Prediction of the Summertime Overheating Risk in UK Homes Using Descriptive Time Series Analysis
Argyris Oraiopoulos, Loughborough University
Introduction
This study is based on an extensive data set of 228 dwellings in Leicester and will attempt to develop an empirical model (a model based on this empirical dataset), capable of successfully predicting the risk of overheating in UK homes based only on measured internal temperatures and external weather data.
Research Questions
This research will address the following questions:
- How can time series analysis be applied to room temperature data to develop a model that is able to predict the risk of overheating in UK homes?
- In what way can the measured internal air temperature, external air temperature and solar irradiation data be used to form the input data in an empirically derived predictive model?
- Can the parameters of the model be grouped according to the characteristics of the dwellings and/or the occupants in the dataset?
- How can an empirically derived model be validated using a restricted dataset?
Aim
The aim of this research project is to use time series analysis for the development of statistical models that enable the prediction of future temperatures and therefore the risk of overheating in UK homes based on past measured values and external climate data.
Objectives
The aim of the project will be achieved by pursuing the following objectives.
- Critically reviewing previous work on measuring and modelling studies with regards to overheating in dwellings both in UK and internationally.
- By capitalising on previous work, to investigate the possibility of explaining the variation of internal temperatures in dwellings with the aid of an empirical model.
- Develop an appropriate methodology to produce a modelling technique that allows an efficient way of predicting the risk of overheating in spaces in homes.
- Using statistical analysis to investigate correlations between the model’s parameters and the characteristics of the dwellings (house type, house age, wall type, etc.) to identify the possibility of grouping models according to dwelling characteristics.
- Investigate the overheating risk predicted by the model for all 228 homes and compare the results against measured data by using different overheating criteria (CIBSE static, BSEN 12521, CIBSE TM 52, ASHRAE 55).
- Using the current dataset to validate the empirically derived model.
- Critically discussing the results and drawing conclusions regarding the usefulness of the developed modelling technique and suggest ways in which it can be further developed, evaluated and improved.
Project Team
Student(s)
Argyris Oraiopoulos
Supervisor(s)
Steven Firth
Tom Kane
Kevin Lomas
Outputs
Conference Presentation – June 2016
Methodology: 4 Funerals and a Wedding
Presented during the 3rd LoLo Student-led Conference, UCL
Conference Presentation – December 2015
Measured Internal Temperatures in UK Homes – A Time Series Analysis and ARIMA modelling approach
Presented during the 14th International Conference of the IBPSA in Hyderabad, India
Journal Paper – November 2015
Measured internal temperatures in UK homes – A Time Series Analysis and modelling approach
Published on Energy Procedia, Vol. 78, pp. 2844-2850
Workshop Presentation – November 2015
Multivariate Time Series Analysis and modelling of measured internal temperatures in UK homes
Presented during the workshop on Time Series Analysis & Inverse Modelling at University of Bath
Conference Poster – November 2015
Mapping the current and future risk of overheating in UK homes – A Time Series ARIMA modelling approach
Presented during the LoLo Annual Colloquium 2015
Conference Presentation – June 2015
Measured internal temperatures in UK homes – A descriptive Time Series Analysis and modelling approach
Presented during the 6th IBPC in Turin, Italy
Conference Presentation & Poster – June 2015
Mapping the current and future risk of overheating in UK homes – The Persistence of Memory (Time Series)
Presented during the UKIEG (UK Indoor Environments Group) Annual Conference at the London School of Hygiene and Tropical Medicine
Conference Poster – April 2015
Mapping the current and future risk of overheating in UK homes – The Persistence of Memory (Time Series)
Presented during the LoLo student Conference at Loughborough University
Research Festival Poster – January 2015
Mapping the current and future risk of overheating in UK homes – A Time Series Approach
Presented during the CBE Research Festival 2015 at Loughborough University
Conference Presentation – November 2014
Mapping the current and future risk of overheating in UK homes – A Time Series Analysis and modelling Approach
Presented during the LoLo Annual Colloquium 2014 in London
Conference Poster – April 2014
Mapping the current and future risk of overheating in UK homes – A statistical approach
Presented during the LoLo Stakeholders Meeting at Loughborough University