Copyright 2005, Society of Petroleum Engineers This paper was prepared for presentation at the 2005 SPE Production and Operations Symposium held in Oklahoma City, OK, U.S.A., 17 – 19 April 2005. This paper was selected for presentation by an SPE Program Committee following review of information contained in a proposal submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society of Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the Society of Petroleum Engineers is prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300 words; illustrations may not be copied. The proposal must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435. Abstract The South Umm Gudair (SUG) oil field located in the Neutral zone between Kuwait and Saudi Arabia has produced since 1968 from an active water drive carbonate reservoir of Lower Cretaceous age. The lower zones are homogenous intervals of higher permeability which appear to be sufficiently swept by natural water drive over a period of time. The upper zones of the reservoir are more heterogeneous and have lower permeability in the range of 50-150 millidarcies. These upper zones are relatively thin and are bound by tighter intervals that act as effective barriers to the natural water drive system. Due to the presence of barriers and low permeability intervals, these zones have been poorly swept resulting in significant volumes of by-passed oil remaining in these parts of the reservoir. The new approach of exploiting these reserves by drilling and completing 4 horizontal and 2 horizontal side track (HST) wells targeting the lower permeability portions of the reservoir in the SUG field since January 2004; have yielded considerable success in extracting significant incremental oil production with the added benefit of very low water cut. This success has led to field development plan to recover un-swept oil reserves from these low permeability zones. This paper summarizes the various aspects of field development plan taking into consideration geology, reservoir data and production data while highlighting the successes of the new horizontal and HST wells in the low permeability reservoir portions of the SUG field.

Introduction The SUG field was discovered in 1966, and put on production in 1968. The primary recovery mechanism is a combination of

edge and bottom water drive aquifers. The field has been developed initially by vertical wells targeting all productive zones; which have been perforated and produced commingled. The SUG field today produces approximately 70,000 Barrel of Oil Per Day (BOPD) and 80,000 Barrel of Water Per Day (BWPD) from 64 active oil wells, out of which 23 are horizontal and HST wells. All wells are produced by artificial lift using Electrical Submersible Pump (ESP). Location and structure maps of the field are shown in Fig.1 and Fig.2.

Carbonate Reservoir The Ratawi Oolite carbonate reservoir is an anticlinal structural trap. The Early Cretaceous reservoir section was subjected to folding during the Late Cretaceous and Tertiary times associated with compressional events. The dominant carbonate lithologies consist of pelloidal/skeletal grainstones with lesser amounts of packstones, wackestones and minor mudstones. The Ratawi Oolite section at SUG is considered to have been deposited on a very broad, shallow, carbonate platform. Deposition occurred in inner ramp tidal flat, lagoon, and higher energy ramp crest environments. Deeper water more micritic lithologies, associated with flooding events at the base of these depositional cycles, form important reservoir barriers across large portion of the field. Low energy, tight, shallow shelf mudstone and wackestones of Upper Ratawi Oolite formation overlie the porous reservoir section. The upper and the lower part of the reservoir are fairly different with respect to formation characteristics. The lower zones (M4 to M12) are relatively homogeneous with permeabilities ranging from 300 to 400 millidarcies. The upper zones (M1 to M3b) of the reservoir are more heterogeneous and have lower permeabilities ranging from 50-150 millidarcies (called Low Permeability Reservoir). These intervals have a fining upward lithologic signature, as evidenced from the open hole logs of the wells. The reservoir layers are illustrated in the cross-section shown in Fig. 3. Several potential barriers are known to affect the vertical communication within the reservoir. Most of these barriers appear to be local, but at least two of these tight vertical barriers of fairly wide lateral extension have been identified in the reservoir. Barrier-1 is within the Low Permeability

SPE 94249

Development of Low-Permeability Carbonate Reservoir by Use of Horizontal Wells in Mature South Umm Gudair Field in the Neutral Zone–A New Approach M. Jha, SPE, T. Tran, B. Hagtvedt, M. Al-Haimer, and M. Al-Harbi, Joint Operations, Saudi Arabian Texaco/ Kuwait Oil Co.

