production simulator for the simulation of a mature hydrocarbon field

专利摘要:
PRODUCTION SIMULATOR FOR THE SIMULATION OF A MATURE HYDROCARBON FIELD. The present invention relates to a production simulator (2), for the simulation of a mature hydrocarbon field, which supplies the quantity produced per phase, per well, per layer (or group of layers) and per cap depending on parameters production (PP), and said production simulator (2) identifies historical data (HD) of the said mature hydrocarbon field and verifies a Vapnik condition.
公开号:BR112012032159B1
申请号:R112012032159-5
申请日:2011-06-15
公开日:2020-11-17
发明作者:Jean-Marc Oury；Bruno Heintz；Hugues de Saint- Germain
申请人:Foroil；
IPC主号:

专利说明:

[001] The technical field of the invention is the exploration of hydrocarbon fields. More particularly, the invention relates to a method of building a reliable simulator capable of predicting quantities produced vs. production parameters, in the particular case of mature fields.
[002] Mature hydrocarbon fields represent a special challenge both in terms of investment and in the allocation of human resources, because the net present value of any new investment decreases with the degree of maturity. Consequently, less and less time and effort can be invested in studies on reservoirs to support the exploration of fields. Still, opportunities remain to improve production on the so-called “standard” or “routine” behavior of an entire mature field, even with little investment. The strategic choices made in the way of operating hydrocarbon fields created a certain heterogeneity in pressure and saturation. They can be drastically reconsidered and the production parameters changed accordingly. With regard to mature hydrocarbon fields, many production routes have been explored in the past, and a learning process can be applied: refurbished parameters can be implemented with low risks.
[003] Two approaches to the state of the art are currently known to model the behavior of a hydrocarbon field, and to predict an expected quantity produced in response to a given set of production parameters.
[004] A first approach, called “mesh model” or “finite element modeling” breaks the reservoir into more than 100,000 elements (cells, flow lines, etc.), with each cell comprising different parameters (permeability, porosity, initial saturation, etc.), and applies laws of physics to each cell to model the behavior of fluids in the hydrocarbon field. In this case, the so-called Vapnik-Chervonenkis h-dimension of the solution space S, from which the simulator is selected, is very large. Consequently, the available number of m measured data in historical data remains relatively small, even for mature fields, and the ratio — appears to be much greater than 1. As a result of Vapnik's theory of learning, which will be mentioned again below, the expected risk R predicted is not properly defined (because of the term G), and such a simulator cannot be considered reliable, even though it presents a very large identification with historical data. In practice, it is widely recognized that, for such mesh models, good historical identification does not guarantee a good forecast: there are billions of ways to compare the past, which leaves great uncertainty about which one provides a good forecast.
[005] A second approach, on the contrary, uses over-simplified models, such as, for example, declining curves or material balance. However, these are too simplified to properly take into account the relevant physics and geology of the reservoir, particularly complex interactions and phenomena. In this case, the expected expected R risk is not minimized, as no good identification can be achieved (the term empirical risk Remp remains large).
[006] In summary, the invention represents an appropriate balance between very complex and very simplistic approaches. It is valid only for mature fields, which provide sufficient past information, in the form of historical HD data, to allow the construction of an S space of candidates to become field production simulators, without becoming too complex and requiring so that many historical data to be calibrated.
[007] Vapnik's theory of statistical learning defines under what conditions such a simulator can be conceived. Such a simulator can be designed to satisfy these conditions for reliable predictive capability.
[008] The object of the invention is a production simulator, for the simulation of a mature hydrocarbon field, which supplies the quantity produced per phase, per well, per layer (or group of layers) and for time depending on parameters of production, being that such a production simulator identifies sufficiently historical data of the said mature hydrocarbon field and verifies the condition of Vapnik. This Vapnik condition ensures that the calculated quantities to be produced per phase, per well, per layer (or group of layers) and over time are accurate. This allows simulator users to rehearse different production scenarios, according to different production parameters, each providing reliable quantities, so that the scenarios can be properly compared to each other and a suitable scenario can be selected according to specific criteria. . As all production scenarios will provide reliable quantities, the selected scenario will also provide a reliable production forecast, so it will be a low-risk and preferred path for field production.
[009] According to another characteristic of the invention, an identification with historical data is obtained when:
for p% ends of the existing values of | Q <* ftHD ll [r1-Xy, Ti], and the accumulated oil produced over the same time interval [Ti-Xy, Ti] is accurate up to 82, being that: Q are quantities produced per phase, per well, per layer (or group of layers) and by time, determined by the production simulator, Q ^ ihimsao the same quantities produced per phase, per well, per layer (or group of layers ) and by time, found in historical data, [Ti-Xy, Ti] is the time interval that comprises the most recent X years before time Ti, | z | ^ denotes a norm of Z along the interval of time [Ti, T2], d is a small positive number in relation to 1, and 82 is a small positive number in relation to 1, and p is a positive number close to 100%.
[0010] This feature practically defines the situation when a simulator can be considered a supplier of satisfactory historical identification, which is a necessary condition for being a candidate to demonstrate satisfactory forecasting skills.
[0011] According to another characteristic of the invention, X = 5, 8i = 0.2, and 82 = 0.15 and p = 90%.
[0012] This characteristic allows framing the conditions in which the historical identification can be considered accurate; it offers a degree of magnitude of accuracy, which can be expected for satisfactory predicted quantities.
[0013] According to another characteristic of the invention, the condition of h Vapnik is expressed as - <0.1, where: mh is the Vapnik-Chervonenkis dimension of the S space of solutions, from which the simulator is selected, em is the number of independent measurements available in historical data.
[0014] Such a characteristic allows framing the conditions, when they can be calculated, to achieve a satisfactory forecasting capacity.
[0015] According to another characteristic of the invention, the condition of Vapnik is verified by performing a satisfactory blind test over N years, defined by the following steps: determination of a production simulator that identifies historical data over a time interval before T-Ny, forecast of quantities produced over a time interval [T-Ny, T] given the production parameters over the same time interval [T-Ny, T], being the test satisfactory blind when:
for more than p% of the existing values, and the accumulated oil produced over the same time interval [T-Ny, T] is needed in up to 82, being that: Qçitbsã ° quantities produced per phase, per well, per layer (or group of layers) and by time, determined by the production simulator, Qg / abHD are the same quantities produced per phase, per well, per layer (or group of layers) and by time, found in historical data, [T- Ny, T] is the time interval comprising the most recent N years before time T, with T being the last date for which historical data are available, Wlrrrl L17 denotes a norm of X over the time interval [Ti , T2], £ I is a small positive number with respect to 1, and 82 is a small positive number with respect to 1, and p is a positive number close to 100%.
[0016] This characteristic allows defining a discriminating test that decides whether a simulator has a satisfactory forecasting capacity, without considering the way it was configured, but considering the real results that it can achieve. This feature allows the test to be performed without the need to wait for years and measure the results, but instead use data already available to the third party who conducts the test, and not available to the part that configures the simulator (the latter being “Blind”).
[0017] According to another characteristic of the invention, 8i = 0.1, and 82 = 0.1, p = 90% eN = 3.
[0018] This characteristic allows framing practical conditions for the blind test, which can be accepted for the appropriate discrimination between a satisfactory simulator and an unsatisfactory simulator.
[0019] According to another characteristic of the invention, the condition of Vapnik is a property of predicted stability verified when, if
where: PP are production parameters, PP 'are slightly different production parameters, QgiabHD are quantities produced per phase, per well, per layer (or group of layers) and by time, found in historical data, Q ^ HD are different quantities slightly variable produced per phase, per well, per layer (or group of layers) and time, found in different historical data slightly variable, Q <pktb are quantities produced per phase, per well, per layer (or group of layers) and by time, determined by the production simulator, Q ^ 'are quantities produced per phase, per well, per layer (or group of layers) and by time, determined by another production simulator sufficiently close determined from slightly slightly different historical data , [To, T] is the time interval from time To to time T, which are respectively the start and end date for which historical data are available, [T, T + Ny] is the time interval that co includes Ny years after time T, | z | ^ denotes a norm of Z over the time interval [Ti, Tz], ε is a small positive number in relation to 1, and n is a small integer less than that 5.
[0020] This feature allows to define whether a simulator has properties that are a prerequisite for the reliable calculation of predicted quantities. This test does not require comparison with another existing simulator, nor does it require knowledge of how this simulator was configured, since only its resulting properties are tested.
[0021] According to another characteristic of the invention, ε = 0.05, n = 2 and N = 3.
[0022] This characteristic allows to practically fit the expected stability, over a period of three years.
[0023] According to another characteristic of the invention, the production simulator is built following the steps of: detailed initial definition of the reservoir partition, rock properties, laws of reservoir physics and laws of well physics, expansion of said reservoir partition, rock properties, reservoir physics laws and well physics laws until said Vapnik condition is verified, and calibration of said production simulator by choosing, among the candidate solutions for production simulator, the solution candidate, defined by α and that minimizes an expected so-called “empirical risk” Remp (a).
[0024] Expansion means reducing complexity. The expansion is carried out in such a way that the space of candidate solutions to become one (h 1 simulator is such that 0 -, δ is the smallest possible, where: δ is a positive number close to zero, and 1 - δ defines a probability , and <J> is a positive function defined by:
where: h is the Vapnik-Chervonenkis dimension of the solution space, m is the number of independent measurements available in historical data, and is equal to exp (l).
[0025] The calibration of the production simulator among the candidates in the solution space means taking the “empirical risk” R ^ a) as small as possible, while choosing the appropriate α parameters that completely define a solution within the solution space. This empirical risk Remp (a) measures a distance (positive) between actual past data and the corresponding data calculated by the simulator.
[0026] This expansion process is a way to achieve a {h 1 balance between the values of
which must be kept together as small as possible, since the objective is to minimize
with a given probability 1 - δ.
[0027] By minimizing the sum mentioned above, the expected risk of the forecast, R (a), is minimized, according to Vapnik's inequality:

