Survey (EFUS) for 823 nationally representative dwellings are compared against predictions made by EnergyPlus simulations. Due to lack of information on the characteristics of individual dwellings, the models struggle to predict maximum temperatures in individual dwellings and performance was worse on days when the outdoor maximum temperatures were high. This research indicates that unknown factors such as building characteristics, occupant behaviour and local environment makes the validation of models for individual dwellings a challenging task. The models did, however, provide an improved estimate of temperature exposure when aggregated over dwellings within a particular region.

Inhabitant actions and summer overheating risk in London dwellings

ABSTRACT An indoor overheating assessment study of 101 London dwellings during summer 2009 is presented. The study included building surveys, indoor dry bulb temperature monitoring and a questionnaire survey on occupant behaviour, including the operation of passive and active ventilation, cooling and shading systems. A theoretical London housing stock comprising 3456 combinations of building geometry, orientations, urban patterns, fabric retrofit and external weather was simulated using the EnergyPlus thermal modelling software. A statistical meta-model of EnergyPlus was then built by regressing the independent variables (simulation input) against the dependent variables (overheating risk). The monitoring and questionnaire data were analysed to explore the relationship between self-reported behaviour and overheating, and to test the meta-model. The monitoring data indicated that London homes and, in particular, bedrooms are already at risk of overheating during hot spells under the current climate. Around 70% of respondents tended to open only one or no windows at night mainly due to security reasons. An improvement in the coefficient of determination (R2) values between measured temperature and meta-model predictions was obtained only for those dwellings where occupants reported actions that were in line with the modelling assumptions, thus highlighting the importance of occupant behaviour for overheating.

Development of an England-wide indoor overheating and air pollution model using artificial neural networks

With the UK climate projected to warm in future decades, there is an increased research focus on the risks of indoor overheating. Energy-efficient building adaptations may modify a buildings risk of overheating and the infiltration of air pollution from outdoor sources. This paper presents the development of a national model of indoor overheating and air pollution, capable of modelling the existing and future building stocks, along with changes to the climate, outdoor air pollution levels, and occupant behaviour. The model presented is based on a large number of EnergyPlus simulations run in parallel. A metamodelling approach is used to create a model that estimates the indoor overheating and air pollution risks for the English housing stock. The performance of neural networks (NNs) is compared to a support vector regression (SVR) algorithm when forming the metamodel. NNs are shown to give almost a 50% better overall performance than SVR.

Assessing urban population vulnerability and environmental risks across an urban area during heatwaves – Implications for health protection

Heatwaves can lead to a range of adverse impacts including increased risk of illness and mortality; the heatwave in August 2003 has been associated with ~70,000 deaths across Europe. Due to climate change, heatwaves are likely to become more intense, more frequent and last longer in the future. A number of factors may influence risks associated with heat exposure, such as population age, housing type, and location within the Urban Heat Island, and such factors may not be evenly distributed spatially across a region. We simulated and analysed two major heatwaves in the UK, in August 2003 and July 2006, to assess spatial vulnerability to heat exposure across the West Midlands, an area containing ~5 million people, and how ambient temperature varies in relation to factors that influence heat-related health effects, through weighting of ambient temperatures according to distributions of these factors across an urban area. Additionally we present quantification of how particular centres such as hospitals are exposed to the UHI, by comparing temperatures at these locations with average temperatures across the region, and presenting these results for both day and night times. We find that UHI intensity was substantial during both heatwaves, reaching a maximum of +9.6°C in Birmingham in July 2006. Previous work has shown some housing types, such as flats and terraced houses, are associated with increased risk of overheating, and our results show that these housing types are generally located within the warmest parts of the city. Older age groups are more susceptible to the effects of heat. Our analysis of distribution of population based on age group showed there is only small spatial variation in ambient temperature that different age groups are exposed to. Analysis of relative deprivation across the region indicates more deprived populations are located in the warmest parts of the city.

Comparison of built environment adaptations to heat exposure and mortality during hot weather, West Midlands region, UK

There is growing recognition of the need to improve protection against the adverse health effects of hot weather in the context of climate change. We quantify the impact of the Urban Heat Island (UHI) and selected adaptation measures made to dwellings on temperature exposure and mortality in the West Midlands region of the UK. We used 1) building physics models to assess indoor temperatures, initially in the existing housing stock and then following adaptation measures (energy efficiency building fabric upgrades and/or window shutters), of representative dwelling archetypes using data from the English Housing Survey (EHS), and 2) modelled UHI effect on outdoor temperatures. The ages of residents were combined with evidence on the heat-mortality relationship to estimate mortality risk and to quantify population-level changes in risk following adaptations to reduce summertime heat exposure. Results indicate that the UHI effect accounts for an estimated 21% of mortality. External shutters may reduce heat-related mortality by 30-60% depending on weather conditions, while shutters in conjunction with energy-efficient retrofitting may reduce risk by up to 52%. The use of shutters appears to be one of the most effective measures providing protection against heat-related mortality during periods of high summer temperatures, although their effectiveness may be limited under extreme temperatures. Energy efficiency adaptations to the dwellings and measures to increase green space in the urban environment to combat the UHI effect appear to be less beneficial for reducing heat-related mortality.

Can the choice of building performance simulation tool significantly alter the level of predicted indoor overheating risk in London flats

The accurate prediction of building indoor overheating risk is critical in order to mitigate its possible consequences on occupant health and wellbeing. The Chartered Institution of Building Services Engineers issued Technical Memorandum 59 (TM59) with the aim of achieving consistency in the modelling processes followed for the prediction of overheating risk in new dwellings. However, as each tool’s prediction may depend on its inherent assumptions, an inter-model comparison procedure was used to assess whether the choice of building performance simulation tool influences the overheating assessment. The predictions of two popular tools, IES VE and EnergyPlus, were compared for nine variations of a naturally ventilated, purpose built, London flat archetype, modelled under the default algorithm options. EnergyPlus predicted a high overheating risk according to TM59 criteria in seven out of the nine model variants, contrary to the low risk of all the IES VE variants. Analysis of heat transfer processes revealed that wind-driven ventilation and surface convection algorithms were the main sources of the observed discrepancies. The choice of simulation tool could thus influence the overheating risk assessment in flats, while the observed discrepancies in the simulation of air and heat transfer could have implications on other modelling applications. Practical application: Technical Memorandum 59 issued by the Chartered Institution of Building Services Engineers may be widely adopted within the industry to assist the prediction of overheating risk in new dwellings. This work suggests that the choice of building performance simulation tool can greatly influence the predicted overheating risk. Furthermore, the differences identified in the modelling of heat transfer processes could also impact other modelling applications. Following these results, the need for detailed empirical validation studies of naturally ventilated homes has been highlighted.