The northern part of the Western Desert of Egypt represents the second most promising area of hydrocarbon potential after the Gulf of Suez province. An artificial neural network (ANN) approach was used to develop a new predictive model for calculation of the geothermal gradients in this region based on gravity and corrected bottom-hole temperature (BHT) data. The best training data set was obtained with an ANN architecture composed of seven neurons in the hidden layer, which made it possible to predict the geothermal gradient with satisfactory efficiency. The BHT records of 116 deep oil wells (2,000–4,500 m) were used to evaluate the geothermal resources in the northern Western Desert. Corrections were applied to the BHT data to obtain the true formation equilibrium temperatures, which can provide useful constraints on the subsurface thermal regime. On the basis of these corrected data, the thermal gradient was computed for the linear sections of the temperature-versus-depth data at each well. The calculated geothermal gradient using temperature log data was generally 30 °C/km, with a few local high geothermal gradients in the northwestern parts of the study area explained by potential local geothermal fields. The Bouguer gravity values from the study area ranged from −60 mGal in the southern parts to 120 mGal in the northern areas, and exhibited NE–SW and E–W trends associated with geological structures. Although the northern Western Desert of Egypt has low regional temperature gradients (30 °C/km), several potential local geothermal fields were found (>40 °C/km). The heat flow at each well was also computed by combining sets of temperature gradients and thermal conductivity data. Aerogravity data were used to delineate the subsurface structures and tectonic framework of the region. The result of this study is a new geothermal gradient map of the northern Western Desert developed from gravity and BHT log data.
All Science Journal Classification (ASJC) codes