Minimum stress and pore fluid pressure in sedimentary basins

Abstract

Leak-off pressures (LOPs) recorded during leak-off tests (LOTs) conducted down boreholes are often used to estimate the magnitude of the minimum stress (usually assumed to be horizontal – S(_h)) in the subsurface. However, the reliability of these tests has previously been questioned in the literature and the accuracy of the data obtained from them has been in doubt. Using original LOT data from Mid-Norway, this study has shown that through stringent quality control, good LOT data can be used to accurately constrain the magnitude of S(_h). Knowledge of the relationship between in-situ stress and pore pressures (Pp) in basins provides insights into their structure as well as having implications for well design and drilling safety. Using stress-depth plots to display S(_h) measurements from Mid-Norway and six further basins from around the world reveals a variability in the magnitude of Sh at all depths. Analyses show that rock mechanical properties or differences in the way LOTs are performed cannot explain this variability. Separate analysis of extended leak-off test (XLOT) data from Mid-Norway shows that variability in the magnitude of the LOP (most often used to calculate S(_h)) is inherent in the testing procedure. This inherence suggests either the variations in Sh are real (they represent basin heterogeneities) or that they result from a combination or rock mechanical and/or pumping pressure test parameters. Further use of multiple cycle XLOTs shows that using LOPs and instantaneous shut-in pressures (ISIPs) to calculate S(_h) produces similar results. Considering re-opening cycles of tests and those tests from greater depths shows the difference between the magnitude of Sh calculated using the LOP and ISIP is reduced. These same high quality data have been used to calculate the magnitude of the three principal stress from Mid-Norway and show the contemporary stress situation to be S(_h)<S(_v)<S(_H).Compilations of S(_h) and P(_p) have also been used to calculate the lower bound to LOPs and the upper limit to P(_p) as means of predicting Sh. Results show that using the lower bound to estimate the maximum P(_p) (or the upper limit to estimate S(_h)) will lead to large errors within normally pressured zones but successful estimates at overpressured depths. Analyses also show that there is no systematic relationship between the magnitude of the lower bound to LOPs and the contemporary stress situation. The S(_h) and P(_p) data were normalised to a "hypothetical unconstrained basiri' and/or depth to investigate pore pressure in-situ stress coupling and quantify the change in S(_h) with overpressure. Results show that coupling can be inferred in three of the seven basins studied. Poro-elasticity or frictional limits to stress are the most likely coupling mechanisms because regional S(_h) magnitudes do not relate to tectonic regime. Coupling is not inferred for Mid-Norway. An explanation for the lack of coupling is the late timing of overpressure following normal compaction of the rocks. It is suggested that "inflationary mechanism" overpressures produce lower ∆S(_h)/ ∆P(_p) values than overpressures that developed synchronous with burial. Using the poro-elasticity equation to back-calculate the Poisson's ratio (v) of the rocks reveals high values thus establishing a relationship between high v and higher levels of compaction.