A quick analysis approach for early-time well test data of vertically fractured wells: Applications in Changqing oilfield, China

Abstract

The increasing development activities for low-permeability even ultra-low permeability reservoirs worldwide present many challenges to well test interpretation. On the one hand, the increasing well number puts higher requirements on the reduction of interpretation time. On the other hand, much test data locates at the early-time regimes, so incomplete flow regimes pose serious non-uniqueness. To solve the problems above, we first introduce an analytical solution with simple form, fast calculation speed, and accurate results for finite-conductivity vertical fracture model on the basis of our previous work, which is further used to shorten the inversion time and obtain short-time asymptotic solution in the real space. We verify its accuracy by the semi-analytical model proposed by Cinco-Ley et al. and numerical model built in KAPPA. Results indicate that the analytical solution is valid when FcD>1. Then, the wellbore storage (CD) and skin (S) are extracted by the short-time asymptotic solution of proposed analytical solution to construct a novel type curve. The type curve can aid the conventional interpretation to obtain more reliable CD and S, thus reducing the non-uniqueness. We mainly divided the type curve into three typical regimes. The first regime is a straight line with zero-slope at early time which is dominated by CD. Under the larger CD, the duration of straight line lasts longer and the position is lower on the log-log plot. The second is a transition between two straight-line regimes which reflects the S. The transition appears the later under the larger S. The third is also a constant-slope straight line at late time and we think it is no use for well test inversion. Finally, the methodology is applied in Changqing oilfield cases, which illustrate its applicability. This work can help for better understanding of its potential in improve the interpretation efficiency and reduce the non-uniqueness.