[0028] This feature minimizes the expected risk of the R (a) forecast, while starting from high values of
(since initial models are complex) and low empirical risks R ^ fa) (since such complex models can adequately apprehend past data) and gradually decreases the complexity, and consequently
while maintaining a good identification of past data, thus keeping the empirical risk R ^ fa) low.
[0029] According to another characteristic of the invention, the said reservoir partition is enlarged following the steps of: partition of a reservoir G into elementary parts Gab, AB so that G = UUGβbcom = 0 for (a, b) ψ (a ', b'), where aG {l..A} describes an area xy, and eb € {l..B} describes one or more layers ^ 5 grouping of adjacent elementary parts that have homogeneous rocky properties, in subgeologies Gc where cG {l..C}.
[0030] This feature provides a practical way to expand the geology of the field by identifying a reasonably small number of Gc subgeologies.
[0031] According to another characteristic of the invention, the rock properties are expanded following a step of calculating the average rock properties through each subgeology, according to the formula:
, where Vc is the volume of the Gc subgeology.
[0032] This feature explains a way to practically define the properties to be used in a given Gc subgeology.
[0033] According to another feature of the invention, the laws of physics dr »c rí» cí »rxzatAri / -» c cõr »am-nliadae HAta! manpira miA «p a-nlimipm r * nm functional parameters of the subgeology, and that the space and time scales associated with the subgeology are determined so that the associated solution space is congruent with the complexity of the historical data at the well level.
[0034] This characteristic provides the rules to be observed for the laws of the physics of the expanded reservoirs, by emphasizing their behavior measured at the well level.
[0035] According to another characteristic of the invention, the production simulator is built following the steps of: initial gross definition of the reservoir partition, rocky properties, laws of reservoir physics and laws of well physics, reduction of said reservoir partition, rock properties, reservoir physics laws and well physics laws while maintaining the verified Vapnik condition, until the production simulator identifies historical data, and calibration of the referred production simulator by choosing from among the candidate solutions for production simulator, the candidate solution that minimizes an expected expected risk.
[0036] Reduction means increasing complexity, starting from a simplistic description of the field, and adding relevant reservoir and / or well phenomena, which will properly apprehend the behavior of the entire field, well by well. The reduction is carried out in such a way that the space for candidate solutions to become a production fhj simulator remains such that
as small as possible, where: δ is a positive number close to zero, and 1 - δ defines a probability, and is a positive function defined by
where: h is the Vapnik-Chervonenkis dimension of the solution space, and m is the number of independent measurements available in historical data, and is equal to exp (l).
[0037] The calibration of the production simulator among the candidates of the solutions space means taking the “empirical risk” R „np (c () as small as possible, while choosing the appropriate“ a ”parameters that completely define a solution within the This empirical risk measures a distance between actual past data and the corresponding data calculated by the simulator.
[0038] This reduction process is a way to achieve a (h} balance between the values of
and ReJa), which must together be as small as possible, since the objective is to minimize their sum
with a given probability 1 - δ.
[0039] By minimizing the sum mentioned above, the expected risk of the forecast, R (a), is minimized, according to Vapnik's inequality:

