Simulation Study Of The LETC TS-2C In Situ Combustion Test In Utah Tar Sands

Abstract

Abstract A field in situ combustion test on a Utah tar sand featuring reverse combustion followed by forward combustion was simulated with a two-dimensional mathematical simulator. The model provided answers to some operating problems and gave indications for improving such a test in the future. Introduction From August 1977 through February 1978 the Laramie Energy Technology Center (LETC) of the United States Department of Energy conducted an in situ combustion field test, designated LETC TS-2C, on a tar sand deposit near Vernal, Utah. A mathematical model study was subsequently conducted to match the performance of this test to help interpret the effects of field parameters and operating procedure upon the test and answer certain questions procedure upon the test and answer certain questions that arose about the test and thus help refine the design and operation of any possible future tests. The study was conducted using a multi-dimensional in situ combustion model, to attempt to calculate the field performance using the normal history matching procedure. Since the model was limited to about 120 grid blocks by computer storage size, it was felt that most of the questions posed could be answered by modeling only a portion of the experiment to allow finer definition of the reservoir. The northwest quadrant (bounded by wells 2I1-2I2-2P1-2P2,) (Fig. 1) was chosen as the best candidate for the simulation, since it appeared that this quadrant had most closely operated as planned, and yet most of the questioned phenomena were also observed in this quadrant. Review of the Field Test Details of the LETC TS-2C test have previously been reported by Johnson, et al. but a brief review is included here. Figure 1 shows the test pattern with the formation top locations of the wells. LETC planned to ignite in well 2P2 and then draw reverse planned to ignite in well 2P2 and then draw reverse combustion fronts from well 2P2 to wells 2P1 and 2P3. Upon breakthrough of the fronts, injection was to be switched to wells 2I1, 2I2, 2I3, 2I4, 2I5, and 2I6 to draw two broad combustion fronts along the pattern, wells 2P1 and 2P3 being converted to pattern, wells 2P1 and 2P3 being converted to production to help achieve this purpose. Upon arrival production to help achieve this purpose. Upon arrival of the reverse combustion fronts at the 2I1-212-2I3 and 2I4-2I5-2I6 lines the injection rates were to be increased to help the front "echo" into a forward combustion mode. The forward combustion front was then to be propagated back to the center line of wells in the line drive pattern. Wells numbered 201 through 2014 were drilled as observation wells to measure bottom hole temperatures, pressures, and compositions during the experiment. Before the test began, air rate and tracer testing was conducted within the pattern as described in Ref. 1 with the result of a directional permeability or channel being identified, particularly permeability or channel being identified, particularly strongly between wells 2I2 and 2P1 but also between other pairs and generally trending in a SW to NE direction. The testing, however, only examined connectivity of injectors with producer sets, and not with other injectors. All wells were cored with permeability, porosity, and saturations determined on each core. The combustion test was operated generally as planned, but unforeseen circumstances caused the four planned, but unforeseen circumstances caused the four quadrants of the pattern to respond differently to injection. Temperature responses were observed and recorded in all observation wells in the northwest quadrant, some of the wells in the southwest and northeast quadrants, and none of the wells in the southeast quadrant. Pressure, rate, and gas analysis data were also recorded during the test. The details are presented and LETC's analysis is discussed by Johnson et al. Because of the manner in which the reservoir responded to injection, not only between the different quadrants but also within the quadrants where temperature responses were different than has been observed in previous tests, it was decided to attempt to model study of the project to see if the performance could be resolved and to determine if performance could be resolved and to determine if altering the operating procedure would improve future tests.