2 SPE 94249

Reservoir (Intra M3b) of upper zone, and Barrier-2 is in the Lower zones (M5/M6) of the high permeability intervals. These barriers have been identified and confirmed through the use of formation pressure data from Repeat Formation Testers (RFTs) and Modular Dynamic Testers (MDTs). Monitoring this data through time, show a growing pressure difference across these identified barriers, as shown in Fig. 4. These two barriers can be traced all over the field based on open-hole logs as illustrated in Fig. 5. The lateral extension and sealing effect of these barriers across the field are still not understood completely, however, the significant pressure difference in certain areas around the field seem to indicate quite an effective seal or barrier to flow. Typical core data for a SUG well is shown in Fig. 6. Reservoir Development The SUG field was initially developed using vertical wells only. All productive zones in the vertical producers were perforated and produced together. Production Logging Tools (PLTs) run to evaluate production have shown generally poor contribution from the upper reservoir zones. Recent open hole logs from new wells indicate that the lower part of the reservoir with higher permeability, have mostly been swept, however the upper zone with lower permeability has by-passed oil remaining, due to its poor sweep efficiency. Due to the combination of low permeability and isolation by barrier, oil reserves in the upper layers could never be recovered effectively and economically by existing or additional vertical wells. The original reservoir pressure of 4,050 psi has dropped to the current average reservoir pressure of 2,200 psi. This reservoir pressure is not sufficient enough to produce oil from low permeability zones by existing vertical wells. This phenomenon is observed through out the field regardless of structural positions. The current reservoir pressure of 2,200 psi is still high above the bubble point pressure of 900 psi. This is due to, there being sufficient aquifer support thus not necessitating implementation of pressure maintenance through the use of water injection. Further more, water injection in this type of heterogeneous reservoir often affects the parts of the reservoir with the least resistance to flow, the layers having higher permeability. Thus, the Low Permeability Reservoir which has by-passed oil, would still not get much benefit from water injection. The vertical wells were drilled on 80 acres spacing in SUG field till the end of 2003. The review of the field performance indicated some signs of production interference between the wells. Based on plots and analyses of various well test and reservoir data, it has been determined that 80 acres is the optimum spacing for vertical well development to minimize production interference between wells. Therefore, additional vertical locations within SUG field can not be considered, as the productive area is already filled with wells drilled on a density of 80 acres spacing.1 The drilling of horizontal and HST wells began in year 2000, with the objective to mitigate the water conning

problems prevalent across the field in the vertical wells. The first horizontal wells were initially designed to target both the low permeability and high permeability layers. These wells were very successful in increasing the field oil production and reserve recovery; however, water began to show up after a phase of good production period. The subsequent reservoir surveillance data indicated that the source of water is primarily from the high permeability layers. This information and analysis resulted in a new approach of development of the SUG field which began in January 2004. The new approach was to drill horizontal and HST wells targeting only the Low Permeability Reservoir. These new horizontal wells were drilled mostly along the flank of the SUG field’s anticlinal structure on about 120 acre spacing. The justification for targeting the flank portion of the reservoir is primarily to recover oil reserves in the upper low permeability layers in areas, where vertical wells had been producing with high water cuts from the lower high permeability layers. Candidates for these re-entry HSTs are selected from existing vertical wells showing history of producing high water cuts and low oil production from the lower zones. As the target Low Permeability Reservoir layer is relatively thin (thickness varying from 15 to 30 feet) and having proximity to bottom water fronts, special horizontal well drilling applications using Geo-steering technology had to be employed to assure keeping drain hole within the targeted layer.2 Real Time Measurement While Drilling (MWD), Logging While Drilling (LWD) accomplished with the latest Geo-steering modeling services were used during lateral placement of horizontal well bores. Horizontal Well Performance Prediction The performance prediction study of the SUG field's horizontal wells in this Low Permeability Reservoir was carried out using a simplified simulation model.3-4 The model was built using the application of Eclipse to be able to focus on uncertainties related to development of a reservoir with limited formation thickness. The objective of the model was to evaluate the dynamics around the new producer well bores during the first few months after putting the well on production and to verify the level of protection from identified barriers against vertical water movement. Average formation characteristics, relative permeability data and the fluid model are taken from an abandoned full field simulation model of the SUG field. This model represents a dynamic unit that includes the target Low Permeability Reservoir level and the higher permeability reservoir unit directly below as shown in Fig. 7. The model grid contains 6 layers, Layer-1 to Layer-6 from top to bottom. Layer-1 is just above the target pay, Layers 2, 3 & 4 are within the target pay and Layers 5 & 6 are below the target pay. The reservoir characteristics of the various layers are based on available field data and are shown in Table-1. The layers below the barrier, have experienced water breakthrough, though not fully swept. Illustration of how the water cone builds up under the horizontal section of the well after 6