[0040] Such a feature allows to minimize the expected risk of the fh forecast R (a), while starting from small values of
(since initial models are simplistic) and high empirical Remp risks (a) (since such a simplistic model does not adequately apprehend past data) and
in order to better identify past data, thus reducing the empirical risk Remp (cc).
[0041] According to another characteristic of the invention, said reservoir partition is reduced following the steps of: starting from the entire reservoir, partition of said reservoir into subgeologies, with a substantial change in the properties of the reservoir being present around the boundary between these subgeologies.
[0042] This characteristic describes the process of moving from the global description of the field to a more refined understanding, in which local behaviors are identified on a finer scale Gc.
[0043] According to another feature of the invention, rock properties are reduced by defining new rock properties in the field across each subgeology.
[0044] This characteristic describes the principle of allocating different physical properties at the Gc level.
[0045] According to another feature of the invention, the laws of reservoir physics are reduced in such a way that they are applied with functional parameters of the subgeology, and that the scales of space and time associated with the subgeology are determined so that the space associated solutions is consistent with the complexity of historical data at the well level.
[0046] This characteristic provides the rules to be observed for the laws of reduced reservoir physics, by emphasizing their behavior measured at the well level.
[0047] According to another characteristic of the invention, the laws of reservoir physics are derived from the (Navier-) Stokes moment conservation and mass conservation equations for a fluid or gas evolving in a rock modeled as a porous medium , characterized only by its porosity and relative permeability by time, phase and subgeology.
[0048] This feature explains how the general laws of reservoir physics can be simplified at the level of subgeologies, while averaging some key parameters across the entire Gc subgeology.
[0049] According to another characteristic of the invention, the laws of the physics of reservoirs comprise the formula:
where: is the velocity of phase ψ in well k, in subgeology c, px is the viscosity of phase ψ, in subgeology c, Pipc is the pressure of phase <p, in subgeology c, is the density of phase ψ, in subgeology c , g is the gravitational vector, k is a permeability coefficient, and krψc is a relative permeability coefficient of phase <p, in subgeology c. krcpc is a function of time, through its dependence on phase saturation.
[0050] This feature explains which law is considered in practice for the transport of fluids and how it is parameterized.
[0051] According to another characteristic of the invention, the laws of the physics of reservoirs further comprise laws of thermal transfer between a fluid or a gas and a rock, given by the following formulas:

