Subsea hydraulic jet pump optimizes well development offshore Tunisia

The operator considered the use of seabed booster pumps, but after further study, Lundin chose hydraulic jet pumps for its lift needs on the Oudna well.

Successful application and optimization of artificial lift systems, for marginal offshore fields, require considerable evaluation prior to implementation. If the wrong lift choice is made at the outset, an economically favorable field can quickly become unprofitable. This is due largely to factors, such as the high cost of intervention, once the wells are on production, or the need for water injection to maintain reservoir pressure.

Such considerations factored heavily into the development planning of Lundin Petroleum’s Oudna prospect, offshore Tunisia, in 2005. Initial field economics at Oudna were based on an oil price of $27/bbl and trouble-free production for at least four years, just to break even. Therefore, the field was likely to be marginal under any production scenario. Reservoir modeling indicated that both reservoir pressure support and an artificial-lift production mechanism would be required to keep oil output at economically viable levels. However, because the field is in 886 ft of water and roughly 50 mi from shore, production handling and artificial-lift options would be limited.

The production handling challenge was resolved fairly easily by deploying an FPSO that was on long-term lease to Lundin Tunisia. However, the reservoir management problem would prove more complex and require a unique approach to make the field economics viable.

GAS LIFT NOT VIABLE

Gas lift is often the preferred approach for remote subsea wells, because the required equipment contains few moving parts, and can be varied quickly and easily to suit changing reservoir conditions. However, this approach also requires a readily available source of gas, which would prove difficult for the Oudna reservoir. Oudna did not contain a viable gas cap or appreciable levels of dissolved gas in the produced oil.

Lundin considered several options to bring in alternative sources of gas, including running a pipeline from the nearest gas field (12 mi away), generating nitrogen gas on the FPSO, or using flue gas from burners on the FPSO. Each option was reviewed carefully, but ultimately rejected because of cost and processing concerns.

CONSIDERING OTHER LIFT OPTIONS

Having eliminated gas lift as a possibility for Oudna, the operator considered positive lift mechanisms, including electric submersible and seabed booster pumps. Electric submersible pumps (ESPs) for subsea completions have gained more widespread acceptance over the past decade. Operators off the coast of Australia and in the UK sector of the North Sea have adopted this lift approach.

However, expecting more than two years of continuous run life from an ESP was considered too risky and unrealistic, because the subsea application of ESPs was still a relatively new development with a short track record. Pumps fail from a combination of constant use and the breakdown of electric insulation in the relatively hostile environment in which they operate. The use of a second pump in tandem may improve the longevity of the system, but the second pump is exposed to the same chemical, vibrational and thermal influences as the working pump, even if it does not run for a prolonged period.

An intervention would likely be required after two years of ESP service. With a regional spread cost of roughly $250,000 per day, a 10-day intervention would cost between $5 million and $10 million. This intervention cost was unacceptable, given the already marginal field economics, and it effectively eliminated ESPs as an option for the field.

The operator considered the use of seabed booster pumps, which offer a reliable means of boosting production for reservoirs with sufficient drive to lift commercial amounts of hydrocarbons as far as the seabed. However, concerns about the Oudna reservoir’s drive, coupled with a relatively high upfront capital equipment outlay, prompted the operator to keep searching for alternatives.

FIRST SUBSEA JET PUMP

After further study, the operator chose hydraulic jet pumps for its lift needs on the Oudna well. Jet pumps have a long track record of providing simple, trouble-free artificial lift in onshore applications. They work by using the fluid—typically oil or water produced from the reservoir—that is pumped down the tubing or down the annulus, between the tubing and casing via a surface positive-displacement pump. The fluid then travels through the downhole jet pump, which is equipped with a nozzle-throat combination. The fluid travels through this nozzle, creating a low-pressure (Venturi) effect at the discharge of the nozzle. This low pressure allows the reservoir to push the produced fluid into the jet pump. The produced fluid and power fluid then enter the throat and are mixed together. The mixed flow then goes to a diffuser, where sufficient static pressure is recovered to lift the combined fluids to the surface, Fig. 1.

Fig. 1. Reverse circulation, high-volume jet pump.

