Experience of Drilling the Horizontal Well VLD-1152 in Lagunillas Formation, Block IV, Lake Maracaibo Basin, Venezuela.
The main objective of the horizontal well VLD-1152, located in Block IV, was to improve recoverable reserves which was impaired by pressure depletion and reservoir heterogeneities. The well represents an important challenge because it is the first horizontal well drilled in a depleted pressure area and it was drilled within a small productive interval of 25 feet thick only.
A pilot area was selected after a detailed multi-disciplinary study by geologists. petrophysicists and reservoir engineers. New 3D seismic interpretation revealed a structural model that conformed well with pressure behavior of the area. New information from well VLD-1112 were utilized to update the petrophysical properties and the volumetrics These data were input to develop improved reservoir description and build a reservoir model for flow simulation.
The results indicated that Layer VII is the most important drainage target. The principal reasons for selecting this unit were, good mechanical stability of the rock, absence of a water front and a secondary gas cap and the presence of a regional shale ar the top that might be used to navigate drilling.
Despite some operational problems encountered in drilling, the results were mostly satisfactory. The entire pay was penetrated and the geology and petrophysics of the drilled area came in line with our model predictions.
A pilot area, containing approximately 30 wells located in Block IV in Lake Maracaibo Basin,was selected as the site for a horizontal well (Fig.1). The target reservoir, VLC-52/VLD-192 Lower Lagunillas, commenced production in 1957 with well VLD-192.
The reservoir, which is stratigraphically divided into L, M and N sands, has not been uniformly drained. Since 1960, most of the wells have been completed in the L and N sands; therefore, the M sand has been less depleted. (Fig. 2) Production declination was very intense and was partially controlled by a gas injection program in 1967. Dropping pressure however continued until getting 1000 psi (Fig. 3)
The drilling of VLD-1112 well at south of the selected area, contributed with valuable information to validate the petrophysical parameters and calculate a new OOIP number which was found to be 20 % greater than the initial estimate of 264 MMSTB.
Once the new geological model and petrophysical parameters were defined, the more prospective area with less operational risk was selected. The output was used as information to develop a dynamic model for the simulator. The results provided well defined boundaries conditions and indicated the absence of an independent aquifer.
The selection of the zone was based on a combined evaluation of different criteria of orientations, lengths and restrictions for the horizontal well. This zone showed low values of porosity and permeability and a depleted reservoir character, justifying drilling of horizontal wells in order to improve oil recovery and maximize the production rate. The recommended location was GOF-3.
The main target is Unit VII of M Sand Lower Lagunillas Member. This sand has an potential of 1200 STBOD, and is expected to have a water cut of 5% and a GOR of less than 1000 SCF/STB.
Accurate Reservoir Evaluation from Borehole Imaging Techniques and Thin Bed analysis. Case Studies in Shaly Sands and Complex Lithologies in Lower Eocene Sands, Block III, Lake Maracaibo, Venezuela.
Computer aided signal processing in combination with different types of quantitative log evaluation techniques is very useful for predicting reservoir quality in complex lithologies and will help to increase the confidence level to complete and produce a reservoir. The Lower Eocene Sands in Block III are one of the largest reservoirs in Block III and has been producing light oil since 1960. Analysis of Borehole Images shows the reservoir heterogeneity by the presence of massive sands with very few shale laminations and thinly bedded sands with numerous laminations. The effect of these shales creates a low resistivity that has been interpreted in most of the cases as water bearing sands. A reduction in porosity due to diagenetic processes has produced a high resistivity behavior. The presence of bed boundaries and shales is detected by the microconductivity curves of the Borehole imaging Tool which also allows for estimation of the percentage of shale in these sands. Interactive computer aided analysis and various images processing techniques are used to aid in log interpretation for estimating formation properties. Integration between these results, core information and production data was used for evaluating producibility of the reservoirs and to predict reservoir quality. A new estimation of the net pay thickness using this new technique is presented with the consequent improvement in the expectation of additional recovery. This methodology was successfully applied in a case by case study showing consistency in the area.
The Lower Eocene Sands in Block III (Fig. 1) were discovered in 1962 by well VLC-363 in the eastern part of Lake Maracaibo. Early estimates suggest that 1.6 Billion stock tank barrels of oil were initially in place in the Lower Eocene Sands in Block III (Ref. 1).The initial recovery factor was calculated at 42 % and has produced 270 MMSTB while the remaining estimated reserves are calculated at 290 MMSTB. It was interpreted at the beginning of field exploitation that Lower Eocene sands were homogeneous (Ref. 1 ). Initially, the wells in the southern part had an average production of 4000 barrels per day of light oil. After ten years, the northern and eastern parts of the field where developed. These wells were troublesome, very costly and with very low production rates (Fig. 2). It became apparent that the reservoir was highly heterogeneous and complex. Production logging performed in some wells showed that only small intervals in the lithologic column were productive and that in some cases those intervals did not appear on standard logs as having the best potential. Any improvement in reservoir evaluation would require significative changes in the standard procedure followed for a normal homogeneous field. Effective reservoir characterization requires a successful integration of varied geological and petrophysical data that provides multiple benefits including increasing accuracy of reserves and well productivity estimates. Although high resolution well logging instruments have significantly contributed to improved reservoir description, the use of these increasingly remote and sophisticated methods makes it easy to forget that integration is the key to solving problems. High resolution electrical resistivity imaging is an important new tool in the field of petrophysics and is an important piece of information for core-log calibration. There are heterogeneous properties that can not be detected with a normal suite of log. These properties could control production of an interval due to restrictions in vertical flow through the sands. This paper describes how measured electrical images correlate with the porosity and permeability of the samples, and, in turn, with their petrographic characteristics.
