Conformance Control and Proactive Reservoir Management Improve Deepwater Production

The K field is one of the more developed deepwater fields currently going through development in Malaysia.

The K field is one of the more developed deepwater fields currently going through development in Malaysia. It has an excellent data set from which to optimize future development activities. In hindsight, it is clear that much more complexity exists than initially thought. As a result, uncertainty does not necessarily diminish at the start of production and a comprehensive collection and analysis of dynamic performance data are required in order to optimize recovery further.

Field Background and Geological Setting

The K field is a deepwater development located in 1330-m water depth offshore Sabah, Malaysia. The field was discovered by the K-1 well drilled to a depth of 3600 m on 30 July 2002 and marked the start of deepwater development in Malaysia. Five years after first oil, the field has more than 30 active wells including producers and injectors.

The K field is located in Block K (Fig. 1) and comprises the outbound tract of a major northwest/southeast-trending foreland fold-thrust belt that extends from Brunei to the Philippines and forms the margin of the North Sabah trough. Block K is dominated by fold-thrust structures. The reservoir section of the K field is dominated by mass-transport deposits with interspersed complex reservoir-bearing turbidite deposits.

jpt-2013-9-conformancefig1.jpg
Fig. 1—K field location in Malaysia.

Subsurface Development Challenges

The deepwater fields under development and study phases in Malaysia are usually considered to have more or less similar subsurface complexities and uncertainties. Among these uncertainties are level of heterogeneity, thinly-bedded-to-blocky sands, compartmentalization, fault intensity and behavior, reservoir connectivity, pressure and flow communication across the field, injection requirement from early production time, sand/fines production and reactive shale, wellbore stability, and commonly inadequate available data at the time the development decision is made.

The typical type log in the K field shows the reservoir has been subdivided into eight distinct reservoir packages labeled H110 through H150. In this example, the gross reservoir thickness (h) is 492 m with a net sand thickness of 50 m, giving an overall net/gross ratio of 0.102. Apparent from the type log is a large percentage of thinly bedded reservoirs, characterized to be beds that are less than 30 cm thick.

Core data indicate that the thin beds ranging from 2 to 30 cm in thickness have porosity and permeability (k) comparable to those of the thick beds, an observation further supported by well-test kh comparison with log-derived kh and numerous production-logging tool logs that have been run in the field.

The initial field-development-plan strategy was to develop the reservoir in three packages (H110–H115–H120, H130– H136, and H140–H145–H150) with updip production and downdip water injection.

Despite being in 1330-m-deep water, reservoir horizons are as shallow as 2400-m true vertical depth subsea. The development strategy had to employ multiple drill centers to access the oil, including wells drilled from the dry-tree unit or spar, and required subsea manifolds for both production and injection.

Production- and Injection-Performance Evaluation

The field has a comprehensive data-monitoring system and reservoir-management strategy (Fig. 2). All wells have downhole pressure and temperature gauges with real-time data access provided back to the office. The early interference detected in the downhole-gauge data proved invaluable in confirming producer/injector connectivity, particularly in the blockier sands. The reservoir-management strategy was based on the full voidage replacement by waterflooding and gas-cap gas injection (only in H150).

jpt-2013-9-conformancefig2.jpg
Fig. 2—Reservoir-monitoring and -management workflow. DHG=downhole gauge; VRR=voidage replacement ratio; GOR=gas/oil ratio; FBHP=flowing bottomhole pressure; PTA=pressure transient analysis; MDT=modular formation dynamic tester.

The implementation of the reservoir-management strategy has been based on weekly reviews including production, subsurface, and operations teams. The reviews incorporate voidage calculations, production performance, well-test data, and dynamic-model history-match updates as available.

The actual field performance after Phase 1 and during Phase 2, however, was not completely in line with the initial 3D dynamic-model predictions. The timing of water breakthrough was one of the key uncertainties because of unknown thin-bed extension and potentially uneven injection. The earlier-than-anticipated water breakthrough in some wells caused sand/fines instability and severe sand production, which led to catastrophic well failures and production loss after Phase 1.

Another event that added more complication to the dynamics of the field was out-of-zone water injection, which happened in several injectors. Although corrective measures were taken instantly, the extent of the healing process is uncertain and some crossflow might result within the field.

Considering high-pressure depletion and the unexpected-water-breakthrough pattern observed in some areas in the field partly because of heterogeneity, subseismic geological features, water-injection distribution, and thin-bed extension, the connected sand volumes in the initial models were proved to be overestimated. Many examples now exist in the field where sands disappear or thin or thicken dramatically within 100 m of well control. On the basis of the observations so far, the classic view that uncertainties will decrease through time with more wells and dynamic data does not necessarily hold for all deepwater turbidite fields. The extension of the thin beds and the dynamic complications thereof have always been and will remain a major source of uncertainty in this deepwater field development.

Injection-Performance Analysis

For analyzing water-injector performance, the Hall plot and modified Hall plot in combination with other plots such as the Hall-plot derivative, well-performance-analysis plot, and injectivity index were used. Because all the wells were equipped with downhole gauges as part of the reservoir-monitoring plan, the Hall plot was generated on the basis of both tubinghead pressure and bottomhole pressure.

