RUSSIAN VENTURES-2 EVALUATING OIL, GAS OPPORTUNITIES IN WESTERN SIBERIA-RESERVOIR DESCRIPTION

William Connelly
Pangea International Inc.
Golden, Colo.

Jack A. Krug
Questa Engineering Corp.
Golden, Colo.

Part one of this five part series discussed core and log data used for evaluating oil and gas properties in western Siberia.1 In this article, we discuss how to use the subsurface data to describe hydrocarbon reservoirs and estimate the original oil in place (OOIP).

Initially the evaluation should use only the "research" wells because they include relatively complete data sets and are adequately located over the structures (4-8 sq km spacing), These preliminary reservoir models can be refined later by using data from "production" wells.

The methodology for describing a reservoir and estimating the OOIP in western Siberia is similar to the approach for most reservoirs:

Establish stratigraphic correlations across the field;
Construct structure maps on key horizons;
Construct porosity isopach maps for significant reservoirs;
Construct net pay maps;
Determine reservoir parameters; and
Calculate pore-volume estimates of OOIP.

STRATIGRAPHY

Most production in western Siberia comes from the thick sequence of sandstone and shale deposited during the Jurassic-Cretaceous subsidence of the Western Siberia basin.2 These clastic sediments often contain glauconite and marine fauna evidencing deposition under shallow marine conditions.

Occasional coals and root tubules evidence periods of emergence. Porosity isopach maps of these sand bodies usually contour as prograding deltas, estuaries, marine bar sands, and occasional channels.

The first step in describing a reservoir is to correlate strata in all "research" wells and identify any faults. The "electric log" (SP and lateral log displayed at 1:500 scale) is an excellent correlation log. The known productive zones have distinct characteristics and are easily recognized throughout the basin.

Once the logs are correlated, construct a series of restored (fault corrected) stratigraphic cross-sections hung from reliable regional shale markers. The Bezhenov shale is an excellent hang-horizon for Jurassic objectives (Table 1). These cross-sections aye used repeatedly to study reservoir continuity and paleoenvironments and to construct isopach maps.

STRUCTURE MAPS

Most western Siberia fields are located on gently dipping four-way seismic closures. Seismic data quality varies but usually is adequate to define the gentle structural closures typical of the region.

Seismic data are multi-fold and common-depth-point (CDP); most seismic data acquired since the mid-1980s are digitally recorded. In some instances, reprocessing is beneficial. Because most Siberian ventures available to western companies involve fields already delineated with subsurface control, seismic interpretation is much less important than subsurface data.

Before mapping can begin, suitable base-maps must be located or constructed. This task is complicated by the absence of any grids or x-y coordinates for wells. Occasionally the latitude and longitude of wells are available, but often it is necessary to trust the wells are accurately posted on the Russian maps. We recommend digitizing the base-maps and creating an x-v coordinate system for each project. This x-y coordinate system is necessary for a simulation study and thus has a dual purpose.

The Bezhenov shale is an excellent seismic reflector and overlies several productive Jurassic sands, therefore the first structure map should be on this horizon. Russians usually construct seismic structure maps on this horizon.

Because the Bezhenov seismic structure maps commonly have not been updated to honor subsequently drilled delineation wells, we update these maps with the newer subsurface control. Russians are good structural mappers, and it is unlikely you will make significant changes to their interpretations.

Subsurface structure maps of shallower horizons are made as needed. Many structures are syndepositional, therefore it is wise to map several horizons and study growth history (graphing a series of isopach thicknesses versus datums will establish periods of growth; alternatively, construct interval isopach maps). Differential compaction structures are also common and will show no evidence of stratigraphic thinning over structure.

POROSITY ISOPACH MAPS

The preferred measure of porosity thickness used on isopach maps is microlog separation. If the suite of logs does not include a microlog, then the SP in combination with an induction-conductivity log can be used to estimate the porosity thickness. A third alternative for picking porosity thickness values is the caliper log. In permeable zones, the caliper shows mud cake build-up and generally agrees with microlog separation.

Russian geologists conventionally core many "research" wells and annotate the logs with brief core descriptions. While picking isopach values from logs, it is important to study these core descriptions and/or the log character for indications of depositional environments.

The map should be extended beyond the area defined by productive wells in order to develop regional depositional models. Flow test data also is commonly posted on logs and is the best source of information about fluid contacts.

NET PAY MAP

Construction of the net pay map is the integration of structural, fault, and fluid level information from the structure map in combination with the sand distribution information from the porosity isopach map.

