One of the key duties of a reservoir engineer is extracting information and knowledge about flow patterns and reservoir architecture. This paper will address the combination of empirical statistical analysis techniques with conventional field surveillance analysis and analytical forecasts to determine reservoir fluid flows. Generating these mental reservoir models will greatly assist in determining upside potential for oil fields.
Analyzing production and injection data to formulate production forecasts and to understand oil field mechanics is a critical skill that petroleum engineers need. Typically, reservoir engineers use decline analysis and analytical models to understand base case forecasts. Engineers commonly use analytical and analogy techniques to determine upside potential on a field wide basis. To forecast upside potential we need to understand what the remaining oil in place (ROIP) is, where this ROIP is located, and the conditions that the ROIP is in (pressures, saturations, etc). In order to formulate an answer to the above questions it is crucial to understand individual producer-injector connections.
The methodology explored in this paper uses empirical statistical analysis as well as common waterflood surveillance techniques and analytical methods to develop a mental model of fluid flow in reservoirs. All cases used are field cases.
Empirical statistical approaches, although innovative in their inception, have to be tempered with classical reservoir surveillance multiphase techniques to yield practical answers and eliminate false correlations.
Some of the key conclusions are;
Purely empirical statistical approaches such as Spearman ranking are very useful in identifying possible flow paths.
The degree of connectivity or communication is a function of waterflood maturity and primary depletion. Therefore classical analytical techniques should be used to determine when statistical approaches are used.
Analogy or analytical models should be applied to check the physicality of the connectivity and more definitely identify flow paths.
Using production data to identify trends enhances the flow simulation understanding.
The paper's objective is to gain preliminary understanding of communication amongst wells, using only production data and translate that communication knowledge into geology and reservoir engineering concepts.
Reservoir surveillance on existing waterfloods and any EOR injection process is key to optimizing oil recovery and determining production potential. Geological studies alone cannot conclusively quantify the reserve and oil rate increases that can be achieved by optimizing existing waterfloods or EOR injection process. The usual way to incorporate geological knowledge in understanding reservoir flow is flow simulation.
But before flow simulation the authors strongly recommend to do surveillance techniques to build mental models of the reservoir flow. There are six basic interrelated questions must be addressed in any EOR/waterflow surveillance and monitoring programs;
What is the original oil in place (OOIP)?
Where is the current OIP?
What is the remaining oil in place (ROIP)?
How is it distributed and what condition is it in?
What are the factors that limit recovery?
How can oil recovery be improved/ what is the production potential?
Can oil rates and reserves be economically increased?
Previous papers and talks have focused on determining the quantity of ROIP at a field or unit basis.
