Abstract
The development of pressure transient analysis was originally based on single phase flow for slightly compressible fluids. It was later extended to include dry gas flow and multiphase conditions.
In this work, we explore the accuracy of the three most commonly used multiphase pressure transient analysis methods [single phase analysis using composite model, Perrine Martin (P-M), and multiphase pseudo-pressure method] under various conditions. Our approach included using appropriate numerical simulation techniques to generate drawdown and buildup data for gas-oil system at variety of conditions of GOR, CGR. Then, each test was evaluated with the 3 well test analysis methods, and error between calculated and actual results was computed for each method. In each case, we compared the well test derived properties (e.g. permeability, skin, completion information) with actual properties (given from numerical reservoir simulation). We tested various oil and gas properties spanning different fluid types of gas condensates and volatile oils. The testing also included different common well testing models such as radial, limited entry, hydraulically fractured, and horizontal wells. We then embarked on detailed analysis of these errors to derive conclusions regarding which methods give more accurate results under which conditions.
As an example, application of the three analysis approaches to horizontal well models in gas condensate reservoirs using different actual fluid samples with varying composition, at different levels of drawdown; it was found that (1) single-phase with composite reservoir method underestimated the permeability by 8% for very lean gas-condensate and by 18% for rich gas condensate fluid sample, (2) Perrine Martin method was able to predict the permeability with error <17% at low gas production rate, and (3) multiphase pseudo-pressure functions (steady state and three-zone) produced best results with permeability error less than 4% for lean gas-condensate samples, and by 16% for very rich gas condensates. Many other useful conclusions for different systems were also derived. These conclusions show the expected level of error for properties predicted from well test analysis under different conditions of multiphase and for different well test models. Using these conclusions, a list of guidelines was recommended for well test analysis in multiphase tests.