where: T = T (x, y, z) is the temperature and can vary across the reservoir, 0 is the rock porosity, Sç, is the saturation of the cp phase, pψ is the density of the phase ψ, uψ is the speed of the phase ψ , E is the internal volumetric energy (indices s Q f corresponding respectively to the solid and liquid phases), Uf = Ug + U0 + UW is the enthalpy flow, with Up = hψQa specific enthalpy of phase ψ, hg = hw + A, where A is the specific heat of water vaporization, the g, w, o indices are respectively for gas, water and oil, 2 (T) is the conductivity coefficient of the reservoir.
[0052] This feature describes which thermal transfer laws must be considered and how they can be parameterized.
[0053] According to another characteristic of the invention, the laws of the physics of reservoirs also comprise phase laws given by: the viscosity of the phase ψin subgeology c, a function of the local pressure P and the local temperature T, Pvc = pvc (P, T) the density of phase ψ in subgeology c, a function of local pressure P and local temperature T, krfC ° relative permeability coefficient of phase ψ in subgeology c, a function of local pressure P and local temperature T.
[0054] This characteristic as the laws of fluid propagation and the laws of thermal transfer are coupled, through their dependencies of pressure and temperature.
[0055] According to another characteristic of the invention, the laws of well physics comprise the formula:
where: Qçktc is the quantity produced from phase ψ, in well k, at time t, in subgeology c, is the transfer function of well k, in subgeology c, PPklc are the production parameters applied to well k in subgeology c, at time t, u ^ c the speed of the phase in well k, in subgeology c.
[0056] This feature explains how well behavior can be modeled, for example, through transfer functions. Such functions have the same "fineness" as the reservoir modeling: if a subcology Gc includes several layers b, these will have their averages taken together at the well level.
[0057] The invention also relates to an ideal method of exploration of a mature hydrocarbon field, comprising the steps of: building a production simulator as defined in any of the preceding embodiments, iterating over several executions of said simulator of production in order to find the ideal production parameters that optimize a gain value derived from the referred quantity produced, application of the referred ideal production parameters thus obtained for the exploration of the hydrocarbon field.
[0058] This feature describes a generic optimization process, aiming at maximizing a gain function over the field. A feature of the production simulator mentioned above is fully used, that is, its relatively low complexity, which allows a very short calculation time for the calculation of a given production scenario, defined by its corresponding production parameters. This short computation time makes it possible to test a vast number of scenarios, usually more than 100 of them.
[0059] According to another feature of the invention, said optimized gain value is a net present value of said hydrocarbon field.
[0060] This characteristic illustrates a practical use of the invention, which comes to generate large financial gains, against non-optimized scenarios.
[0061] According to another characteristic of the invention, said net present value can be determined using the formula:
where: Píc is the oil production (in barrels) for the kea well and subgeology c, Rik is the rate and royalties for the keo well year i, Si is the selling price of oil (per barrel) for year i , d ca percentage discount rate, / ik is the investment made in well k during year i, OCik is the operating costs for well k during year i, Zkc is the net production (in barrels) for well k and subgeology c, TOi is the treatment cost (per barrel of oil), for year i, TLi is the treatment cost (per barrel of liquid), for year i.
[0062] This feature offers a practical way to define a gain function, which can be completely calculated using the production of wells, by phase, by time and by subgeology, which in turn is calculated by the production simulator for any given set of production parameters.
[0063] Other features, details and advantages of the invention will be more evident from the detailed illustrative description given below with reference to the drawings, in which: Figure 1 shows an illustrative hydrocarbon field, Figure 2 is a block diagram of a whole simulator according to the invention, Figure 3 is a block diagram that details a production simulator according to the invention.
[0064] According to Fig. 1, a hydrocarbon field is an underground (or submarine) G reservoir that comprises rocks, gases and fluids with oil. Said reservoir G is drilled from the surface 10 by several wells 11, in a known manner. Said wells 11 can be injectors in charge of injecting a fluid, generally water, gas or steam, into the reservoir in order to change its properties, for example, to take its pressure or its modified temperature. Said wells 11 may instead be producers in charge of extracting gases, oil and associated fluids (and perhaps other components) from the reservoir. Appropriate fluids or gases can also be injected into these producers, in order to enhance their production efficiency or take modified local reservoir properties.
[0065] The purpose of the invention is to calculate the future production of hydrocarbons per well, per phase and per subgeology over time.
[0066] Another objective of the invention is to determine and apply the ideal PP production parameters, that is, PP production parameters that optimize a gain value, such as, for example, production, NPV, or field reserves mature hydrocarbons.
[0067] The invention relates to an ideal exploration method of a mature hydrocarbon field, based on an optimizer 1. Said optimizer 1 uses a production simulator module 2 that provides quantities produced Qçxkj per phase <p, per well k, 11, by subgeology c, and by time t (this time is usually expressed in months) as a function of the production parameters PP. Phase ψ can be oil, gases, water or other relevant fluids, as chemical additives (the oil can be considered as a single phase or divided into many different hydrocarbon phases). There may be several phases of oil or gas, depending on the hydrocarbon chains. The gas can be hydrocarbon, steam, or any other element.
[0068] PP production parameters include all control parameters that impact production. These production parameters include: injection rates that define the behavior of injectors or producers, production rates, for example, in terms of pumping rates, that define the behavior of producers, but also interventions, such as drilling, finishing, acidification, fractures, etc., or other impactful choices, such as a possible conversion of a producer into an injector, or the drilling of a new additional well (be it a producer and / or an injector). PPs may be dependent on subgeologies.
[0069] Such optimizer 1 allows the testing of new production strategies that would imply the modification of production parameters in relation to the standard (“routine” production strategy).
[0070] As the aforementioned production simulator 2 is simple enough for quick computation of the result, the optimizer 1 can iterate over many executions of the aforementioned production simulator 2, thus testing various strategies or scenarios, in order to find the ideal production parameters PPque optimize a gain value 5 derived from said amount produced Q ^ t.
[0071] As illustrated by the block diagram of Fig. 2, the essential block of optimizer 1 is the production simulator 2. For a given set of PP production parameters, it provides a forecast of quantities produced.
[0072] From the said quantities produced Q </ Xkl, a gain value 5, to be optimized, can be computed by a module 4.
[0073] In optimizer 1, a module 3 is in charge of proposing the referred scenarios in the form of sets of PP production parameters. Several heuristics / algorithms can be used in that module 3, from random Monte-Carlo methods to more efficient and context-aware methods, to propose candidate sets relevant to PP production parameters.
[0074] As illustrated by a feedback arrow, said module 3 can take advantage of the feedback value 5, to determine new sets, and to drive towards new scenarios, according to known optimization methods.
[0075] When iterating in this way, optimizer 1 can determine an ideal set of production parameters that optimize the gain value 5.
[0076] The method also comprises the application of the referred ideal parameters of PPassim production obtained for the control of the exploration of the hydrocarbon field. According to the simulation, the expected gain or revenue may be ideal and at least higher than the standard.
[0077] To be efficient, according to an essential feature of the invention, without reproducing the drawbacks of state of the art solutions, said production simulator 2 must first identify historical HD data, which provide a result relatively well comparable to the quantities produced for the past, when the PP production parameters applied historically are applied, thus differentiating themselves from much simpler models. In addition, said production simulator 2 must satisfy a Vapnik condition.
[0078] Production simulator 2 is designed to produce estimated data for the future, starting from time T, as precisely as possible. A necessary condition to aim for such precision is that the aforementioned production simulator first identifies historical HD data, which reproduces data known in the past. The identification of historical HD data then means that production simulator 2 is accurate enough to reproduce produced data, such as quantities produced per cp phase, per well k, subgeology c and time t, and the total accumulated oil produced, when known , that is, before T. The quality of the referred identification is evaluated by comparing the known produced quantities registered in the historical HD data with the produced quantities Q ^ kt computed through the production simulator 2, to which the past production parameters are applied. The accumulated oil produced is also compared between the sum of the quantities produced recorded in the historical HD data, and the sum of the quantities produced Q ^ kt computed using the production simulator 2. Such comparison can be expressed by
, for% of the quantities and
Q <pktbs are quantities produced per phase ψ, per well k, per subgeology c, and per time t, determined by production simulator 2, QçktbHD are the same quantities produced per phase q>, per well k, per subgeology c, and per time t, found in historical data (HD), [Ti-Xy, TJ is the time interval comprising the most recent X years before time T /, with Ti being the last date for which historical data (HD) is available , | z | ^ 7] denotes a norm of Z over the time interval [Ti, T2], εi is a small positive number with respect to 1, £ 2 is a small positive number with respect to 1, and p is a positive number close to 100%.
[0079] The length of the interval considered for the identification condition can be adapted to the available length of historical data. However, if possible, a length of five years is considered satisfactory. This length is congruent with a comparable horizon of the precise forecast expected for the five-year production simulator 2.
[0080] The particular values of X - 5, p - 90%, £ i - 0.2 and £ 2 - 0.15 were considered satisfactory for the detection of good identification.
[0081] The standard used here for comparison can be any, such as, for example, weighted least squares or smallest squares.
[0082] In addition to the identification condition, an accurate production 2 simulator must also check a Vapnik condition. Such a Vapnik condition can be expressed in different ways.
[0083] One of the results of the theory of learning as developed by Vapnik shows that, from a given set of HD historical data of dimension m, that is, that comprises m independent measurements, one can derive, by learning, a model within a space of dimension-VC h, with an expected risk of reliable prediction R (a) associated with parameters a, delimited superiorly by the sum of two additive terms Remp (a) + 0. The first term Remp (cc) is a empirical risk associated with parameters a, indicative of the quality of identification for historical HD data provided by said parameters a. The second term G is characteristic of the model, and can be expressed by
where δ is a positive number close to zero, with 1 - δ defining a probability that the referred risk R (a) is actually delimited by Remp (a) + Q>, where h is the mentioned dimension- VC of the space S of solutions, also called Vapnik-Chervonenkis dimension or VC dimension, em is the already mentioned number of independent measurements available in historical HD data.
[0084] According to the referred result, all state-of-the-art oversimplified models suffer from a high risk Remp (a) because of the excessive simplification of the model. Conversely, state-of-the-art complex mesh models can provide a low Remp (a), but h suffer from a high O value because of a high reason - since the VC h dimension of space number of solutions is very large in relation to the number of independent measurements available in historical HD data.
[0085] A balance must then be achieved. The value of m is constant and given by historical HD data for a given hydrocarbon field. The size h of the solution space must then be adapted so that the ratio - remains small. With the average objective of a five-year forecast and a reliability around +/- 5%, a value of h - <0.1 was considered compatible with the objective reliability and with the possibility of changing scales, as it will be detailed below to obtain a model.
[0086] Another way of expressing and checking Vapnik's condition is to perform a blind test. A blind test over N years can be performed by breaking historical data over two time intervals. A first “past” interval ranges from an initial To time for which historical data is available, to a T-Ny time that precedes the final time T in TV years. A second “blind” interval runs from T-Ny to the time T corresponding to the last time for which historical data is available. T is usually the current time. The data from the “past” interval are considered known, and are used to build a production simulator 2 identifier, by learning from the said data. The “blind” interval data is considered unknown or at least is omitted during the construction of the production simulator. They are then compared to the data predicted by the production simulator 2 over the aforementioned “blind” interval, given the production parameters along the aforementioned “blind” interval. A blind test is considered satisfactory when the predicted data reproduces the historical data over the omitted interval with sufficient precision.
[0087] A blind test therefore comprises the following steps: determining a production simulator 2 that identifies historical HD data over a time interval prior to T-Ny, forecasting the quantities produced over a time interval [T -Ny, T], the blind test being satisfactory when:
parap% of existing values, and the accumulated oil produced over the same time interval [T-Ny, T] is needed up to £ 2, where: Q <pktbs are quantities produced per phase <p, per well k, per layer or group of layers b, and by time t, determined by production simulator 2, QçhbHD are the same quantities produced by phase <p, by well k, by layer or group of layers b, and by time t, found in data historical, [T-Ny, T] is the time interval comprising the most recent N years before time T, T being the last date for which historical data are available, | z | ^ T] denotes a norm of Z over the time interval [Ti, T2], ei is a small positive number in relation to 1, £ 2 is a small positive number in relation to 1, and p is a positive number close to 100%.
[0088] The particular values of p = 90%, εi = 0.1 and ε2 = 0.1 were considered satisfactory for performing a good blind test. An AT zone - Q o-risvc tamkám A cof-ic-TofArm
[0089] The standard used here for comparison can be any, such as, for example, weighted least squares or smallest squares.
[0090] Another way to verify the Vapnik condition is expressed by measuring a property of predicted stability. This property tests the stability of the production simulator 2 in relation to its output data in response to a slight variation in the input data. The input data are historical HD data that comprise PP production parameters and the quantities produced in the past, that is, over a period of time [To, T]. Based on these nominal PP data and a nominal production simulator 2 is built. These data are then slightly varied to obtain PP 'and corresponding data. Based on this slightly varied input data, another production simulator is built, which is expected to be close enough to the nominal production simulator 2.
[0091] The output data, that is, the quantities produced Q ^ h and are predicted respectively by the nominal production simulator 2 and by another production simulator, over a future time interval [T, T + My], and are then compared.
[0092] The expected stability property is verified when
imply that
where: PP are production parameters, PP 'are slightly variable production parameters, Q ^ HD are quantities produced per phase ψ, per well k, per layer or group of layers b, and for time t, found in historical data, QçMb are quantities produced per phase ψ, per well k, per layer or group of layers b, and for time t, determined by the production simulator, Q ^ th are quantities produced per phase ψ, per well k, per layer or group of layers b, and by time t, determined by another production simulator sufficiently close determined from slightly variable data, [To, T] is the time interval from time To to time T, respectively the start date and the last date for the what historical data are available, [T, T + My] is the time interval that comprises Manos after time T, | z | rT, denotes a norm of Z along the interval [Typhus], ε is a number small positive with respect to 1, and n is a small integer less than 5.
[0093] The important value here is the ratio n between the allowed input variation e, and the output variation obtained. The values of ε = 0.05, n = 2 were considered satisfactory. A horizon of M = 3 years is also satisfactory in view of the precision objectives.
[0094] The standard used here for comparison can be any one, such as, for example, weighted least squares or smallest squares.
[0095] The three described ways of checking Vapnik conditions can be used separately or concurrently.
[0096] As already mentioned, the heart of the invention is production simulator 2. Said production simulator 2 can use behavioral laws of physics, such as the laws of reservoir physics 6 and the laws of well physics 7, in order to accurate enough to accurately identify historical HD data. However, the VC h-dimension of the solution space S must remain small enough for the aforementioned Vapnik condition.
[0097] Two approaches are proposed here to achieve such a result. Both approaches include the use of laws of physics and its application to a reservoir model.
[0098] In a first approach, also called enlargement, both the reservoir model and the laws of physics begin with detailed formulations that are gradually simplified by approximations in the place and time when some homogeneity can be observed, until a condition of interruption is achieved. This condition is achieved by expanding a reservoir partition, the rock properties RP, the laws of reservoir physics 6 and the laws of well physics 7 until the Vapnik condition is verified.
[0099] According to a second approach, also called reduction, both the reservoir model and the laws of physics begin with coarse formulations that are gradually made more complex, detailing where and whenever any substantial variation can be observed, until an interruption condition is reached. Such a condition is achieved by reducing a reservoir partition, the rocky properties RP, the laws of reservoir physics 6 and the laws of well physics 7 while maintaining the verified Vapnik condition, until production simulator 2 identifies data HD historical.
[00100] Both approaches can be used alone or can be used interchangeably. For example, you can enlarge the geology of a thin model by joining parts, and then locally reduce a subgeology that cannot be equated, thus revealing a substantial change in properties within it.
[00101] In this step, whatever the approach used, the structure of production simulator 2 is defined. However, said production simulator 2 depends on several parameters a, which must still be tailored for production simulator 2 to become operational. Depending on parameters a, several production simulators 2 can be candidate solutions. These various candidate solutions are checked against an expected expected risk R (a) for each possible set of parameters α until the ideal set of parameters α is found to minimize that risk R (a).
[00102] According to Fig. 1, a reservoir G can be modeled considering its geology, that is, the closure of the rocky volume that contains oil, and the partition of this using a reservoir partition.
[00103] When using an enlargement approach, a first thin partition can be defined by crossing x-y areas indexed by the letter a, each, for example, comprising at least one well 11, and z areas indexed by the letter b. In the example in Fig. 1, four areas xy (a = 1 ... 4) and five areas z (b = 1 ... 5) are thus used to initially break reservoir G. Reservoir G is thus initially broken at elementary AB parts Gab, so that G = UUG, ífcwith GabπGa, b, = 0 for (a, b) # a = lfe = l (a ', b') (two different parts are disunited), where aC {l ... A} describes an xy area, and bC {l ... B} describes a z layer.
[00104] Starting from this initial partition of the reservoir, the said partition is enlarged considering the elementary parts Gab, and the adjacent parts are gradually joined when they present RPhomogeneous rock nroDruits. The result is a new, more coarse articulation, comprising Gc subgeologies in which cC {l ... C}. In Fig-1, three of these subgeologies are shown. Gi gathers Gn, G12, G13, θi4 and Gis, which present a homogeneous behavior. G2 gathers G21, G22, G23, θ24, G25, G31, G32, G33, G34 and G35. G3 brings together G31, G32, G33, G34 and G35. In real cases, a subgeology usually includes 3 to 50 wells. The referred expansion of the geology partition is a first way to reduce the h dimension of the solution space S. The size of the solution space then depends on the number of geologies, which is in the order of a few units instead of more than 100,000 cells. in a state of the art mesh model.
[00105] When using a reduction approach, the process starts from a coarse partition, for example, composed of a single part corresponding to the entire reservoir G. Said coarse partition s is then broken into Gc subgeologies where cC {l ... C}, where and whenever a substantial change in properties is present around the boundary between the referred subgeologies. In this way, homogeneous behavior can be expected in each subgeology, while keeping the number of such subgeologies as small as possible.
[00106] The rock properties RP comprise mainly the porosity of the rock, the permeability k and the relative permeability krψc, used to determine the dynamics of fluids through the rock, but also other properties, such as net charge, thermal capacity or conductivity, when relevant. As the subgeology becomes the new elementary volumetric unit, all of these RP rock properties are considered to be substantially constant through a given Gc subgeology.
[00107] When using an enlargement approach, the RP rock properties are homogenized through each Gc subgeology, according to the average formula:
, where you see the volume of the Gc.K cc subgeology
[00108] When, on the other hand, a reduction approach is used, the rock properties RPs are reduced by defining new rock properties RPC separated through each new subcology Gc obtained breaking along a discontinuity.
[00109] With respect to Fig. 3, the content of production simulator 2 will now be described in greater detail. Said production simulator module 2 can be divided into two main modules 6, 7. A first module implements the laws of reservoir physics 6. Given a set of PP production parameters, this module provides the dynamic characteristics of the fluids through the reservoir, at least at the entrance of each well 11 and in each subgeology b. These dynamic characteristics are, for example, expressed as a speed field.
[00110] The referred laws of the physics of the reservoirs 6 are enlarged or reduced in such a way that they are applied with functional parameters of the Gc subgeology. In addition, the space and time scales associated with the Gc subgeology are determined so that the associated solution space is congruent with the complexity of the historical HD data at the well level. For example, if an insufficient wealth of details is available in the production related to layers (very small m), the production of the well will be added, partially (some layers are grouped) or totally (all layers are grouped):

