Effective Application of Steam Assisted Gravity Drainage of Bitumen to Long Horizontal Well Pairs

Abstract

Abstract A recent field test of the Steam Assisted Gravity Drainage (SAGD) process at the Alberta Oil Sands Technology Research Authority (AOSTRA) Underground Test Facility (UTF) has proven that the process mechanisms work in the field as expected from previous laboratory, analytical, and numerical work. The UTF test used well pairs that were completed for only about 5 m, however, as compared to lengths of several hundred metres which will be required for commercial application. Accordingly, the UTF Phase B test currently under construction will use wells with a full 500 m of completion. This paper reviews the design and operating requirements that have been developed to ensure that the full lengths of the Phase B well pairs contribute to bitumen recovery, i.e. that the production rates will scale up in proportion to length. The discussion proceeds from the conditions required to initiate drainage at a given location along the well pair, to the means of establishing and maintaining such conditions at every location. Numerical examples show that, for typical formation permeabilities, start-up can be achieved with thermal conduction and gravity alone, two highly predictable phenomena. This leads to consideration of design and operational requirements to ensure that conduction heating is maintained on the one hand, and gravity is not defeated (by adverse pressure gradients) on the other. In summary these are:the well pairs must be drilled so that the interwell spacing remains within specific tolerances;the liner and tubing must be large enough, relative to the completion length and expected formation productivity, so that pressure gradients are kept below certain values, anda full length tubing string must be combined with correct operating policy to ensure that the full lengths of both injection and production liners are kept hot throughout the life of the well. Introduction Butler et al(1), and Edmunds et al(2) have previously described the mechanics of two dimensional SAGD to a pair of horizontal wellbores. The recent Phase A test of SAGD at the AOSTRA UTF(3) showed that the process mechanisms work in the field as expected from previous laboratory, analytical, and numerical work. Phase A well pairs were completed for only about 55 m, however, whereas economic commercial application requires several hundred metres of active completion. AOSTRA is accordingly proceeding with a Phase B Test of three long well pairs, with 500 m completed length. In a reservoir of immobile bitumen, fluid mobility must be artificially FIGURE 1: Three dimensional start-up of a long well pair. created between a well pair before SAGD can proceed. With reference to Figure 1, it should be easy to force communication between an injector and producer over at least one section of a long well pair, but it is less clear how the remainder of the pair can be made to follow in a reasonable time, say a few months. Long delays will incur very significant economic penalties; not only will the Calendar Day Oil Rate (CDOR) be reduced, but also, because a longer time will be required to recover the same amount of oil, the Stearn/Oil Ratio (SOR) will be increased.