Prediction of Soil Temperatures for Underground Heat Exchanger Applications in Manisa Turkey

Abstract

This present study deals with the prediction of soil temperature’s as a function of time and depth. Soil temperature values, especially near the surface, depend on different types of parameters such as seasonal weather variations, moisture content, thermal conductivity, etc. The main drawback is that even though these values are really important, they cannot be reached easily. Due to this reason we developed theoretical model for prediction. Measured data between 1960 and 2014 that was taken from the Izmir State Meteorological Station, and predicted soil temperatures at depths of 5cm, 10cm, 20cm, 50cm, and 100cm were compared for city of Manisa. Additionally, at depths of 5cm, 10cm, 20cm, 50cm and 100cm, the maximum average percentage errors were 13.4% , 12%, 8.5% , 9.4% ve 8.7% respectively. In summary, we investigated the relationship between ambient temperature fluctuations and soil temperatures in terms of depth from 5cm to 3000cm.