[00111] In such a case, production simulator 2 describes only production per well, and not production per well and per layer (or group of layers).
[00112] The enlargement / reduction process also applies to laws. A variation in the number and complexity of the laws of physics and the well can be obtained using simpler or more detailed models of behavior and formulas instead. For example, Darcy's law is a simplification of the more general Navier-Stokes law. The order of the formulas can also be changed, increased or decreased, depending on the desired Vapnik condition. Some parameters can be constant, or refined by a variable function from another parameter or variable. Depending on how the field has been explored so far, thermal transfer may or may not be considered, thus employing or disregarding thermal transfer laws.
[00113] Similarly, the transfer functions of the wells can sometimes be linearized or approximated around their functional points.
[00114] Such agreement must be considered in partial time intervals or in partial spatial areas of the hydrocarbon field, according to the richness (dimension m) of details of the data available in historical HD data.
[00115] The laws of the physics of reservoirs 6 are, in a known way, derived from the equations of (Navier) -Stokes of conservation of the moment and conservation of the mass for a fluid, being that said fluid successively corresponds to the phases, evolving in rock modeled as a porous medium characterized only by its average porosity O, permeability k and relative permeability fcrpc.
[00116] The usual laws of reservoir physics 6 are derived using a general flow model such as the Navier-Stokes equations coupled to a multiphase transport considered on the microscopic scale. The fluids evolve in a domain obtained as the complement of the rock, which can contain a wide variety of topologies and geometries of obstacles. The enlargement / reduction process consists of determining the effective dynamics of fluids in the porous medium in which the rock structure is only described by its average porosity, permeability and relative permeability at a given time t.
[00117] Fluid propagation in reservoir G is derived from the general Navier-Stokes equations, which describe moment conservation for a fluid:
where u is the speed of the fluid, p is the density of the fluid, T is the tension tensor and f is the bodily forces acting on the fluid (for oil in a reservoir, / is gravity: / = pg).
[00118] This is complemented by the conservation of fluid mass:
And for each phase:
where S, denotes the saturation of phase ψ, and Jψ the diffusion flow. Evidently, there are:

[00119] As the flows in the reservoir are usually characterized by small Reynolds numbers, that is, Stokes flows, the Navier-Stokes equations can generally be simplified in the Stokes equations, assuming a stationary and incompressible flow:
where u is viscosity, ut is velocity in direction i, gi is the component of gravity in direction i, and p is pressure.
[00120] Using a volumetric average calculation procedure, Stokes' equations can be homogenized so that the effective viscous resistance force is proportional to the speed and in the opposite direction. It is therefore possible to write, in the case of porous isotropic media:
where k denotes the permeability of the porous medium and O is the porosity. This gives the speed in terms of the pressure gradient:
which provides Darcy's law. In the case of anisotropic porous media, we have:
where K denotes the symmetric permeability tensor. The resulting fluid mass conservation equation is given by:

[00121] In fact, the fluid is generally made up of gases (g), oil (o) and water (w), the composition of which can be complex, depending, for example, on its salinity and hydrocarbon chains. The conservation of the phase mass ψ € {o, w, g} at the extended level can be written as:
where uψ, pcp, Sipdenotam, respectively, the velocity, density, saturation and diffusion flow of the cp phase.
[00122] In any case, closing laws must be determined for phase velocity u in terms of medium flow characteristics, such as velocity u, pressure p and phase saturation <p. Congruence relationships must be satisfied, such as:

