Liwan Gas Project: South China Sea Deepwater Case Study

The Liwan gas reservoir was discovered in the South China Sea Block 29/26 in September 2006. A sixth-generation, deepwater semisubmersible platform was contracted that subsequently drilled 26 deepwater wells (700- to 1600-m water depth) between November 2008 and November 2011. During the 1,011 days of drilling and testing activity, the rig and onshore supporting teams incurred no lost-time incidents. The rig team drilled 61 593 m of hole below the mudline and ran the associated strings of casing. Nonproductive time consumed 265 days (26.6 %), comprising 65 days waiting on weather, 147 days of rig repair, and 53 days of other unscheduled events.

Introduction

In April 2006, a dynamically positioned drillship drilled exploration Well Liwan 3-1-1 in 1500 m of water to a depth of 3843 m, discovering the Liwan 3-1 sandstone, a gas-bearing reservoir. The Liwan 3-1 field is approximately 320 km southeast of Hong Kong in 1300- to 1500‑m water depth, as shown in Fig. 1. The nearest field was 64 km northwest. The main reservoir in the Liwan 3-1 gas field (Sand 1) is the Zhujiang formation.

Weather and Ocean Environments. The South China Sea presents significant weather conditions that affect offshore operations. Tropical cyclones, typhoons in the western Pacific, occur throughout the year over the northern South China Sea. China’s offshore operators evacuate personnel from permanent and moored offshore facilities when typhoons approach. For dynamically positioned drilling rigs, the typhoon is monitored as it forms and a judgment is made as to when to suspend operations. While drilling, the average time required to suspend operations and recover the lower-marine-­riser package (LMRP) or blowout preventer (BOP) is approximately 72 hours. Nonessential crew are evacuated, and the rig moves out of the path of the typhoon, generally in a southwest direction. There are two monsoons that affect the South China Sea: Northeast (between November and March) and Southwest (between May and September) monsoons, the nomenclature deriving from the predominant wind direction.

Solitons, solitary waves that move across the ocean surface, can cause problems with station keeping and the ­dynamic-positioning (DP) system. Solitons can occur at any time of the year, but are more prevalent with the spring tides. Little is known about the solitons in the area, but they are known to produce westerly currents exceeding 2.0 m/s in the upper current layers and easterly currents exceeding 1.5 m/s in the lower current layers. Eighty-three percent of all solitons are westerly (coming from the east), and 17% of solitons come from other directions. The DP operators were able to observe solitons with the rig’s surface radar or by vision and react accordingly.

Drilling Program

Liwan 3-1-1 Discovery Well. The primary objective of exploration Well Liwan 3-1-1 was the Zhujiang formation. The ­shallow-hazard survey indicated two buried slumps or potential saltwater flows at 30 and 140 m below the mudline, respectively. The well was spudded in April 2006, and the 36-in. conductor was jetted to 82 m below the mudline. With the inability to predict shallow water flows, a pump-and-dump system was used to drill the 26-in.-hole section. The 20-, 13⅜-, and 9⅝-in. casings were set at 2081, 2387, and 3048 m, respectively, and the 8½-in. hole was drilled to total depth at 3843 m.

A planned sidetrack was performed with a 9⅝-in.-casing whipstock to core the sand reservoirs penetrated in the original well. The 8½-in. hole was sidetracked because of a hole-opener failure, with the root cause being hole instability. A drillstem test was not planned. The well data combined with the seismic interpretation provided justification to announce a significant gas discovery and future investment in an appraisal program.

The rig experienced one driftoff and two typhoons. The first typhoon and the driftoff delayed operations by 20 and 9 days, respectively. The second typhoon caused the marine riser to part below the rotary table, dropping 52 joints of riser and the BOP stack. The BOP stack was recovered after 32 days.

Liwan 3-1 Appraisal Program. The appraisal program helped in estimating the original gas in place, supported the substantial commercial investment, and provided fluid and production data for the front-end-engineering and design team. The Liwan 3-1 appraisal program comprised three wells, with the following data-acquisition program:

• Logging-while-drilling tools in 26-, 17½-, and 12¼-in. hole sections
• Wireline logs in 17½- and 12¼-in.-hole sections
• Cores in target reservoirs
• Drillstem tests

The drilling experience from Liwan 3-1-1 supported the following well-design modifications:

• Use oil-based mud below the 20-in. casing.
• Set 13⅜-in. casing above the reservoir.
• Set 9⅝-in. casing at total depth.

