Summary
The use of 3D streamline methodologies as an alternative to finite-difference (FD) simulation has become more common in the oil industry during the past few years. When the assumptions for its application are satisfied, results from streamline simulation compare very well with those from FD and typically require less than 10% of the central processing unit (CPU) resources. The speed of 3D streamline simulation (3DSM) lends itself not just to simulation, but also to other components of the reservoir simulation work process. This characteristic is particularly true of history matching.
History matching is frequently the most tedious and time-consuming part of a reservoir simulation study. In this paper, we describe a novel method that uses 3D streamline paths to assist in history matching either 3D streamline or FD models. We designated this technique Assisted History Matching (AHM) to distinguish it from automated history-matching techniques. In this manuscript, we describe this technique and its application to three reservoir simulation studies. The example models range in size from 105 to 106 gridblocks and contain as many as several hundred wells. These applications have led to refinements of the AHM methodology, the incorporation of several new algorithms, and some insights into the processes typically employed in history matching.
Introduction
The advent of powerful geostatistical modeling techniques has led to the development of very large (>107 cells) geocellular reservoir models. These models capture, in greater detail than before, the heterogeneity in porosity, permeability, and lithology that is critical to accurate simulation of reservoir performance. Three-dimensional streamline simulation has received considerable attention over the past several years because of its potential as an alternative to traditional FD methods for the simulation of these very large models.
While 3DSM is a powerful simulation tool, it also has a number of other uses. The speed of 3DSM is ideal for such applications as geologic/geostatistical model screening,1 reservoir scoping, and history matching (the focus of this paper). In this manuscript, we describe the technique and present three example reservoir applications that demonstrate its utility.
The AHM Technique
The models used in reservoir simulation today contain details of structure and heterogeneity that are orders of magnitude greater than those used just 10 years ago. However, there is still (and probably always will be) a large degree of uncertainty in the property descriptions. Geologic data are typically scattered and imprecise. Laboratory measurements of core properties, for example, often show an order of magnitude variation in permeability for any given porosity and several orders of magnitude variation over the data set. Upscaling replaces geologic detail with estimates of effective properties for aggregated data, placing another level of approximation on the resulting model. It is unlikely that any geologic model will match the observed reservoir performance perfectly, and history matching continues to be the technique by which the adjustments are made to the geologic model to achieve a match between model and historical reservoir performance. Ref. 2 provides a good presentation of traditional history-matching techniques.
History matching by definition is an ill-posed problem: there are more unknowns than there are constraints to the problem. Indeed, any reservoir simulation engineer knows that there is always more than one way to history match a given reservoir model. It is the responsibility of the simulation engineer to make only those changes that are consistent with the reservoir geology. AHM was designed to facilitate these changes.
As defined here, AHM is different from automated history matching and traditional history-matching techniques. Generically, traditional history matching involves five key steps:Simulation and identification of the difference between model predictions and observed performance.Determination of the gridblocks in the model that require change.Designation of the property(ies) that requires change and what those changes are.Implementation of the changes in the simulation input data.Iteration on the above steps until a satisfactory match is achieved.
The two principal uncertainties in this process lie in Steps 2 and 3, both of which are empirical and tedious and frequently involve ad hoc decisions that have an unknown impact on the ultimate results. AHM is designed to simplify this process and uses 3DSM to facilitate Steps 2 and 3 and thus minimize the ad hoc nature of the process. AHM uses an underlying 3DSM model to determine the streamline paths in the reservoir. These streamlines describe the principal flow paths in the model and represent the paths along which the fluids in the model flow from source (injector or aquifer) to sink (producer). By tracing all the streamlines from a given well, the gridblocks through which the fluids flow to that well are identified. This process, in essence, replaces Step 2 with a process that is rooted in the fluid-flow calculation. Once these gridblocks are identified, changes can be performed according to any (geologically reasonable) algorithm desired. Here, a simple program that largely replaces Step 4 carries this out. Fig. 1 illustrates the concept.
The AHM process is based on the assumption that history matching is achieved by altering the geologic properties along the flow paths connecting a producing well to its flow source. The source may be a water injector, gas injector, aquifer, or gas cap; however, the drive mechanism must be a displacement along a definable path. Because the technique relies upon identification of the flow paths, it is assumed that the grid is sufficiently detailed to resolve the flow paths. In very coarse grids, a single gridblock may intersect the flow to several wells, and satisfactory history matching in this case may not be possible with AHM.
For streamline-simulation models, the calculation model provides the path directly. For FD simulation, a streamline model incorporating the same structure and geologic parameters as the simulation model is used to calculate the streamlines defining the flow paths.
Data-Driven Dispatching Rules Mining and Real-Time Decision-Making Methodology in Intelligent Manufacturing Shop Floor with Uncertainty