SPE 94249 3

months of production is shown in Fig. 8. The flow barrier in the upper section of the reservoir is modeled by a transmissibility reduction between grid Layer-3 and Layer-4. Based on available core and pressure data, Barrier-1 does not seal completely and is in pressure communication with the lower reservoir layers. Water from the lower layers could invade Barrier-1 if the pressure difference became high enough. However, this has not happened since the inception of production in 1966 and is unlikely to happen in the future. The retained transmissibility of the barrier was also verified by comparing simulated pressure differences across the barrier with pressure differences observed on recent MDT data. The model was initialized to match average static bottom-hole pressure from recent monitoring data. The simulation grid was made big enough to ensure adequate pressure support during production to avoid artificial depletion effects. Lateral water movement caused by an edge aquifer was not taken into consideration as it would require a history matched model. The model includes a well with a horizontal section of 1,500 feet landed in the middle of Layer-2. In order to test the impact of several uncertain parameters, various runs were made to understand the evolution of water production during the initial months of production. The two most important cases (Case-A and Case-B) are based on the expected formation characteristics given in the Table-1. The only difference between the two scenarios (Case-A and Case-B) is the sealing capacity of the Barrier-1. The Case-A has a vertical transmissibility estimated from core data, while Case-B has ten times higher transmissibility than Case-A in the same barrier. The simulation model predicted early water coning breakthrough. However, the model also indicated that horizontal wells recovered oil reserves more efficiently with a high oil production rate as compared to a low oil production rate which delays water conning. Well Planning and Geosteering Modeling The typical plan for the new horizontal well in SUG field is to first drill a high angle (about 50-60 degree inclination) pilot hole to verify the lateral target and oil water contact. The open hole logging data of the pilot hole and off-set wells are utilized to build a Geosteering model, taking account of apparent dip, stratigraphic position, electrical anisotropy, polarisation and shoulder bed boundary, until best fit of modeled log response, verses actual log data, is achieved. The final out put is predicted Resistivity, Density, Porosity and Gamma Ray log curves along the proposed horizontal trajectory of the well bore. The most common geological uncertainty in the SUG field is structural dip. Therefore, a Geo-steering model is built with variable structural dip change scenarios (+/- 0.5 to +/- 1.0 degree). The lateral placement strategy for the horizontal well bore in new and HST wells, is to put the lateral in the most porous and permeable part of the target reservoir while staying above the barrier. The primary objective is to obtain optimum oil production rate and delay the production of water from