The first appraisal well, LW 3-1-2, experienced more than 30% nonproductive time because of rig repairs and the winter monsoons. Operational performance improved quickly, and total days on location for the three appraisal wells (including drillstem testing) were 115, 72, and 53, respectively.
Liuhua 34-2 Exploration Well. In September 2009, after completing the Liwan appraisal program and two exploration wells, the rig was moved to exploration Well LH 34-2-1. On 14 September 2009, typhoon Koppu formed west of the Philippines, requiring immediate suspension of operations. The LMRP was disconnected; however, the storm’s force was greater than the rig’s propulsion system and pushed the rig into shallower waters. Weather conditions prevented disconnecting the direct-acting tensioners (DATs) from the tension ring, so the riser could not be recovered. Attempts were made to slow the rig’s movement with anchors, but the mooring system had never been used. Two of the anchors could not be deployed, and a third anchor was lost. The LMRP struck the seabed, damaging the six DAT cylinders, which required 45 days to be replaced. This incident led to a review of the marine thrusters, power systems, and to the installation of a DAT-fastening system that provides a means of controlling DAT movement during storms.

Operations resumed, and a sandstone reservoir was penetrated, discovering gas with high liquid content. The test measured an equipment-restricted flow rate of 55 MMscf/D, with indications that the well could deliver 140 MMscf/D.

Liuhua 29-1-1 Exploration and Appraisal Program. Exploration Well ­Liuhua 29-1-1 was drilled in December 2009, approximately 43 km northeast of the Liwan 3-1 field. The well was tested at 57 MMscf/D, with indications that it could deliver 90 MMscf/D. The complexity of the LH 29-1 reservoir required drilling four appraisal wells and running one additional drillstem test. The design of Well LH 29-1 was similar to that of the Liwan 3-1 appraisal wells. Three of the five appraisal wells were grouped at a future well-manifold location. The target distance from the surface location required wellbore inclinations greater than 75°.

Liwan 3-1 Development Program. The Liwan 3-1 development program identified nine targets in three sand reservoirs. Three appraisal wells were used, and six development wells were drilled. Three production-manifold locations were selected, with two manifolds adjacent to appraisal Wells LW 3-1-3 and LW 3-1-4 and a third location for the central pipeline-end manifold. Four wells were drilled at the west manifold, three wells at the east manifold, and two wells at the pipeline-end manifold. Well LH 34-2-1 would be connected to the east manifold. The wells were capable of producing 50 MMcf/D, with a field plateau rate between 350 and 400 MMcf/D.

Development-Well and Completion Design. The ­completion-selection process commenced during the appraisal program. The main focus was on sand control and flow assurance. The deep­water infrastructure shown in Fig. 2 would combine the following:

• 10 wells
• Manifolds to gather production from individual subsea trees
• Compression and mono-ethylene-glycol reclamation on a shallow-water platform approximately 75 km from the manifolds
• A shore-based gas plant 300 km from the platform

The completion objectives that required analysis were:

• Sand control
• Subsea trees
• Surface location of wells (i.e., drilled vertically above target vs. directionally from the manifolds)
• Intelligent wells

Generally, development-well design incorporated the following completion elements:

• Cased-hole frac-pack completions
• Subsea trees 30 m from manifold locations
• 10¾-in. casing to 600 m, for surface-controlled subsurface safety valve

Economic analysis justified positioning wellheads near the manifolds vs. placing them above the targets. Drilling directionally from the manifold to the target, connecting the well to the manifold with a rigid jumper, and future well operability were less expensive than drilling vertically and connecting the subsea tree to the manifold with a long pipeline that could not be pigged.
Drilling Fluids. A flat-rheology oil-based drilling-fluid system improved results by increasing penetration rates and reducing the need for wiper trips before running wireline logs or casing. It also resulted in only one stuck-pipe incident in 26 wells. A cuttings drier was installed to reduce the amount of oil on the cuttings to approximately 4%. Excess mud was stored in the pontoons during operations.

Casing and Cementing Design. The 36‑in. conductor was jetted to 70 m below the mudline, leaving the wellhead top 3 m above the mudline. Pore-pressure and formation-strength separation provided the opportunity to eliminate one casing string between the 36-in. shoe and the reservoir. However, a conservative approach was used by setting the 20-in. and 13⅜-in. casings at 600 and 1500 m below the mudline, respectively, providing additional hole protection for the directional wells. The 9⅝-in. casing was set across the reservoir.

Standard subsea-cementing operations were conducted on the 20-, 13⅜-, and 9⅝-in. casing strings. One lightweight-cement formulation (13.5 lbm/gal) was used for the 20- and 13⅜-in. casings. The 9⅝-in.-casing string was cemented with a 15.8-lbm/gal blend.

Wellhead System. A 15,000-psi, weight-set, metal-to-metal-seal wellhead with a rigid-lockdown system that preloads the wellhead connection for higher fatigue life was used on all wells. The 36‑in. low-pressure wellhead was set 3 m above the mudline. The 18¾-in. housing was locked into the 36-in. housing, and the rigid-lockdown system was actuated after the cementing operations. The external lock ring was not run on the 13⅜‑in. seal assembly but was installed on the 9⅝-in. seal assembly.

The following adjustments were made to the wellhead during the program, to improve operation.

