"Quality Interpretation for Quality Prospects"

REJUVENATING A MATURE SUPERGIANT FIELD, VLC-363, BLOCK III FIELD, LAKE MARACAIBO, VENEZUELA

by MARAVEN PETROLEOS DE VENEZUELA, S.A.

Emir Arzola, Bice Cortiula, Gedi González and Luis Rondón

TCA RESERVOIR ENGINEERING SERVICES, INC.

Michael Todd

LOREN AND ASSOCIATES, INC.

John T. Kulha and J. Dennis Loren

ROBERT M. SNEIDER EXPLORATION, INC.

John S. Sneider and Robert M. Sneider

EXPLOITECH SERVICES, INC.

Dan Shaughnessy

John R. Farina

 

This poster session consists of seven related poster contributions:

INTRODUCTION (BACKGROUND AND SCOPE)

J. Dennis Loren, Robert M. Sneider, Luis Rondón*

GEOLOGICAL FRAMEWORK

John S. Sneider, Dan Shaughnessy, Robert M. Sneider*, Gedi González and Bice Cortiula

PETROPHYSICAL EVALUATION

J. Dennis Loren*, John T. Kulha and Carolina Coll

RESERVOIR CHARACTERIZATION

John Sneider*, Gedi González and Bice Cortiula

RESERVOIR PERFORMANCE AND OBSERVATIONS

John R. Farina*, John T. Kulha and Emir Arzola

DEVELOPMENT OPPORTUNITIES

John S. Sneider, Gedi González*, John T. Kulha and Dan Shaughnessy

NUMERICAL SIMULATION

Michael R. Todd* and Emir Arzola



BACKGROUND AND SCOPE


The Eocene Lower "C" Reservoir is a supergiant field in Lake Maracaibo, Venezuela. The field is developed by 63 wells.

A multidisciplinary team of geoscientists and engineers undertook an integrated multidisciplinary study of the Lower "C" Reservoir to provide a detailed reservoir description for reservoir simulation and to develop additional new well and workover opportunities to increase production. A reservoir simulation study was performed to determine processes and methods to improve future production. The team making the studies consisted of thirteen staff from TCA Engineering, Loren and Associates, Robert M. Sneider Exploration, Exploitech and Maraven (Petroleos de Venezuela, S.A.).

This paper consists of six parts:

1.Geologic Framework is based on 3D seismic and detailed well log correlation. The C440-460 reservoirs are composed of deltaic and marine deposits, which are subdivided into 16 flow units. Complete to partial barriers to vertical and horizontal fluid flow result from stratigraphic facies changes and/or faults. The deltaic-marine depositional model was used to guide the mapping of geological and petrophysical parameters, sand/shale continuity, and facies distribution.

Page 2, Background and Scope

2.Petrophysical Evaluation includes documentation of the methodology used to calculate petrophysical variables in each of the 63 wells over the C440 - C460 intervals. Average values of porosity, water saturation, permeability, net feet pay and hydrocarbon pore volume were summarized for each of four pay categories for the sixteen flow units. The Lorenz coefficient, which expressed heterogeneity, was calculated for each flow unit.

3.Reservoir Characterization includes maps of porosity, permeability, water saturation, net pay, hydrocarbon pore volume, Lorenz coefficient, structure and depositional facies. These parameters are the major input data for the reservoir simulation model. A network of cross sections illustrate the structure, stratigraphy, and continuity of reservoir units and flow barriers.  

4.Reservoir Performance and Observations includes an analysis of some of the production performance behavior of the C440, C455 and C460 reservoir units.

5.Development Opportunities consist of seven types of workovers, two field extension areas and one new drill opportunity to increase production in a newly defined fault block.

6.Reservoir simulation study covers the reservoir engineering processes and methods, including a water alternating with gas (WAG) scheme to improve future production.

GEOLOGIC FRAMEWORK

The Eocene "C" reservoirs and flow barriers are controlled primarily by depositional processes and to a lesser extent by diagenesis. The major depositional system for the C-440 to the Guasare is an overall transgression.  The lower C-4 interval, including the C-455 and C-460 is low energy deltaic, and the upper interval including the C-440 to C-452 reservoirs are shallow marine.

Detailed structure maps were constructed using seismic and well log control for the Top Guasare (unconformity), and four stratigraphic horizons.

