SPE 110605

A Methodology to Design Exploitation Plans through the Application of Thermal Process, Orocual Field, Venezuela I. Anaya, M.M. Hernandez, A. Luces, PDVSA; A.J. Serna, Computer Modelling Group

Copyright 2008, Society of Petroleum Engineers This paper was prepared for presentation at the 2008 SPE/DOE Improved Oil Recovery Symposium held in Tulsa, Oklahoma, U.S.A., 19–23 April 2008. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Contents of the paper have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of SPE copyright.

Abstract Orocual is the most complex Field in Monagas, eastern Venezuela. Due to its complexity it has been divided vertically into two zones: shallow (heavy oil) and deep (light oil and condensate). The OOIP in shallow reservoirs is reported as 3500 MMSTB. The reservoir pressure is close to original conditions due to low historical production. Therefore, the current oil recovery factor is less than 1%. To date, cyclic steam injection pilot was applied to 5 wells and the initial oil rate showed to be as high as five times the cold production in vertical wells and six times the cold production for horizontal wells. According to the results of the pilot test, the future development of shallow Orocual will be mainly based on the application of thermal processes. This paper shows a methodology to design an exploitation plan through the application of thermal process. The first step involves analyzing the reservoir by sectors in order to determine which thermal process is appropriate to the reservoir; the second step is building sector models to simulate each process and optimize operational parameters. For the first time cyclic steam injection, steamflooding and steam assisted gravity drainage were simultaneously simulated in the same model. According to this study it is possible to maximize the production of these reservoirs, accelerate the exploitation of its reserves and optimize operational parameters in thermal recovery, as well as determine critical factors for each process. This study shows that numerical simulation of complex process can be efficiently carried out in FullField scale. Introduction Heavy oil has become an important theme in our industry with an increasing number of operators getting involved or expanding their plans in this market around the world. Venezuela has many heavy oil reservoirs, including the Orinco belt - the world's largest accumulation of heavy and ultra heavy oil. For these reasons many efforts should be focused in the definition of the production strategy of the heavy oil. The strategy affects the reservoir behavior, which influences future decisions and consequently, attractiveness of projects. An adequate plan of recovery for heavy oil reservoirs has a great economic importance in the oil industry and the interest for optimum management has increased, promting several studies to develop efficient methodology for optimization problems. Thermal processes has proved to be a good alternative to improve the recovery of heavy oil but the application of these processes should be studied in the contex of the reservoir. The number of the processes and its placement are basically related to static and dynamic characteristics of the reservoir. Modeling adequately several thermal processes under the context of the reservoir allows to design an exploitation plan to increase the recovery factor. This paper describe a methodology to select thermal processes and its placement, study each process and simulating them simultaneously in the same model. It is important to indicate where the proposed methodology is most appropriate. A reservoir characterization must have been completed in a preliminary phase and a numerical simulation model exists calibrated through a history matching process so that model saturation and pressure distributions are representative of field conditions. Alternatively, a model could also be directly initialized to field conditions if there is enough reservoir data and sufficient confidence to build it.

2 SPE 110605

Thermal Processes

There are a number of processes which can be used in this methodology. In the case study thermal processes which improve the recovery factor were considered. Cyclic Steam Stimulation (CSS)

To use this EOR method, a predetermined amount of steam is injected into wells that have been drilled or converted for

injection purposes. These wells are then shut in to allow the steam to heat or "soak" the producing formation around the well. After a sufficient time has elapsed to allow adequate heating, the injecting wells are put back to production until the heat is dissipated with the produced fluids. This cycle of soak-and-produce, or "huff-and-puff," may be repeated until the response becomes marginal because of declining natural reservoir pressure and increased water production.

Steam Flooding

This is a thermal oil recovery process which presents improved recovery technique for shallow heavy-oil reservoirs by

pumping high-temperature steam into injectors wells to heat up the formation and reduce the viscosity of the oil.

Steam Asisted Gravity Dranaige (SAGD) The most common implementation of SAGD consists of two parallel horizontal wells, the first drilled near the bottom of

the reservoir and the second located a short distance above it. The top well provides continuous steam supply into the reservoir and the lower one allows for continuous production of oil, gas and condensate water, developing a steam chamber.

In-Situ Combustion

The oil in the reservoir is ignited and oxygen is inyected in the wells: part of the oil is burned in the formation to generate

heat. The combustion zone created moves through the formation toward production wells, providing a steam drive and an intense gas drive for the oil recovery.

