Casing Deformation in Ekofisk

Summary Casing deformation resulting from reservoir compaction occurred in the Ekofisk field operated by Phillips Petroleum Co. Norway and is a serious problem in three of the fields. This study established a relationship between reservoir compaction and casing failure by statistical analyses, finite-element modeling (FEM), and the analyses of deformed casing and logs run through collapsed casings. Ekofisk casing deformation is related primarily to the near-well incremental strain, well inclination, and casing diameter. Useful correlations to estimate future probabilities of casing deformation as a function of reservoir variables and well parameters were also obtained. We concluded that casing failure induced by reservoir compaction can be minimized through a pressure-maintenance program to reduce strain by drilling with the highest practical angle and by using the largest possible casing in the well. Introduction Casing failure at Ekofisk was first discovered in 1978. Subsidence and reservoir compaction as the result of pressure depletion were detected in late 1984. Subsidence and reservoir compaction resulted in two challenging problems: the jackup project carried out during Summer 1987 and casing deformation, which is discussed by Anvik and Gibson. Casing deformation was first noticed while routine workover and wireline operations were performed. Currently, about two-thirds of Ekofisk well casings reportedly have failed. Casing failures at Ekofisk were discovered in the producing formation and in the overburden. The overburden deformations were located in the lower 1,500 ft [457 m] of the overburden, with the highest concentration between 400 and 900 ft [122 and 274 m] above the reservoir. A casing failure in the formation has no noticeable effect on the well's productivity. It can go undetected until a wireline or workover operation is performed. The consequences of a casing deformation in the overburden, on the other hand, are more severe. An overburden deformation typically results in a casing leak, a tubing leak, or both. Either case would cause an unacceptable annulus pressure problem, resulting in the well being shut in and worked over. To understand why the Ekofisk casings were failing, two parallel approaches to the problem, a statistical analysis of field data and FEM of the reservoir in the vicinity of the wells, were undertaken. Failure Mechanism Casing failure can result from compaction of an overpressured, undercompacted reservoir (e.g., Houston/Galveston area), from earthquakes (e.g., Wilmington and Siberia), from salt shift (e.g., Cedar Creek anticline in Montana and North Dakota), and from permafrost (e.g, Prudhoe Bay). permafrost (e.g, Prudhoe Bay). It is often difficult to identify, the actual casing failure mechanism. Fig. 1 illustrates casing loads resulting from compaction of reservoir rock. The variety of mechanical failure modes that a well casing can undergo can be divided into two areas: structural instability (e.g., Euler-type axial buckling) and failure modes related to the inherent strength of the casing (e.g., collapse and axial yield). Wilson et al. showed that axial buckling in casing occurs only if radial restraining forces are not present. Most casing strings not cemented continuously throughout their length are probably buckled when subjected to large axial loads resulting from reservoir compaction. At Ekofisk, excessive matrix acidizing could result in lack of lateral support around the casing and consequently lead to buckling as the casing is loaded in compression. The failure modes related to the casing strength are collapse (radial stresses), tensile break (axial tension), and thread jump (axial compression/tension). The casing can fail in one or more of these modes. The factors that cause casing failure are interrelated. A factor that reduces the occurrence of one kind of failure can promote another kind of failure; e.g., casing in tension is usually considered more stable to buckling, but it is more susceptible to radial collapse. Furthermore, when the effects of wear, corrosion. and fatigue are added to the stresses on the casing, the potential for failure increases. Caliper logs were run in failed casing to help determine the failure mechanism. We were unsuccessful in running the caliper logs through deformed casing in the producing formation because deformations were too severe to permit the logging tool to pass. We were, however, able to pass caliper logs through the overburden. The results from Well 2/4 B-10 indicated that deformation, or ovaling, occurred in the overburden, above the reservoir. The overburden section of the casing string close to the top of the reservoir is under tension as a result of the reservoir compaction. The increased tension will considerably reduce the collapse strength of the casing. Insight into the reservoir compaction and casing deformations is provided by logs run in Well 2/4 C-11, a vertical nonproducing well drilled in 1986 to monitor subsidence and reservoir compaction. To measure the subsidence and reservoir compaction, radioactive markers were placed at various depths in the well. Highly accurate casing-collar logs were used to measure changes in casing length. Logs were used to detect movement of the markers relative to each other. Monitoring of casing deformation consists of periodic logging with cement-evaluation and acoustic-caliper logs. By comparing logs run at different times for changes in the compressive strength of the cement or changes in diameter or length of the casing, we could estimate stresses in the casing. The base log was run in Oct. 1986 and a periodic log was run in March 1987. The comparison of the two logs showed that during that period the reservoir and casing were compacted simultaneously. In the same areas where the casing was shortened, the ID was reduced. A 1-ft [0.3048-m] compaction of the casing was observed in this well over the interval from 9,610 to 10,554 ft [2929 to 3217 m]. This compaction of the casing is consistent with the amount of compaction in the reservoir. Monitoring producing wells over longer periods of time did not show simultaneous strain (see Frictional Slippage section). Statistical Analyses Effect of Reservoir Stress and Strain. A rigorous analysis of the compaction-induced casing failure would require voluminous reservoir and overburden data as functions of time, many of which are impractical or difficult to obtain. This problem can be circumvented by applying statistical methods to analyzing the large number of failed wells at Ekofisk to arrive at an empirical correlation of the main variables affecting casing deformation. The statistical analysis was performed on all 7-in. [17.8-cm] -casing wells in Ekofisk. Reservoir stresses and strains are the dominant factors in casing deformation. For Ekofisk rock, the strain is a nonlinear function of stress. Because the formation stress is directly related to reservoir pressure distribution, it is a very useful variable in correlating casing pressure distribution, it is a very useful variable in correlating casing collapse. For new wells drilled and completed in a partially depleted reservoir, however, the use of reservoir pressure will lead to inaccuracies because of the nonlinear relationship between formation stress and strain at high effective-stress levels (low pore pressure). Therefore, there is a definite advantage to correlating casing failure to near-well formation strain rather than stress (reservoir pressure). JPT P. 729