posted on 2019-06-19, 08:12authored byArgyris Oraiopoulos
As global temperatures rise and the climate becomes more unstable, heatwaves will be a more common phenomenon. This could result in an increase of energy consumption in UK homes during summer periods due to a higher demand for cooling, but it could also have a substantial impact on heat related morbidity and mortality rates and produce a series of challenges for the emergency services and the national health system. The risk of overheating in domestic buildings is typically predicted using modelling techniques based on assumptions of heat gains, heat losses and heat storage. Often dynamic thermal simulation software is used in which the modeller is required to decide a number of input assumptions upon which the result depends on. These assumptions often lead to modelling errors and reduce confidence in the results. Recent large-scale data collection studies allow empirical approaches based on measurements alone.
This thesis presents the development of an empirical model for the prediction of summertime overheating risk in UK homes, using the hourly measured internal air temperatures from the living rooms and main bedrooms of 228 dwellings in the city of Leicester, recorded between the 1st July 2009 and 31st August 2009, as well as the external air temperature and global solar irradiation data that were measured by a weather station in Leicester, during the same period.
Descriptive time series analysis is used to identify the mechanisms that shape room temperatures, in rooms that are neither mechanically heated nor cooled, and to develop building-specific empirical models of room temperatures for use in predicting future temperatures based on past measured values and on future weather conditions.
The results show that the internal air temperature time series data can be decomposed using descriptive time series analysis, in two separate components, based on the long-term and short-term impact of the external climate on the internal air temperature. Even though the external climate conditions during the monitoring period were characterised as a rather cool summer, the occurrence of overheating according to the CIBSE static criteria, was largely existent. The newly developed Internal Trend and Cyclical Component (ITCC) model is shown to predict the overheating occurrence with an accuracy of 84-92%, with a mean R2 value of 0.83 and 0.84 for living rooms and bedrooms respectively.
The ITCC model was developed using only internal air temperatures and external weather data and no information regarding the dwellings or the households, therefore it is a model that could potentially be applied in any free-running building worldwide, provided that is first tested and validated. By applying the ITCC model to national datasets, this can provide significant insights for the developments of future policies to mitigate overheating, and enable overheating risk alerts for home-owners and public authorities to be more accurately estimated and targeted.