The hydrocarbons are trapped by a three-way closure on the upthrown side of a major down to the northeast fault. Six major faults dominate the overall structure pattern. These six faults separate the field into four production zones. A north to south fault with 100 to 150 feet of throw separates the field into an eastern and western half that produce differently. Twenty-two small faults break up the western half of the structure. The eastern half of the field contains one minor fault. Except for the major east-west bounding fault on the north side of the field, no appreciable growth occurs across any of the faults in the C-4 section. Cross sections indicate a general thickening of the upper C-4 interval to the northwest. 

Page 2, Geologic Framework

The C4 to C460 reservoirs are subdivided into sixteen-reservoir flow units that are mapped across the field.  A reservoir flow unit is a reservoir zone that is continuous laterally and vertically, has similar ranges of porosity, permeability, and capillary pressure properties, and has similar position in a sedimentary sequence and bedding characteristics.

The field area contains numerous barriers and baffles to horizontal and vertical flow. Both horizontal and vertical flow barriers and flow baffles were recognized throughout the field. Both continuous and discontinuous shale laminations form barrier and baffles to vertical flow. Barriers to vertical flow are easily recognized on RFT data.  Barriers and baffles to horizontal flow are more difficult to recognize due to limited pressure information.  Facies changes and faults form the barriers and baffles to horizontal flow.

Three scales of vertical flow barriers (Mega, Macro and Micro) are recognized in the field. The mega barriers separate shales/siltstones that cover wide areas. RFT data suggest that mega flow barriers support pressure differences across the field. The macro barriers are recognized on logs because they are thick enough to be recognized. These barriers or baffles can not be correlated in nearby wells with certainty. The micro barriers or baffles are too thin to be recognized on logs. The continuity of these individual shales is probably local, but the high number of these barriers makes the effective vertical permeability low.

RFT data indicate that mega boundaries are field wide fluid barriers. Recognition of this fine scale of flow barriers/baffles has a large impact on completion practices and simulation results. The entire zone must be perforated to get effective flow through a flow unit. Pressure differences within flow units indicate that the entire section is not being drained.

Capillary pressure measurements for six shale/siltstone core samples show that the sealing capacity of the flow barrier shales ranged from 90 feet to 3,892 feet for gas and 71 feet to 3105 feet for oil. The permeability estimates from capillary pressure indicate permeabilities <0.001 md for all samples except for one that has a permeability of 0.044 md to gas. The shale layers will be effective barriers to fluid migration during any fluid injection project.

PETROPHYSICAL EVALUATION

Petrophysical evaluation of the C440 through C460 interval is based on log response interrelationships established by reservoir unit and calibrated to rock information obtained from petrographic analysis of cores, drill cuttings and thin sections. A lithology fraction variable was first calculated from the normalized gamma ray curve, and this variable was then utilized in the subsequent calculation of petrophysical characteristics. A reservoir quality index (RQI) derived from the deep resistivity log and the lithology fraction curve was useful in the early stages of the project as an aid in differentiating bar and channel facies and in correlation of the flow units.  Porosity was based on the density log response, and was augmented by a relationship between density porosity and lithology fraction in order to solve problems associated with log calibration and invalid log response in rugose boreholes. The cementation exponent was defined as a function of the porosity and RQI using the observed dependence of cementation exponent upon porosity established by petrographic analysis.

The Waxman-Smits cation exchange capacity model was selected for the water saturation calculation. Clay volume was determined from the lithology fraction curve, and Qv (cation exchange capacity per unit pore volume) was determined from clay volume and porosity using a clay activity level representative of kaolinite. Permeability was calculated as a power function of porosity and water saturation, with a correction technique applied in order to account for the variation in water saturation within the transition zone. Selection of appropriate parameters for the permeability equation was guided by measurements on cores and by estimates on drill cuttings.

The pay counts and petrophysical properties of each of the sixteen reservoir units are summarized for each well. Within each of the zones the net feet of pay and average porosity, water saturation and permeability were calculated in each of four pay categories. In addition, the Lorenz coefficient to describe reservoir heterogeneity was calculated for the combined pay categories selected for reservoir mapping and the simulation model. The software utilized for the petrophysical work was used to generate files of the pay summary results in the format required by the mapping software system. The mapping process sometimes highlighted values of reservoir properties for specific wells, but not that unusual when considered as a single well. Investigation of the potential problems sometimes led to discoveries of errors associated with the upstream processing calculations or with the original digital log data. All significant problems associated with the data anomalies have been addressed prior to generating the final maps.

