A majority of the electric submersible pumps (ESPs) used today are installed with jointed tubulars using workover rigs. The ESP is installed as part of the tubular string. The cable that provides power to the ESP is banded or clamped to the tubular. This method has been in use since the first ESP installations back in the 1930s.
Since the 1970s, several types of alternate deployment have been developed. Systems using the alternate technologies include cable suspended submersible pump systems1; cable internal, coiled tubing (CT) deployed submersible pump systems2; and umbilical systems3. Much was learned from the use of these alternate deployment systems; however, each of these systems had drawbacks that prevented widespread adoption of the technology. Major drawbacks of the current technologies include safety concerns, problems finding material able to endure corrosive environments, the uncertain reliability of downhole electrical wet connectors, and deployment and retrieval concerns.
Saudi Arabia’s experience with several of the systems led to a rethinking of ESP design and the development of a new ESP system that could operate in a high hydrogen sulfide (H2S) environment in the long term. A major challenge was designing the compact, H2S resistant cable. The material used had to be strong enough to support the weight of both the copper power cables and the ESP system, and also provide resistance to high H2S levels. The cable itself had to be a compact size to minimize pressure drop along the length of cable. Finally, the technology had to be simple enough to be reliably and economically manufactured in large quantities.
CABLE DESIGNS
The initial plans were to brainstorm potential cable design solutions. A review of several design possibilities identified some key technology gaps. As expected, the most critical aspect of the cable deployed ESP was the power/support cable. Finding a material for the cable able to withstand H2S concentrations up to 15% in the vapor phase presented a consid- erable challenge. (In addition, the material had to be strong enough to support the weight of the cable, but with a small enough diameter to minimize pressure drop across the length of the cable.) H2S permeates elastomers and thermoplastics, and even if it does not cause deterioration directly to the elastomer, any H2S reaching a copper conductor will eventually destroy the conductor. So, a reliable method for sealing the conductor from H2S exposure was a key requirement.
The effect of tension on the power/support cable was the next most critical concern. Since materials have different elastic characteristics, any stretching of the cable will have a major impact on the various other components in the ESP system. Some of the materials will stretch and recover completely, while others will not. Special attention therefore had to be paid to how the components and materials interact. Given that some type of traction method will be used to deploy the cable, it was also important to consider how the tension applied when pulling the system would mechanically affect each cable component.
A third key element of the design was the termination of the power/support cable. It would have to accommodate the electrical requirement of the applications, along with high temperatures and high H2S levels, and support all the hanging weight.
Taking the above issues into consideration, a feasibility study was devised that focused on three solutions:
1. Wire-wrapped copper conductor cable (conventional cable);
2. Non-copper conductor cable. 3. Metal-jacketed cable.
Based upon the design information and requirements, it was determined that the metal-jacketed cable had the most merit. It is impermeable to H2S. Its size and weight are within reasonable limits, making transportation less of an issue. It uses fewer different materials, so differential thermal and mechanical force considerations are simplified; and electrical and mechanical connections are more straightforward.
The manufacture of a metal-jacketed cable is a multi-step process. To fabricate the metal jacket, a rolled flat coil is slit into strips. Each strip is about 300 ft in length. To achieve the production lengths required, they are welded together using a specialized in-line laser that produces welds in a biased pattern — a bias weld provides greater overall strength than a butt weld. During the process of welding, complete full production lengths with the cable inside, the smaller coils are welded together in an in-line welding process, as needed. After the bias welding, the weld is ground and polished to the exact same thickness as the strip. The joined strips are wound onto a large diameter spool, once they are long enough to produce a full production length in a single longitudinal pass. To complete the manufacture of the metal-jacketed cable, the flat strip and a completed ESP cable are run through a forming line to shape the strip around the cable and then to continuously weld the longitudinal seam using a specialized laser. After this, the completed tube is drawn down (or cold drawn) onto the cable to finish the process.
One factor was especially critical to the manufacture of the cable. Unlike regular CT, the tubing seam weld could not be annealed after welding and cold working because the high heat required to anneal the weld would damage the cable. Little information was available on the use of unannealed materials in a welded tube and in a high H2S well environment. NACE standards set a material’s maximum hardness level based upon testing and historical information, and welded and unannealed materials typically exceed the NACE maximum allowable hardness level.
WELL CONDITIONS
ESPs are used in a number of wells in Saudi Arabia. ESPs are installed in wells with varying fluid properties, from sweet to sour, depending on the field and reservoir. Chloride levels in the formation water in many wells are extremely high. H2S levels can vary from 0% to 15%. Carbon dioxide (CO2) levels vary from 0% to 10%. Water cuts vary from 10% to 50%. Bottom-hole temperatures range up to 220 °F. Separate metallurgy solutions offer economical solutions for the varying H2S levels in different fields. It was decided to identify three levels of H2S for metallurgy testing. Samples would be tested in a H2S solution of 1%, 5% and 15%.
TESTING
Information was sought before testing the new unannealed laser welded alloys in these low to severe sour well conditions. Literature searches and a review of industry standard specifications (NACE, API, etc.) revealed, however, very little testing history or useful guidance related to this new material.
Ward et al. (2014)4 reports on qualifying a CT velocity string in sour brine service. Their article discusses some of the testing challenges faced when using nonstandard material geometry as well as providing some insight into methods of sample construction and testing in a fit-for-service test.
A collaborative team worked together to develop a list of potential materials and to design a qualification test that would simulate well conditions related to actual field conditions (a fit-for-service test). Three materials were selected as likely to have the best success in the 1%, 5% and 15% H2S test conditions. UNS N06625, a nickel alloy, was considered for the conditions of high H2S, chloride levels, and temperatures; however, UNS N06625 was too costly. So, UNS N08825, another nickel alloy, was chosen as a trial option. Similar to UNS N06625, N08825 provides exceptional resistance to many corrosive environments. It has a lower overall tensile strength, which could limit the total length of the cable. Lower cost materials for use in a less corrosive well environment were researched and selected. A high performance austenitic grade stainless steel, S34565, and a duplex stainless steel, S32205, were chosen for testing. The cost of S32205 is less than half the cost of N08825. S34565 has a higher cost, roughly 75% of the N08825 cost, but it was expected to have improved corrosion resistance and provide a mid-tier resistance to H2S compared to S32205.
The qualification process was broken down into environmental or chemical testing, electrical testing, and mechanical testing. These included tests of the cable’s H2S corrosion resistance, load, bend/fatigue, electrical integrity, thermal growth, and shearability.
Testing involved subjecting a high nickel alloy, a super austenitic stainless steel and a duplex stainless steel — all unannealed laser welded alloys — to conditions at different percentages of H2S, at 5% CO2 and at 300°F. This testing was developed to follow procedures similar to NACE testing. The main objective was to compare the super austenitic and duplex stainless steel alloys to the high nickel alloy. Electrical, cable thermal expansion and fatigue testing was completed after the H2S testing.
石油圈
