Stochastic Modelling as a Design Technique for Predicting Internal Heat Gains in Buildings

Knbr, Asaad Mohamed Ali (1988) Stochastic Modelling as a Design Technique for Predicting Internal Heat Gains in Buildings. MSc(R) thesis, University of Glasgow.

Abstract

Building internal heat gains from people, lighting and equipment have been identified as a source of uncertainty in estimating building heating and cooling loads and energy requirements. The common practice for estimating the internal heat gains relies on a single worst case design value which is usually assumed constant over the occupation period (constant profiles). In reality the building occupancy and the use of lighting and equipment are characterized by being partially time dependent and partially random. A preliminary study using simple variable profiles has shown that the consideration of the time variation of internal heat gains can lead to significant effects on the estimates of the heating and cooling loads and the total energy requirements when compared with the constant profiles. This study justified the development of more accurate methods of estimation of the internal heat gains. One suitable approach for modelling the random variation of internal heat gains is the use of stochastic techniques. Hence a stochastic model has been developed using real-life data to predict the likely building occupancy patterns of office buildings and the likely use of the lighting and equipment in different time steps. Arrival and departure information for staff and visitors was collected from two large office buildings. Statistical analysis of this information provided the input data on occupancy for the stochastic model. Good agreement between predicted and observed patterns was obtained within the available data. The lighting use has been related to the occupancy patterns, daylight levels and type of control. A stochastic model has been developed for predicting the use of manually controlled lighting for a probabilistic model described by BRE which was based on field studies of artificial lighting . Models were also developed for localised and photoelectric lighting control The use of equipment has been predicted based on the simulated occupancy patterns and the average probability of use. The model of equipment offers the possibility of simulating the use of equipment according to its function and type (personal or general). The results of the predictions of internal heat gains for each time step of the day (the stochastic profiles) have been used as an input to Strathclyde University's building thermal modelling program ESP to predict the overall heating and cooling loads and the total energy requirements for a well defined building. The results were compared with those obtained from using the constant profiles. The summarized results have shown that the use of the constant profiles of internal heat gains leads to inaccurate estimation in the building total heating and cooling energy requirement. The difference depends on the type of light control and is in the range of 12 to 42%. Differences of this order may be important when the thermal modelling programs are being used to compare design alternatives.

• Item Type: Thesis (MSc(R))
• Qualification Level: Masters
• Keywords: Mechanical engineering, Energy
• Date of Award: 1988
• Depositing User: Enlighten Team
• Unique ID: glathesis:1988-76799
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 14 Jan 2020 09:39
• Last Modified: 14 Jan 2020 09:39
• URI: https://theses.gla.ac.uk/id/eprint/76799