Numerical Simulation

The reservoir simulation was conducted with Computer Modeling Group’s (CMG) thermal reservoir simulator STARS. The geological model imported into the simulator is the result of an upscaling geoestatistical model integrating structural, petrophysical and sedimentological models. The oil-water and gas-liquid relative permeability curves were obtained from a detailed history match of an Orocual Cyclic Steam injection pilot project. Some of the values used in the simulation are: a top of the model 2383 ft, a rock heat capacity of 20 Btu/cu.ft-°F, a thermal oil conductivity 1.8 Btu/ft.day.°F, 8.6 Btu/ft.day.°F for thermal water conductivity, 0.64 for thermal gas conductivity, initial temperature 117 °F, initial oil saturation 77% and initial water saturation 23%. The reservoir simulation fluid model components consisted of oil, solution gas and water (liquid and steam). The bubble point pressure is 1200 psi saturated at initial condition. The characterization of the viscosity vs. temperature table was obtained by using six existing laboratory analysis.

Full-Field Grid

The usual sequential stochastic modeling approach (1) was applied for the characterization of this heavy oil reservoir. The

total number of the cells is 580000 (i, j, k) with 240000 active cells. The grid size is 85mx85m in i and j directions, and 6m in the k direction.

Sector-Grid Models

Most of the data and parameters were taken directly from the full field model. However, information from new wells was

included and considered relevant for the sectors. The grid size is 20m x 20m in i and j directions and 1 m in the k direction.

Methodology The design of an exploitation plan includes several stages from visualization to evaluation. The six stages of this methodology are described as follow:

SPE 110605 3

Preselection of Areas and Processes

In order to study and select the most appropriate area for each process under the context of the reservoir, several sectors encircle among existing wells were analyzed to determine if the rock and fluid properties are within the ranges for applying a specific thermal processes. Properties like API gravity, reservoir depth, pressure, Vsh, net pay, porosity, permeability, water cut, etc. were analyzed to determine the most optimal processes for the reservoir according to the above metioned parameters. The processes considered for the preselection were cyclic steam stimulation, steam flooding, SAGD and in-situ combustion. In the case study the analysis was applied to 15 sectors of the lower sands of Las Piedras Formation.

This analysis indicates that cyclic steam stimulation could be applied in all the sectors studied, SAGD processes are

proposed for four areas and ten sectors were selected for steam flooding. No sector was selected for in-situ combustion because of the reservoir conditiond is close of its original state. Figure 1 shows the areas selected for SAGD processes with the reservoir properties for one of the areas.

Figure 1.Selected areas for SAGD and stratigraphic column.

Sectorial Evaluation of Processes

The success of the start up stage in a thermal process depends on the operational parameters. It is very important to find out the proper injection rate, distance between producers and injectors, the length of horizontal section of the well length, preheating period, well configuration, etc.

Evaluation of Cyclic Steam Stimulation

Cyclic steam stimulation pilot project has been applied successfully to 5 wells with three cycles each one, whose initial oil rate proved to be five times the cold production in vertical wells and six times the cold production for horizontal wells. The results of the pilot proyect proved that the future development of the shalow Orocual will be mainly based on a widespread application of cyclic steam injection. Because of the limited operational experience with horizontal wells a sector model was built from the full field (Figure 2) to study the operational parameters of this process.

4 SPE 110605

Figure 2. Full field model and sector of the CSS study.

The volume of cumulative steam injected was evaluated with 5000, 7500, 10000, 12500, 15000, and 17500 Tons of steam injected. The results shows in the Figure 3 indicate that increasing injected volume, improve the recovery of the oil up to a maximum of 750 MSTB is reached when 14000 tons of steam is injected.

Figure 3.Optimization of cumulative injected volume and injection rate for CSS.

Sensitive analysis for steam injection rate was made from 125 to 1000 Tons/Day, indicating that the maximum

cumulative oil recovered is reached with 250 Tons/Day of injected steam. The soak time was studied in an interval between 15 and 60 days showing that there are no differences in cumulative oil production at several soak times.

Evaluation of SAGD

The optimization of operational parameters for SAGD processes was made with the most common configuration which consists of two parallel horizontal wells, the first drilled near the bottom of the reservoir and the second located a short distance above it. The selected areas for SAGD processes and the sector-model for numerical simulation are shown in the Figure 4.