Carefully review flow tests, core descriptions, and log analyses for oil/water contacts (OWC), gas/water contacts (GWC), and gas/oil contacts (GOC). When a fluid contact is established, post the contact datum on all logs in the "control area" to be certain there are no contradictions ("control area" refers to each reservoir in pressure communication). Many reservoirs contain stratigraphic barriers, so there may be multiple fluid levels in what appears to be a continuous sand. Planimeter each reservoir control area to determine the bulk reservoir volume. If suitable core data are available, construct a "porosity times net pay" map in addition to the net pay map.

RESERVOIR PROPERTIES

The following reservoir properties are used in the volumetric calculations: formation volume factor (Bo), porosity (f), oil saturation (So), and recovery factor (RF). Reservoir fluid properties are obtained from PVT analyses of fluid samples and from measurements taken during flow tests.

Typically oil, water, and gas samples are collected during testing operations and field measurements are made of oil density, gas-oil ratio, and salinity of produced water. Other samples are sent to laboratories where the bubble-point pressure, density, viscosity, and bulk modulus are measured at reservoir temperature and pressure, and oil density, gas density, gas-oil ratio, and relative volume factor are measured for flashed and differential conditions.

These data are analyzed and average values calculated which provide the fluid properties for volumetric calculations, flow test analyses, and material balance calculations.

In addition to these single condition measurements, the oil density and viscosity are also measured at 1) reservoir pressure for decreasing temperatures beginning at reservoir temperature and ending at 20 C., and 2) reservoir temperature for decreasing pressures beginning at reservoir pressure and ending at 0.1 Mpa. It is recommended a comparison of the analytical results with Vasquez and Beggs correlations be made of the measured data.3 Using the oil density, gas-oil ratio, and reservoir pressure and temperature, the formation volume factor and bubble point may also be determined using the correlations. If there is a large difference between the reported formation volume factor and the calculated value, the reason for of the difference must be identified. Normally this is not a difficulty; however inconsistencies do occur, so the data should be checked and independent calculations made.

ORIGINAL OIL IN PLACE

The original oil in place (OOIP) and recoverable reserves are determined using volumetric calculations. The oil volume is reported by Russians in tons rather than barrels so the volumetric calculation includes oil density (Yo).

[SEE FORMULA]

Recoverable reserves = OOIP * RF

The reservoir volume terms f, h, and A, can be estimated through mapping or through a combination of map, core, and log data. The mapping approach is recommended because it averages the data across the entire reservoir.

The most difficult item to estimate is the oil saturation.

We recommend selecting several key wells with complete log, core, and test data located at least several meters above the OWC for rigorous oil-saturation calculations.

Russians classify OOIP in a fashion similar to our own. Table 2 summarizes the classifications used in Russia and compares it to Society of Petroleum Engineers and U.S. Geological Survey classifications. OOIP and recoverable reserve pore-volume estimates are reported by reservoir, by category, and/or by well. These estimates periodically are updated and provide the basis for reserve certification.

Table 3 is an example of an and recoverable reserve computation as it might appear in a table on a Russian net pay map. Metric tons of oil are calculated from the table by horizontally multiplying all of the parameters in the columns.

To convert from thousands of metric tons (MT) to thousands of barrels of oil, divide MT by the oil density (= thousands of cubic meters) and multiply by 6.29.

Table 4 summarizes common Russian terms and abbreviations dealing with reservoir descriptions and volumetric calculations.

To calculate recoverable reserves, the recovery factor must be estimated. Sands in the Western Siberia basin respond well to pressure maintenance through waterflooding because most are moderately permeable volumetric reservoirs rather than extensive reservoirs with active water drives. Pressure maintenance generally is started two to five years after production begins.

Ultimate recovery factors range from 14-35% for mature waterflooded fields; thus waterflood sweep efficiencies range from ineffective to excellent. From our experience, recovery factors range from 9% for primary recovery (thin discontinuous sands) to 35% for floods with good sweep efficiency. Average recovery factors for well-managed waterfloods are expected to range between 20-27%.

Part three of this series is a discussion of well tests and production forecasts.

REFERENCES

Connelly William, and Krug, J.A., Evaluating oil, gas opportunities in western Siberia-log and core data OGJ, Nov. 23, 1992, p. 97.
Peterson, f.A., and Clark, J.W., Geology and hydrocarbon habitat of the West Siberia Basin, AAPG Studies in Geology No. 32, 1991, 96 pp.
Vasquez, M., and Beggs, H.D., Correlations for fluid physical property predictions, JPT, 1980, pp. 986-970.