RESERVOIR CHARACTERIZATION

Seven attributes were mapped for each of the 16 flow units of the Eocene "C" reservoirs (112 total maps) .

The maps for each reservoir flow unit were mapped as follows: Facies maps were constructed using information from core, cuttings, calibrated well log shape, reservoir quality index, RQI and mapped distribution of petrophysical attributes. Structure maps had structural contours that follow well values. Where well control was sparse, contouring follows the depth converted seismic Guasare interpretation. Seismic data and well cuts positioned fault polygon. For the simulation maps, contours were only offset along faults with large throw. Net Feet of Pay/Net Sand used depositional facies models as a guide for contouring.

Porosity data came from core-calibrated log determinations. Porosity values show very little variation across the field. To ensure a uniform distribution of contours across the maps, porosity was contoured at 0.5% or less. Permeability came from cross plot data from cores calibrated with well logs. The distribution of permeability for each flow unit was hand edited based on porosity, facies and net pay thickness trends. 

Water saturation was determined from well logs and saturation contours follow well values. However, away from well control, the relationship between structure and porosity guided contouring of water saturation. Structure and porosity control water saturation by the equation:

Sw=O. 0386*ř-1.526*d-0.464

Where Ř = Porosity

Sw = Water Saturation

d = Depth above free water level (FWL)

FWL = (C453/455/460) = 13,860 feet

FWL = (C440 to 452) = 13,400 feet.

Hydrocarbon Pore Volume: The water saturation, net feet of pay and porosity maps controlled the hydrocarbon pore volume (HPV)mapping. Contours were based on the calculated HPV found by multiplying the hand edited porosity, water saturation and net feet of pay maps using equation:

(I-Sw) x (Net Feet Pay) x (Porosity)

Lorenz Coefficient is a measure of reservoir heterogeneity. Lorenz Coefficient contours follow well values. No clear relationship between facies and Lorenz Coefficient was observed in the data.

RESERVOIR PERFORMANCE AND OBSERVATIONS

Reservoir performance of the Eocene "C" was monitored using wireline formation tests, pressure buildup analysis and production logging with flowmeter surveys. Data from these tools provided an understanding of the horizontal barriers. reservoir lateral continuity, formation damage, correlation with log-observed pay categories and water production and its effect on well performance.

The wireline formation pressures are the most accurate in determining the amount of depletion in specific producing intervals. Buildup pressures measured after the well completion are difficult to use for this purpose because the completion intervals usually commingle several different reservoirs. The extrapolated pressure is a weighted average of the reservoir pressures and is therefore not definitive for determining the pressure and depletion in the individual layers. It is difficult to establish layer continuity from RFT data, but these measurements clearly identify depletion occurring in the various layers. However, where depletion exists, lateral continuity must exist with offset wells responsible for the depletion at the infill well location.

Page 2, Reservoir Performance and Observations

Production logging is an invaluable tool for quantifying the contribution to production of the individual layers in a multi-zone field, such as the Eocene "C". The profile is necessary to monitor reservoir performance, to insure all perforated intervals are producing and to design and evaluate stimulation treatments. Based on a review of the
flowmeter logs in this field, many of the perforated intervals in the wells are not contributing to production. Many of these non-flowing, perforated intervals are zones that have reasonably high reservoir pressure and have good rock quality based upon petrophysical analysis. The explanation for this problem is probably inefficient
perforations and/or formation damage from completion/stimulation fluids or scale.

Water production in this field is a serious concern because of the impact on well performance and the possible reduction in oil reserves within the affected reservoirs. All wells in the study area completed in the C-455/C-460 reservoirs are producing water at rates ranging from 1 to 300 BWPD. Many of the wells in this area have been abandoned within the C-455/C-460 due to water production (sometimes at relatively low water production rates) and have been recompleted in shallower horizons. Water production appears to be due to breakthrough in an interval or communication behind casing. Two mechanisms that may account for the water production with poor pressure support are water cusping from the aquifer and expansion of the connate water as pressure declines.

Production rates could most likely be increased by workovers and selective stimulation in the C-460 and the intervals between the C-455 and the C-440. Seven wells have been selected for a proposed workover program.

DEVELOPMENT OPPORTUNITIES

Development opportunities consist of new wells in new reservoir compartments and workovers of existing wells.

