Method for Assessing Production Strategy Plans |
|||||||
申请号 | EP13382215.5 | 申请日 | 2013-06-06 | 公开(公告)号 | EP2811108B1 | 公开(公告)日 | 2017-11-22 |
申请人 | Repsol, S.A.; International Business Machines Corporation; | 发明人 | Embid Droz, Sonia; Rodriguez Torrado, Rubén; Hegazy, Mohamed; Echevarria Ciaurri, David; Mello, Ulisses; | ||||
摘要 | The invention relates to a method that generates well location plans and field development plans assessing and ranking the potential of the different plans with a small number of parameters or initial conditions, thus considerably reducing the decision time for taking a particular strategy when compared with the techniques described in the art. | ||||||
权利要求 | |||||||
说明书全文 | The invention relates to a method that generates production strategy plans and field development plans assessing and ranking the potential of the different plans with a small number of parameters or initial conditions, thus considerably reducing the decision time for taking a particular strategy when compared with the techniques described in the art. A typical state of the art hydrocarbon reservoir production strategy provides production decisions for a given planning horizon on a drilling schedule to maximize production. A typical planning horizon may locate production and injection wells. The drilling schedule may indicate which wells are to be drilled and when, and the production rate at which the wells are to operate. Varying the position, schedule and/or control of each wells may vary production to have a multimillion-dollar impact. Thus, evaluating reservoir production potential and economic performance over a wide range of alternative oil and gas production strategies is crucial. Also, because there are a large number of variables in selecting the strategy, it has been a time-consuming activity. Frequently, available information is limited, and based on uncertain reservoir geological and petro-physical properties. Typically, major investment decisions must be made on this limited information, especially when subterranean-flooding (e.g., with water) is the main production strategy. Previously, experienced reservoir engineers heuristically defined and ranked complex production development plans, using trial-and-error to deal with problem components, separately and sequentially. After selecting drilling locations, for example, engineers heuristically defined a well drilling schedule. However, these ad hoc heuristic solutions, frequently have application only within the limited framework for which they were developed. Additionally, rather than arriving at the best overall or optimum realization, separately and sequentially dealing components may have optimum realization, separately and sequentially dealing components may have discarded the most attractive or optimal solutions. Thus, there is a need for dramatically reducing the number of drilling configurations that must be considered for a comprehensive reservoir production strategy and more particularly for rapidly generating and ranking several representative development plans under uncertainty to quickly converge on plans that jointly encompass all available production aspects. The present invention solves the problems identified above by providing method of generating production strategies suitable for the exploitation of a reservoir of hydrocarbon in a natural environment, wherein said natural environment is limited in its surface by a domain (Ω). It should be noted that if hydrocarbon reservoir is surrounded by water then the domain (Ω) may be defined in such a way the hydrocarbon reservoir is completely located within the domain (Ω). The method comprises the following steps:
The publication "Exploitation Plan Design Based on Opportunity Index Analysis in Numerical Simulations Models", A. Molina et. al., Society of Petroleum Engineers, SPE, discloses Opportunity Index Analysis as an intelligent method that uses numerical simulations models to identify zones with high potential of production. Determining the location of production wells as the production strategy according to this method reduces the number of potential development plans in a certain domain (Ω) necessary to achieve an accurate reservoir simulation. For that purpose, an opportunity index is defined as a function that quantifies, for every location, the hydrocarbon production potential taking into account the local properties of every location - for example, as a function proportional to the amount of hydrocarbon trapped in that location and inversely proportional to the ability of hydrocarbon to flow thorough the rocks in that location. The information about the local properties of every location retrieved from the collected data may be obtained by averaging a set of geological models, named as "reservoir realizations" taking into account the uncertainty wherein each model may be simulated using CFD (Computational Fluid Dynamic) codes. Departing from deterministic data, tools like interpolation, Design of Experiments and others, provide a set of reservoir realizations taking into account the uncertainty. Statistic variables as average values or dispersion measures may be evaluated over the whole set of reservoir realizations. In the particular case of the opportunity index, the value taken in a predetermined location is the average measured on the whole set of reservoir realizations computed by means of simulations. The locations with a similar opportunity index are grouped in zones called clusters, each one thus having similar behavior in terms of hydrocarbon production potential. By 'locations with similar OI', it must be understood locations whose OI is within a certain range of values. It is possible to provide only one cluster if the whole domain has a similar behavior in terms of hydrocarbon production potential. A new function, the