Geostatistical Analysis of Bottom-Hole Temperatures in the Denver and Williston Basins: North America

摘要

Bottom-hole temperatures (BHTs) obtained from oil and gas wells have never been completely reliable due to the formation temperature disturbance caused by the influence of the drilling mud on the formation rock during drilling. A correction method must be applied before any BHT data can be used. The source and method of the correction, however, has been a topic of dissention since the early 70s, when BHTs began to be used for estimates of temperature at depth to determine such things as hydrocarbon maturity, thermal history, and geother-mal energy assessment. Several correction methods are currently used: the Harrison (Harrison et.al., 1983), Kehle (Kehle et.al., 1970), and Forster (Foerster et.al., 1996) are among the most prevalent. None of these methods yield a correction that represents a statistically accurate distribution of BHTs, although the Harrison and Kehle have been found to be a much better approximation (Crowell and Gosnold, 2013). All of these methods were developed using a top-down approach, where an equilibrium temperature profile has been obtained and a correction equation was developed to attempt to shift the best fit line of the data points to the best fit line of the data obtained at equilibrium. In addition, formation data are not always included with the bottom-hole temperature data. This makes resource assessment based on formation difficult, if not impossible. We therefore hypothesize that by using two geostatistical methods, Moran's Ⅰ and Getis-Ord, we will be able to evaluate if a better correlation exists between a depth-interval well parsing versus a geochronological unit well parsing, and if a correlation exists, is it strong enough to indicate that a correction factor is possible.