CASING SHOE DEPTHS ACCURATELY AND QUICKLY SELECTED WITH COMPUTES ASSISTANCE

D. Mattiello, M. Piantanida, A. Schenato, L. Tomada
Agip SpA
Milan

A computer-aided support system for casing design and shoe depth selection improves the reliability of solutions, reduces total project time, and helps reduce costs.1-5

This system is part of ADIS (Advanced Drilling Information System), an integrated environment developed by three companies of the ENI group (Agip SpA, Enidata, and Saipem). The ADIS project focuses on the on site planning and control of drilling operations.

The first version of the computer-aided support for casing design (Cascade) was experimentally introduced by Agip SpA in July 1991. After several modifications, the system was introduced to field operations in December 1991 and is now used in Agip's district locations and headquarters. The results from the Validation process and practical uses indicated it has several pluses:

The reliability of the casing shoe depths proposed by the system helps reduce the project errors and improve the economic feasibility of the proposed solutions.
The system has helped spread the use of the best engineering practices concerning shoe depth selection and casing design.
The Cascade system finds numerous solutions rapidly, thereby reducing project time compared to previous methods of casing design.
The system finds or verifies solutions efficiently, allowing the engineer to analyze several alternatives simultaneously rather than to concentrate only on the analysis of a single solution.
The system is flexible by means of a user-friendly integration with the other software packages in the ADIS project.

Other advantages include the critical evaluation and formalization of company expertise and experience. The best practices can then be standardized and transferred by a computerized system to specific areas or field offices where the methods can be further tailored to problems unique to the area.

The system was developed using knowledge-based techniques (expert s),stems). These expert systems can explain each design choice by explicitly representing the shoe depth selection and casing design methodology. The knowledge-based approach has proven its worth in tackling casing design problems and other well site problems. 6 7

DESIGN METHODOLOGY

Fig. 1 is the Cascade system's logical structure showing the most important elements of the shoe depth selection and casing design methodology adopted.

The following information is needed for examining a particular well:

General information such as total depth, well type (vertical or directional), classification, location (onshore or offshore)
Well bore trajectory data
Forecasted lithostratigraphical series
Expected pressure gradients in the drilled formations
Information about completions and production tests
Drilling targets.

This information is referred to as "context." Cascade uses the information given in the context (made available within the ADIS environment) to position shoe depths and choose corresponding geometrical and mechanical casing characteristics.

The overall elaborated information about the shoe depths and casing design will be subsequently referred to as "project."

The Cascade knowledge bases contain a set of criteria representing the conditions and constraints for positioning shoes or choosing certain types of casings. The system develops the project in the following incremental manner:

Initially the project only contains the information regarding the shoe depths identified by the system.
Information about the drilling phases and casing diameters is then added.
The project is finally completed with the generation of information about the thickness and steel grade of each casing.

The approach to find the best solution consists of the subsequent application of generation and verify stages for the positioning of the shoe depths and for the casing choice. During the generation stage, the system produces a set of solution hypotheses. During the generation stage, the system analyzes the hypotheses generated during the previous stage and chooses the best one.

During the first stage (Generate shoe depths in Fig. 1), Cascade generates all the possible settings for shoe depths for the well being examined in accordance with the criteria in its knowledge base (insulation of overpressure levels, insulation of high filtration surface levels, protection of hydrocarbon targets, protection of levels presenting drilling problems). At the end of this stage, the project has many more possible shoe depths than technically applicable (Fig. 2). In any case, the best solution is certainly represented by a subset of these solutions.

During the second stage (Verify shoe depth in Fig. 1), the system verifies the shoe depth hypotheses generated and rejects the unnecessary casing strings. This process is carried out in accordance with a set of constraints (for example, the minimum allowable choke margin and the maximum allowable differential pressure) that can be specified by the user or generically tailored to a specific area. At the end of this stage the project has a number of optimal shoe depths for practical application (Fig. 3). The engineer, however, can change the shoe depths in the project or the constraints. Then the engineer can repeat the generation and verify stages to explore other scenarios until the results are satisfactory.

