| Title | Improving Bottom-Hole Temperature Corrections on the Basis of Statistical Correlations |
|---|---|
| Authors | Thorsten AGEMAR |
| Year | 2020 |
| Conference | World Geothermal Congress |
| Keywords | temperature, correction, BHT, statistics, data base, information system, GeotIS, cylinder source model |
| Abstract | In Germany, most available data on subsurface temperature distribution come from hydrocarbon exploration activities. So-called bottom-hole temperature measurements (BHT) represent the largest part of all measured subsurface temperatures. These BHTs are available for approximately 11.000 industrial boreholes. Some wells have BHTs for more than one depth. Raw BHTs provide the temperature only a few hours after drilling has stopped. The temperature field around a borehole is usually disturbed by mud circulation related to the drilling process for several weeks. A number of methods to extrapolate the true formation temperature have therefore been developed based on various assumptions about the cooling effect of the circulating mud and the thermal behaviour of the borehole. The choice of the most appropriate correction method depends on the availability of data such as the circulation period, number of subsequent measurements with shut-in times, and the well radius. Despite such corrections, these results still have errors of up to ± 8 to 10 K, and are therefore much less accurate than equilibrium temperature logs. Generally, BHT correction requires a time series of two or more measurements for optimal fitting. However, many well reports provide only one BHT value with shut-in time. A statistical analysis of 385 temperature records revealed a correlation between well radius and initial temperature deviation. This correlation has been used to estimate the initial temperature difference at t=0 for a correction based on the cylinder source model. For a number of sites, BHT values corrected with this new method were compared with equilibrium temperature logs or temperature recordings during hydraulic production tests from neighbouring wells. This comparison proved that the new method is robust and reliable. |