Casing wear modeling integral to well life cycle assessment

Making use of existing downhole assets

Owen Jenkins
Shell Expro

Predicted burst pressure (API) CWEAR5 Modeling (after Excel Processing) [25,316 bytes]
Predicted burst and collapse from Grant Jardine Model [21,853 bytes]
Cost comparisons for different weights of casing [54,117 bytes]
Typical costs for cutting and pulling casing

Twenty years ago, when North Sea development activity was at its peak, few would have expected that many of the wells then being drilled would see productive lives running into the 21st century. The drive was to drill and complete wells quickly, and get them producing. Little thought was given to optimizing reservoir management or what would happen to the wells during their lives.

There was no expectation that wells would be re-used. In many Shell Expro wells, three, four and more major sidetracks are common. An example well, illustrated in this paper, was sidetracked, for geological and production reasons, five times over fifteen years.

Each sidetrack was performed from within the same 9-5/8-in. production casing. Yet, with the strategic use of casing wear logs and computer wear prediction modeling, the fifth sidetrack was drilled and the well successfully completed and put on production. It continues to produce from an installation which, as recently as 1991, was expected to have ceased production during 1997.

In many modern field developments, drilling and completion costs are around 50% of total capital expenditure. As an investment and asset, therefore, the wells are as important as the facilities above ground. When wells cease to be economically productive, much of their asset value can be retained through re-use.

Well life cycle

Let us compare a typical exploration or appraisal well with a typical production well. The first important difference is that the anticipated life of the latter may be one or two orders of magnitude greater than the former. Secondly, much more activity takes place inside the casing in the production well. Thirdly, the long life means that long-term corrosion can occur in the latter.

At a first pass it can be seen that four times as much activity has taken place inside the 9-5/8-in. casing in the production well, yet its original design criteria were essentially the same as the exploration well, having been based primarily on anticipated burst and collapse pressure ratings.

The traditional design factors provided enough "meat" to allow for the originally unconsidered, later activities. It could also be argued that the original casing was considerably over-designed for its initial duty. The original designers, however, have provided us with a relatively low-cost well operation today.

Probabilistic casing design is now being considered by many operators and has gained acceptance among some. Where it is used, it is important that an assessment of the probable life cycle of the casing is also considered. In so doing, casing wear and corrosion should be considered. By modeling the likely wear, it should prove possible to refine the design factors used in the algorithms to produce a more cost-effective design, which still offers the possibility of re-use.

Also, the casing comes in standard sizes, strength grades, and wall thicknesses (usually expressed as weight/ft). Material selection becomes a series of quantum leaps. Thus, careful refinement of a design may or may not yield any saving, but it may demonstrate that a tighter control of tolerances may be necessary to ensure that the design desiderata are met.

In a recent Shell example, in order to meet the design requirements, a tolerance of -10% on wall thickness was stipulated (API Spec. 5CT allows -12.5%). All the casing was inspected and sorted into batches by wall thickness groups (100%+, 95 - 100%, 90 - 95%) prior to shipment offshore. The strongest (thickest) casing was selected for the most highly stressed sections of the well.

Casing integrity

Casing integrity is vitally important throughout the whole well life. The production casing provides a containment barrier should the production tubing fail. In gas lift wells, it is a primary barrier. The production casing is under greater pressure than the tubing, thus the intermediate casing performs the duty of a secondary barrier. The outer casings continue to provide structural support to the wellhead and the other casing strings.

Leaking casing can be patched. It can also be cut or milled and replaced, but such operations are expensive in time, services, and materials. Moreover, the additional non-availability of the well can have serious knock-on cost implications (lost production revenue, failure to meet gas nominations, etc.).

The material cost of the casing is but a small proportion of its installed cost. Rig time and rates are usually the critical cost factors. An accompanying table gives a view of the range of costs likely to be encountered in cutting, pulling and replacing a worn or damaged string.

Drilling capabilities

Since most of Shell Expro's platforms were installed, drilling capabilities have increased enormously. Rigs originally intended to drill no further than 16,000 ft are now, with relatively minor modifications, being used to drill 50-80% further, thanks to the development of extended reach drilling techniques. It is anticipated that drilling out to a reach of 10 km or more should become commonplace in the next few years.

BP at Wytch Farm in their Mll well have already gone beyond 10 km. It is this type of capability which has extended the life of many. It has opened up an era of drilling wells using comparatively cheap platform mounted rigs, which otherwise would have demanded expensive mobile drilling units and additional subsea facilities.