The third stage, casing design, follows the identification of the shoe depths. During this stage (Generate casing in Fig. 1), the engineer chooses the desired sequence of drilling phases and casing diameters, as well as the hypotheses used to calculate the burst, collapse, and tensile stresses. The system then calculates the stresses and generates a set of all the possible casinos capable of resisting the expected stresses. At the end of this stage, the system produces the shoe depths, the diameters needed for each drilling stage, and a set of casings that can be used for each drilling phase.

In the fourth stage (Verify casing in Fig. 1), the system analyzes the set of possible casings identified during the previous stage and selects the casing based on safety factors and on an economic evaluation. (The system can also propose tapered strings to optimize costs.) At the end of this stage the project is complete.

The engineer can change the projects, propose casings which are different from those chosen by the system, propose liners or any tieback string in place of the casing, and verify them in different load conditions.

If the system does not find a suitable solution that complies with the chosen shoe depth or if the solution does not satisfy the engineer, it is possible to go back to the shoe depth project stage and use different shoe depth hypotheses or create new ones for the casing project.

The characteristics regarding the casing stresses can be documented in a report or displayed graphically (Fig. 4).

VALIDATION CASES

The validation of the Cascade system included comparisons of actual well casing depths to solutions proposed by the system. Of the 40 wells considered for historical comparison, 24 were deeper than 5,000 m, and 31 wells had overpressured zones. With regard to shoe depth selection, it is important to note that any project for a well does not have one unique solution. Engineers may give different solutions for a particular well because personal experience and knowledge of similar cases play an important role in the project.

The solutions proposed by Cascade were compared to those actually used in the historical cases according to the following criteria:

The shoe depths actually used represented a subset of the shoe depths solutions proposed by Cascade at the end of the generation stage.
The shoe depths used in the actual wells were those chosen by Cascade at the end of the verification stage.
The shoe depths used in the actual wells were different than those proposed by the system, but the solution proposed by Cascade was deemed by the engineer to be just as valid.

SHOE DEPTHS

The following are the results from the validation process for the shoe depths for the 40 wells:

The criteria in the system's knowledge bases allowed precise identification and analysis of all the Conditions that determine the positioning of a casing shoe. In 33 wells (82%), the field solution was a subset of the solution proposed by the system at the end of the generation stage.
The system was able to identify with precision the shoe depths needed to insulate the overpressure levels. In Fig. 2, these shoes are positioned at depths of 3,700 m and 5,250 m.
The quality of the shoe depths project generally depends on the quality of the available input data. If the lithostratigraphical series to be crossed, the depths where drilling problems are possible, and the depths which represent the drilling targets are all described accurately, then the system is able to generate a high quality project. Fig. 2 shows how different depths representing drilling targets inside a single formation have been defined.
In 14 wells (35%), the shoe depth hypotheses verification stage was able to produce a project identical to the actual one. However, in another 17 cases (43%), the chosen shoe depths were equivalent to the field project's shoes (for example, if a shoe which delimits an overpressure zone is near a shoe which protects a hydrocarbon target, either of the two shoes could be rejected because the other can carry out both functions). Thus, the system's success rate rises here to 78%.
In four cases (10%), the system generated projects different from the actual wells, but the projects were still deemed valid by the engineer. These cases occurred in shallow wells without overpressure where there were no particular constraints on the positioning of the shoes.
In five cases (12%), the system was not able to generate automatically a valid project without assistance from the engineer. In these cases the constraints not considered by the system regarded the length of the open hole section and the time needed to drill the open hole. Subsequently, a set of approximate criteria was introduced to overtake this limitation. However, such input requires a good knowledge of the area.

CASING DESIGN

The comparison of the casings actually used and those proposed by the system is more objective and is based on reduced project time and reduced project costs.

Project time is reduced by an order of magnitude ranging from days per project before the use of Cascade to hours per project after its implementation. Project costs are lowered via the system's ability to propose tapered casings and the possibility of analyzing different scenarios. Of the 164 casing strings designed, 43 were optimized along these lines.

