Deepwater flow assurance best practice: Keeping the network moving

Modern simulation technology informs important network decisions, but other factors, such as fluid composition and interaction, hydrates, scaling, corrosion, sand, gel, emulsion, foaming, and wax and asphaltene deposition, also must be understood.

Pg74 Chart Sm

Simulation technology can improve design, operations decisions

Hongkun Dong
Schlumberger

Modern simulation technology informs important network decisions, but other factors, such as fluid composition and interaction, hydrates, scaling, corrosion, sand, gel, emulsion, foaming, and wax and asphaltene deposition, also must be understood. The benefits of digital flow assurance simulation are illustrated by a deepwater well restart benchmark and optimization study in theGulf of Mexico.

The classic exploration, appraisal, and development stages of the E&P lifecycle all build toward one thing—sustained hydrocarbon production for as long as possible. Significant efforts are made to extend the life of productive wells to achieve this. In the same vein, operators must pay attention to maintaining the constant flow of hydrocarbon fluids from pore space to processing plant and, ultimately, on to the balance sheet.

Flow assurance indeepwater is even more critical. Appropriate fiscal returns on such expensive and risky projects must be ensured, and with a million-dollar-per-day average spread rate, non-productive time is a serious issue for deepwater operations. It is a constant consideration for every field—even long producers can develop flow assurance issues over time. However, planning for successful flow assurance management can begin with modern simulation technology long before any hydrocarbons are produced.

Optimal deepwater flow assurance involves analysis and modeling of fluid interactions within the reservoir, well, pipeline, surface facilities, and the surrounding environment. Regardless of the development scenario, accurate reservoir characteristics and fluid property information must be established to produce the ideal production system design, from the reservoir to the topside facilities, and from exploration to abandonment.

Pg74 Chart Sm
Click to Enlarge
Simulated oil flow rate at the separator compared with actual field data. (All images courtesy Schlumberger)

Fluid sampling and analysis

Correct understanding of fluid behavior and interaction is of the utmost importance for flow assurance analysis, and representative reservoir fluid samples are essential to provide quality data. High-quality, single-phase, downhole samples for accurate flow assurance characterization can be collected in open and cased-hole environments. Modern downhole reservoir testing uses wireless toolstring communication to allow a clean fluid sample to be acoustically triggered, in real time, from a surface computer during cased-hole drillstem testing (DST).

Samples must then be maintained at reservoir conditions to ensure the fluid remains intact for laboratory analysis—for example, maintaining the pressure and temperature to ensure waxes and asphaltenes remain in the fluid. Samples can then be validated and quality checked, with the best selected for flow assurance analysis.

Next, the gas/oil ratio can be established, as well as its composition and saturate, aromatics, resins, and asphaltene (SARA) contents. This is important for sample validity checking.

Particular attention is paid during the analysis to understand when the waxes and asphaltenes will drop out of solution, important for quantifying potential flow issues. It is also possible to predict when hydrates will form and cause blockages—all important considerations for system design.

Optimal flow assurance characterization defines phase boundaries, and establishes the likelihood and extent of depositions in the pipeline, and the severity of the resultant blockages over time. Digital simulation tools for transient and steady-state conditions also can represent thermal hydraulic behavior to determine whether, and what kind of, thermal management will be needed.

The ability to test organic and inorganic deposits in live reservoir fluids at field conditions is the most accurate way to determine fluid behaviors and can help reduce both capex and opex. Quantifying the effects of chemical additives on actual deposits under representative conditions contributes to efficient spending and reduces cost. Realistic organic solids deposition measurements improve the accuracy of systems modeling and completion designs. In turn, production operations can be optimized through better system design, chemical selection, dosage, and treatment. Pigging frequencies and remediation strategies can be improved, too.

The latest equipment can independently vary test parameters to quantify the effects of pressure, temperature, composition, surface type, flow regime, and shear on the deposition behavior of organic deposits such as waxes and asphaltenes. Deposits can then be collected for testing and quantification. The deposit mass is used to calculate the deposition rate, based on the cell surface area and test run time, which can be scaled up to the field conditions through modeling.

Deepwater flow assurance should always be considered from an integrated standpoint, taking into account the well, reservoir, and production angles to make sure the full range of fluid scenarios and compositions are examined, and to avoid costly resampling and re-evaluation after systems are built. Production fluids interact with the reservoir, well, pipeline, surface facilities, and the environment. All these impact flow assurance, leading to potential issues with hydrates, wax, asphaltenes, scales, slugging, emulsion, foam, sand, and corrosion. Deepwater flow assurance requires a full understanding of these interactions and a multi-disciplinary approach to managing them. Modern simulation software allows such an approach to be integrated efficiently into asset team workflows.

