Development of a Stranded Tight Gas Field in the North Sea With Hydraulic Fracturing
There are more than 100 accumulations in the southern North Sea that are flagged as stranded fields. One of these stranded tight gas fields, the Kew field, has been developed successfully with the use of a subsea well, horizontal drilling, and hydraulic fracturing.
There are more than 100 accumulations in the southern North Sea that are flagged as stranded fields. One of these stranded tight gas fields, the Kew field, has been developed successfully with the use of a subsea well, horizontal drilling, and hydraulic fracturing. Because this was a subsea development well, all the hydraulic-fracturing operations had to be performed with the rig in place. The utmost efficiency of the operations was paramount; otherwise, the economics of the project would be affected negatively.
The Kew field is a gas field, with small volumes of associated condensate, located in Blocks 49/4a, 49/5a, 49/5b, and 49/4c of the UK continental shelf. It lies 120 km east of the English coast and 5 km west of the UK/Netherlands median line. The field location is pictured in Fig. 1 above. (For further geological and geophysical details of the Kew field, please see the complete paper.)
The planned Kew 49/04c-7Y subhorizontal development well was drilled along the crest of the Kew structure and is a sidetrack of the existing (suspended) Kew appraisal well 49/4c-7z.
The planned sidetrack 49/04c-7Y was initially intended to target the Lower Carboniferous units. To maximize reservoir contact, the well was initially planned to be completed with four to five hydraulic fractures, with a minimum of one per target unit. Because of the proximity to the gas/water contact, the decision was made to complete the well with a cased-and-cemented liner and plug-and-perforation technique for placement and isolation of the hydraulic fractures. Previous experiences with openhole uncemented multistage systems have positively affected the efficiency of hydraulic-fracturing execution in the North Sea. Also, previous experience of spalling and out-of-gauge hole was another driver toward a cemented system.
Difficulties encountered during the drilling phase caused total depth (TD) to be called early at 5225 m, meaning that a stimulation treatment could not be placed as planned in the E2 sand in the southern section at the other side of the main fault. This resulted in a change in fracture placement, with only one fracture placed in the E2 sand in the northern section.
The design of the well called for a subsea completion that could be hydraulically fractured and cleaned up and would allow for the installation of a downhole pressure/temperature gauge (DHPG) to provide an accurate understanding during the stimulation treatment. The tubing hanger could not be used while fracturing because the tubing-hanger/subsea-production-tree design did not allow a through port for the DHPG cable. For details of the proposed design, as well as specific components in the string, please see the complete paper.
The design basis for Kew was to engineer a maximum of 200-psi pressure drop during the injection phase. This pressure-drop number was the governing factor in determining charge design, for both the required entrance hole in the tunnel and the shots per foot.
The design basis used in the perforation design assumed a discharge coefficient of 0.75 (i.e., newly created perforations in the liner). The charge design was also validated by use of the selected contractor’s in-house perforating design tool, to ensure that the required entrance hole met the requirement to prevent proppant bridging and, ultimately, early near-wellbore screenout with the planned 16/30 resin-coated ceramic proppant.
The planned fracture placement, subject to the final logging-while-drilling data and dedicated repeat formation-tester points, was to initiate fractures along the wellbore in all of the Carboniferous sandstone units to provide maximum reservoir coverage. The original planned first fracture in the E2 sand had to be canceled, as previously mentioned, because TD of the well was called early, resulting in a reduced number of fractures being placed along the wellbore.
The initial well design contained five hydraulic fractures pumped at rates of up to 40 bbl/min. A fracture gradient of 0.7 psi/ft was forecasted, which indicated an expected maximum effective stress of just over 5,000 psi. Because Kew was to be developed as a subsea well, proppant flowback was seen as a significant risk for the overall project, and to mitigate this risk, resin-coated proppant (RCP) was used. Extensive laboratory testing was performed in the design and preparation phases to ensure compatibility between the RCP and the fracturing fluid. Correct functionality of the coating in the specific well conditions and with the selected fracturing fluid was tested. All of the tests were successful, with the proppant having little to no effect on the fracturing-fluid properties and the fluid not influencing the ability of the proppant grains to bond together under stress and remain unbonded when no stress was applied. The proppant sizing was also designed to reduce any non-Darcy effect.
The fracturing fluid used for the treatments was a freshwater-based borate-crosslinked guar with a polymer concentration of 35 lbm/1,000 gal.
To acquire more information on the reservoir and decide on the fracture-initiation points, a formation-pressure-acquisition-tool run was attempted to determine the pressure gradient and to confirm the permeability in the reservoir section; unfortunately, this was canceled because of tool communication failures. Low mobility and tight points were expected to confirm the low-permeability Carboniferous sands. This formed the basis for choosing fracture-initiation points in combination with the petrophysical-generated computer-processed interpretations. Overall, there was a wide variation of permeability in the tested intervals.
During the execution phase, only four of the five hydraulic fractures initially planned were placed because of the shorter length of the horizontal section. The four hydraulic fractures were placed successfully in the Kew reservoir using more than 1,000,000 lbm of 16/30 resin-coated intermediate-strength ceramic proppant and more than 800,000 gal of fluid. The isolation between consecutive stages was achieved by use of soluble-fiber-enhanced sand plugs set at the end of each individual treatment. For the execution of the fracturing treatments, a purpose-built stimulation vessel was used to provide the versatility and storage of the material volumes required for a successful stimulation campaign.