[00123] Closing the phase velocity ψinvolves the so-called relative permeability coefficient kr (p, taking into account the differential mobility of the phases:
where pv denotes the viscosity of the phase <p.
[00124] Finally, module 6 implements the laws of reservoir physics using the following formula:
where: u ^ is the velocity of the phase ψin well k, in subgeology c, Pp is the viscosity of the phase in subgeology c, Pp is the pressure of phase ψ, in subgeology c, Pp is the density of the phase in subgeology c, g is the gravitational vector, fc is a coefficient of permeability, and krψc is a coefficient of relative permeability of the cp phase, in subgeology c.
[00125] The laws of physics of reservoirs 6 may also comprise, if the corresponding data is available in historical HD data, thermal transfer laws between a fluid or a gas and a rock, given by the following formulas:
T = T (x, y, z) is the temperature and can vary across the reservoir, O is the rock porosity, Sç, is the saturation of the phase ψ, pψ is the density of the phase ψ, uψ is the speed of the cp phase, E is the internal volumetric energy (indices s Q f corresponding respectively to the solid and liquid phases), Uf = Ug + U0 + UW is the enthalpy flow, being U = p ^ u , hpé the specific enthalpy of phase q>, hg = hw + A, where A is the specific heat of water vaporization, the g, w, o indices are respectively for gas, water and oil, z / T ) is the conductivity coefficient of the reservoir.
[00126] In addition, some variables in the production simulator 2 can be constant or can be detailed considering them as functions, for example, of local pressure P and local temperature T. So the laws of the physics of reservoirs 6 can still understand phase laws given by: viscosity of phase ψ in subgeology c, function of local pressure P and local temperature T, Pψc = Pψc (Phase density ens in subgeology c, function of local pressure P and local temperature T, krψc = krψc (P, T ^ the relative permeability coefficient of phase ψ in subgeology c, a function of local pressure P and local temperature T.
[00127] From the said speeds, known at least in a layered drilling of subgeology c of each well k, a second module 7, implementing well models in the form of well physics laws 7, determines the quantities produced from the phase ψ in subgeology c, in well k at time t, depending on the PPktc production parameters applied to each well k, in subgeology c, at time t, and the speed u ^ h of phase fase in layer b in well k, using the formula: Q ^ lc = Tfc (PPtó., wífc), in which Tkc is a transfer function defined for each well k and each subgeology c. Such transfer functions are functions that depend, among other factors, on the local pressure in the reservoir, geology, skin effect, well / pump design and finishing. They can vary over time as a result of aging or damage. They are usually tabulated functions, which are derived from past production data.
[00128] The space for candidate solutions to become a production simulator 2 of the hydrocarbon field is the combination of T (well transfer physics) and u (reservoir propagation physics). Each of these candidates is associated and defined by a set of parameters a. These α parameters are those that define the candidate simulator that will link the PPte input production parameters to the produced quantities Q ^. The space of candidate simulators in which they are defined is the result of the expansion / reduction of the reservoir, wells and laws. The simulator, defined by such parameters a, will fully characterize the behavior of the field, well by well, subgeology by subgeology, in response to certain PPktc. These are valid for a given range of values allowed for input production parameters. This space of allowed values is usually less than the envelope of all the passed production parameters, with an additional variation of +/- 30%.
[00129] Based on historical HD data, each candidate solution for production simulator is compared with past data. The candidate solution whose set of parameters minimizes the expected risk R (a) is determined by iterating through said sets of parameters a.
[00130] In the case of mature fields, the quantity of past production or historical HD data is large enough to allow the definition of a space of solutions S that incorporates the appropriate physical wealth capable of capturing all the crucial phenomena that (imported) and will import for future production from the same wells. Such a situation allows a balance for which the VC h dimension of the solution space is adequate in relation to the complexity of the available data, consequently minimizing the ratio - and therefore the values of G and R (a). The expected expected risk m R (a) is minimized, and the forecast provided by production simulator 2 can be considered reliable.
[00131] For each candidate solution, the expected risk expected R (a) is delimited superiorly using the Vapnik formula:
where: R (a) is the expected expected risk associated with parameters a, Remp (a) is an empirical risk associated with parameters a, determined by an identification process with historical HD data, δis a positive number close to zero, 1 - δ defining a probability that the inequality is satisfied, and O is a function defined by:
, in which: h is the Vapnik-Chervonenkis dimension of the solution space, m is the number of independent measurements available in the historical HD data, and e is equal to exp (l).
[00132] The best production simulator 2 retained is determined and defined by a set of ideal parameters aopt that minimize the expected risk R (a).
[00133] Once the production simulator 2 is thus determined and optimized in order to respect the identifier condition, the Vapnik condition and minimize the expected risk R (a), it can be used within an optimizer 1. The Optimizer 1 generates PP parameter production scenarios and applies Production Simulator 2 on these scenarios. Many of these runs are then iterated over different sets of PP production parameters. In each iteration, a gain value 5 is calculated, derived from the quantity produced Q ^ fc predicted by the production simulator 2. The resulting gain value 5 can be used to select the next scenario. Thus, an ideal set of PP production parameters can be obtained that optimizes the referred gain value 5.
[00134] Said optimized gain value 5 can be a net present value NPV or value of the RES reserves of the hydrocarbon field.
[00135] Any gain values 5 can be determined by module 4 from quantities produced QçMl. input by optimizer 1, taking into account the necessary economic indexes or parameters, as are well known to those skilled in the art.
[00136] Said net present value NPV can, for example, be determined using the formula:
where: Pkc is the oil production (in barrels) for the kea well and subgeology c, Rik is the fee and royalties for the keo well year i, Si is the selling price of oil (per barrel) for year i , gives the percentage discount rate, / ik is the investment made in well k during year i, OCik is the operating costs for well k during year i, Lkc is the net production (in barrels) for well kea subgeology c, TOi is the treatment cost (per barrel of oil) for year i, TLi is the treatment cost (per barrel of liquid) for year i.
[00137] An alternative choice is the value of the RES reserves of the hydrocarbon field, defined as the accumulated oil produced over a certain period of time. Other choices are possible.