Inevitably, it will be necessary to shutdown and restart a system, whether for repair, maintenance, or to unfavorable weather. A well-designed startup procedure, informed by precise simulation, is therefore important. Deepwater well intervention is particularly difficult. Even on land re-entering wells is expensive and time consuming. Intervention expense, risk, and complexity are amplified in deepwater.

Pg76 Chart Sm
Click to Enlarge
Upstream and downstream wellhead choke pressure comparison: Simulated and actual.

Flow assurance simulation

The power of accurate simulation in deepwater flow assurance scenarios is illustrated in a GoM project1. It tested the accuracy of transient multi-phase flow simulations against field data for a full-cycle well restart operation, and proposed optimizations for future procedure and project design.

The study simulated the actual offshore production system to include one well, two flowlines, and one riser. Well reservoir pressure was around 620 bara (9,000 psia) and its gas/oil ratio was 116 Sm3/Sm3 (650 scf/stb). The analysis was done using the OLGA dynamic multi-phase flow simulator, taking system fluid thermodynamic and transport properties from PVTsim software.

Benchmarking evaluated the accuracy of the transient model simulating a cold earth well restart after 35 days of shut-in. The simulation took in the entire process, from steady state production prior to shut-in, to the restart and to the return to steady state (which were benchmarked with actual field data).

The simulation assumed a cold restart, in which the entire system including the wellbore reached ambient conditions enabled by a "fastwall" option in the software. To minimize the manifold pressure before the restart, gas lift administered through coiled tubing unloaded the riser. After this the pressure in the riser dropped to about 59 bara (850 psia). During the restart simulation, the wellhead choke was opened to match field data. Flow coefficient information was not available for the topside valve, so the outlet boundary was modeled at topside valve with the pressure changing according to field data.

The simulation results closely matched the field data, with less than 4% differential for steady-state operation in terms of flow rate, pressure and temperature, and they correctly reflected the transient process during the well restart. Compared with the steady-state measurement, the subsea manifold pressure tended to be lower than the field data and the temperature matches at wellhead, and was about 2% higher at outlet. A 5% increase in oil viscosity and density gave a better match with the field data pressure measurement.

During restart benchmarking, the simulation caught transient flow behaviors in terms of terrain slugging, pressure variation, etc. The production rate trend at separator matched the field data. Subsea pressures tended to be slightly lower than the field data, which agreed with steady state results, and the simulation showed lower temperature during the restart if the "fastwall" option was used during shutdown. This was because 35 days were not enough for the fluid in the wellbore to cool, so the period was re-simulated without using this option to improve the match.

Using the benchmarking results, optimization of the future well restart procedure was done to identify the most efficient, economic, and safe process. This included the well and topsides choke opening schedule to allow restart at a higher rate, while minimizing slugging. It was known that the flow coefficient did not increase significantly until the valve opened more than 25%. The valve was opened rapidly to 20% before the restart, and then linearly up to to 50% within 13 hrs. of restart. This significantly reduced the initial slug and eased flow rate fluctuations.

In another scenario, the topside valve was held at 35% open for 9 hrs. into the restart until the second slug was processed through, and then opened linearly to 100% over the next 5 hrs. Here, the initial slug was larger compared with the first scenario, but it was still much smaller than without any control. Trailing slugs were also smaller. It was concluded that a small topside valve opening through the first 9 hrs. of the restart is important for slug control.

In addition to the manual control scenarios tested, two others were done using a PID controller to maintaining a constant pressure upstream at the topside valve. The first scenario saw pressure held at 51.7 bara (750 psia) at the beginning of the restart and gradually dropping to 27.6 bara (400 psia) after 9 hrs. The topside valve was also kept below 30% open for the first 9 hrs. Slugging was very low with this configuration. The flow rate spike was as low as 3,975 Sm3/d (25,000 STB/D) for the initial slug. The second PID scenario kept pressure on the upstream topside valve at 68.9 bara (1,000 psia). This almost eliminated the slugging problem throughout restart. The flow rate fluctuation also was negligible.

Subsea pressure downstream of the wellhead choke was considerably higher than either with the manual control or no control. If the system allows, a higher pressure control upstream the topside valve would help minimize slugging during restart.

Reference

1 Benchmark and optimization study of a deep water well restart in Gulf of Mexico. H. Shi, BP America Production Co., USA, H. Dong, Schlumberger Technology Corporation, USA, R. Berger, Manatee Inc., USA. Paper given at the 16th International Conference on Multiphase Production Technology, June 12-14, 2013, Cannes, France.

The author

Hongkun Dong is Flow Assurance Business Development Manager for Europe and Africa at Schlumberger. He has over eight years' oil and gas industry experience, with a key focus on flow assurance and multi-phase flow. His responsibilities include project execution, and technical and business development.

More in Production