Detection of production and well events is crucial for planning of production and operational strategies. Event detection is especially challenging in mature fields in which various off-normal events might occur simultaneously. Manual detection of these events by an engineer is a tedious task and prone to errors. On the other hand, abundance of data in mature fields provides an opportunity to employ data-driven methods for an accurate and robust production event detection. In this study a data-driven workflow to automatically detect production events based on signatures of events provided by experts is demonstrated. In the developed workflow, state-of-the-art data-driven methods were integrated with the domain knowledge for an accurate and robust detection. The methodology was applied on several case studies of mature fields suffering from production issues, such as scaling and liquid loading. It was found that the workflow is accurate, robust and computationally efficient which could detect new events (verified by the expert). The demonstrated method could be implemented both in the real-time or offline fashion. Such a workflow is sufficiently generic which can be applied for detection of different events and anomalies than tested and verified in this paper, such as leakage, production losses, …

Reliable forecasting of production rates from mature hydrocarbon fields is crucial both in optimizing their operation (via short-term forecasts) and in making reliable reserves estimations (via long-term forecasts). Several approaches may be employed for production forecasting from the industry standard decline curve analysis, to new technologies such as machine learning. The goal of this study is to assess the potential of utilizing deep learning and hybrid modelling approaches for production rate forecasting. Several methods were developed and assessed for both short-term and long-term forecasts, such as: first-principle physics-based approaches, decline curve analysis, deep learning models and hybrid models (which combine first-principle and deep learning models). These methods were tested on data from a variety of gas assets for different forecasting horizons, ranging from 6 weeks to several years. The results suggest that each model can be beneficial for production forecasting, depending on the complexity of the production behavior, the forecasting horizon and the availability and accuracy of the data used. The performances of both hybrid and physical models were dependent on the quality of the calibration (history matching) of the models employed. Deep learning models were found to be more accurate in capturing the dynamic effects observed during production – this was especially true for mature fields with frequent shut-ins and interventions. For long-term production forecasting, in some cases, the hybrid model produced a greater accuracy due to its consideration of the long-term reservoir depletion process provided by the incorporated material balance model.

No abstract