The importance of gas production has increased as gas represents a clean source of energy. We studied different multiphase flow correlations for gas wells. We collected large database for bottomhole flowing pressure for different flow conditions and well configurations. In total, 32 gas wells were selected and our target was to study the effect of multiphase flow correlations input parameters on the accuracy of the predicted pressure drop. Several important multiphase correlations input parameters were selected for this study. These include condensate to gas ratio (CGR) and water to gas ratio (WGR) which represent the production conditions, API and specific gravity of surface gas (Ɣg) which represent PVT properties and the tubing roughness (ε) which represents the tubing condition. Our method was based on changing the values of these selected parameters by a percentage from its original value and determining the new predicted bottomhole flowing pressure. Consequently, we determined the new error compared to the actual measured bottomhole pressure. We performed 352 cases, and we could obtain the effect of the different parameters on both pressure drop calculations and the selection of the best correlation. Guidelines were developed to explain which parameters are more important to be measured accurately for different conditions.

Our target was to develop an expert system to help petroleum engineers in selecting the most suitable multiphase flow correlation in the absence of measured flowing pressure. A large database of pressure points was collected and analyzed using many multiphase flow correlations. The expert system was developed with a set of rules to identify the best correlation for variety of well, flow, and PVT conditions. The expert system is based on the idea of clustering the data and finding the best multiphase flow correlation(s) for each cluster. The error associated with the selected correlation is also quantified for every correlation in each data sub-cluster so the engineer would expect the accuracy of pressure drop prediction when utilizing this approach. Over the entire database, if one multiphase flow correlation is selected, the overall mean absolute percent error ranges between 12.7 and 57.5%, while the range of errors for best correlation(s) in different data sub-clusters range from 0.01 to 3% for most cases with accurate PVT. The expert system was validated by a new set of data. It succeeded in identifying the best correlation(s) 70% of the times, and the calculated pressure was more accurate than using one correlation by a factor of 2. Use of the expert system in the validation database gave a mean absolute percent error of 8.8%. This represents approximately one-third of the error value when any single correlation is used over the entire validation dataset. (Error for using single correlation ranges from more than 21 to 29%.) Keywords

In absence of downhole permanent gauges, multiphase flow correlations are widely used to predict well flowing pressures and production under variety of conditions. They also have wide applications in nodal analysis and artificial lift optimization. In our previous work (SPE-175805), we collected a large database of more than 3000 multiphase pressure points and tests. We used this large database to develop guidelines to select the most accurate correlations for different pipe and flow conditions. Typically, there are close to 20 input parameters used in the calculations of multiphase flow correlations. In this study, we used sensitivity analysis techniques with our large database to determine which input parameters affect the multiphase flow correlations results for different pipe and flow conditions. Among these important input parameters are producing gas-oil ratio (GOR), API, gas specific gravity and tubing roughness. For each of the important input parameters, we change the value of the input parameter within the measurement accuracy in the field and find out their effect on the calculated pressure in comparison with the measured pressure. We categorized the large database into clusters representing different pipe and flow conditions, then used a commercial pipe flow software to model these wells. We then performed hundreds of sensitivity studies on almost all input parameters for oil wells. It was observed during the study that several parameters have significant effect on multiphase correlations results. These effects were different for low and high production rate wells and low and high GOR wells. For example, accuracy of API was found to have low effect on results. However, GOR had significant effect on results. Pipe roughness was found to be much less important in oil wells especially at low production rates and at low GOR. Based on the results of the sensitivity studies for the large database, we developed guidelines for which parameters are important for which wells and under what conditions. The guidelines will help engineers working on developing multiphase flow models for wells focus on obtaining good values for the most important input parameters.

Oil and gas production represents an essential source of energy. Optimization of oil and gas production systems requires accurate calculation of pressure drop in tubing and flowlines. Many empirical correlations and mechanistic models exist to calculate pressure drop in tubing and flowlines. Previous work has shown that some correlations provide more accurate results under certain flow conditions, PVT data, and well configurations than others. However, the effects of errors in input data on the selection of which correlations to use have not been investigated. This paper studies different multiphase flow correlations to determine the effects of their input parameters on (1) the accuracy of calculated pressure drop and (2) the selection of best correlation. A database consisting of 33 oil wells and 32 gas wells was selected, and a commercial software was used to build different well models. A total of 715 well models were constructed and used to investigate the effects of errors in correlations inputs on both the calculated bottomhole pressure and the selection of best correlation(s). The methodology was based on perturbing the values of the selected input parameters and calculating the new predicted bottomhole flowing pressure. Then, the effects of error in input parameters on how the calculated bottomhole pressure was different from observed data were quantified. The effect of this error in input parameters was also checked against the algorithm that selects the best correlation(s). It was found that errors in input GOR have the greatest effects for oil wells, while gas specific gravity and the tubing roughness are the most effective parameters for gas wells. The results were integrated into a rule-based expert system. A new set of data, consisting of 220 cases from 10 new oil wells and 10 new gas wells, was used to validate the expert system. The expert system was found to predict the best correlation(s) with a success rate of 80%, and it also identifies the input parameters whose error would affect the value of calculated bottomhole pressure significantly. Finally, the rules of the expert system were programmed into a VBA-Code to ease its use.

No abstract