From start to finish, the stimulation operation for the four stages took 18 days, which translates into one hydraulic fracture pumped every 4.5 days. This is considered extremely efficient for multistage plug-and-perforation fracturing treatments in the southern North Sea, considering the interventions required between stages, vessel sailing times, and waiting on weather.
The first fracturing treatment from the Kew stimulation campaign targeted the E2 sands. The treatment was started with an extensive suite of diagnostic injection tests consisting of formation breakdown, step-rate tests (SRTs), step-down tests (SDTs), and calibration injection. The decline from the initial breakdown was used to perform post-closure analysis (PCA) with the mini-falloff software.
From the PCA, the reservoir pressure was estimated to be 5,598 psi (hydrostatic-pressure offset from gauge to perforations=647 psi). The transmissibility of the reservoir is 2,747 md-ft/cp, and taking the assumptions that the zone has a net height of 11 m and a reservoir-fluid viscosity of 0.023 cp, the permeability can be estimated at 1.75 md. From the logs, the permeability ranges from 0.1 to 1 md, which gives a good indication that the analysis provides an accurate estimate. The previously described investigation was over a radius of 7.9 m from the wellbore.
The reservoir-pressure estimate of 5,598 psi matches closely with the Horner-plot interpretation, from which the reservoir pressure was estimated to be 5,721 psi. The lower bound of closure is 6,450 psi, less than the determined pressure from the G-function and SRT analysis, which is expected.
The SDT showed the total near-wellbore (NWB) friction to be approximately 2,300 psi. Tortuosity was the main contributor, with 67% of the total NWB friction. The remainder was made up of perforation friction, at approximately 759 psi.
Following the SDT, the decision was made to pump a 100-mesh-sand slug at 1 lbm added/gal (ppa); this was pumped before the calibration-injection test to remove some of the NWB pressure drop. Following the sand slug, the resulting instantaneous shut-in pressure showed a total friction decrease of 900 psi at 35 bbl/min.
Following the diagnostic injections, a redesigned pump schedule was generated with a planned total proppant of 301,061 lbm pumped at 35 bbl/min and with a maximum concentration of 8 ppa. Because of bottomhole-pressure increases in the 6-ppa stage, the decision was taken to extend the 6-ppa stage and not continue on to the 8-ppa stage. Following this, 243,125 lbm of proppant was placed in Zone 1. The estimated conductivity of the executed fracture is presented in Fig. 2.
Zonal isolation was achieved by setting sand plugs between the subsequent fracturing stages. To ensure the success of the plugs, dissolvable fibers were added to the sand-plug slurry mix to enhance the suspension and transport of the sand plug in the horizontal section.
To evaluate the cleanout efficiency of the hydraulic fractures in each of the sand bodies and to provide an overall contribution (qualitatively) of each stimulated zone, chemical tracers were used for each of the four fracturing treatments. Two tracers were used for every stage, one in the pad and one in the slurry stage, which would help in identifying the flowback efficiency and breaker designs for each fracture, post-stimulation.
A rigorous sampling schedule was designed at an early stage of the planning phase, with a total of 29 samples taken during the cleanup phase of the well. The tracers themselves are chemicals found in nature and from the family of fluorobenzoic acids (i.e., nonradioactive).
The samples showed good breaking of the fracturing fluid, with tracers recovered from slurry and pad stages of all the fracturing stages. As was expected, a higher volume of tracer was recovered from the slurry stages than from the pad. This is in agreement with the volume and amounts of breaker added to the fluid.
Once all fracturing operations were complete, a rig-based cleanup and well test were conducted. Per platform requirements, there was a maximum limit to both fluids handling and proppant production before the well was considered clean and could be handed over to the operations group.
During the cleanout, the well was opened gradually, to limit the drag force and prevent proppant production. At the end of the cleanout, the choke size was 52/64-in. fixed at a flowing wellhead pressure of 3,370 psi. The final gas rate was approximately 45 MMscf/D, with a condensate rate of 420 B/D and 179 B/D of water/fracture fluid. The proppant-rate traces died off at the end, and the basic-sediment-and-water content was at 38%. The cumulative condensate was 2,470 bbl, with a cumulative water volume of 5,325 bbl. The total proppant that flowed back was 639.3 lbm. In addition, pressure/volume/temperature sampling was performed.
The minimal amounts of proppant produced during the test phase are notable. The rig-based cleanup also showed the deliverability of the well exceeding expectations and flowing on a constrained choke setting of approximately 45 MMscf/D. The tubing performance relationship using the cleanup data indicated that the Kew well was capable of prolific instantaneous gas rates, with a potential absolute open flow ranging from 100 to 123 MMscf/D.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 170852, “Development of a Stranded Tight Gas Field in the UK Southern North Sea With Hydraulic Fracturing Within a Subsea Horizontal Well: A Case Study,” by Marc Langford, SPE, Douglas Westera, SPE, and Brian Holland, SPE, Centrica Energy, and Bogdan Bocaneala, SPE, and Mark Norris, SPE, Schlumberger, prepared for the 2014 SPE Annual Technical Conference and Exhibition, Amsterdam, 27–29 October. The paper has not been peer reviewed.