Evaluating Reservoir Connectivity and Compartmentalization Using aCombination of Formation Tester and NMR Data

摘要

The detection and evaluation of reservoir connectivityand compartmentalization continues to be a significantissue in reservoir characterization, especially inoffshore fields that must commit significant upfrontcosts before production can begin.It is generally accepted that with most of thepetrophysical and formation testing measurementsmade in the borehole we are able to detect reservoirseparation but we are not able to prove connectivity.That is, we can prove the negative but not the positive.However, compartmentalization (proving the negative)can economically doom a project. Therefore it isincumbent upon the practitioner to incorporate andintegrate all measurements at hand before passingjudgment.In the example discussed in this paper, multiple wellsare being drilled as injectors. Connectivity to theproducer wells is critical. We evaluate acquiredpressures and gradients from formation tester tools inseveral of these wells. In one of the injector wells wehave acquired NMR data including gas/oil ratio (GOR)and viscosity measurements.Compartmentalization within a reservoir can be eithervertical, lateral, or both. Evaluation of verticalconnectivity can be done by either evaluating pressurediscontinuities or Fluid Profiling*1. The use of pressurediscontinuities is well established. Fluid profiling1 Mark of Schlumbergerassumes that fluids in flow communication overgeologic time will equilibrate to a predictable fluidgradient. Fluid distributions that are not predictable canbe suspected to be not in flow communication andfurther investigation is warranted.In our injector example, we show how an inverseresponse of the viscosity and GOR measurements fromthe NMR data can be used to indicate a reservoirdiscontinuity. Additionally we have formation testerpressure gradient data over this same zone and weprovide a statistical analysis of this data to allow us tointegrate it with the NMR data.For lateral connectivity, we evaluate well-to-well data.Formation pressures and samples were obtained fromthe producer and the injector well. We show how tocalibrate the response between different tools indifferent wells and then describe how the differencescan be resolved. Again a statistical analysis of thepressure gradient data is applied to ensure that anyinferences made are sustainable given the accuracy,resolution, and repeatability of the acquired data.