A model for predicting the yearly load in district heating systems

• Published in: Proceedings of the Institution of Mechanical Engineers, Part A: Journal of Power and Energy
• Main Author: Jónsson, G. R.
• Format: Article in Journal/Newspaper
• Language: English
• Published: SAGE Publications 2002
• Subjects: Mechanical Engineering; Energy Engineering and Power Technology; Iceland
• Online Access: http://dx.doi.org/10.1243/095765002320183595; http://journals.sagepub.com/doi/pdf/10.1243/095765002320183595

This paper presents a model that can be used to predict the yearly load in district heating systems some years ahead. To demonstrate the model, data from Reykjavik, Iceland, were used in the analysis. The model is based on the three main factors affecting the consumption. The first factor is the increase in the number of houses, called the connected volume, that need to be heated. The trend in the volume during the last decade forms the basis for predicting future values using regression analysis. The second factor is the climate, which is an important factor since approximately 85-90 per cent of the hot water consumption is weather dependent. The estimated variability in the weather is based on data for the period from 1949 to 1999 using bootstrap techniques. The third factor is the behaviour of the consumers, which is analysed by looking at past consumption values after isolating the effect of the first two factors from the consumption. A prediction of the consumption is obtained given some values of each of the three factors. This process is repeated several hundred times, and from those values an estimate of the expected consumption is obtained. Furthermore, the two-sided (100 — α) per cent confidence interval for the consumption is determined by cutting off α/2 per cent of the lowest and highest predicted consumption values.