We study a new vertical sweep efficiency correlation for reservoirs with non-uniform layers and compare the effect of permeability autocorrelation on these estimates. Vertical sweep efficiency correlations are important because, in spite of sophisticated numerical simulators now available, there still remains a need for rapid estimates of oil recovery. The vertical sweep efficiencies are calculated using the chemical flooding simulator UTCHEM for a two-dimensional vertical cross-section with one injector and one producer, both also vertical.
As measures of heterogeneity we use the Dykstra-Parsons coefficient (VDP) in the case of a strictly layered reservoir and the Gelhar-Axness coefficient (HI) in the case of a nonlayered reservoir. We validate a strictly layered case and predict the vertical sweep efficiency in cases with nonuniform layers. We show that for mobility ratios greater than one, the effluent concentration of the injected fluid (in this case, polymer) oscillates because of viscous instability.
For unlayered cross sections, we show that the Gelhar Axness coefficient is a better vertical sweep efficiency estimator than the Dykstra-Parsons coefficient, because it accounts for spatial correlation (autocorrelation) as well as for heterogeneity. The vertical sweep efficiency decreases as we increase HI for the same VDP.
The Laguna water injection project was initiated by Maraven, S.A. in Jan. 1967 and consist of 145 wells of which seven are injectors located in the low part of the reservoir. Since 1973 two attempts to reproduce the project performance using three dimensional reservoir simulation performance using three dimensional reservoir simulation models have failed due to difficulties in matching the reservoir performance and the advance of the water injection front. Both studies were temporarily suspended for acquisition and analysis of additional field data to resolve ambiguities which affect the simulation results. In 1986 a sedimentological study of the reservoir and flow units allowed the identification of the main flow channels within each sedimentary unit observed in the main reservoir. Based on the results of this study a new three dimensional reservoir simulation model was built, integrating the geological and sedimentological models, allowing proper characterization of reservoir heterogeneities and reproduction of the project performance on a well by well basis through the end of 1987. The use of sedimentological facies and the flow units concept for the reservoir description of the Laguna project was the key for the numerical modeling of the water injection front advance in each flow unit within the accumulation, allowing to establish the actual distribution of the remaining oil saturation and the pressure profile. Finally, the future performance of the reservoir was predicted under different injection schemes, and an infill predicted under different injection schemes, and an infill drilling plan, employing ten new production wells in regions being poorly drained will permit to increase the recovery for the field by 40 MMSTBO (3% STOIIP).
The Laguna reservoir is located in Lake Maracaibo, Venezuela (Fig 1.). The initial oil in place was estimated in 1458 MMSTB, and cummulative production for January 1987 was 387 million barrels of 28 API gravity oil (26.5% of the STOIIP) leaving a remaining oil volume of 1072 MMSTB. Since 1951 to 1967 the reservoir produced by natural depletion with little support from an aquifer. The pressure had declined from the original of 4050 psia to 2060 psia at the depth of 9500 feet subsea. During August, 1967 secondary recovery was initiated injecting gas at the reservoir crest which was reforced in 1968 with water injection at the aquifer. Furthermore in 1974, the water injection was extended to the north-central part of the field. During 1975, the gas injection was part of the field. During 1975, the gas injection was discontinued since the expected reverts pressure support in the north area was not observed. The actual reservoir production rate is 16.5 MSTB/D and water is being production rate is 16.5 MSTB/D and water is being injected at a rate of 50 MSTB/D through five injectors in the south area and two in the north area. In the first part of the study a 3-D model was developed to reproduce the historical reservoir performance. The results of the reservoir match allowed to identify regions poorly drained under the actual production-injection front; observing that the injected water at the south of the reservoir has been channeling up dip along the main axis of the litoral bars identified in the sedimentological reservoir study (Ref.1) invading the zones with the best sand development in the reservoir. In the second part of the study the model developed was used to carry out sensitivities aimed to evaluated the strategy that allow to obtain the optimum scheme for the exploitation of the reservoir remaining reserves, and examine the effect on recovery of in fill drilling in regions indicated by the simulator as poor drained by the actual well configuration.
Maraven has implemented a Pilot Project in the Zuata area of the OrinocoBelt to evaluate the following parameters in relation to a commercialdevelopment:
Twelve inclined wells (7 producers and 5 observers) have been drilled in acluster configuration, using a slant rig with a well spacing at surface of 15metres and 300 metres in the reservoir.
Electricity, pumps and rig time are the major costs associated with beam pump artificial lifting method in heavy crude oil.
