Numerical simulation for the determination of hydraulic fracture initiation and breakdown pressure using distinct element method

Abstract

Hydraulic fracturing technique has been widely used in many cases to enhance well production performance. In particular, this technology is proven to be the most viable technique for the oil and gas production from unconventional reservoirs. Accurate prediction of fracture initiation and breakdown pressure is vital for successful design of Hydraulic Fracturing operation. Methods of predicting these pressures include analytical analysis, field experiments, laboratory experiments and numerical simulations. Despite great achievements in the area of analytical analysis, they often failed to represent the true reservoir case, and consequently are found to be erroneous. Field tests such as mini-frac test are the best method for prediction of initiation and breakdown pressure. However these tests are very limited due to their costs and are not very suitable for sensitivity analysis. Controlled laboratory tests seem to be the best option for predicting initiation and breakdown pressures. Test parameters such as fracturing fluid properties and principal stresses can be controlled with great precision to achieve accurate results. However, same as field tests, laboratory experiments are expensive. Core samples are limited and are expensive. Coring operation can take 4-5 days of rig time to take a 90 ft core. Geo-mechanical tests can take up to three days of a laboratory technician's time per sample. Consequently, this will limit the number of tests to be done, and as a result it causes limitations on the conclusions that can be drawn from these tests. Simulation studies on the other hand do not have these limitations and can be used for as many times as desired to perform sensitivity analysis.This paper presents a simulation model that is based on distinct element method. It is used to study the fracture initiation and breakdown pressure during hydraulic fracturing tests. The accuracy of the model was justified through comparison between laboratory experiments and numerical simulation. Four sandstone samples from two different sandstone types and a synthetic cement sample were used in the experimental studies. The tests were performed in True Tri-axial Stress Cell (TTC) with the capability to inject fluid into the samples. Simulation results demonstrate good agreement with experimental results. Fracture propagation path was found to be very similar. Fractures propagated in the direction of maximum horizontal stress.