|
Minimum Miscibility Pressure (MMP)Also see the optimization example given at Generalized Automatic Global Predictive Optimization. The optimal producing bhp for gas flooding in the optimal producing strategy for SPE5 (3755.3 psia) is just below the first-contact miscible pressure (3874.8), and demonstrates that the given lab MMP (3000) is meaningless in the reservoir, and that "miscible flood" models assuming a fixed miscibility pressure do not apply in general. That example also demonstrates the non-existence of any competent AI/ML method in engineering for design, optimization, or forecasting. Laboratory slimtube tests have conventionally been used to determine the minimum pressure required for some solvent (gas) to attain multiple-contact miscibility (MCM) with some oil (called the minimum miscibility pressure, or MMP). Compositional simulators determine miscibility as a function of composition, pressure, and temperature from thermodynamic equilibrium flash and saturation pressure calculations and do not use the slimtube test MMP at all, except possibly for (mis)guidance regarding the setting of production and injection pressure constraints in the reservoir model at or above the MMP so that miscible recovery is (expected to be) achieved. The problem is that in most "miscible flood" simulators, MMP is an input variable and miscibility is assumed in all blocks with pressures greater than or equal to MMP. MCM may not be achieved in the reservoir as a simple function of lab MMP because of extreme mixing due to multiphase 3D flow, gravity (sweep), capillary pressure, and heterogeneity. So, fully compositional models are generally required in order to robustly simulate miscible or near-miscible or partially miscible recovery (or whenever hydrocarbons are injected), except that our 2 component FCM option applies to the first-contact miscible case (when the reservoir is operated above the FCM pressure in the single-phase hydrocarbon region). MMP as measured in the lab has absolutely no significance with respect to the achievement of miscibility in real reservoirs. Any model that assumes otherwise is simply wrong. The SPE5 problem1 is a good example. Scenario 2 is a primary wag flood. Sensor data and output files are spe5b.dat and spe5b.out. The first-contact miscibility pressure is 3874.8 psia, which is the maximum saturation pressure on the solvent/oil phase diagram that is printed in spe5b.out. The given slimtube test indicates an MMP of about 3000 psia. The wag flood is operated with an injection BHP of 4500 psia and a producing BHP of 3000 psia. The pressure map printed at end of run in spe5b.out shows that the entire reservoir is well above MMP, probably leading the author of SPE16000 to his conclusion that "The discussion of the previous section indicates that for scenario two, in which minimum miscibility conditions were exceeded during the entire simulation for most grid blocks, four-component results with complete mixing gave excellent agreement with compositional results."
P TIME = 6186.1 DAYS DATE: 0
0 0 MAP WINDOW 1 However, the oil saturation map shows that miscible recovery occurred mostly only in the top layer and near the injector in layers 2 and 3, with residual oil saturations of about .3 (Sorw from water/oil relperm table) remaining in most of the rest of the reservoir. This is due to gravity override of the solvent and the fact that the top layer is the high-perm layer (layer areal perms are 500, 50, 200, respectively):
SO TIME = 6186.1 DAYS
DATE: 0 0 0 MAP WINDOW 1 We believe that it is impossible for any model assuming miscible conditions based on MMP alone to reproduce similar values of final oil saturations indicating where miscible and immiscible recovery (to waterflood) has been achieved.
1. Killough, J. E., "Fifth Comparative Solution Project: Evaluation of Miscible Flood Simulators", SPE 16000 presented at the Ninth SPE Symposium on Reservoir Simulation held in San Antonio, Texas, February 1-4, 1987
|
© 2000 - 2022 Coats Engineering, Inc.
|
