Reservoir Simulation of Horizontal Wells in the Helder Field

Abstract

Summary. Results of a reservoir simulation study of horizontal wells in a field that is entirely underlain by water and has a highly unfavorable mobility ratio are compared with about 4 years of actual field data. Finely gridded single-well models were used to investigate the short-term performance of the horizontal wells and to generate pseudofunctions for performance of the horizontal wells and to generate pseudofunctions for use in the field model. Introduction The Helder oil field, on the Dutch Continental Shelf, was brought on production 1982 and developed with 13 conventional wells. To improve oil recovery and to minimize water coning, the field was redeveloped during 1987 and 1988 with the drilling of 10 horizontal wells by sidetracking from existing wells. The simulation study described here was initiated to answer the following questions posed as a result of observing the performance posed as a result of observing the performance of the horizontal wells.Is the water-cut performance adversely affected by producing the wells at high gross rates?What effects do various well and reservoir parameters have on short-and long-term well parameters have on short-and long-term well behavior? Parameters to be considered are horizontal-well length, height above the water/oil contact (WOC) reservoir heterogeneity, localized vertical permeability barriers, and drainage area shape.Is water production distributed evenly along the horizontal wellbore, or does it enter preferentially at one or more intervals?Can production performance be improved by water shutoff treatments on existing wells or by modified completion design in future wells?Can the ultimate recovery be improved with additional infill horizontal wells? Field Description. The Helder field is an Block Q/1 of the Dutch North Sea, about 62 miles northwest of Amsterdam in 85 ft of water. The structure is a slightly faulted anticline at a depth of 4,600 ft. Formation dip along the northwest/southwest axis of the field is generally less than 1 degree. The field is underlain by water over its entire 1,140 acres of closure and has a maximum oil column of 131 ft. The reservoir is supported by a strong aquifer. Two other fields, Helm and Hoorn, brought on production at the same time as Helder, share the same aquifer. The original oil in place us estimated to be 72 × 10(6) STB. Oil gravity is 22 degrees API, and initial viscosity at reservoir conditions is 30 cp. Production is from the Vlieland sand, which has permeabilities ranging from 500 to 6,000 md. Production History With Conventional Wells. Production History With Conventional Wells. Helder was put on production in Oct. 1982. During 1982–84, the field was developed with 12 conventional (vertical/deviated) wells from a centrally located platform. In 1986, a previously drilled appraisal well (Well B1) was tied back to the platform by use of satellite tripod tower. Early water breakthrough and rapidly increasing water cut were expected in the conventional wells because of the unfavorable mobility ratio (M=25), flat field structure, and high vertical permeability. The majority of the wells produced water within a few days of production startup. This performance was predicted by initial computer simulations of single-well models, which showed that the critical rate to prevent water production had to be exceeded if the field was to be produced economically. produced economically. Subsequent simulations and preliminary history matching predicted that water-cut performance was not rate-sensitive for practical performance was not rate-sensitive for practical purposes. This led to the conclusion that purposes. This led to the conclusion that maximizing gross fluid off-take was necessary to improve economics. Individual wells were produced at rates up to 12,000 BFPD with water cuts from 85 to 97%. Some of the implications of the high-gross-fluid/ high-water-cut production strategy were decline in reservoir pressure and increase in workover frequency with conversion to larger pumps. To overcome these extreme coning conditions, Unocal Netherlands investigated the feasibility of redevelopment with horizontal wells. Redevelopment With Horizontal Wells. The first horizontal well (Well A4RD) was put on production in Jan. 1987. After the success of this well, nine more wells were horizontally sidetracked. Figs 1 and 2 show the locations of the original wells and the horizontal sidetracks. Table 1 summarizes the horizontal-well parameters. The performance aspects of parameters. The performance aspects of the Helder horizontal wells were the subject of a previous paper. Model Rock and PVT Properties. Seven layers were used in the field model and 15 layers in the single-well models. JPT P. 906