Three new drill well opportunities are recognized: northeast of the field in a separate fault block, down structural dip to the southwest and to the southeast. Northeast Fault Block: This well has a potential for 6.0 million barrels of moderate risk recoverable reserves in an undrilled structure in a separate fault block. The opportunity is located several hundred feet structurally higher than a poor producer in an adjacent separate fault sliver.

Southwest Extension: this opportunity is located in production Area IV that is separated from the rest of the field by an east-west fault. The opportunity is along the structural strike of four wells that have produced at least 17.8 million barrels with a low water cut. This extension area could have over 100 million barrels of oil-in-place and
potential recoverable reserves of over 20 million barrels from the C-440 through the C-460 reservoirs.

Southeastern Flank: this opportunity is within production Area I and most likely contains significant volumes of undrained oil. The most downdip well in the fault block was never completed, but petrophysical evaluation indicates considerably more pay than well VLC-363, which produced more than 6 million barrels.

Page 2, Development Opportunities

Seven representative workover candidates in the C450/C-460 reservoirs are identified. These seven wells (VLC-586, VLC-642, VLC-916, VLC-952, VLC-955, VLC-964, VLC-1029) will provide critical opportunities and help refine the reservoir simulation model. A successful workover program could add up to 3,600 barrels of oil per day and add reserves of over six million barrels.

There are numerous other downdip opportunities in the C-440 to C-488 intervals. Petrophysical evaluation shows no large increase of water saturation with depth. This suggests that the field water level is considerably down structure than previously mapped.

From production performance and water occurrence, four separate production regions are proposed. The north-south fault in the center of the field is probably sealing; production performance is different on either side of the fault. East of the north-south fault, the producing water level is more than 250 feet higher than west of this fault. Structurally higher wells just to the north of the east-west fault are abandoned, yet the wells south of this fault are still productive. This suggests that the east-west fault has a major effect on production.

NUMERICAL SIMULATION

The objective of the simulation effort was to examine the potential for enhance oil recovery in the C455/C460 pay intervals of the Eocene C-Lower, VLC-363 reservoir, Block III, Lake Maracaibo. The study was conducted in two parts. First the displacement efficiencies of alternative processes were examined using one-dimensional simulations. Then the recovery efficiencies of alternative processes were examined using multidimensional simulations of representative reservoir elements.

The oil is very volatile in this reservoir and there is considerable shrinkage below the bubble point. The pressure in the target intervals of the reservoir were well below the bubble point when this study was initiated. The extensive depletion below the bubble point has resulted in 1) significant oil shrinkage; 2) the formation of secondary gas caps underlying barriers separating flow units; and 3) insufficient free gas being available to
significantly swell the oil upon repressuring by means of water injection. The consequences of these circumstances are 1) there is very little waterflood moveable oil saturation; 2) repressuring by means of water injection will not significantly increase the waterflood target; 3) without repressuring, gas injection will resulting very little oil recovery by means of mass transfer, although with repressuring the system could become miscible; and 4) without repressuring, injected gas will stream through existing high gas saturation regions of the reservoir.

A representative cross-section of the reservoir, some 28% COIP, was used to study alternative EOR processes. An eleven grid-layer model was used to represent seven geological flow units comprising the pay interval of interest. The two dominant, highly stratified flow units were described by three grid-layers each by a technique that utilizes Lorenz functions.

Thus, the reservoir mapping provided by the geoscience study provided not only the normal petrophysical parameters for input to the simulation model, but also a mapping of the Lorenz function which represents the areal variation of the reservoir heterogeneity which controls vertical sweep efficiency. 

Without an initial reservoir fill-up, the predicated poor displacement efficiency for water and gas injection resulting from the extensive depletion below the bubble-point is predicated to be exacerbated by poor sweep efficiency resulting from reservoir heterogeneity and the high free-gas saturations existing at the start of injection. Because so much prior shrinkage and limited free gas remaining in place at the start of injection, a waterflood is predicted to recover less than 0.05% OOIP additional oil over continued primary. Similar results are found for gas injection. However, if target flow-units are filled up via water injection, resulting in substantially increased displacement pressure for subsequently injected gas, the oil recovery is predicted to improve markedly. Straight gas injection after fill-up is predicted to recovery 0.11 OOIP additional oil over primary. For a 2:1 WAG process, the predicted additional recovery is 0.16 OOIP. And for an infill program which would result in much better pattern configurations, the predicted additional oil recovery is 0.23 OOIP.