Thermal Profiles in Water Injection Wells: Reduction in the Systematic Error of Flow Measurements during the Transient Regime.

This article presents an analytical solution for calculating the flow rate in water injection wells based on the established thermal profile along the tubing. The intent is to minimize the intrinsic systematic error of classic quasi-static methodologies, which assume that all thermal transience on well completion has passed. When these techniques are applied during the initial hours of injection well operation, it can result in errors higher than 20%. To solve this limitation, the first law of thermodynamics was used to define a mathematical model and a thermal profile was established in the injection fluid, captured by using distributed temperature systems (DTSs) installed inside the tubing. The geothermal profile was also established naturally by a thermal source in the earth to determine the thermal gradient. A computational simulation of the injection well was developed to validate the mathematical solution. The simulation intended to generate the fluid's thermal profile, for which data were not available for the desired time period. As a result, at the cost of greater complexity, the systematic error dropped to values below 1% in the first two hours of well operation, as seen throughout this document. The code was developed in Phyton, version 1.7.0., from Anaconda Navigator.

Diagnosis of Operating Conditions of the Electrical Submersible Pump via Machine Learning.

In wells that operate by electrical submersible pump (ESP), the use of automation tools becomes essential in the interpretation of data. However, the fact that the wells work with automated systems does not guarantee the early diagnosis of operating conditions. The analysis of amperimetric charts is one of the ways to identify fail conditions. Generally, the analysis of these histographics is performed by operators who are often overloaded, generating a decrease in the efficiency of observing the well operating conditions. Currently, technologies based on machine learning (ML) algorithms create solutions to early diagnose abnormalities in the well's operation. Thus, this work aims to provide a proposal for detecting the operating conditions of the ESP pump from electrical current data from 24 wells in the city of Mossoró, Rio Grande do Norte state, Brazil. The algorithms used were Decision Tree, Support Vector Machine, K-Nearest Neighbor and Neural Network. The algorithms were tested without and with hyperparameter tuning based on a training dataset. The results confirm that the application of the ML algorithm is feasible for classifying the operating conditions of the ESP pump, as all had an accuracy greater than 87%, with the best result being the application of the SVM model, which reached an accuracy of 93%.

Diagnostic of Operation Conditions and Sensor Faults Using Machine Learning in Sucker-Rod Pumping Wells.

In sucker-rod pumping wells, due to the lack of an early diagnosis of operating condition or sensor faults, several problems can go unnoticed. These problems can increase downtime and production loss. In these wells, the diagnosis of operation conditions is carried out through downhole dynamometer cards, via pre-established patterns, with human visual effort in the operation centers. Starting with machine learning algorithms, several papers have been published on the subject, but it is still common to have doubts concerning the difficulty level of the dynamometer card classification task and best practices for solving the problem. In the search for answers to these questions, this work carried out sixty tests with more than 50,000 dynamometer cards from 38 wells in the Mossoró, RN, Brazil. In addition, it presented test results for three algorithms (decision tree, random forest and XGBoost), three descriptors (Fourier, wavelet and card load values), as well as pipelines provided by automated machine learning. Tests with and without the tuning of hypermeters, different levels of dataset balancing and various evaluation metrics were evaluated. The research shows that it is possible to detect sensor failures from dynamometer cards. Of the results that will be presented, 75% of the tests had an accuracy above 92% and the maximum accuracy was 99.84%.

Study of Flow Rate Measurements Derived from Temperature Profiles of an Emulated Well by a Laboratory Prototype.

The objective of this work is the study of the procedure for flow rate measurement derived from temperature profiles. This method is deemed appropriate because of the inability to mount conventional flow meters in multiple-zone oil wells. In this work, a reduced-scale prototype well with horizontal geometry was developed to study and validate a method of flow profiling by temperature measurements in the well column based on a heat transfer mathematical model studied by Ramey in 1962. Flow sensors were installed at key points to provide validation data for the flow measurements derived from the temperature. The plant was automated and all the tests were managed from a workstation. It was possible to test different situations to provide a variability of evaluation scenarios. The initial experiments used injected fluid flow rates of 15 and 18 L/min in the well inlet. The results of the calculated flow values in different work conditions were compared with a relatively low error reference meter.

Flow Rates Measurement and Uncertainty Analysis in Multiple-Zone Water-Injection Wells from Fluid Temperature Profiles.

This work is a contribution to the development of flow sensors in the oil and gas industry. It presents a methodology to measure the flow rates into multiple-zone water-injection wells from fluid temperature profiles and estimate the measurement uncertainty. First, a method to iteratively calculate the zonal flow rates using the Ramey (exponential) model was described. Next, this model was linearized to perform an uncertainty analysis. Then, a computer program to calculate the injected flow rates from experimental temperature profiles was developed. In the experimental part, a fluid temperature profile from a dual-zone water-injection well located in the Northeast Brazilian region was collected. Thus, calculated and measured flow rates were compared. The results proved that linearization error is negligible for practical purposes and the relative uncertainty increases as the flow rate decreases. The calculated values from both the Ramey and linear models were very close to the measured flow rates, presenting a difference of only 4.58 m³/d and 2.38 m³/d, respectively. Finally, the measurement uncertainties from the Ramey and linear models were equal to 1.22% and 1.40% (for injection zone 1); 10.47% and 9.88% (for injection zone 2). Therefore, the methodology was successfully validated and all objectives of this work were achieved.