Estimation of reserves in lenticular reservoirs consisting of many thin-bedded sand/shale sequences is complicated by an overly pessimistic evaluation of sand count and hydrocarbon in place when conventional log interpretation techniques are used. It is probable that thin clean sand lenses have connected permeability. Their contribution to production should be considered in the estimation of reserves. An approach has been devised to improve the evaluation of thin clean sands by introducing accurate bed boundaries between sand and shale laminae as identified clearly on the dipmeter microresistivity curve processing presentation (GEODIP). Dipmeter data are integrated presentation (GEODIP). Dipmeter data are integrated into conventional computer log analyses to yield more realistic estimates of porosity and hydrocarbon saturation throughout the reservoir. The method and the results attained to date are described.
The subject of this paper is a producing field in the Tarakan basin of northeast Kalimantan, Indonesia. It was discovered in late 1975 and has produced since 1977. Reservoirs are upper-middle Miocene sandstones that were deposited in a lower delta plain to front tidaldominated environment. The thin (less than 5 ft [1.5 m]) and interbedded nature of the oil reservoirs throughout this field has led to inaccurate oil in place (OIP) estimates on the basis of conventional log analyses. This paper demonstrates (1) how the thin sands were overlooked in the calculation of reserves in the initial log analysis, (2) the means by which the correct interpretation of this reservoir has been accomplished on 12 producing wells in this field, and (3) how the newly producing wells in this field, and (3) how the newly developed quantitative method of evaluating these thin reservoirs has been accomplished. The inaccuracy of conventional log analysis and its effect on reserve calculations did not appear significant until a later date when cumulative production data were interpreted. These production data, in conjunction with drillstem tests (DST's), which have produced oil from thin sands, have now been evaluated with respect to the log and core data previously gathered. The standard suite of logs was run on each well in this field. However, not until recently did we develop a technique to adapt accurately the log interpretation to evaluate the thin-bedded nature of these reservoirs. This adaptation uses logs and extends normal interpretation with the aid of recently developed computer programs. The interpretation of the reservoir geology has become the most important aspect of the evaluation and analysis of reserves in this producing field.
The early logging program included the induction electic log (IES), sonic, formation density log/compensated neutron log (FDC/CNL), MLL/Microlog, and dipmeter (HDT). In later wells, the (induction spherical focused) ISF/sonic replaced the IES and the dual laterolog (DLL) was added also. The oil producing interval is within a gross sand/shale sequence 1,000 to 1,400 ft [305 to 427 m] thick. Initial log analyses indicated relatively clean sands ranging from 5 to 20 ft [1.5 to 6 m] thick. The logs showed the hydrocarbon-bearing sands to have resistivities of more than 10 and commonly to 30 0 in the thicker sands (Fig. 1). Porosities range from 18.5 to 27.7% across the field in the well-developed thicker sands. These sands have produced individually more than 2,000 B/D [318 m 3 /d] on a DST. These clean sands commonly add up to more than 100 ft [30.5 m] of net estimated oil pay in the structurally high wells. (See Fig. 9 for original net pay calculation.) Conventional cores taken over a portion of the reservoir section showed that in addition to major sand bodies there also were sequences that consisted of thinly interbedded clean sand and shale. Early log analyses of these thinly interbedded sequences indicated that throughout the producing section, in addition to the clean sands, there are many other sand intervals that range in their log response from shaly to very shaly. These sands usually exhibit resistivities from 4 to 10 (Fig. 1). Use of standard V-clay (clay volume) corrections of Coriband generally yields average porosities in the range of 5 to 12% in these sands. porosities in the range of 5 to 12% in these sands. JPT
p. 1535