Multilateral wells are now increasing in popularity, allowing production from several locations through a single wellbore. By definition, these require drilling several holes from out of the same casing, thus greatly increasing the potential for wear occurring in the primary bore.

Wear modeling tools

Casing wear occurs as a result of the drillstring rubbing against the casing. To a lesser extent, damage can be caused by wireline activity. This is more normally seen inside completion tubing than casing, but it can be disastrous.

Wear depends upon the contact force, the wear track length (distance one surface moved across the other), the nature of the surfaces in contact, the material strength and hardness, and the presence of third bodies or lubricants between the wearing surfaces.

The main generator of wear track length is drillstring rotation. Consider drilling a 5,000 ft section of hole, for example: If an average 50 ft/hr were maintained and the drillstring rotated at 100 RPM, the rotational distance traveled would be over 1 million feet with 6-5/8-in. OD tool joints.

As part of the work of the DEA-42 project, Maurer Engineering developed a casing wear program. The current version of the program is CWEAR5 v. 5.01. This program, is available to members of the project and others who wish to purchase it commercially. It is a useful tool, capable of computing wear using one linear volumetric wear model and two different non-linear corrections.

The program is most useful as a predictive tool, although it has been used successfully for back-modeling and determination of wear factors. It contains an expert system that recomments what factors to use, if no better data is available, and contains databases of tubular goods dimensions.

Like much specialist software, it is more suited to the regular user who is used to its peculiarities. It assumes that the drillstring follows the Johancsik, Friesen and Dawson soft-string model. This calculates the highest contact loads as being at the points of highest dogleg severity (kinks in the well path) and handles a maximum of 600 depth/deviation survey points. Statoil reported upon their successful experiences using the program in planning and drilling ERD wells in the Statfjord and Gullfaks fields.

In Shell Expro and most other Shell Opcos, the Landmark program, WellPlan for Windows, is in general use for modeling the drilling of wells. This program, also uses the soft string model. Shell RTS in Rijswijk have developed a number of programs to work in conjunction with WellPlan. The casing wear program is WEAR2000. The current version, 2.06, is being introduced within the Shell group.

When planning a well, the first advantage of WEAR2000 is that it utilizes the same data files the engineer will already have created for other purposes (drillstring- casing scheme, mud and deviation survey). The deviation survey allows up to 2000 depth/deviation input points. It does not have the database of tubular goods on offer, although both user-generated and program-supplied catalogs of components are available.

The program does not advise on choice of wear model in the same way as CWEAR5, but it does have an updateable library of wear models from which to choose. Provided that the creator of the wear models gave enough information about the wear system in use, it provides a good guide as to whether the model being considered is likely to be relevant to the user.

The greatest strength of the program is its back-modeling capability, allowing the generation of probabilistic casing wear models of varying degrees of complexity and linearity. From the probabilistic models, the user may opt to settle for a simpler, deterministic approach and select wear factors which can be used in conjunction with CWEAR5 or the Grant Jardine tool.

Once again, a note of caution should be sounded. It is important for the user to understand what the program is providing, or the user could, for example, end up with a model which states that wear is constant, regard less of how much activity takes place. For this reason, at present, Shell is recommending restrict ing use of this aspect of the program to "Super-users".

The third tool is the Grant Jardine Excel-based tool. This is being developed by Shell Expro and Grant Jardine primarily as a decision making tool, aimed initially at deciding whether it is worth spending money on wear-prevention measures when drilling the well. Utilizing the same wear formulae, it can estimate wear deterministically, generating similar results to the other programs. It is limited in its data handling only by the maximum permitted size of an Excel spreadsheet and the speed of computer used. Although in principle simple, it too is best employed by an experienced user.

Case history 1

Toward the end of 1996, in order to increase oil production, a Shell business unit wished to increase the water injection pressure on a platform which was well into its second decade of production. In water injection wells, there is a high probability of premature tubing failure. This means that the production casing must be capable of withstanding full water injection pressure.

There were 12 candidate wells being considered for the increased water injection pressure. Prior to this increase, tests were to be undertaken on the 'A' annulus of the wells, to ensure that they would withstand the higher pressure. While these wells were known to be capable of withstanding the existing injection pressure, there was no wish to cause premature failure of the casing in wells that could not take the increased pressure.

