Gas deviation factor (z-factor) and other gas reservoir fluid properties, such as formation volume factor, density, and viscosity, are normally obtained from Pressure-Volume-Temperature (PVT) experimental analysis. This process of reservoir fluid characterization usually requires collecting pressurized fluid samples from the wellbore to conduct the experimental work. The scope of this paper will provide an alternative methodology for obtaining the z-factor. An IR 4.0 tool that heavily utilizes software coding was developed. The advanced tool uses the novel apparent molecular weight profiling concept to achieve the paper objective timely and accurately.
The developed tool calculates gas properties based on downhole gradient pressure and temperature data as inputs. The methodology is applicable to dry, wet or condensate gas wells. The gas equation of state is modified to solve numerically for the z-factor using the gradient survey pressure and temperature data. The numerical solution is obtained by applying an iterative computation scheme as described below:
A gas apparent molecular weight value is initialized and then gas mixture specific gravity and pseudo-critical properties are calculated.
Gas mixture pseudo-reduced properties are calculated from the measured pressure and temperature values at the reservoir depth.
A first z-factor value is determined as a function of the pseudo-reduced gas properties.
Gas pressure gradient is obtained at the reservoir depth from the survey and used to back-calculate a second z-factor value by applying the modified gas equation of state.
Relative error between the two z factor values is then calculated and compared against a low predefined tolerance.
The above steps are reiterated at different assumed gas apparent molecular weight values until the predefined tolerance is achieved.
This numerical approach is computerized to perform the highest possible number of iterations and then select the z-factor value corresponding to the minimum error among all iterations. The proposed workflow has been applied on literature data with known reservoir gas properties, from PVT analysis, and showed an excellent prediction performance compared to laboratory analysis with less than 5% error.