Bar Channel Shale Crevasse

SPE 110605 5

Figure 4. Sectors selected for SAGD processes.

Different sensitivies of vertical spacing between injector and producer (I/P spacing) of 5m, 10m and 15m were tried.

Simulation results showed that increasing the spacing between injector and producer improves the performance of SAGD process and increases the final cumulative oil production (Figure 5). However the distance between injector and producer is limited by the net pay of the reservoir and could be more than 15m. In the current case the optimal I/P spacing is estimated to be 15m out of reservoir net pay of 100 ft.

Figure 5. Optimization of I/P spacing and injection rate for SAGD.

Figure 5 also shows the results of oil production for several injection rates ranging from 80 to 190 Ton/day.

For higher values (160 and 190 Ton/D), the ultimate recovery decreases indicating that SAGD process is not efficient at high steam injection rate. While injecting 110 Ton/D the production is more estable and the ultimate recovery is higher indicating that it is the optimum value for steam injection.

A sensitivity analysis on horizontal length was carried out using lengths of 700, 1000 and 1500 ft of horizontal section. The results show that there is no difference between the oil rate behavior of horizontal wells with 700 and 1000 ft of length, in both cases the oil production is stable for about five years. The optimal horizontal length was estimated to be 1500 ft due the oil production lasting for seven years.

The preheating period was evaluated circulating steam, in both wells during a month and the production was compared to

the model without steam circulation. The results show that for the Orocual Field crude of 10000 to 15000 Cp of viscosity, the preheating period is not necessary. The thermo-hydraulic communication between the producer and the injector is created without initial circulation phase.

1

23

Bar Channel Shale Crevasse

6 SPE 110605

Several well configurations were evaluated using the optimal operational parameters above mentioned. The Figure 6

shows the five wells configuration considered for the evaluation with steam injection in horizontal or vertical wells (Blue) and oil production in horizontal wells (Green). A sensitivity analysis was made for the distances between producers from 100 to 200 m and the result shows that the optimum value is 140 m.

Figure 6. Evaluation of SAGD for several well configurations.

One of the most useful parameter to determine efficiency of a SAGD process is the steam oil ratio (SOR). Figure 6 also

shows the SOR for the evaluated well configuration, indicating that during a period of ten years, pattern number three is the most efficient because of its low SOR. The less efficient configuration seems to be the number four.

Figure 7.Cumulative oil production for several SAGD configurations.

The ultimate oil recovery for several SAGD configuration is shown in the Figure 7 indicating the pattern number three as

the best configuration with the highest value of cumulative oil. Based on the SOR and cumulative oil production, the configuration with injection in two horizontal wells and productions through three horizontal wells, gives the best SAGD performance.

Evaluation of Steam Flooding To optimize the operational parameters of steam flooding a fine scale sector model with geostatistic in more details was

built in one of the areas selected. The sector-model built to simulate and the others four areas selected are shown in the Figure 8.

SPE 110605 7

Figure 8. Areas selected for Steam Flooding.

In order to determine the most appropriate position of the producer and injector wells, distance between producers and

injectors, injection rates and preheating periods, several configurations were studied combining vertical and horizontal wells. The well configurations for the analysis can be observed in the Figure 9 showing in blue the injector wells and in green the producers.

Figure 9.Evaluation of Steam Flooding for several well configurations.

Figure 9 also shows the temperature profile for each evaluated well configuration. According to the results largest area is

affected by steam when the well configuration number four (4) consisting in four horizontal and two vertical wells, is used. Its cumulative oil production higher than the others patterns as is shown in the Figure 10.

1

23

4

N1

Bar Channel Shale Crevasse

8 SPE 110605

Figure 10. Cumulative oil production and injection rate for steam flooding.

These results are obtained for the following operational parameters: 250 Tons/day of injected steam (Figure 10), a

distance between 140 and 200 m from the producer to the injector. The preheating period is not necessary for the Orocual Field crude between 200 and 500 Cp of live oil viscosity.

Evaluation Scenarios

To design the exploitation plan of the field, the results of the processes sector evaluation were used as input data, taking

into account areas and processes selected with their optimal operational parameters. Different scenarios were studied from the most simple to multiple processes numerical simulation, adding complexity to the model. The base case to estimate the efficency of recovery processes consist in one simulation with no new well perforation (drilling) in the field, but workover of the existing wells in order to simulate a case in the current condition of the reservoir, this is called the “Base Case”. It is very important to metion that the “Base Case” is a reference scenario to compare the efficiency of the recovery processes, it is not presented as a choice. All scenarios were evaluated at twenty years of production and the workflow followed to design the exploitation plan of the field is shown in the Figure 11 and explained in the next section.

