Monitoring and Analysis of Gravel-Packing Procedures To Explain Well Performance

Harry O. McLeod

DOI: https://doi.org/10.2118/27356-pa

1994-10-01

Journal of Petroleum Technology

Abstract:Summary Gravel-packed gas wells completed in the Gulf of Mexico since 1980 were reviewed to build a selective database for a completion-effectiveness study. Gas wells with clean, uniform sands were selected for analysis. Significant monitoring data identified were injectivity tests at different points during the completion and fluid loss rates (barrels per hour). Injectivity before gravel packing and productivity after gravel packing were classified according to sidewall-core permeabilities. Different gravel-pack preparation and execution techniques were reviewed. Fluid-loss-control pills were identified as the greatest source of damage restricting gravel-packed well productivity. Injectivity tests and sidewall-core permeabilities provide valuable information for monitoring well completion procedures. Introduction There have been many improvements since 1970 in gravel-packing design and execution. Well performance improved as a result of changes made in the years since, yet many wells do not perform up to the expected level. McLeod and Crawford showed that voids would not be well filled with gravel when severe damage existed before gravel was placed in the wellbore. Perforated tunnels were not packed, although sand was controlled by placement of a good annular pack between the screen and casing. Improved filtration of completion fluids and underbalanced perforating with tubing-conveyed perforating (TCP) guns were applied with results that doubled the performance of wells completed with previous techniques. Later field work has focused on more effective placement of gravel in tunnels. Work by Brannon et al. in 1986 showed that wells were still damaged, even after using surge perforating and high-quality completion fluid filtration. Many times, this damage could be removed by acetic acid or HCl alone. This implied damage by residue remaining from well-control pills formulated with hydroxyethyl cellulose (HEC) polymer to prevent high completion-fluid losses or HEC-gelled carrier fluids. Others who recognized this problem developed shearing and filtering of HEC gels to reduce this residue of poorly hydrated fisheyes. Such developments resulted in improved gravel-packed well performance; however, even this development was not enough to achieve ideal gravel-pack performance because well-control pills used after perforating and wellbore cleanup are often not sheared and filtered. Subsequently, prepacking perforations with stages of acid and low gravel concentration slurries was introduced by Matherne and Hall with much improved well performance. Much earlier, Penberthy had obtained excellent results using water (no HEC polymer) to prepack perforations before running screens and to transport the gravel into the annulus between the screen and casing. Conoco implemented these new techniques, yet well performance was highly variable, and no technique stood out consistently from the rest. So we conceived a study to analyze completion procedures and well performance of a large group of wells to identify the most significant procedures (good or bad)that affect the performance of a gravel-packed well. The following strategy was devised to accomplish this goal. Strategy for Gravel-Packed Well-Completion Analysis We selected gas wells with clean, uniform sands using openhole logs of the perforated intervals. Gas wells provide more reliable prospects for analysis because fluid properties are well defined. Although high-permeability gas reservoirs are difficult to quantify with standard pressure-buildup tests without downhole shut-in, steady-state drawdowns are available. These are very helpful in estimating the completion pressure drops because so little of the total drawdown is attributed to the flowing pressure drop in the reservoir. There is an abundance of completion and production data that may provide an understanding of well-completion factors if all data can be integrated with the industry knowledge of formation damage and well completion. The condition of the gravel-packed perforations determines the completion pressure drop. Most significant is the degree of damage to the formation sand before gravel packing, whether temporary with removable gelled water polymers or permanent with completion-fluid contaminants and polymer residuals (either poorly hydrated fisheyes or unbroken gel residue). This damage limits the effectiveness of perforation packing with gravel. Any vertical well has only a few minutes for each perforation to be packed before the annulus is filled during placement of the gravel pack. Damage restricts fluid leakoff so that little gravel is carried into the perforation. Subsequently, voids created beyond the cement sheath by perforating and flow are not filled and supported with gravel. Such incompletely filled voids are then subject to filling with lower-permeability formation sand when the well is produced. This causes a severe loss in productivity. When wells are prepacked, the perforations are more likely to be filled and not leave empty voids, but residual damage to the formation will still affect well productivity. Analysis of several wells with reliable data allowed us to identify those procedures that cause the most damage in the perforating, well-control, and gravel-packing processes. Data that quantified or indicated the well condition or extent of damage were (1) fluid loss rates, usually in barrels per hour (for various pill types and sizes); (2) injection rates and pressures (usually measured before prepacking or gravel packing); (3) injection rates and pressures during acid treatments; (4) perforating method, shot density, and gun size; (5) amount of gravel placed outside the casing during prepacking and during the main gravel pack; (6) squeeze rate and pressure-buildup rate during gravel placement; (7) placement of any contaminated fluid in the well either by using unfiltered fluids or by pumping fluids through inadequately cleaned tubing; (8) type and volume of fluids lost to the formation; (9) production performance of the well after initial cleanup; (10) nodal analysis of well performance after initial cleanup; and (11) pressure-transient testing and analysis for completion efficiency and skin factor. Data Acquisition All gas wells chosen were perforated in clean, uniform sands to improve net pay estimates and to reduce the effect of flowing pressure drop in the reservoir. The wells had average sidewall-core permeabilities ranging from 350to 2,000 md. Average in-situ formation permeability was assumed to be one-third the average sidewall-core permeability owing to the absence of accurate pressure-buildup tests. Although sidewall-core permeabilities are variable and we expect considerable scatter in correlations with formation permeability, we still believe that this approach will help improve completion practices. Wells with accurate pressure-transient tests indicated a good correlation in these clean, uniform sands with the estimate of one-third the average sidewall-corepermeability. P. 878^