An assessment was made of the likely maximum pressure the casing could take. This entailed going through the well files, sum marizing all the wear-causing activities that had occurred and then modeling the likely wear using CWEAR5. WEAR2000 was not then available.

Selecting the API burst and collapse algorithms, a notional burst pressure was thus determined for the worst wear "hot-spots" in each well. A view was then taken of whether these numbers were likely to be optimistic or pessimistic for each well. Factors which were considered included the time since the last pressure test, the history of tubing leaks in the well and any other known problems which might affect the casing integrity. The computed burst pressures were then factored appropriately to provide a recommended maximum test pressure. In some wells, the injection pressure could not be increased without further remedial activity.

Case history 2

In some instances, ERD wells have been drilled in re-used well slots. some kicking off from within existing 13-3/8 in. or 9-5/8 in. casing. This has meant that extensive rotating hours have been spent inside old, already worn (and possibly corroded) casing. In BA-27, four sidetracks had already been drilled when, in 1997, it was chosen as the slot from which a new, long reach well would be drilled into outlying structures.

The well had no less than six targets along its trajectory. Once the completion was pulled, the condition of the casing was assessed by running a Schlumberger USIT log and performing a pressure test. It was deemed to be satisfactory for drilling ahead, but, as a check, the additional casing wear was modeled using the Grant Jardine Excel-based tool. At that time, no casing wear model had been calibrated for Brent Alpha, so an intelligent guess had to made as to the likely range of wear factors to use.

After drilling the sidetrack, some difficulties were encountered when running the second liner, which resulted in tools hanging up and, in freeing them, some damage was done to the casing. A further USIT log was run and the burst capabilities of the casing calculated. Again, this showed that the casing should be capable of withstanding the required pressure test. This test was duly performed satisfactorily. The second USIT log now meant that we had an accurate record of the activities which had taken place since the previous wear log had been run and an accurate measure of the wear which had occurred. From this, it was possible to calibrate our wear model for Brent Alpha, using WEAR2000. The base wear factor was found to be very close to that used in the original wear estimate.

Some untoward wear "hot-spots" were found where aggressive tools had been rotated inside the casing to free them. This allowed calibration of some "what-if" factors, which we can use in future well modeling. Further confirmation of these factors was provided by back-modeling wear in two other Brent Alpha wells using WEAR2000. Close agreement was found be tween all three, which has greatly increased our confidence in our wear modeling techniques.

Conclusions

A number of conclusions can be drawn from extending the life of mature fields and determining the remaining life of well casing:

In considering the whole life of an offshore installation, casting the net widely is now essential to maximize its ultimate revenue-earning potential.
Large fields now can be reached from one location.
Several different accumulations can be reached by wells drilled from one location.
Well bores can be re-used extensively, typically saving half the cost of a new well and reducing the number of well slots needed for a phased development.
The use of a heavier weight of casing initially may prove cost-effective in whole-life assessment.
Multilateral wells offer further possibilities for reaching several objectives from a single well slot.
Casing wear modeling is an important tool in planning these wells and ensuring that the throughlife cost savings potential can be achieved.
Casing wear modeling can be used to optimize the selection of drilling tools and materials.

With steadily advancing drilling technology, we should at some point begin to predict the "end of life" for assets.

References

Pasley, P., Talling, A., Onyewuenyu, O., Cernocky, E., "The True Burst Capacity of Ductile Well Casing and Tubing as a Basis for Probabilistic Design," Shell E & P confidential document, August 1997.

Payne, M., Swanson, J., "Application of Probabilistic Reliability Methods to Tubular Design," Arco Oil & Gas, SPE 19556, December 1990.

Von Flatern, R., "Extending the Reach," Offshore Engineer, February 1998.

Hall, R., Garkasi, A., Deskins, G., Vozniak, J., "Recent Advances in Casing Wear Technology," Maurer Engineering, IADC/SPE 27532, February 1994.

Johancsik, D., Friesen, D., Dawson, R., "Torque and Drag in Directional Wells, JPT, June 1984.

Ostebo, E., Andreasson, E., Tjotta, H., Weltzin, T., "Casing Wear in Horizontal Wells - Field Case Histories," Statoil, 1996 IADC Well Control Conference of the Americas.

Acknowledgement

This is an abbreviated version of a technical paper presented at the Cradle to Grave: Whole Life Asset Management in Oil Conference, Marcliffe at Pitfodels Hotel, Aberdeen, 25-26 March 1998.