The Hall plot is a diagnostic tool for monitoring water-injection-well performance. Hall-plot analysis is conducted by plotting cumulative water injected vs. cumulative injection pressure, either bottomhole pressure or tubinghead pressure. A straight line with constant slope indicates the well is injecting consistently. Any deviation from the straight line indicates plugging or fracturing effects. In order to look at the Hall plot more closely, a derivative of the Hall plot vs. cumulative water injection was also generated. For a well performing consistently, the derivative of the Hall plot will be a horizontal line, and any change in this horizontal line vs. cumulative water injection indicates plugging or fracturing.

On the basis of well-by-well injector-performance analyses during the injection-well life, decreased injectivity was observed in some of the water injectors after prolonged shut-in or after being choked back. One possible reason for this kind of behavior is flowback of sand/fines during shut-in, which would block the pores at the sandface and reduce injectivity.

The kh distribution in three WX02 reservoirs was predicted to be quite even; however, the observed water-injection rate into the H110–H115–H120 reservoirs did jnot follow the ratio of the kh. This is thought to be primarily because of fracturing performance in the early part of the injection life of the well. The water-injection split and distribution sensitivity to the injection rate raises questions regarding the water-injection distribution in commingled sands on the basis of kh data. This kind of problem potentially can be addressed by using selective and smart water-injection schemes, which are being considered currently.

Production-Performance Analysis

Because of the uncertainties and unexpected events in the field, different production-performance-analysis techniques were applied to provide a range of predictions of future field production. This analysis was also used to complement and sense check the dynamic 3D simulation-model results. Production-/injection-well performance analysis and decline-curve analysis (DCA) were performed on a well-by-well basis.

Wells with sufficient production history were considered in the category of “existing wells.” DCA for each well was evaluated by two methods: log (water/oil ratio) vs. cumulative oil and oil rate vs. cumulative oil.

Wells with little or no production history were considered in the category of “new wells.” Carrying out DCA for these wells is not as straightforward as for those with sufficient production history; therefore, a customized DCA was carried out.

For existing wells, a correlation was generated between measured oil rate and the capacity (kh) of the well. Capacity of wells in the new-well category was estimated from the reservoir properties from static/dynamic models and by use of the previously mentioned correlation, and initial oil rates were estimated for the new well. To cover the range of uncertainty apart from the most likely initial oil rate, low and high values also were estimated from the correlation. For estimating the production on plateau and decline factor for new wells, statistical analysis of the DCA of the existing wells was used. To estimate the production on plateau, a cumulative distribution function of cumulative production as a fraction of estimated ultimate recovery (EUR) for each well was prepared.

On the basis of the production-performance analysis (DCA), the EUR is lower than the most likely figure of the field prediction based on the 3D dynamic model. This has been translated to an average 8% lower ultimate recovery factor (up to 50 million STB) compared with the reported value from the 3D dynamic model.

The next stage of development is likely to focus on identifying, screening, and ranking missed opportunities, in terms of either unswept or unsupported oil. The use of smart injection wells is being considered to optimize injection conformance and focus pressure support on the known areas of unswept oil.

Smart-Well-Design Methodology

The K field is moving into the brownfield stage, and a Phase-3 redevelopment review is under way. As a consequence of the geological and stratigraphical compartments, more producers might be needed; however, improving the reservoir conformance and sweep efficiency may also assist in optimizing the number of wells. From the operational point of view, it makes more sense to free up the spar slots for producers and relocate injectors to subsea templates. This would serve to create better access to producers for possible interventions or chemical treatments for reducing skin.

On the basis of the subsurface feasibility studies, the following objectives were defined:

  • Improve reservoir conformance and reservoir sweep.
  • Delay water breakthrough in new wells.
  • Reduce/eliminate well intervention for subsea wells to reduce operational expenditures.
  • Accelerate and increase oil-production rate, and maintain or improve reserves.

A comprehensive subsurface-opportunity-framing and well/zone-screening study was performed to arrive at candidate selection for the first field trial. Consequently, a study was conducted on smart-well-completion design on the basis of the performance data and subsea-wellhead configuration and taking account of surface-facility considerations.
The following solution was proposed:

  • Implement multizone selective subsea water injectors.
  • Allocate the required water injection into selected zones by means of choking interval-control valves.

Considering all the operational and development challenges in this field, the project team successfully adopted various technical initiatives and fit-for-purpose solutions in different disciplines such as drilling, completion strategy, sand-control methodologies, selective smart injection, and proactive reservoir-monitoring and -management plan. Fig. 3 shows the short- and long-term production and additional-reserves gains in this field as a result of this study. In addition, the study shows that there is a potential of achieving a 4% incremental recovery factor through increasing the well water-cut limit from 95 to 98%. This will be pursued by enhancing the topside facilities and water-handling capacity.

jpt-2013-9-conformancefig3.jpg
Fig. 3—Short-term production profile after optimization efforts in different disciplines.

This article, written by Editorial Manager Adam Wilson, contains highlights of paper IPTC 16702, “Deepwater Production Improvement Through Proactive Reservoir Management and Conformance Control,” by Rahim Masoudi, SPE, Hooman Karkooti, SPE, Shlok Jalan, SPE, Anndy Arif, Keng S. Chan, SPE, and Mohamad B. Othman, SPE, Petronas; and Steve Burford and Philip Bee, Murphy Sabah Oil, prepared for the 2013 International Petroleum Technology Conference, Beijing, 26–28 March. The paper has not been peer reviewed.