Prediction of Abrupt Reservoir Compaction and Surface Subsidence Caused By Pore Collapse in Carbonates

Abstract

Summary. A new procedure has been developed to predict the abrupt in-situ compaction and the associated surface subsidence above high-porosity carbonate fields that show pore collapse. The approach is based on an extensive laboratory compaction study in which the effects of carbonate @, porosity, core preparation, pore saturant, horizontal/vertical stress ratio, and loading rate on pore-collapse behavior were investigated. For a number of carbonate types, a trendline was established that describes the relationship between the porosity after collapse and the vertical effective stress. This trendline concept, in combination with existing subsidence models, enables reservoir compaction and surface subsidence to be predicted on the basis of wireline porosity logs. Static and dynamic elastic constants were found to be uncorrelated during pore collapse. The position of the trendline depends strongly on carbonate type, pore saturant, loading rate, and stress ratio. Therefore, procedures are given to derive the correct in-situ trendline from laboratory compaction experiments. Introduction In general, the bulk compressibilities of carbonates are much lower than those of sandstones of comparable porosity. Field cases of considerable reservoir compaction and surface subsidence caused by hydrocarbon production from carbonate reservoirs are therefore rare. In high-porosity carbonates, however, the phenomenon of pore collapse can occur in the pressure regime prevailing during production. These carbonates, which are usually well consolidated, exhibit a low compressibility up to a certain stress level but strongly compact at higher stress. This sudden increase in compressibility, coupled with a large irreversible deformation, is called pore collapse. Several authors have observed this phenomenon in the laboratory. In the field, the resultant reservoir compaction and associated surface subsidence can be large. In the Ekofisk field offshore Norway, reservoir compaction caused by pore collapse has resulted in a reduction in the platform airgap of some 3 m [10 ft],3 while up to 7 m [23 ft] of surface subsidence has been predicted for gas-bearing carbonate buildups offshore Sarawak. Objectives and Scope In land operations, subsidence may cause environmental problems, while seabottom subsidence must be taken into account when offshore platforms are designed. Severe reservoir compaction can result in damage to casings and liners, and can have consequences for the perforation policy. Although the productivity of the carbonate reservoir can seriously decline as a result of any accompanying permeability reduction, pore collapse may also act as a drive mechanism. Early prediction of the time evolution and amount of pore collapse to be expected is therefore essential for both onshore and offshore operations. In this study, the relationships between material properties, such as carbonate rock type, porosity, Brinell hardness number, static and dynamic elastic constants, and pore-collapse behavior, have been investigated in detail. In addition, special attention has been given to parameters that could affect the translation of laboratory experiments to field conditions, such as sample preparation, pore content, loading rate, and horizontal/vertical stress ratio. A procedure is proposed for field application of the laboratory results obtained. Experimental Procedures Sample Selection and Preparation. Experiments were carried out on Danian outcrop samples, Danian and Maastnchtian chalk samples from various North Sea fields, chalk samples from a Middle East field, and moldic limestone and dolomite samples from a field in the Far East. (About 190 samples have been studied altogether.) All samples used for uniaxial compaction experiments had a diameter of 50 mm [2 in.] and a length of about 30 mm [1.2 in.]. For yield-strength and hydrostatic loading tests, samples were used with a length-to-diameter ratio of 2 to limit end-plate effects. Scouting experiments showed a considerable influence of sample orientation on the pore-collapse stress level; therefore all samples were drilled vertically with inspect to the earth's surface, simulating field-compaction conditions.uring the initial stages of the study, the samples were not cleaned before the compaction experiments to avoid any influence of cleaning agents on pore-collapse behavior. Instead, the samples were dried in a vacuum oven and subsequently saturated with "chalkified" water, which was obtained by boiling determined water concerning crushed sample material. After the experiment, the samples were cleaned with a calculated of chlorothene and methanol and their porosities were determined. The initial porosity of the sample was then calculated on the basis of the measured stress/strain curve. Once it had been established that sample cleaning does not influence the subsequent compaction behavior (see the section on Influence of Loading Rate), the samples were cleaned before the experiment, resulting in a more accurate determination of the initial porosity. Experimental Setup. Pore collapse is studied by performing compaction experiments in a triaxial compaction cell (Fig. 1). In this cell, the sample is enclosed in an impermeable elastomer sleeve, which is loaded by hydraulic pressure around its circumference. Axial stress is applied by a piston. Axial and radial stresses can be changed independently, and the axial loading rate can be varied between 0.05 and 0.1 × 10(4) MPa/h [0.5 and 10(4) bar/hr]. Piezoelectric crystals enable the measurement of compressional- and shear-wave velocities. The radial deformation induced by the increasing axial stress is measured by a radial displacement transducer. In the uniaxial strain experiments, this signal drives a feedback system that controls the radial stress rate in such a way that the diameter of the sample remains constant during the compaction experiment, thus simulating reservoir compaction at zero lateral strain. In this procedure, end effects resulting from friction SPEFE P. 340^