water conning breakthrough. During the course of landing and lateral drilling inside the target reservoir, Real Time MWD and LWD data supplemented by Geo-steering is transferred and communicated between the Asset Management Team (AMT) and well-site personnel. This is done in order to monitor progress of well bore path so that rapid and real time decisions can be made to ensure success in staying inside the target zone. A typical display of MWD and LWD logs with a Geo-steering Model is illustrated in Fig. 9. Drilling and Completion One of the critical factor for the success of the horizontal new drills and HSTs in the Low Permeability Reservoir is the proper selection of a drilling fluid. This is due to the target zone lithology having much smaller grains and pore throat size, which can be easily damaged by an improper drilling mud system. The drilling fluid system used in the SUG field consists of a water based mud mixed with Calcium Carbonate as a bridging agent and a simple polymer system of XCD and Starch. The Calcium Carbonate particles in the fluid are properly sized based on the permeability and pore throat size of the target pay, in order to form a bridge across the pore throat and to quickly build a filter cake to minimize spurt and fluid losses. During lateral drilling, the fluid loss in the mud is maintained at 4 to 5 cc/30 min and drill solids content is controlled at less than 25 lbs/barrel. The FloPro NT mud system of MI Drilling Fluid with a re-refined XCD polymer and Starch was selected to further minimize fluid loss during drilling of the lateral. After the total depth of the well is reached, the drilling fluid and cuttings are displaced with a clean completion fluid. The open-hole lateral is then stimulated utilizing Coiled Tubing with 15% Hydrochloric acid at about 30 gallons per foot of well bore length and emulsified with diesel as a diverting agent to remove drilling filter cake and near well-bore damage. Subsequently an Electrical Submersible Pump (ESP) is run and the well is put on production. Monitoring Production History From January to July 2004, a total of six 6 horizontal new drills and HSTs (Wells A, B, C, D, E, F) were drilled and completed in the Low Permeability Reservoir of the SUG oil field. The cumulative initial test production from these six horizontal wells is about 20,000 BOPD with about 3% water cut (average of 3,300 BOPD/well IP). These results are very encouraging when compared to production of vertical wells in the same reservoir, which average 300–500 BOPD with about 50% water cut. Although the production history is short and the data being monitored is limited to production rates and bottom-hole pressures from ESP, the comparison of real production data and simulation results are consistent to date. The plot of water-cut versus cumulative oil production (Fig. 10) compares simulated data with real production data. The only anomaly

4 SPE 94249

thus so far has been that, Well-C experienced water breakthrough earlier than predicted by the simulation model. This however, is explained by the fact that, this well is located on the edge of the reservoir and is likely affected by lateral water movement. The production of the wells in the central part of the field fit very well and is very consistent with the model’s simulated water-cut. The consistency between the simulated and real data indicates that, the barrier does form an effective seal and is laterally extensive. This also supports the idea that the upper reservoir interval above the barrier could be treated as a dynamic independent unit. The simplified model indicates that vertical water movement resulting in water conning will occur within the initial few months after putting new horizontal wells on production. This model does not include for lateral water movement that could occur. The lack of robustness of the simulation model consequently makes it inadequate for predicting expected oil recovery and reserves. Therefore, the expected oil reserves per well for these horizontal wells has been estimated by simple plotting techniques and is estimated to about 2 Million Barrels of Oil (MMBO) (see Fig. 11). Conclusion The success of the horizontal drilling and side track program targeting the remaining/trapped/by-passed oil in the Low Permeability Reservoir layers of South Umm Gudair field have resulted in reducing the production decline for this mature field. The success of optimum horizontal well performance has been realized through: 1) Optimized well placement using Geo-steering technology. 2) Use of proper drilling and completion fluids. 3) Team work between Asset Management Team members, Drilling/Work-over teams and Directional drilling service contractors. This innovative method demonstrates that different thinking and strategies utilized and applied to mature fields can result in significant improvements in reserve recovery and production. This success has also led to a development plan and strategy of additional horizontal drilling locations in the flank areas and side tracking of existing wells to maximize recovery of un-swept oil from the Low Permeability Reservoirs in South Umm Gudair field. Acknowledgement The authors thank the management of Joint Operations, Saudi Arabian Texaco-Kuwait Oil Company for their permission to publish the paper. Suggestions of Khaled Al-Ali, Superintendent and review by Kevin Lahay, of Joint Operations are gratefully acknowledged. We acknowledge the modeling support provided by J.Ray of Schlumberger.

