Improved wormhole prediction model considering propagation characteristics of wormhole head in carbonate acidizing

Abstract

Comprehensive studies on wormhole prediction models have been performed to design successful matrix acidizing in carbonate reservoirs. In particular, linear coreflooding experiments have been extensively conducted to predict the optimum injection rate of various acid types under reservoir conditions. Currently, the optimum injection rates are obtained by fitting the linear core-flooding experiment results using a semiempirical model; however, the wormhole propagation behavior has not been considered. This study develops an improved wormhole prediction model to estimate the optimum injection rate and wormhole velocity considering the propagation characteristics of wormhole heads, which is derived by wormhole visualization with Xray computerized tomography (CT). 5 cases comprising 25 linear core-flooding experiments are conducted to determine the optimum injection rates under different acid concentrations and permeabilities. In addition, wormhole visualization is performed using high-resolution X-ray CT to analyze the propagation characteristics of wormhole heads. To validate the improved model, the experimental results of coreflooding are matched with the fitting curves of the improved model and previous model. Furthermore, the absolute average percentage error (AAPE) is obtained to estimate the fitting accuracy of the improved model and previous model. From the results, the AAPE of the improved model was remarkably lower than those of the previous model, confirming that the new model can perform the more preferable prediction of wormhole propagation than the previous model. Finally, the optimum injection rate and the optimum wormhole velocity () are determined using the new model and previous model. The optimum injection rates derived using the new model were lower thanthose of the previous model, but the values obtained by the new model were higher than those of the previous model. It is believed that the lower pump rate may result from the lower optimum injection rate with a faster optimum wormhole velocity. Therefore, a less acid consumption owing to the lower pump rate would be required to create the optimum wormhole approaching the targeted distance from the well to damaged zone for the stimulation. In conclusion, a more economical design of the carbonate acidizing is possible by determining the optimum injection rate for the new model compared to the previous model.