Deepwater reservoirs push frac design and equipment capabilities

Bob Hanna and Dick Ellis with BP provided the next impetus. They pioneered the frac-pack application within BP, which resulted in its application in wells at Mississippi Canyon block 109, Amberjack platform. These frac-packs were considered extreme at the time, with rates of 12 to 15 bbl/min and proppant volumes of 15,000 to 30,000 lbm. As the frac-pack technique gained more acceptance, further tool testing and qualification was required as the limits on rate and proppant volume continued to be pushed. In order to meet surface pumping capabilities, companies are testing at higher rates and proppant volumes.

As a result of this testing and applying the lessons learned, new tool systems have evolved rapidly in the last decade. The industry’s innovations have included a new larger-exit-area crossover design, a solid-steel crossover design to replace the earlier welded designs, improved metallurgy and material coatings, changes in positioning of the crossover during pumping to protect the external production casing from the erosive forces, and improved exit-sleeve design. Additionally, this led to the design of weight-down tools for positive positioning during the frac-pack. The latter was required to counter the ballooning and cooling forces exerted on the workstring during pack placement.

Today, the weight-down system has been further refined with the use of multi-acting collets that maintain both a weight-down circulating and a squeeze position.

The industry’s pumping designs have a direct effect on tool ratings. On smaller tool sizes, these upper limits are fairly well defined. On the larger tool sizes, there still might be additional capability to increase rates and volumes. The existing concentric reciprocating tool designs dating from the 1970s are approaching their limits.

What is the upper rate and proppant limit, and is there a need to continue to go at higher rates and larger proppant volumes? These are questions still to be answered. If history tells us anything, it is likely that the operator engineers will continue to push the limit. The capability to tie two frac vessels together only enhances the chance of further pushing the limit.

Reservoir drivers

Reservoir conditions drive the requirement for higher rate frac-pack tools in prolific deepwater zones characterized by high deviations, high permeability, and long intervals. In deepwater, it is not uncommon to frac-pack zones with permeabilities in excess of 250,000 md ft.

The permeability of these intervals can be on the range of one to two darcies, interval lengths can range from 200 to 500 ft, and deviations greater than 50° are fairly common. Consequently, these zones often exhibit high fracture-fluid leak-off rates.

The frac-pack must cover the entire zone in order to maximize the production rate and the long-term reliability of the completion. Such completions usually require a high rate and large volumes of proppant placed across the interval.

The degree of heterogeneity across the producing zone is the second driver for high-rate frac-packing. In some cases, there may be two or more highly permeable zones that are separated by a shaly or silty zone of poorer quality. Since the shaly or silty sand may act as a barrier to fracture height growth, it is critical to be sure that the frac-pack covers the entire interval.

A third driver for frac-packing long intervals is economics. If it is possible for a company to complete two intervals that are close together with a single completion rather than a stacked completion, cost, time, and risk associated with completion can be minimized.

Some alternatives will be to use other completion types that require lower injection rates. This includes screen-only completions, openhole gravel packs or high-rate water packs in order to complete these long high-permeability intervals. However, these completions typically exhibit higher skins than frac-packs. Furthermore, historical evidence in the Gulf of Mexico indicates that frac packs are currently the most reliable and provide the lowest skins of all completion types. Other alternatives may be to reduce the effective perforated interval, perform a stacked completion, or use more viscous fluids that will have lower leak-off.

However, each of these approaches has disadvantages, such as lost reserves resulting from limiting the perforated interval, stacking completions increasing the cost of the completion (which is a particular concern in deepwater wells), and the use of more viscous fluids resulting in increased proppant pack damage and skin that could impact productivity. The industry’s use of higher-rate tools helps to perform frac-pack operations at lower overall costs, shorter turnaround time, and with increased productivity than might be obtained from conventional tools.

Challenge

Industry’s deepwater completions have constantly challenged placement design. Pumping rates have slowly been increased to handle the longer intervals or to maximize sand placement. The Shell URSA A7 well pushed the record rate to 41.6 bbl/min. This well required a special rig-up of twin 3-1/2-in. co-flex hoses to safely place the pack. The well, when brought on production, turned out to be the record producer for the GoM, producing at a rate of 50,150 b/d of oil. This well was superseded by a 42 bbl/min placement and followed by 45 bbl/min placement.

BP once again raised another challenge. Consulting engineer Mike Mullen stated that for the upcoming Aspen project, the anticipated frac-pack design would require a pump rate of 50 bbl/min with an anticipated proppant volume of 500,000 lbm. Fortunately, the planned casing size was 9-7/8 in., which would allow the use of 5.0-in.-bore packers and service tools.

The 50 bbl/min rate had been reached on a previous Shell Na Kika job in the GoM. The difference between the Na Kika job and the proposed test was the plan to stay at maximum rate, using the larger proppant volume throughout the test.

Rating frac service tools has been done through surface testing, actual job comparisons and modeling. With the variety of proppant types, pumping regimens, exit port designs, limiting flow areas, and the upper extension material and grade defining the ultimate tool pump and volume ratings, one can see by the number of variables that rating a service tool is not an easy task and one has to take into consideration the packer assembly into which it is to be installed. The 5-in service tool (5.0-in.-ID packer bore) had a conservative rating of 40 bbl/min with a maximum proppant capacity of 300,000 lbm carbolite with an internal protective sleeve, although the tool had been tested at a higher proppant volume. These test results were used as the baseline data for setting up the 50 bbl/min and 60 bbl/min tests.