Figure 11.Simulation workflow for the evaluation of scenarios.

SPE 110605 9

Scenario A It consists in applying cyclic steam stimulation in a number of new wells located through geological, geophysical and

petrophysical analysis. Sixty five horizontals wells were located and its trajectories were designed respecting 3°/100ft of dogleg variation and a maximum of 45° of tangent, as trajectory construction parameters. The first simulation run considered only one cycle of steam stimulation (CSS) per well in order to analyse the behaviour of oil production for each of them and according to the results, location, trajectories and operational parameters were checked for the wells with low production and a second cycle was scheduled. The model was run with two CSS in the wells and some of them show low oil production indicating that a second cycle is not efficient. In the case where the wells show high oil production, a third cycle was planed. The oil production of the wells with the third cycle is analyzed and a last cicle is applied or discarded according to the behaviour of production. It is important to mention that in some wells no more than two CSS were planed due to their low production. The procedure described above is show in the first loop of Figure 11.

Scenario B This scenario includes all wells of the before case with cyclic steam stimulation, adding SAGD processes and steam

flooding in the selected areas during the screening stage. The operational parameters to simulate SAGD and steam flooding come from sector evaluation of processes. In the case of SAGD the patterns used in the model were selected according to the evaluated well configuration and geometry of the channel. For the steam flood the injectors wells were placed where the steam affect the largest number of wells in the neighborhood. It is important to mention that the schedule should be designed to evaluate thermal processes through pilot test in the field, and according to its results, the operational parameters for others sectors of the reservoir should be ajusted. In the case studied the execution of pilot project for SAGD and Steam Flooding was planed at second year of the exploitation plan and its application in other areas of the reservoir began during the fourth year after two years evaluation of the pilot test.

SAGD and steam flooding requires drilling 24 new wells, so that a total number of 98 wells are in the simulation model.

The first simulation run allowed to evaluate the behaviour of thermal processes in the full-field model and according to these results location of the well configuration, well trajectories, location of injectors and operational parameters were changed to improve oil production. This procedure is shown in the second loop of the workflow (Figure 11) and is repeated until the behaviour of thermal processes is improved in the simulation model.

Scenario C According to the results of the scenario B the model is run to evaluate the oil production of infill wells, located in high oil

saturation zones to drain the crude not contacted by steam. A number of 46 well trajectories were designed respecting the construction parameters and added to the model to be stimulated with cyclic steam injection and the operational parameters used in the scenario A. This scenario encompasses CSS, SAGD, Steam Flooding and infill wells, with a total number of 135 wells.

As in the scenario B the model is run to evaluated the oil production of infill wells. In the wells with low production, the

location, operational parameters or well trajectories should be changed many times as necessary while oil production of the infill wells is improved. This procedure is illustrated in the third loop of the workflow shown in Figure 11.

The forecast of simulated scenarios and its corresponding number of wells by year is presented in Figure 12. It is very

important to notice the difference in oil production between the Base Case (yellow line) and the rest of the scenarios.

10 SPE 110605

Figure 12. Oil production and perforation activity for the scenarios evaluated.

The green line shows that it is possible to increase the oil production with cyclic steam stimulation up to a maximum of

21000 STB/day accumulating 120 MMSTB which represents an increasing of 3.5% in the recovery factor. Applying SAGD in three sectors and steam flooding in four areas of the reservoir, the oil production showed in the blue line reach a maximum of 30000 STB/day with an ultimate recovery of 174 MMSTB which represent 4.6 % of recovery. The red line shows that it is possible to produce 10000 STB/day aditional to reach a maximum of 40000 STB/day, with infill drilling and accumulating 216 MMSTB, which represent 6% of the original oil in place.

Figure 12 also shows the well planning necessary to generate the oil production for each scenario. It can be seen that the largest drilling activity is concentrated in the first three years and the number of well by year including producers and injectors. This well planning could be acelerated or delayed according to the number of rigs available.

Conclusion

1. The application of this methodology allows optimizing the time of technical analisys during the design of an exploitation plan, generating high impact in the visualization and strategy of the business.

2. This methodology can be used as a powerful tool to calibrate the operational parameters of thermal processes, finding critical factors.

