The goal of acid fracturing operations is to create enough fracture roughness through non-uniform acid etching on fracture surfaces such that the acid fracture can keep open and sustain a high enough acid fracture conductivity under the formation closure stress. A detailed description of the rough acid-fracture surfaces is required for accurately predicting the acid-fracture conductivity. In this paper, a 3D acid transport model was developed to compute the geometry of acid fracture for acid fracturing treatments. The developed model couples the acid fluid flow, reactive transport and rock dissolution in the fracture. We also included acid viscous fingering in our model since the viscous fingering mechanism is commonly applied in acid fracturing to achieve non-uniform acid etching.

Carbonate reservoirs mainly consists of calcite and dolomite minerals but the mineral distribution can be quite heterogeneous. Based on the developed model, we analyzed the effect of mineral heterogeneity on the acid etching process. We compared the acid etching patterns in different carbonate reservoirs with different spatial distributions of calcite and dolomite minerals. We found that thin acid-etched channels can form in carbonate reservoirs with interbedded dolomite layers. When the reservoir heterogeneity does not favor growing thin acid-etched channels, we investigated how to utilize the acid viscous fingering technique to achieve the channeling etching pattern in such reservoirs. Through numerical simulations, we found that thin acid-etched channels can form inside acid viscous fingers. The regions between viscous fingers are left less etched and act as barriers to separate acid-etched channels. In acid fracturing treatments with viscous fingering, the etching pattern is largely dependent on the perforation spacing. With a proper perforation design, we can still achieve the channeling etching pattern even when the reservoir does not have interbedded dolomite layers.

You can access this article if you purchase or spend a download.