Jet pumps have no moving parts and, therefore, no mechanical wear. As a result, they have proven to be a reliable, low-maintenance system that affords flexibility in lift volumes and can operate for extended periods of time without the need for intervention. Jet pumps can use a common system for both the power fluid and the field water injection fluid, when one is present, thus minimizing space requirements and surface facility complexity.

Although hydraulic jet pumps had been deployed reliably in onshore wells since they were introduced to the oil field in 1972, the Oudna well would be the first subsea installation (the wellhead was located on the seafloor). The operator selected Weatherford to develop, build and optimize the subsea jet pump for this application, but with some special considerations unique to the field. For example, the jet pump would have to handle oil and/or water production rates of up to 25,000 bpd in 7-in. production tubing (actually casing). A 5½-in. jet pump had proven capable of producing at this rate in land applications, but the larger tubing size required developing a 7-in. jet pump. This presented a challenge, because only one such jet pump had ever been built—nearly 30 years prior.

SIZING AND DESIGN CONSIDERATIONS

A jet pump must have sufficient capacity to sustain the rate of production that the well is capable of delivering. At the same time, the required surface horsepower must be kept to a reasonable level. The first part of the sizing process is matching the jet pump performance curves with the productivity index (PI) and the inflow performance relationship (IPR) of the well. The rest of the process is staying within the operating limitations for a particular installation. The most common limitations on an offshore installation are power-fluid injection pressure, rate and space limitations. The backpressure, or discharge pressure, imposed on any form of artificial lift must be as low as possible.

Fig. 2. JEMS graphical output for the base case.

The Weatherford design also had to account for, and avoid, the possibility of cavitation, which occurs when the local static pressure at any point is equal to or less than the vapor pressure of the liquid being produced. Production cavitation is encountered when too much produced fluid is forced through the area available for it in the throat (that is, throat area minus nozzle area). Per fluid dynamics, the higher the volume for a given flow area, the higher the velocity and the lower the local static pressure. Power fluid cavitation can occur whenever there is too little production. The shearing action between the power fluid and the produced fluid will generate spherical voids. These voids result in cavitation damage, either in the constant diameter section of the throat or in the diffuser. The presence of gas or oil can mitigate cavitation damage, which is greatest when there is a high percentage of water with little or no gas present.

Fig. 3. JEMS graphical output for reduced-flowing bottomhole.

Weatherford engineers used the company’s proprietary jet pump evaluation and modeling software (JEMS) to predict the performance of the jet pump as a function of its size. The program also allowed engineers to simulate anticipated downhole conditions and performance ranges for different scenarios, and to determine the precise nozzle and throat sizes to provide the most efficient jet-lift system.

Figure 2 is the graphical output for the base case from that program, which shows the PIs of the well at perforation depth and pump depth lines of constant injection pressure; and the two cavitation zones. The specified production rate is 20,000 bopd (depicted by an asterisk on the pump depth PI line). The analyses indicated that the optimum size for the Oudna well, for the specified conditions, was a 7-in. Weatherford reverse-circulation jet pump (Kobe model) with a 17D combination (where 17 is the number of the nozzle and “D” specifies the area ratio), and capable of producing 25,000 bopd. Additional analyses determined how that combination would perform, if the PI was lower than expected, Fig. 3.

DEPLOYMENT

Though a jet pump can be installed using just the power fluid, the most common installation method offshore is via wireline. For the Oudna well, an isolation sleeve was installed in the bottomhole assembly (BHA) while it was on the surface. The purpose of the isolation sleeve was to isolate the casing annulus from the tubing, so that the hydraulic packer could be set using pressure in the tubing. The BHA, with the isolation sleeve in place, was then run into the well as part of the tubing string. After the packer was set, the isolation sleeve was wireline-retrieved, and the jet pump was installed via wireline.

The next step was to determine the most efficient power-fluid delivery system for the well. On land, it is a fairly simple process to access the wellhead and deliver the power fluid required to drive the pump. For subsea wells that are not connected to a fixed installation, problems arise. The large jet pump on Oudna required up to 20,000 bpd of power fluid delivered at up to 5,000 psi.
To meet those requirements, a surface power-fluid pump and prime mover would need to deliver approximately 2,000 hp.

