Integrated Reservoir-Surface Facilities Network Analysis

Abstract

Abstract Accuracy in reservoir performance prediction is a key in successful oil field management. There are methods available for this purpose including a reservoir simulation technique! This technique is normally used for its comprehensiveness and it's ability to design the desired scenarios to exploit the hydrocarbon resources. Modeling of a massive reservoir (average gross thickness exceeds 100 ft) using current reservoir simulation technology is quite attractive. However, modeling multi-layer reservoirs with more than 100 reservoir layers, such as reservoirs in deltaic environments, is prohibitive. This is due to the required cells to model these reservoirs, perhaps beyond the computer memory capacity. In this case, reservoir modeling is limited to a number of reservoirs. Other reservoir performance predictions are obtained by using an analogy method. Another technique to handle a multi-layer reservoir system is to use the Decline Curve Analysis. This paper proposes a method to predict reservoir performance in a multi-layer system where the reservoir simulation technique is impractical. This method integrates reservoir performance and surface facilities network analysis to obtain the field's production rate. The production rate is calculated using an appropriate IPR curve. The curve is validated using existing production data. Using this technique, any surface facilities performance changes affect the reservoir performance. As a result, the predicted flow rate in the determined point (e.g. gathering station) may be calculated immediately. The network analysis calculates pressure or flow rate of any nodes in the network. The algorithm used in this calculation is very simple and may be applied to several pressure traverse correlations such as Hagedorn-Brown, Beggs and Brill, Azis-Govier and Mukherjee and Brill methods. These correlations may be validated using flowing pressure measurements. A complex network is solved by separating independent hybrid pipelines from the whole network and solving them individually without violating the network pressure loss constraints. Using this technique, the nodes pressure or flow rate is calculated effectively. Since this technique requires no matrix computation, it is independent of any initial guesses normally used in current methodologies. This technique may be regarded as a reservoir-network simulator.