References

1. Giger, F.M., "Low-Permeability Reservoirs Development Using Horizontal wells", SPE 16406

2. Arnold, C.W., Djauhari, H.H., "Review of Early Horizontal well Applications in Central Sumatra", SPE 29288

3. Fonseca, C.M., et al, "Reservoir Study of the Largest Oil Field in Argentina – A Two Reservoir 2200 Well Simulation Model", SPE 90952

4. Sherman, B.B.,et al "Case Study: Development of a Comprehensive Simulation Model for a Large Middle Eastern Carbonate in Kuwait and the Partioned Neutral Zone, Kuwait", SPE 68098

Table-1

_______________________________________________________________ Madan Jha works as Asset Management Team Geologist for Joint Operations, Saudi Arabian Texaco- Kuwait Oil Company. He holds a MS degree in Applied Geology from IIT University of Roorkee, India. Madan has 19 years Industry experience, the initial 14 years with Oil India Ltd ex Burmah Oil Company UK as Senior Geologist working for Onshore Oil & Gas fields and Offshore Exploration project before joining Joint Operations in 1999. He is member of Petroleum Exploration Society of Australia (PESA) and Society of Petroleum Engineers (SPE). Thanh Tran works as Asset Management Team Leader and Senior Petroleum Engineer for Joint Operations, Saudi Arabian Texaco- Kuwait Oil Company. He holds a BS degree in Petroleum Engineering from University of Tulsa, USA. Thanh has 23 years Industry experience in Production operation, Reservoir management, Horizontal drilling and Geosteering. Bård Hagtvedt works as Asset Management Team Senior Reservoir Engineer for Joint Operations, Seconded from TOTAL-Kuwait to Kuwait Oil Company. He holds a MS degree in Reservoir Engineering from the Technical University of Trondheim, Norway. Bård has 14 years industry experience as a Reservoir Engineer for TOTAL which includes: Reservoir management, Field monitoring, Field operations and Numerical Simulation. Mohammad Al-Haimer works as Asset Management Team Senior Reservoir Engineer for Joint Operations, Saudi Arabian Texaco- Kuwait Oil Company. He holds a BS degree in Petroleum Engineering from University of Southern California, USA. Al-Haimer has 7 years industry experience in Field operations & monitoring, Reservoir management and Numerical Simulation. Musaad Al-Harbi works as Asset Management Team Geologist for Joint Operations, Saudi Arabian Texaco- Kuwait Oil Company. He holds a BS degree in Geology from Kuwait University. Musaad has 13 years Industry experience as a Development Geologist with Saudi Arabian Texaco. He is member of American Association of Petroleum Geologist (AAPG) and Society of Petroleum Well Log Analysts (SPWLA).

Grid layer Pay Por

(%)

Kh

(md)

Kv/Kh Layer thick-ness

(ft)

Sw

(%)

1 M3a 15 50 0.2 10 0.20

2 10 0.20

3 10 0.20

4

M3b

20

100

0.4

10 0.25

5 10 0.46

6

M4 25 150 0.6

10 0.67

SPE 94249 5

Fig. 1—2D Structure map

Fig. 2—3D Structure map

6 SPE 94249

Fig. 3—Cross-section through reservoir layers

SPE 94249 7

Fig. 4—MDT data: increasing pressure difference across the two identified barriers

8 SPE 94249

Fig. 5—Distribution of Barriers

SPE 94249 9

Fig. 6—Core data

10 SPE 94249

Fig. 7—Phenomenological model: Cross-section Initial conditions

Fig. 8—Phenomenological model: Illustrations how The water cone builds up under the horizontal section of the well after 6 months of production

SPE 94249 11

Fig. 9—Geosteering model: Horizontal display of MWD-LWD logs

12 SPE 94249

Fig. 10—Comparison of simulated data and historical production data for the three first wells

Fig. 11—Expected oil reserves for the three wells With longest production history