Generation of tapered profiles, depending on the availability of casing stocks, allows improvement in the choice of the casing grade and thickness, with consequent cost savings. Fig, 4 shows the diagram of stresses for a tapered casing string designed by the system.

The system also has the following advantages:

It guarantees the use of calculation and verification hypotheses that always conform to company standards. However, the user still has the freedom to choose some parameters used to calculate loads. For example, for an intermediate casing, the system automatically calculates the mud level drop in casing. The engineer, however, can specify a different mud level drop or ask the system to make the calculation considering a specific depth of the thief zone.
The system takes into consideration some casing design criteria that are difficult to verify (for example, stress analysis) and often neglected. These criteria do not limit the engineer's choices--he is free to accept or reject the system's suggestions.
While usual diameter sequences (for example, 7 in., 9 5/8 in., 13 3/8 in., 20 in.) are generally used in normal circumstances or when there are strict time limits, the system allows easy exploration of more unusual diameter sequences that could be more suitable from technical and economical perspectives (for example, slim-hole profiles).
The engineer can explore different alternatives by applying different stress hypotheses. In his opinion, the engineer can choose the best solution from the resulting design choices (Fig. 5).

HARDWARE AND SOFTWARE

The Cascade expert system is implemented on a RISC (reduced instruction set computing) work station with the UNIX operating system and the X Window/Motif environment. The system was designed for the currently available high-power work stations which allow the engineer to use all the tools available in the ADIS environment in an integrated manner. Thus, Cascade has access to data from the well bore trajectory, program, the formation gradient estimation program, and from well hydraulics programs.

Data are exchanged between programs through a common relational data base (accessible through the network) from which the programs access their input data and in which they write their output data. Access to the data base is limited by suitable consistency rules, so that only those output data which are guaranteed to be valid by the program producing them can be read as input data by another program.

The integration of programs is achieved both at the levels of exchanged data and user interface. User interfaces are all based on Motif and share the same tv pe of interaction.

RESULTS

Cascade was installed in Agip's district locations and headquarters during its validation stage, and it continues to be validated through practical use and through the study of increasingly complex historical cases. The speed with which the system can generate the projects, the intuitive nature of the graphical user interface, and the flexibility offered by the system in exploring alternative solutions have proven beneficial to the operating engineers.

Design methodologies for sizing casing strings have recently been improved by including additional criteria that take into consideration effects such as wear, torque, and drag.

Other fundamental aspects regarding safety factors and costs, linked to the correct choice of materials for the reservoir's environment, are under development.

In the future, Cascade will be integrated with a well completion expert system which is currently under development and with an expert system for material selection in hostile environments.

ACKNOWLEDGMENT

The authors wish to thank Agip SpA for permission to publish this article and numerous Agip personnel for assistance with this work.

REFERENCES

Mattiello, D., and Sansone, A., "CASCADE: a knowledge-based system supporting shoe depths selection and casing design," European Conference on Artificial Intelligence in Petroleum Exploration and Production, Paris, 1991.
Mattiello, D., and Sansone, A., "CASCADE: a knowledge-based drilling engineering software tool," SPE paper 24273, presented at the European Petroleum Computer Conference, Stavanger, Norway, 1992.
Rabia II, Fundamentals of Casing design, Graham & Trotman, London, 1987.
Schenato, A., Giacca, E)., Pirocchi, A., and Carli, R., "ADIS: Advanced Drilling Information System Project," SPE paper 22317, presented at the Sixth SPE Petroleum Computing Conference, Dallas, 1991.
Schenato, A., Bertoloni, L., and Brambilla, M., "ADIS Advanced Drilling Information System Project," presented at the Eleventh Petroleum Exploration and Production Conference, Cairo, 1992.
Heinze, L.R., "A knowledge base for designing casing strings," SPE paper 24420, SPE Computer Applications, January-February 1991.
Eckles, W., "Expert system for casing and tubing strings," Petroleum Engineer International, August 1991.