Hydraulic fracturing is practiced to stimulate unconventional and tight oil and gas reservoirs. This technique increases the deliverability of reservoir by inducing reduction in pressure drawdown and exposing more reservoir contact; hence, improving the well productivity index. Fracturing can also help to break through and pass near well-bore severe formation damage and can make low permeability and tight reservoirs economically viable.
Prior to execution of fracturing job, the feasibility of the operation can be accessed through analyzing the effect of skin on overall deliverability of the system. Valuable extent of research has already been done to analytically model or numerical simulate the whole scheme for evaluating this effect. While the orthodox analytical approach to address this problem incorporates simple assumptions, an attempt has been made to build a semi analytical correlation.
This paper proposes a mathematical correlation to determine equivalent skin in hydraulic fractured system for pre-analysis of hydraulic fracturing obtained through numerical simulation modeling. The paper describes two 3D simulation models scenarios; Scenario 1 is without hydraulic fracture and Scenario 2 is with hydraulic fractures. Both scenarios consider the anisotropy (different Kx and Kv) in the system. Scenario 1 has been run for the whole range of matrix permeabilitity values from 0.001 to 1 md. Scenario 2 has been run with different sets of varying fracture's lengths, fracture widths and fracture permeability as well as matrix permeabilities. The two scenarios with corresponding matrix permeability are comparatively analyzed and skin has been determined by calculating fold of increase. Total number of 29040 Data points are obtained by the 29040 number of simulation runs by coupling the commercial reservoir simulator with Matlab. Multivariate regression technique is used to obtain generalized correlation for skin as a function of hydraulic fracture and matrix permeabilities properties. 20000 data points are used to develop proposed correlation and remaining 9040 data points are used to validate and test the proposed correlation.
The newly derived correlation was compared with prevailing analytical approach for skin estimation. It has been found that the degree of absolute percentage error, average percentage error and standard deviation by using of this new correlation has reduced to 3.21%, 3%, and 6.4 respectively.
