Field testing the role of heterogeneity at the inter-well scale during two phase flow
Abstract
Connectivity of rocks with different fluid flow properties (reservoir architecture) is a major source of uncertainty in predicting multi-phase fluid flow. Multi-institution research teams have completed three major DOE-funded test programs in which movement of supercritical CO2 through brine-saturated fluvial sandstones was observed with multiple tools. In each test, closely spaced wells (30 to 100 m) were drilled to reservoir depth so that the amount of reservoir complexity sampled between the wells would be reduced and higher resolution measurements of change could be obtained during time-lapse monitoring. During one test (Frio 1), an exceptionally homogeneous injection zone produced by marine reworking of a fine-grained fluvial sandstone of the Upper Frio Formation produced a classic-wedge-shaped CO2 plume. The maximum area occupied by CO2 was the result of radial expansion of the plume near the injection well; this produced plume down-building. Away from the injection well , the importance of gravity-override increased. A second test at the same well array in a deeper sandstone (Frio 2) was sited in heterogeneous (gravel to muddy sandstone) high permeability (>3 Darcy), weakly cemented fluvial sandstone. A slow injection rate over a small interval at the flow unit base was used to accentuate buoyancy effects. Measurements of plume migration through time using tracers and cross-well seismic documented interaction between near-well radial flow and vertical rise. However, because of discontinuous fast paths in gravel zones and CO2 ponding against muddy sandstone baffles, reservoir heterogeneity was a dominant influence in short-term plume evolution. The third test interval in the SECARB "Early" test documented response of amalgamated gravel and sandstone point bars in which heterogeneity was reduced by cementation. CO2 injection rate was increased incrementally and flow-rate dependent reservoir responses observed. These suggest that capillary-entry pressure and relative permeability evolution guide flow. Plume evolution was highly non-linear, demonstration dominance of preferential flow though fast paths. CO2 continued to access new flow paths as rate increased and through time; pressure was not linear with injection rate. Over a one year test period at the inter-well test scale, reservoir properties seem more important than either pressure or buoyancy in controlling plume evolution. Three intensively monitored two-phase injection experiments across ranges of inter-well reservoir heterogeneity and flow rate provide data to explore methods for bounding uncertainty. More than 20 fluid flow models from these tests have been or are being built to test approaches to history matching.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.H24B..05H
- Keywords:
-
- 1895 HYDROLOGY / Instruments and techniques: monitoring