REAL-TIME DETECTION OF TAR IN CARBONATES USING LWDTRIPLE COMBO, NMR AND FORMATION TESTER INHIGHLY-DEVIATED WELLS

摘要

Detection of tar, which is defined in this paper as theheavy oil phase that does not flow using standardproduction technologies, is critical for reservoircharacterization, reserve calculations and wellplacement. Hydrocarbons that can not be producedmust be classified as resources rather than reserves, andthe distribution of tar must be accurately known for theproper placement water injectors and producers inheavy oil reservoirs.Tar must be detected as early as possible during thedrilling of a well, in order to allow for the timelyexecution of necessary changes to the drilling programand to prevent costly remedial action, such assidetracking. Early detection requires the deploymentof LWD technologies for real-time or near real-timeinterpretation of the data.Lack of invasion observed in resistivity logs orborehole enlargement from the caliper log can becorrelated to the presence of tar, the ideal tool string fortar detection includes triple-combo, NMR and theformation tester. The associated workflow utilizes acombination of tar indicators obtained from triplecomboand NMR logs. Once detected, mobilityreadings from the formation tester are then used toverify the presence of tar.The first tar indicator deployed is the well-knownmissing-porosity. A portion of the NMR signal in tardecays too fast to be detected by the logging tool,resulting in a deficit in NMR porosity when comparedto total porosity obtained from conventional logs. Thesecond and equally effective tar indicator is called theexcess-bound-fluid, defined as the difference of boundfluid porosity from the NMR log and the bulk volumeof water computed using conventional logs. Real-timeLWD T_2 spectra, a recent development with significantfuture potential, can additionally be used to enhance theinterpretation in resolving fluid-property relatedambiguities, when available.Real-time tar characterization demands high accuracyfrom the porosity logs. In the clean carbonateformations typical of the Middle East, this requirementtranslates into the use of an 18-second wait-time toensure complete polarization in the NMR log. The useof such a long wait-time impacts the ROP and verticalresolution of the NMR measurement, necessitating theavailability of certain features in the hardware used.This paper discusses the petrophysical framework for abest-practice real-time tar detection workflow usingLWD measurements. Results from the implementationof the methodology in two different carbonate fields inSaudi Arabia demonstrate that with proper technologiesand robust interpretation algorithms in place, real-timetar identification can be done efficiently and accurately.