radius of drainage, is determined from the OI and, in some cases, also as a function of other variables to be explained in the detailed description of the invention. The radius of drainage is a measure of the optimal radius of extraction for every well, since it provides for each well the radius of extraction at the end of the life of extraction under ideal conditions in such a way that the circumference determined by such radius with center in every well are tangent one each other. This association between a function of the potentiality of a location for hydrocarbon production and the radius of extraction of every well is an advantageous way of generating a well location plan with a number of parameters low enough as to not needing important computational resources. For every cluster, a representative value of the OI is taken (for example, the arithmetic mean of the extremes of the range that defines a cluster) and the subsequent rd is then calculated. The number of parameters to generate a well location plan is relatively small, and therefore the every plan can be quickly obtained from a set of parameters. Some of these parameters, as has been stated, relate the OI and the rd, as a function of the local properties. Others define a reference system and the planned well location with reference to this system such as the location of a first point of each patterned grid and the angle of a reference line of such grid. Once the reference system is defined, since the radiuses of drainage are tangent one each other, the well must be located at a distance of 2rd to comply with this condition. Keeping this condition, and starting at the origin of the first point, the grid is generated to discretize the cluster - the nodes are the possible location of the wells themselves, separated as has been said a distance 2rd. The grid is therefore a patterned grid, and its orientation with respect to the reference system is given by an angle α and is one of the parameters of the well location plan taking a reference line of the grid. Once this angle α is provided, the well location is given by the position of the nodes of the patterned grid with respect to the reference system. These and other features and advantages of the invention will be seen more clearly from the following detailed description of a preferred embodiment provided only by way of illustrative and non-limiting example in reference to the attached drawings.
The present invention proposes a well location plan which is advantageous with respect to those of the state of the art since it provides accurate heuristic solutions that need less design parameters and therefore a less complex (and less time-consuming) forecast. The opportunity index defines the hydrocarbon production potential of certain location of a domain (Ω), and then the radius of drainage rd for such location is defined as a function of the OI so that the higher the OI the higher the rd. In a particular embodiment of the invention, the relation between OI and rd is rd=a*OIb. a and b, positive constants based on local properties of every cluster, are, in this example, two of the parameters used to calculate the potential well location plans. In a further particular embodiment, the well location plan is controlled by five parameters per cluster, a number small enough as to allow that the domain (Ω) can be explored by means of experimental design techniques in a relatively exhaustive manner in a matter of a few hours. Apart from a and b, in a particular embodiment other parameter are space parameters referred to the reference system (the coordinates of the first point of the patterned grid i and j and the aforementioned angle α of every cluster). As it is shown in The domain (Ω) is divided in clusters (C1, C2), as in As it is shown in Once the different clusters, with its OI and rd, have been defined, the location of the production wells (P) for this particular plan (calculated with a particular set of initial parameters) is obtained for each cluster defining a patterned grid in which the distance between closest nodes is twice the radius of drainage (rd) - this can be seen in the examples of Since a reduced number of parameters characterizes every well location plan, and each plan provides a good proposal for the exploitation of the reservoir, a much smaller number of well location plans suffices as opposed to prior approaches that very often required several thousands of plans or more. As a result, a reduced number of computational flow simulations are required reducing the total computational effort. As for selecting the set of parameters (five per cluster in the particular example) that gives as a result a particular well location (P) plan, in a particular example the technique known in the art as Design of Experiments is used. Each set of parameters determine a well location plan. The use of Design of Experiments provides a plurality of different plans according to the disclosed method. Some development plans comprise, aside from production wells (P), injection wells (I) through which water is added to sweep the different regions of the domain (Ω). The injection wells (I) are, in one particular example, placed at the centroid of the pattern of the grid, for instance the square pattern formed by every four neighbor nodes, that set the locations of the production wells (P) for a certain cluster, as can be seen in Alternatively, if the reservoir is susceptible to peripheral injection, injectors (I) can be located at a strip-shaped region (S) extending along a boundary (Γ) of the interface between water (W) and hydrocarbon (O) phases of the reservoir and located in the water side of the interface, as shown in As in the case of the 01, the locations of the strip-shaped region (S) with II within a determined range of values, which is to say locations with a