Testing

A complete frac flow test had to be performed to verify that the 5.0-in. service tool could handle the desired pump rate and proppant volume requested by BP for the Aspen project. Prior to testing, it was decided to push the tool design by increasing rate and volume if initial test results were positive. An existing service tool frac-test fixture was set up in a flow loop fed by frac pumps (in parallel); manifold to provide the necessary flow volumes and rates within a closed-loop system.

The test was set up to represent a simulated frac job with the capability to perform tool inspections between pump stages. Since the test system was a closed loop, the sand and gel slurry was re-circulated through the tool to effectively simulate cumulative volumes. This has been the practiced method used in frac-flow testing, as it provides a cost-effective method limiting the amount of proppant required to simulate a 2 million-lbm flow test. In addition, it is more environmentally friendly since additional amounts of exhausted gel and proppant are not created.

Several design changes were made to meet the demand for higher flow rates and volumes. These changes combined new ideas and verified tool enhancements realized from frac testing in other tool sizes. Most of these changes revolved around minimizing wear on components while limiting the directional change of the slurry at high velocities.

Tool inspection

The company performed tool inspections based on set sand volumes pumped. At each tool inspection, the pumps and test fixture were flushed with clean fluid, and the test fixture was disassembled. Photos of each critical component were taken, and ultrasonic measurements of both the heavy-walled upper extension and the casing extension were also taken. Once the inspection was completed and the test fixture reassembled, the pumps were brought back on line and the concentration taken up to the desired stage. At this point, the pumps were bumped up from idle (15 to 18 bbl/min) to the desired pump rate. (Idle time through the tool system was not counted against total proppant volumes.)

The company pumped 1,250,000 lbm of proppant at 60 bbl/min with an average erosion for the upper extension of 21.95% of original wall thickness and the average erosion for the casing extension figuring to be only 2.61% of original wall thickness.

The first closing sleeve test, after pumping 2,003,700 lbm of proppant at 50 bbl/min, resulted in an average erosion of 35.72% for the upper extension of the original wall thickness. The average erosion for the casing extension was 36.03% of original wall thickness.

At the request of the BP America representative, a pressure test was performed on the closing sleeve to verify the capability to close and hold pressure. Their concern was the capability to prevent fluid loss through the sleeve after closing. The closing sleeve was shifted closed, and pressured up to 3,230 psi. At the end of 30 minutes, the pressure had bled off to 2,850 psi. Based on this test, it was determined that the 5.0-in. closing sleeve is capable of preventing fluid loss after pumping over 2,000,000 lbm of 20/40 proppant at 50 bbl/min.

Case history

The well’s interval to be frac-packed consisted of a zone with a measured length of 370 ft at a bottomhole deviation of 53°. The interval consisted of several clean high-permeability lobes that were separated by shaly to silty mudstone zones. The measured depth of the well was in excess of 19,000 ft, which added to the complexity of the design. Due to the high permeability and interval length of this zone, it was proposed that high injection rates would be needed to effectively cover the entire interval. A 30 lb/M-gal borate crosslinked fluid was used in order to optimize leak-off and conductivity. The minifrac was performed at 50 bbl/min.

The recorded trend of the three bottomhole gauges is of particular interest. The lowermost temperature gauge shows a very different trend as compared to the upper and middle gauges. The lowermost gauge essentially stays flat for approximately 5 to 7 min before showing a decline, whereas the other gauges show an immediate cool down. This could indicate that an effective frac across the lower and middle gauges had taken place but that the fracture had not grown to the position of the lowermost gauge. Finally, at 18:06 the lowermost gauge started to cool down, which also coincided with a decline in the bottomhole pressure. This would indicate that the fracture had broken across a barrier and was now starting to grow downwards. This trend was further confirmed by the accelerated temperature decline seen in the upper and middle gauges and the early data from the main frac-pack treatment.

Based on the analysis of the treatment, it was decided to pump the frac-pack at 60 bbl/min. As in the minifrac, the lowermost temperature gauge initially shows a flat profile while the other gauges show a cooldown from the fracture. However, after approximately 7 minutes of pumping at 60 bbl/min, the bottomhole temperature gauge showed a sharp decline, indicating that the frac-pack had now grown to the bottommost gauge position. The data indicated that the high injection rates were instrumental in ensuring that the frac-pack was capable of covering the entire interval and that rates as high as 60 bbl/min may be necessary in the future for some of these deepwater zones. This treatment resulted in a total of 287,000 lbm of proppant being pumped, of which 268,787 lbm were placed into the formation. The peak horsepower used was 16,292 hydraulic horsepower and achieved a rate of 60.1 bbl/min. The maximum treatment pressure was in excess of 12,300 psi, and the average treating pressure was 11,060 psi.

As frac-packing continues to push upper design limits on service tools, new equipment will have to be developed. Upper limits have been reached on some tool sizes, and new designs in proppant delivery system will be required if pump rates and proppant capacity continue to increase.•