Phoenix Reservoir Engineering Consulting
PVT lab data varies widely in quality. According to a recent round-robin investigation of quality of results from standard oil and gas PVT laboratory tests (SPE 116162),
The authors did not investigate the quality of special laboratory procedures used to study volatile oils and gas condensates. However, the lack of accurate fluid property correlations for volatile oils and gas condensates makes the quality of lab data more, not less, important. One test in particular, the constant volume depletion (CVD) test, provides data essential for numerical reservoir simulation, production forecasting, reserves estimation, and reservoir management.
The CVD test consistents of a variety of measurements at a series of pressure steps, representing the process of production under depletion drive. In a valid CVD test, the various measurements must both 1) be internally consistent and 2) honor specific mass balance, mole balance, and equilibrium constraints.
In the present work, we have developed a workflow to provide an intensive quality check for CVD test data, ensuring that the data are internally consistent and honor rigorous mass and mole balances and equilibrium constraints.
This work has not yet been published.
Nonlinear regression is often used to find values for the parameters that describe an analytical or numerical model that give the best "fit" between measured and predicted pressure responses. The best fit is typically defined in terms of minimizing a least-squares objective function. For nonlinear regression to be useful, the parameters of the model function must provide accurate estimates of reservoir properties. If the underlying assumptions are satisfied, confidence intervals may be used to quantify the accuracy of the parameter estimates obtained through nonlinear regression.
Although the limitations of confidence intervals are well-understood and have been recognized in many other industries, in the field of pressure transient analysis, naive application of confidence intervals has given rise to gross misconceptions and misapplications. A number of these misconceptions have appeared in industry publications and to date not been here-to-fore adequately challenged.
This work addresses two vital questions that have have not been addressed in the well testing literature: 1) What are the effects of violations of the standard assumptions on the resulting confidence limits? 2) Can the validity of the standard assumptions be verified for a particular field data set, and if so, how?
This work has not yet been published.
Most books in the field of well testing have focused on the mathematical problems of defining new reservoir geometries, developing new reservoir models, and deriving new solutions for the diffusivity equation.
///*** UNDER CONSTRUCTION ***///
This project culminated in the publication of a new textbook by the Society of Petroleum Engineers:
The standard oilfield correlations for estimating brine density, formation volume factor, and solution gas-water ratio cover only a limited range of pressures and temperatures. Most of these correlations were developed decades ago and are based on limited experimental data. There are a lot of experimental data on brine density and solubility in the general literature that had never been incorporated into correlations for the properties of interest in the petroleum industry prior to this study.
Using a comprehensive database of brine PVT data compiled from the general literature, we developed new correlations for brine density, specific volume, coefficient of isothermal compressibility, and methane solubility. The new correlations fit the available experimental data within measurement accuracy over the range of pressures from 0 to 200 MPa (0 to 29,000 psi), temperatures from 0 to 300oC (32 to 572oF), and NaCl content from 0 to 5.7 moles/kg H2O (0 to 25 weight% NaCl).
We also developed equations for calculating standard oilfield PVT properties for brines with dissolved methane, including brine density, water formation volume factor, solution gas-water ratio, and coefficient of isothermal compressibility from the new correlations.
To make the correlations useful to the client, we implemented the new correlations in a C++ class library.
This work was published in a technical paper; an updated version was later published in a reference book.
A major oil and gas operator had a number of wells producing from a multilayer, low-permeability gas reservoir. Wells were completed with 10 to 20 frac stages, over a 7,000 foot gross interval. Because of the wide range of depths (and the accompanying wide range of pressures), the operator had used a variety of different proppant types of different strengths, in their frac designs. To optimize future frac treatments in this area, the operator began a study of the effect of proppant type on fracture treatment effectiveness.
Our role in this study was to estimate effective fracture half-length, in-situ permeability, and drainage area for each frac stage by history matching production and production log data. Each well had from one to three years of daily production, and from two to five production log runs.
We also performed statistical analysis of frac half-length results to determine whether differences in frac length for different proppant types were statistically meaningful.
This work was presented in two technical papers:
Gas storage wells often experience a decline in productivity and injectivity over time. The causes of this decline are not well understood. Does the damage accumulate gradually over time? Or can the onset of damage be correlated with specific events, such as switchover from withdrawal to injection operations?
We developed a module for processing high frequency electronic flow measurement (EFM) data from gas storage wells to identify shutin periods, resulting from normal storage operations, that might be analyzed as pressure transient tests. We also developed a module to analyze pressure transient test data quickly and easily, and to graph the results in a variety of ways to assist the operator in understanding how damage accumulates over time.
This work was discussed in a technical paper:
Copyright © 2006-2017, Phoenix Reservoir Engineering. All rights reserved.
Last updated July 24, 2017.