Abstract
This paper aims to estimate surge pressure resulting from the flow of a Herschel-Bulkley fluid through concentric annuli during tripping operations. A semi-analytical model which includes new constraints to avoid non-physical solutions was developed then solved numerically. Moreover, a numerical model was implemented using finite difference method for which accurate solutions were obtained. Besides, the commercial software (Ansys-Fluent 19R3) was employed. The results are compared with existing experimental data from literature where a good agreement is observed with a maximum average relative error of 4% and 7.8% for the two studied drilling fluids, respectively. These models were successfully extended to power-law fluids. On the other hand, numerical results were collected by varying the relevant drilling parameters. The outputs indicated that the increase in the pipe tripping velocity causes an increase of the surge pressure independently of the parameter of interest, however, the rate of increase changes from one parameter to another. Based on this parametric study, a surrogate model using the Random Forest algorithm is constructed. This model predicts surge and swab pressures without requiring cumbersome numerical calculations. The model predictions have shown a satisfactory matching with an R2 equals to 0.99 for both training and test data.
Original language | English |
---|---|
Pages (from-to) | 243-264 |
Number of pages | 22 |
Journal | Petroleum and Coal |
Volume | 64 |
Issue number | 2 |
Publication status | Published - 2022 |
Externally published | Yes |
Keywords
- Concentric annuli
- Numerical model
- Semi-analytical model
- Surge and swab pressures
- Surrogate model and machine learning
- Yield power law (ypl) and power law (pl)