• Eliminated the mud mat that required bullseyes to be attached to the wellhead
• Used a rotary-table housing for the 36-in. low-pressure wellhead to assist the jetting and running of the 26-in. drill-ahead assembly

Directional-Drilling Operations. Rotary-steerable systems were used in the 17½‑in.- (in later wells, this hole section was reduced to 16-in. diameter) and 12¼-in.-hole sections in vertical and directional wells to maintain directional control. The development did not require horizontal wells, although wellbore inclinations ranged from 45 to 75°. The 26-in. hole was drilled directionally toward the target at a build rate of 1°/30 m by use of a steerable motor, obtaining 10 to 20° inclination at casing point. The remaining build section was obtained in the 16‑in. section with ­rotary-steerable systems and was maintained through the 12¼-in. section.

Learning Curve. Learning-curve theory describes organizational improvement mathematically. The drilling per­formance was analyzed with a drilling-­performance curve (as detailed in Brett, J.F. and Millheim, K.K., 1986. The Drilling Performance Curve: A Yardstick for Judging Drilling Performance. Paper SPE 15362 presented at the 1986 SPE Annual Technical Conference and Exhibition, New Orleans, 5–8 October. http://dx.doi.org/10.2118/15362-MS). The tool can be used to assess the drilling performance in an area where consecutive wells are drilled. The analysis indicated an effective learning rate.  The drilling team conducted weekly lesson reviews, devoting measurable attention to improvements. Rig-based work guidelines and drilling/testing programs were adapted quickly to new learnings. Fig. 3 is a chart of the overall drilling performance for wells in Block 29/26.

Rig-Operation Efficiencies. The team used the simultaneous capabilities of this sixth-generation deepwater rig as follows.

• The 20-in. casing was picked up and hung in the moonpool from the subsea-tree cart while jetting the 36-in. conductor.
• The 13⅜- and 9⅝-in. casings were racked in the derrick while drilling associated hole sections.
• A bucking machine was installed to assemble bottomhole assemblies off the critical path of operations.

Additional rig supervisors were employed to manage the simultaneous operations and perform deck management. The need for more supervision led to the need for a high level of accommodation space on the platform. The platform had 180 bunks, and most were filled each day. Although the extra supervision may seem expensive, it was found that efficiency was high and simultaneous operations were safer.
Safety. The rig experienced no lost-time incidents during the 1,011 days of operation, accumulating more than 3,000,000 man-hours. The annual total recordable incident rates (TRIRs) (per 200,000 man-hours) for the years 2009 through 2012 were 0.409, 0.648, 0.28, and 0.19, respectively. Sixth-generation deepwater rigs use derrick-and safety-zone-­management systems to minimize the number of personnel in the derrick and at the rotary table. Specialized-­systems personnel are required to maintain the software and the mechanical elements of the system.

Nonproductive Time. Between 4 November 2008 and 18 October 2011, the rig experienced 265 days of nonproductive time as follows:

• Waiting on weather—65 days (5.7%)
• Rig repair —147 days (15.4 %)
• Other —53 days (5.5%)

Between April and November, typhoon activity in the South China Sea reduces productive time on a deepwater rig by 20 to 30 days/year. This time loss assumes that activities are suspended and the LMRP is recovered 3 to 4 times during the typhoon season.
Approximately 100 days were required for rig repair during the first year.

Only one stuck-pipe event occurred in the 26 wells. The 17½-in. assembly became stuck, and recovery attempts were not conducted; the hole was sidetracked around the assembly. It was believed that poor hole cleaning contributed to this stuck-pipe event. Therefore, the hole size was changed from 17½ to 16 in., increasing annular velocity by 10%.

Redundant systems of all critical-path equipment, including a spare remotely operated vehicle, were stored at the onshore base. This philosophy reduced the unscheduled-event time when equipment repair or replacement was required.

No tools or drillstring elements were lost because of string failures. Three remedial cement operations were required—two low-formation-­integrity tests of the 20-in. shoe that indicated poor isolation and a squeeze of the 9⅝‑in. casing before testing.

Post-Macondo Upgrades. The rig contractor performed the following upgrades:

• Acquired a second acoustic surface-control module
• Installed a hydraulic dead-man system
• Increased the shear-rams closing pressure from 3,000 to 5,000 psi

A thorough review of oil-spill contingency plans was conducted with the other operators in the South China Sea and government agencies.

Conclusions

Key performance indicators for Husky Energy’s 26-well drilling campaign in the South China Sea Deepwater Block 29/26 included the following:

• Operating time=1,011 days
• Distance drilled=61 593 m
• Lost-time incidents=0
• Four-year annual TRIR=0.409, 0.648, 0.28, 0.19
• Nonproductive time=26.6%
• Average days/10,000 ft=17.34

This article, written by Senior Technology Editor Dennis Denney, contains highlights of paper IPTC 16722, “The Liwan Gas Project: A Case Study of a South China Sea Deepwater Drilling Campaign,” by David Triolo, SPE, and Tracy Mosness, SPE, Husky Energy, and Rana Khalid Habib, Schlumberger, prepared for the 2013 International Petroleum Technology Conference, Beijing, 26–28 March. The paper has not been peer reviewed. Copyright 2013 International Petroleum Technology Conference. Reproduced by permission.