Testing models for bottom of hole temperature recovery, Cooper Basin, South Australia

Abstract

True formation temperature (TFT) is a key parameter in geothermal resource evaluation. Measurements of TFT require deployment of a thermometer to depth within the crust, most commonly down a small diameter borehole. Unfortunately, the act of drilling removes heat from the region where temperature is to be recorded. This induced thermal anomaly may persist for months or even years depending upon both in situ conditions and the style and depth of the drilling. In the absence of sufficient time for full down hole thermal re-equilibration, measured bottom of hole temperature (BHT) may be significantly less than TFT. Numerous models have been proposed to correct disturbed BHT measurements. To date, very few of these models have been adequately tested. A unique opportunity to examine models of BHT recovery exists in data from the Cooper Basin, South Australia. Extensively drilled for both petroleum exploration and production, a large database of post-drilling BHT measurements is available for wells in this region. Included within these data is a subset of 335 wells for which temperature data are available from post-suspension completion logs. These data, collected several weeks to months after the end of drilling, are a close approximation to TFT for the bottom of hole. This combination of short and long-term temperature measurements has allowed an assessment of the performance of BHT recovery models for a subset of 61Cooper Basin wells. Four models were tested: the Horner plot (as derived from the line source model of Bullard, 1947), the zero-circulation model of Cooper & Jones (1959), the empirical exponential model of Perrier & Raiga-Clemenceau (1984) and the empirical semi-log plot of Pitt (1986). In all cases model predictions were found to be biased when compared with the TFT. The magnitude and direction of this bias is found to be dependent upon assumptions implicit in each model.