In reducing operation costs, which are mainly associated with sand production, high viscosity, high GOR and rod flotation, Ma raven has successfully tested the progressive cavity pump in Bolivar Coast progressive cavity pump in Bolivar Coast fields.
The application of the method has reduced energy consumption over 60%, increased well production an average of 12% and reduced rig production an average of 12% and reduced rig activity sore than 40% in extraheavy and heavy type crude oils. This equipment was also tested in extraheavy oil slanted wells with excellent results.
The application of this method through an in house computer program has saved 10 million dollars/year in Maraven operational costs. It is expected to extend this program to other areas.
The Bolivar Coast fields are located in the east side of lake of Maracaibo. These fields are characterized by a variable viscosity crude (100 - 60000 cps) with gravity in the range of 9 deg. - 21 deg. API. The reservoir produces by depletion and compactation mechanism and is normally lifted by a conventional beam pumping system. Intensive activity is required to pull and run
plunger due to problems associated with the plunger due to problems associated with the high viscosity of the oil and sand content.
Numerous solutions to handle these production problems have been attempted with production problems have been attempted with various degree of success. The most prevalent remedies have been the use of heavy pony rod and gravel-pack completion, which increase the stroke of the travelling pump and reduce the influx of loose sand with adequately selected gravel held in place by screens.
Production problems attributed to high viscosity are broken rods and polish rod, pump and gear box damage by hitting, low pump and gear box damage by hitting, low production, etc. that normally cause production, etc. that normally cause premature hoist entry which considerably premature hoist entry which considerably increased production costs.
In order to reduce operating cost, alternative lifting methods were investigated. The progressive cavity pump being one of them. This equipment consist basically of a rotary type positive displacement pump driven by standard API sucker rods. A motor installed at the wellhead transmits the rotational movement to the subsurface pump.
The advantages of progresive cavity pump system are related to its capacity to move fluid continously, contrary to beam pump which produces only on up-stroke c.g. or one half the time. This results in better efficiency in handling sand and gas problems.
PROGRESSIVE CAVITY PUMP SYSTEM DESCRIPTION PROGRESSIVE CAVITY PUMP SYSTEM DESCRIPTION A diagram of the complete equipment is shown in Figure 1.
Formation compaction, if present, can have an important influence on thermal recovery methods, as observed in Western Venezuela, and elsewhere. This paper discusses the effect of formation compaction on oil production by cyclic steam stimulation and steamflooding, using a fully implicit steam injection simulator. The simulator accounts for three-phase mass and heat transport occurring in steam injection processes, for a wide variety of operating conditions. It employs an implicit formulation together with a Newtonian, direct solution approach, and was shown to be stable for large time steps.
It was found that oil recovery in a compacting reservoir increases with an increase in the uniaxial compaction coefficient. However, whereas cyclic steam stimulation yielded a favorable response in a compacting reservoir, the opposite was true for a continuous steamflood. A delay in implementing a steamflood in a non-compacting reservoir can lead to a considerable loss of recovery, in the range of 10 to 40% of oil-in-place, depending on the value of the uniaxial compaction coefficient. This finding has far-reaching implications for steamflooding subsequent to intensive depletion by cyclic steaming, or primary production.
Although formation compaction can be beneficial from the standpoint of cyclic steam stimulation response, there is strong dependence on the compaction coefficient. Furthermore, it was found that if the oil in question exhibits non-Newtonian flow behavior - reported for some Venezuelan oils - it must be accounted for in numerical simulations, otherwise the oil production rates may be in error by as much as 100%.
In a number of oil-producing regions in the world, fluid withdrawal has resulted in formation compaction, giving rise to surface subsidence, and the accompanying environmental problems. This was first discussed by Geertsma and later by van der Knaap and van der Vlis. The latter also discussed the causes of compaction. The more notable regions where formation compaction (and land subsidence) has been observed over a considerable period include the Bolivar Coast, Western Venezuela, and the Long Beach area, near Los Angeles, California. Large heavy oil reservoirs occur in both of these areas. In particular, steam injection has been a commercial oil recovery method in Bolivar Coast, since the early sixties. Until recently (1978), the oil production method was exclusively cyclic steam simulation, except for a brief steamflood project. Currently, a major steamflood is underway, described recently by van der Knaap, the early performance of which is "remarkably similar" to that of the earlier, much smaller steamflood.
It has been noted by a number of authors that formation compaction provides an important oil expulsion mechanism. For example, Ref. 2 gives an approximately one-to-one relationship between gross liquid production and subsidence volumes, for conventional oil production. Compaction is also known to be beneficial from the standpoint of cyclic steam stimulation. For example, de Haan and van Lookeren attribute 40% of the oil production to compaction. The role of compaction is not so clear in the case of a steamflood, especially as regards the extent of prior cyclic stimulation. de Haan and Schenk note that upon the completion of the compaction process, conditions would still be favorable for a process, conditions would still be favorable for a steamflood, viz. the oil saturation would still be reasonably high, in view of low recovery by cyclic stimulation, and the reservoir pressure would be low, facilitating steam injection.