Russia Report: Sakhalin Island - Drill cuttings re-injection in challenging environments

Since there are only about six months of the year when the area is ice-free, other drilling waste management and onshore treatment options would limit the drilling operational window. Additionally, ship-to-shore options for cuttings transport are expensive and not considered a realistic option because of the environmental and logistical constraints imposed by the remote location of the field. However, drilling waste management using CRI would allow year-round drilling operations.

Although CRI is often the most effective approach for cuttings disposal in remote and environmentally sensitive areas, there are associated risks and uncertainties. Accidental releases of injected slurries, as well as plugging and loss of injection wells, have occurred in some CRI operations. These problems can still occur if injection operations are not engineered correctly or performed properly.

Some of the important risks and uncertainties specific to this CRI project offshore Sakhalin Island include lack of historical operational data and experience and the potential for solids dropout and consequent injector plugging.

Even though the oil industry has extensive CRI experience in such areas as the North Sea, Alaska’s North Slope, and the GoM, CRI operations in remote areas such as Sakhalin Island are still new.

Although a previous annulus injection well had become plugged, a new well was drilled for cuttings injection. This well was to be used in the future as a development well, and was highly deviated and of large diameter. As a result, there was a high risk of cuttings settling on the low side of the injection tubing and subsequent plugging of the well. In addition, the surface facilities were not capable of grinding the cuttings small enough at required rate or of pumping the slurry at high enough rates to generate turbulent flow.

The risk and uncertainty management measures were undertaken to avoid plugging and loss of the injection well. The CRI operation contract was awarded while the dedicated injection well was being drilled, and the surface facilities were already installed and commissioned. This mandated designing suitable slurries and appropriate operational procedures to match the facilities, as well as ensuring that the CRI operators understood the importance of strictly adhering to specified quality control and parameter requirements.

It was essential to establish successful CRI operations as soon as possible. This well provided the only option for drilling waste management, and loss of the well would delay the drilling program for at least a year. Given SEIC’s commitment to cuttings injection and to cost control of operations, risk management was a crucial factor for success.

For these reasons, the implemented risk management and assurance plan was initially extremely conservative so that CRI could be reinitiated and drilling operations could be resumed as soon as possible, all without plugging the well. Subsequently, it would be feasible to optimize the operations with monitoring and analysis of acquired injection data.

The monitoring and assurance optimizations focused on slurry design and optimization, pump- ing procedure design and optimization, solid trans- port modeling, and assurance of appropriate shut-in intervals between batches.

The well trajectory and completion are among the reasons why the risk of cuttings dropout and plugging is high. There were two cuttings injection options - the primary injection option through tubing into the perforations, and the backup option down the annulus into the perforations at 1,756-1,766 m MD. The injection tubing is 5 1/2-in. tubing from surface to 1,756 m MD and 4 1/2-in. tubing from 1,756 m MD to the injection zone.

The injection tubing volume is approximately 150 bbl. It takes three batches to displace the entire tubing because the slurrification tank capacity allows no more than 100 bbl of slurry per batch. Consequently, an individual batch can remain in the string for a substantial period of time.

The long residence time and the deviated nature of the well may cause the cuttings-laden slurry to form dune beds along the low side of the injection tubing. These dunes may slide down and plug the well during the shut-in intervals between batches. Such sand movement is well known in gravel and fracpacking operations.

Slurry rheology

Because of the high risk of plugging and no fallback options, slurry rheology quality control requirements were higher than normal. Marsh funnel viscosity is almost always used in CRI operations for slurry quality control. There was concern that the commonly adopted slurry viscosity criterion of 60 to 90 sec/qt Marsh funnel viscosity might not be adequate for this well. Gel strength and low shear rate viscosity (LSRV) are more important in characterizing slurry’s solid suspension capacity. Since it was impossible to establish slurry rheology requirements before the first injection, an extremely conservative approach was initially adopted while slurry properties were measured from batches that became available for injection.

Based on the slurry property measurements and analyses, it was possible to confidently reduce the slurry Marsh funnel viscosity requirements to 120 sec/qt initially, and finally to 90 sec/qt. Additionally, the target slurry density was generally 1.3 S.G., although the slurry density could be as low as 1.1 S.G. during drilling of sections when the cuttings generation rates were low.

Cuttings suspension capacity

One critical requirement was to determine how long slurry could be left in the injection tubing without solids dropout and plugging. Considering the large tubing volume (~150 bbl), displacing the injection tubing after every cuttings-laden slurry injection cycle was not an economical or operationally preferable option. Thus, the slurry needed to have adequate cuttings suspension capacity between batches. Otherwise, the cuttings-laden slurry would need to be displaced with a solids-free fluid to avoid plugging of the injection well if the suspension between injections or residence time was too long. For this reason, both gel strength and LSRV were measured and numerical modeling was performed to predict cuttings settling velocity and the maximum allowable residence time.