A power-fluid delivery system for subsea well applications requires a flexible conduit, between the power-fluid supply and the jet pump, capable of handling the pressures and volumes involved, and a mechanism for transmitting the power fluid through the subsea wellhead and into the annulus. For the FPSO, the flexible transmission conduit was simply a hose—a fairly standard oilfield product that could be specified easily to meet pump requirements and get fluids to the subsea wellhead.

For the wellhead, the operator chose a dual-bore subsea Christmas tree, with one bore for production and the other for injection of the power fluid. Such trees were not readily available, so a bespoke tree approach was selected. After a tendering and review process, Dril-Quip Aberdeen was selected to design and build the Christmas tree in a standard 18¾-in. bore and 15,000-psi rating, together with the fluid—bypass-facilitated hangers.

The operator’s requirement to use 7-in. production tubing, a 7-in. jet pump, and annular power-fluid delivery of an additional 20,000 bpd at 5,000 psi, required a considerable review process for the production casing. The typical use of 95/8-in. casing would not work in this instance, because the OD of the jet pump was larger than the ID of the casing. And even if the 7-in. pump would fit, the annular clearance between the 7-in. tubing and 95/8-in. casing would create too much friction loss at the planned power-fluid pump rates. Ultimately, the decision was made to use 113/4-in. Q125 65 ppf VAM (because this pipe was available in stock and met all the well needs) and 7-in. tubing, to minimize friction losses and allow the surface injection pressure to stay under 5,000 psi.

Downhole gauges were a desirable feature for reservoir management. Since there was a reluctance to run umbilical lines in the production annulus, the operator decided to use Metrol acoustic gauges, which work by transmitting data up the wall of the production tubing. This solution required clamping repeater stations to the tubing every 300 m (984 ft), which created minimal annular restrictions and a low pressure drop in the annulus. A transmitter mounted on the subsea Christmas tree allowed the well operators on the FPSO to collect downhole data in real time.

The final step before placing the jet pump in operation was to fully open the bypass in the power fluid supply line, at the surface, and start the power-fluid pump. The bypass was then closed partially until an injection pressure of approximately 1,500 psi to 2,000 psi was obtained, then maintained for at least 30 min. The bypass was then closed slowly until the operating pressure increased by approximately 500 psi, a setting which was also held for at least 30 min. These steps were repeated until the bypass was either completely closed, or until the design operating pressure was obtained.

Returns from the Oudna well were monitored and compared against the calculated returns from the JEMS-generated charts. Adjustments are made as necessary, which could include refining the inputs for the analysis or changing the nozzle and/or throat.

The successful installation and operation of the industry’s first subsea jet pump has allowed the operator to produce the well above its expected rate. Production, to date, has provided income of several hundred million dollars, while intervention costs have been minimal. Only routine daily operational costs have been incurred.

The novel approach to developing Oudna field called for the design and manufacture of considerable bespoke equipment. In the future, efforts should be made to use more standard equipment. The industry will now likely consider jet pumps more seriously for this type of application in the future.

The system has operated trouble-free for several years without the need for intervention. At this stage, its long-term benefits are indisputable.  

ACKNOWLEDGMENTS

The authors thank the Lundin Group, Weatherford, Dril-Quip, Petrofac and Norwell for their support in developing the technologies and sharing the data presented in this article.

REFERENCES

1. Weatherford Jet Pump Operating Guidelines.
2. Norwell Drilling Project Management System, ISO 9001-2000.
3. Anderson, J., R. Freeman and T. Pugh, “Hydraulic jet pumps prove ideally suited for remote Canadian oil field,” SPE 94263.
4. Fraser, K., Managing Drilling Operations, Elseviers, 1991.
5. Cunningham, R. G., “Jet pump as a lubrication oil scavenge pump for aircraft engines,” Wright Air Development Center Technical Report 55-143, 1954.
6. Muller, N. H. G., “Water jet pump,” Journal of the Hydraulics Division, Proceedings of the American Society of Civil Engineers, 1964.
7. Pugh, T., Overview of Hydraulic Pumping, 2013.