| Popis: |
In thinly bedded sandstone reservoirs, hydraulic fractures are required in horizontal wells to connect isolated pay intervals and to improve the volumetric sweep efficiency during waterflooding. This study presents a new, more robust way to evaluate gel damage and cyclic stress in the laboratory. Results from the laboratory evaluation are validated with field production data. Standard ISO/API tests are adequate at comparing proppant types but do not accurately predict resultant conductivity in a well as they do not account for several in-situ damage mechanisms. With a limited number of cores available it is important to clearly define the scope of the laboratory testing and decide which damage mechanisms to investigate. Testing all variables in the laboratory is not practical. For this study, the primary objectives were to 1) compare ceramic proppant to the natural sand, 2) investigate the impact of thinly-bedded sandstone on the fracture conductivity, and 3) determine the minimum required proppant concentration (the cutoff concentration for interpreting the effective fracture half-length in numerical hydraulic fracture model results). The laboratory testing was designed to simulate as realistic the in-situ condition by 1) using actual formation core, 2) performing cyclic stress cycles to mimic multiple shut-in and production periods, and 3) placing the gel and allowing it to cross-link and break in the fracture. During the conductivity experiments, the following steps were taken: 1) oil injection with cyclic stress applied, 2) dynamic cross-linked gel injection and shut-in for gel breaking, and 3) oil injection with cyclic stress applied. Variables investigated include fluid-rock interaction, gel residual, cyclic stress, proppant type, concentration, and size distribution and time dependency of conductivity. Discount factors are derived from the test results which provide a more realistic and repeatable conductivity prediction. This study discovered that for hydraulic fracturing of weak rocks in the shallow formation, the baseline fracture conductivity from API tests should be reduced by 22% first to account for the proppant-rock interaction. After applying the aggressive cyclic stresses, the cumulative conductivity loss increases to 38%. After the cross-linked gel cleanup, a total of 72% fracture conductivity is lost for a proppant pack at 2 lbm/ft2 and 91% conductivity loss for proppant pack at 1 lbm/ft2. It is also found in this study that each large-scale stress cycle reduces an approximate 1% fracture conductivity of the loosely packed proppant until a tighter and stable proppant pack is formed. The cyclic stress effect becomes insignificant when the proppant pack porosity decreases to ∼0.2. Well production history was matched by varying fracture properties in the transient inflow performance analysis. For two wells under the same fracture design, the matched fracture conductivities resulted in less than 25% error compared with the retained conductivities from the laboratory tests. This validated the laboratory findings and method. In summary, this study investigates a critical completion design variable and well performance modeling input, i.e., fracture conductivity, in low-to-moderate permeability, thinly bedded sandstone reservoirs. It breaks down the fracture conductivity degradation into various components and enables further fracturing design optimization, such as proppant selection, fracturing fluid qualification, pump schedule design, well shut-in frequency, frac sleeve spacing, etc. It provides an unbiased estimate of retained fracture conductivity after considering the major impairment mechanisms. It also prevents fictitious and overly optimistic fracture conductivities which originate from the fracturing practices in unconventional reservoirs and the continuous drive for cost savings. This results in calculations of completion skin factors that more accurately represent the fracture conductivity for longitudinal fractures in openhole sleeve completions, reinforcing the importance of fracture design optimization on well productivity. |