relatively similar behavior, are grouped in injection clusters (S1, S2, S3) in the strip-shaped region (S). A II representative for each injection cluster (S1, S2, S3) is taken, for instance the average value of the II in such cluster. Likewise, a radius of injection (ri) is calculated from the II for every injection cluster (S1, S2, S3), so that the higher the II the higher the rd, that is, the bigger surface that a single injector (I1, I2, I3) of said cluster (S1, S2, S3) is able to sweep. In a particular example, the ri is expressed as ri=c*IId wherein c, d are positive constants depending on the local properties for each injection cluster. The spacing between consecutive injectors (I1, I2, I3) in the strip-shaped region (S), starting from a first injection well location (I1) of the strip-shaped region (S), is calculated as twice the radius of injection (ri1, ri2, ri3) of the injection cluster (S1, S2, S3) where the injection well (I1, I2, I3) is, as can be seen in In a particular example, this generation of injection wells (I1, I2, I3) is continued this way until all the clusters in the strip-shaped regions are exhausted or until the first injection well (I1) is reached (when the strip-shaped region (S) is a close region). In a further example, the strip-shaped region (S) is the width of a cell of the discretized domain (as the cells in In a further example, the width of the strip-shaped region (S) is a fraction of the distance between a neighbor producer (P) well and the center of its corresponding pattern. With respect to determining the first injection well (I1) of a strip-shaped region (S) for a peripheral injection, in one example this first location is calculated as shown in In a further example, the location of the first injection well (I1) of the strip-shaped region (S) for a peripheral injection is calculated determining the orthogonal projection of the production well (P) having higher opportunity index OI with respect to the boundary between the hydrocarbon (O) and the water (W). Similarly to the well location plan, other parameters are used to control the well drilling schedule. The drilling schedule comprises generating a list comprising the production wells (P); or both, production wells (P) and injection wells (I) wherein such list is sorted according to three criteria. In a particular example, three input parameters are used for this task and the list comprises both, production wells (P) and injection wells (I). In a further example, two of these parameters define the sequence of production and injection followed to complete the exploitation of the domain (Ω) (for example, a basic pattern of drilling two producers (P) followed by one injector (I), repeated until all wells (P, I) are considered), and the remaining parameter indicates the time interval between drilling two consecutive wells (assuming it is the same for all the drilling sequence). The order of drilling for both the production wells (P) and the injection wells (I) is predetermined according to different criteria. In a particular example, this criterion is as follows: the order is given by a list in which the wells (P, I) first in the list are those with higher index (OI and II), those closer to the outer boundary of the domain (Ω) or to the interfase boundary between hydrocarbon (O) and water (W) in the domain (Ω), or those having a lesser average distance with precedent or antecedent wells (P, I). With this criterion, there is an adequate choice in the exploitation of the wells (P, I), since the first to be drilled are the ones with more oil potential, the ones more easily reachable from the boundary, and the ones closer to each other. The three conditions can be taken into account at the same time, if weights are given to each one of them. For the particular example in which both the well location plan and the well drilling location are taken into account, the n parameters (eight in this particular example) are selected by means of a technique such as the Design of Experiment to obtain a certain well location plan and drilling plan. In a further example, well controls are also provided based on estimations of the average potential recovery factor of the reservoir, on usual injection procedures, on standard economic constraints, etc. The number of well location plans and drilling locations, that is, the number of development plans (N) estimated according to the method which is the main inventive aspect of the invention, each one with a set of (n) parameters (for example, eight), may then be ranked, to select the most appropriate options, according to techniques such as the net present value (NPV). The ranking measure is a measure averaged over all reservoir realizations, for instance those reservoir realizations used for the determination of the opportunity index. For example, if NPV is the ranking measure, for each field development plan the ranking measure is the average of all NPVs over all realizations. The computational cost for the evaluation of a development plan mainly depends on the computational cost of the flow simulation. In this case, the Design of Experiments only needs a reduced number of plans because each plan provides well distributions and drilling schedules selected in an efficient manner thanks to method of the main aspect of the present invention and therefore the Design of Experiments does not need to explore a large amount of well locations in order to reach the efficient ones. In the prior art, the well distribution is entrusted to the Design of Experiments therefore the number of proposals need to be large enough to obtain a reasonable result. Because each proposal requires a flow simulation the computational cost of the present invention is drastically reduced. For this example, the field development plan with the highest average NPV ranks highest or first. |