3. Studying adequately several thermal processes under the conditions and characteristics of the reservoir is possible to design an exploitation plan to drain the reserves and increase the recovery factor.

4. The sectors models are useful to study and understand different thermal processes but it shouldn’t be used to conclude about the efficiency or success of the processes until it would be evaluated in full field scale.

5. Through this methodology it is possible to efficiently model complex thermal processes simultaneously in a full-field scale.

6. The methodology shows that several thermal processes could be applied for different sectors of the reservoir and an exploitation plan can be designed with more than one thermal process.

7. This methodology can be used to study thermal processes as the first step in the design of pilot project in oil fields. 8. The exploitation plan designed through this methodology for Orocual Field involves the application of cyclic steam

stimulation, steam assisted gravity drainage, steam flooding and perforation of infill wells.

SPE 110605 11

Acknowledgement The authors wish to thank PDVSA for supporting this work and for permission to publish this paper. Nomenclature CSS = cyclic steam stimulation HW = horizontals wells I/P = injector-producer OOIP = original oil in place SAGD = steam-assisted gravity drainage SOR = steam-oil ratio VW = verticals wells References

1. Humberto A. Mendoza, Jose J. Finol, and Roger M. Butler “SAGD, Pilot Test in Venezuela”, paper SPE 53687: SPE Latin American and Caribbean Petroleum Engineering Conference, Caracas, Venezuela, 21-23 Abril 1999.

2. I. Anaya, R. Porjesz “Uncertainty Analysis on SAGD Process”, paper 2006-685: 1st World Heavy Oil Conference,

Beijing, China, 13-15 Noviembre 2006.

3. I.D. Gates, J. Kenny, I.L. Hernandez and G.L. Bunio “Steam-Injection Strategy and Energetics of Steam-Asisted Gravity Dranaige”, paper SPE97742/PS2005-332; SPE International Thermal Operations and Heavy Oil Simposium, Calgary, Alberta, Canada 4-7 November 2005.

4. J.W. Venegas Prada, L.B.Cunha, and F.J.S. Alhanati “Impact of Operational Parameters and Reservoir Variables

During the Startup phase of a SAGD Process”, paper SPE97918/PS2005- 414; SPE International Thermal Operations and Heavy OilSimposium, Calgary, Alberta, Canada 4-7 November 2005.

5. L. Marquez, J. Hurtado, PDVSA; J. Wang, F. Iwere, E. Gomez, P. Hudgens, E. Ponce, J.L. Bashbush, O. Gurpinar ,

A. Aly, Schlumberger Source “Application of Large Thermal Sector Models in a Field Performance Optimization Study of a Mature, Heavy Oil Field” SPE Annual Technical Conference and Exhibition, 5-8 October, Denver, Colorado.

6. L. Nakajima, D.J. Schiozer, “Automated Methodology for Field Performance Optimization Developed with

Horizontal Wells” UNICAMP Source SPE Latin American and Caribbean Petroleum Engineering Conference, 27-30 April, Port-of-Spain, Trinidad and Tobago.

7. M. Tamim, J.H. Abou-Kassem, and S.M. Farouq Ali “Recent Developments in Numerical Simulation Techniques of

Thermal Recovery Processes” SPE 54096-MS; SPE International Thermal Operations and Heavy Oil Simposium, Calgary, Alberta, Canada 4-7 November1999.

8. S. Das, “Steam-Fingering Phenomenon During SAGD Process”, paper SPE97729/PS2005-324; SPE International

Thermal Operations and Heavy Oil Simposium, Calgary, Alberta, Canada 4-7 November 2005.

9. Singhal, S. Das, S. Leggitt, M. Kasrale and Y. Ito “Screening of Reservoirs for Exploitation by Application of Steam Assisted Gravity Drainage/Vapex Processes”, SPE 37144 1996.

10. Y. Ito and G. Ipek, “Improving the Performance of SAGD”, paper SPE97921/PS2005-416; SPE International Thermal Operations and Heavy Oil Simposium, Calgary, Alberta, Canada 4-7 November 2005.

11. V. Ramlal, “Enhanced Oil Recovery by Steamflooding in a Recent Steamflood Project, Cruse ‘E’ Field, Trinidad”

paper SPE 89411-MS: Petrotrin Source SPE/DOE Symposium on Improved Oil Recovery, 17-21 April, Tulsa, Oklahoma.