The lower reservoir (SCA) is perforated by nine wells in total, which are grouped in three bundles: SCA-A, SCA-B, and SCA-C, each consisting of three wells. Two of the bundles, SCA-A and SCA-B are to the S−E of the main fault that separates the SCA reservoir, while the third is N−W of the fault, where it targets the best quality sands. The second group of wells penetrates the upper reservoir (SFR) horizontally in two bundles, one East and one west of Weld fault which splits the SFR reservoir. Again, each bundle contains three wells, each of which completes a single layer of the SFR reservoir horizontally.
The reservoir simulations are performed for a period of 10 years, split in three phases designed to support specific geophysical objectives. In the first phase, only the SCA reservoir is being produced. The goal during this phase is to induce maximum amount of pressure drop in the deeper reservoir so that its seismicsignature can be observed without interference from other producers. Mechanical effects throughout the formation, including those manifested in the upper reservoir can also be linked directly to the producers in the deeper reservoir.
During the second phase which is three years, the six shallower wells are also activated in production mode, while production in the deeper reservoir continues. The design objective during this phase is to operate the six upper wells at low rates so that they can induce pressure drop, however the pressure drop should not cross the bubble-point pressure in each respective SFR layer. This is done so that no gas is produced and the observed seismic signature is due only to pressure drop and mechanical effects.
During the final third phase, which is also lasts for three years the size shallower wells (SFR) are produced at higher rates which cause accumulation of gas in their vicinity. During the same phase, one of the deeper wells (SCA-B2) is converted from a producer to a gas injector. The objective during this phase is to observe the seismic signature of gas, injected in the SCA reservoir and out gassed in-situ in the upper SFR reservoir.
The initial oil-water contacts and initial pressure and bubble point pressure are given in Table 2. Note that the bubble point pressure in the SCA reservoir is initially constant, while in the remaining six reservoirs(SFR-1E through 3W) it initially follows the fluid pressure gradient, e.g. initially, it is 10 MPa lower than the fluid pressure.
First the computational mesh is separated into zones with different properties. We have three zones: the non-conductive rock (over/under/sideburden), the reservoirs and the salt. The non-conductuve rock needs to be populated with mechanical properties - Young's modulus, Poisson ratio and density. In addition to these, a static water pressure should also be specified. The reservoirs also need to have a Biot coeffcient A specified, permeability K and a porosity field and relative permeability curves. The salt, unlike the non-conductive rock and the reservoir sands is modeled by a plastic material.