Loss of injectivity could result from solids settling and plugging along the well or fracture plugging in the target formation. Solid particles in the slurry must be small enough to adequately reduce the solids dropout rate along the well and to avoid screenout at the perforations. Fracture analysis indicated that the fracture width at the perforations could be as small as ~0.1 in. (~2.5 mm) because of near wellbore tortuosity.

Injection rate

Due to the limitations on pump capacity, it was not possible to achieve turbulent flow down the 5 1⁄2-in. tubing during slurry injection. Nevertheless, it was desirable to inject at as high a rate as was practically possible to ensure cuttings transport and to keep induced fractures from closing. At the start of the pumping operation the pumping rate was 6 bbl/min, the highest rate possible. However, there was a practical drive to reduce this pumping rate because pump failure would otherwise likely occur quickly.

This risk was identified early, thus leading to the establishment of process and data requirements for assessing the impact of reducing the injection rate even before the CRI operations began.

It was decided that injection pressure monitoring and diagnostic evaluations would be used to decide whether the pumping rate could be reduced and by how much.

Injection pressure monitoring and evaluations at 4 bbl/min showed that the net pressure slope was greatly increased on a Nolte-Smith plot and that fracture efficiency was substantially reduced. The pressure data at 4 bbl/min also showed a rapid trend of injection pressure increase from batch to batch. Based on these observations, the recommendation was made to not reduce the pump rate any further.

In an ideal world, it is desirable to displace the entire wellbore or tubing volume with each batch injection. However, in reality, the wellbore volume is approximately 150 bbl, and the working capacity of the slurry holding tank is no more than 100 bbl. Since it was impossible to displace the entire tubing volume during one batch injection, it was determined that the optimal batch volume was 80 bbl.

This solution would provide the best balance between cuttings generation and cuttings injection rates and would largely minimize the residence time. It also meant that the tubing (~150 bbl) would be completely displaced after three batches with some allowance for inefficient displacement. In this three-batch cycle, the second batch moves the first batch from the top half to the bottom half of the tubing. The third batch moves then the entire first batch into the formation.

The shut-in time or the slurry residence time must be limited and controlled to avoid solids dropout. Otherwise, the solids-laden slurry in the tubing would need to be displaced with a solids-free fluid. Initial modeling, using Fann 35 data and settling tests, showed that the critical residence time was at least eight hours. Later, when LSRV data became available, it was determined that the residence time could be extended to at least 12 hr.

A solids- transport numerical model was developed during the course of the project. The simulator became available after injection of cuttings had been completed from drilling of the first well. The simulator was used to quality assure the slurry rheology and operational procedures, particularly assurance of solids suspension during shut-ins between batches.

As with similar numerical simulators, the injection string/well is divided into small segments, and fundamental physical relationships are used to numerically determine the local solids concentrations, cuttings settling, bed formation, bed sliding and erosion, as well as solids accumulation at the bottom of the well and/or solids ingress into fractures.

Avoiding plugging

It was assumed that 80 bbl was pumped in each batch at 4 bbl/min and that the shut-in time between batches was four hours. The numerical results clearly demonstrated that injection at 4 bbl/min will erode away the solids bed formed in the 4 1/2-in. tubing, and that after four hours of shut-in between batches the upper solids bed (over the 5 1/2-in. tubing) will slide down from 1756 m to 1935 m along the tubing. Since this is still approximately 125 m above the top of the perforations, shut-in interval of four hours between batches is adequate for avoiding perforation plugging from solids dropout.

The monitoring and optimization measures implemented were effective in managing the risks and uncertainties associated with this critical cuttings-injection well.•

Acknowledgements

The success of this CRI project was a team effort. The authors would like to thank the team members for their contributions, particularly, Dr. Reg Minton, Dr. Steve Marinello, Andy Bowman, Jim Dwyer, Iain Sanderson of M-I Swaco, and John Buckee, Denis Vasiljev of SEIC, to name a few. The authors also wish to thank the management of SEIC, the management of M-I Swaco and the Administration of the Sakhalin Oblast for permission to publish this paper as well as Jim Redden of M-I Swaco and Dr. John McLennan of ASRC Energy Services for reviewing this paper.

Editors Note: This is a summary of the SPE 93781 paper presented at the 2005 SPE/EPA/DOE Exploration and Production Environmental Conference held in Galveston, Texas, Mar. 7-9, 2005.