权利要求:
Claims (25)
[0001]
1. Production simulator that implements the laws of reservoir physics suitable for the simulation of a mature hydrocarbon field, providing quantities produced per phase, per well, per layer or group of layers, and for time as a function of production parameters , characterized by being defined by a set of parameters within a space of candidate solutions of production simulator, by identifying historical data of said mature hydrocarbon field and exhibiting a good prognostic reliability, which can be described, according to the Theory of Statistical Learning, as the expected low risk expected R (a) associated with said parameters a, and the said space of candidate solutions of production simulator for taking a production simulator of the hydrocarbon field consists of the combination of the physics of the transfer of wells and the propagation physics of the reservoir, with each candidate solution of a production simulator associated with and defined by a set of parameters a, those parameters a defining the candidate production simulator solution that will link input production parameters PPktc to the quantities produced Qψktc-
[0002]
2. Simulator according to claim 1, characterized by being defined by a set of parameters within a space of candidate solutions of production simulator, being able to demonstrate a low expected risk R (a), minimizing a fh 1 condition of border
[0003]
3. Simulator according to claim 1, characterized in that an identification with historical data is obtained when:
[0004]
4. Simulator according to claim 3, characterized by: X = 5, εi = 0.2, £ 2 = 0.15 and p = 90%.
[0005]
5. Simulator according to claim 1, characterized in that the Vapnik condition is expressed as - <0.1, where: mh is the Vapnik-Chervonenkis dimension of the candidate production simulator solutions space, and m is the number of independent measurements available in historical data.
[0006]
6. Simulator according to claim 1, characterized in that the Vapnik condition is verified by performing a satisfactory blind test for N years, as defined by the following steps: determining a production simulator identifying historical data over a period of time prior to T-Ny, predict the quantities produced over a time interval [T-Ny, T] given the production parameters in the same interval, the blind test being satisfactory when:
[0007]
7. Simulator according to claim 6, characterized by: p = 90%, εi = 0.1 and 82 = 0.1 and N = 3.
[0008]
Simulator according to any one of claims 1 to 7, characterized in that the Vapnik condition is a prognostic stability property verified when, if
[0009]
9. Simulator according to claim 8, characterized by: ε = 0.05, n = 2 and N = 3.
[0010]
10. Simulator according to claim 1, characterized by being built following the steps of: defining the initial detailed reservoir partition, rock properties, laws of reservoir physics, and laws of well physics, expanding said partition of reservoir, rock properties, laws of reservoir physics and laws of well physics until the condition of Vapnik is verified, and to optimize said production simulator by choosing between candidate production simulator solutions, candidate solution for simulator of production production that minimizes an expected expected risk R (a).
[0011]
11. Simulator according to claim 10, characterized in that said reservoir partition is enlarged following the steps of: partitioning a reservoir G into elementary parts Ggh, AB so that G = ULlGai, with G ^ r ^ G ^ , = 0 for (a, b) ψ (a ', b'), where € {l ... A} describes an areax-jy, and eb € {l ... B} describes a layer, grouping elementary parts Adjacent Ggh that exhibit homogeneous rock properties, in Gç subgeologies where c € {l ... C}.
[0012]
12. Simulator according to claim 10, characterized in that the rock properties are increased following a step of averaging the rock properties RPç in each subgology Gg, according to the formula: JJJ RP (x, y, z) dxdydz, where Vc is the volume of the Gc subgeology.
[0013]
13. Simulator according to claim 10, TO / ln nni * oc loic / 4o / 4r c rαoαnzα + Annc cαrαrn o 1 n / 4 oc / 4 o so that they apply to functional parameters of the subgenology Gç and that the scales of space and time associated with the Gc subgeology are determined in such a way that the associated space of candidate production simulator solutions is congruent with the complexity of historical data at the well level.
[0014]
14. Simulator according to claim 1, characterized by being built following the steps of: defining the initial coarse reservoir partition, rock properties, laws of the physics of the reservoirs, and laws of the physics of wells, reducing said partition of reservoir, rock properties, laws of reservoir physics and laws of well physics as long as the condition of Vapnik remains verified, until the production simulator identifies itself with historical data, and optimize the said production simulator by choosing among the solutions production simulator candidates, the candidate production simulator solution that minimizes an expected expected risk R (a).
[0015]
15. Simulator according to claim 14, characterized in that said reservoir partition is reduced following the steps of: starting from the entire reservoir, starting said reservoir in Gc subgeologies where c € {l ... C} , where a substantial variation of properties is present around the boundary between said subgeologies.
[0016]
16. Simulator according to claim 14, characterized in that the rock properties are reduced by the definition of new separate rock properties in each Gc subgeology.
[0017]
17. Simulator according to claim 14, characterized in that the laws of the physics of the reservoirs are reduced in such a way that they apply to functional parameters of the Gc subgeology and that the space and time scales associated with the Gc subgeology are determined in such a way that the associated space of candidate production simulator solutions is congruent with the complexity of historical data at the well level.
[0018]
18. Simulator according to claim 10, characterized in that the laws of reservoir physics are derived from the Navier-Stokes equations of moment conservation and mass conservation for a fluid evolving in a rock modeled as a porous medium characterized only by its average porosity, permeability and relative permeability kr <pc per phase ψby subgeology c.
[0019]
19. Simulator according to claim 18, characterized in that the laws of reservoir physics comprise the formula:
[0020]
20. Simulator according to claim 19, characterized in that the laws of the physics of the reservoirs additionally comprise laws of thermal transfer between a fluid or a gas and a rock, given by the following formulas:
[0021]
21. Simulator according to claim 20, characterized in that the laws of reservoir physics additionally comprise phase laws given by: / Ç = / Ç (P, T) the viscosity of phase ψ in subgeology c, depending on the local pressure P and the local temperature T, pψc = P ^ (P, T) the density of phase ψ in subgeology c, a function of the local pressure P and the local temperature T, krψc = krψc (P'T ') The relative permeability coefficient of phase ψ in subgeology c, function of local pressure P and local temperature T.
[0022]
22. Simulator according to claim 10, characterized in that the laws of well physics comprise the formula: = rfe (PPWc, w, * c), where: is the quantity produced from phase ψ, in well k, at time t , in subgeology c, Tkc is the transfer function of well k, in subgeology c, PPto are the production parameters applied to well £ in subgeology c, at time t, u <pkc is the speed of phase ψin well k, in subgeology c.
[0023]
23. Simulator for the ideal exploration of a mature hydrocarbon field characterized by understanding the steps of: building a production simulator as defined in any of claims 1 to 10, iterating through several executions of said production simulator in order to find the ideal production parameters that optimize a gain value derived from said quantity produced, apply the said ideal production parameters thus obtained for the exploration of the hydrocarbon field.
[0024]
24. Simulator according to claim 23, characterized in that said optimized gain value is a net present value or reserves of said hydrocarbon field.
[0025]
25. Simulator according to claim 24, characterized in that said net present value NPV is determined using the formula: where: Pkc is oil production (in barrels) for well k and subgeology c, Rik is the fee and royalties for well k and year i, Si is the selling price of oil (per barrel) for year i, c percentage discount rate, / ik co investment made in well k during year i, OCik are the operating costs for well k during year i, £ kc is the net production (in barrels) for well k and subgeology c, fOi is the treatment cost (per barrel of oil), for year i, / Xi is the treatment cost (per barrel of liquid), for year i.

类似技术:

---

公开号 | 公开日 | 专利标题

---

BR112012032159B1|2020-11-17|production simulator for the simulation of a mature hydrocarbon field

---

BRPI0706805A2|2011-04-05|computer readable methods, systems and media for rapid upgrade of oil and gas field production models with proxy and physical simulators

---

Fernández-Nieto et al.2008|Extension of WAF type methods to non-homogeneous shallow water equations with pollutant

---

Hajibeygi2011|Iterative multiscale finite volume method for multiphase flow in porous media with complex physics

---

Salehi et al.2019|A comprehensive adaptive forecasting framework for optimum field development planning

---

CN108005644A|2018-05-08|A kind of inclined seam Dynamic Permeability Forecasting Methodology and device

---

Izgec et al.2009|Establishing injector/producer connectivity before breakthrough during fluid injection

---

Lobato-Barradas et al.2002|Integrated compositional surface-subsurface modeling for rate allocation calculations

---

Vignati et al.2009|Innovative implementation of compositional delumping in integrated asset modeling

---

Kechut et al.2017|Advanced Methods for Determining the Value of Information of Surveillance Data Acquisition Using Predictive Analytics

---

LF Bittencourt Neto et al.2020|Laplace Domain Pressure Behavior Solution for Multilayered Composite Reservoirs

---

US20190186238A1|2019-06-20|Parallel production simulation of a mature hydrocarbon field

---

Reverón et al.2021|Numerical Comparison Between ES-MDA and Gradient-Based Optimization for History Matching of Reduced Reservoir Models

---

Al-Ajmi et al.2008|Estimation of storativity ratio in a layered reservoir with crossflow

---

Sanyal2003|One Discipline, Two Arenas–Reservoir Engineering In Geothermal and Petroleum Industries

---

Yudin et al.2021|Efficiency of Using a Proxy Model for Modeling of Reservoir Pressure

---

da Silva Pereira et al.2017|Numerical Validation of Analytical Estimates for Propagation of Thermal Waves Generated by Gas-Solid Combustion

---

Tian2021|Compositional Modeling of Gas Injection in Tight Oil Reservoirs

---

Evseenkov et al.2021|Short-Term Forecasting of Well Production Based on a Hybrid Probabilistic Approach

---

Wang2007|Implementation of a Two Pseudo-Component Approach for Variable Bubble Point Problems in GPRS

---

Pamungkas et al.2020|Integrated Reservoir Model History Matching Workflow: Unlocking Reservoir Opportunities, Onshore Abu Dhabi, UAE

---

Khait et al.2021|A GPU-based integrated simulation framework for modelling of complex subsurface applications

---

Kumar Khataniar et al.2021|Aspects of Multiscale Flow Simulation with Potential to Enhance Reservoir Engineering Practice

---

Md Yusof et al.2022|Experimental Investigation, Porosity-Permeability Modelling, and Artificial Neural Network Prediction of CO2 Injectivity Change for Sequestration

---

Itthisawatpan2013|Upscaling of Two-phase Flow with Capillary Pressure Heterogeneity Effects

同族专利:

---

公开号 | 公开日

---

EP2583121B1|2015-01-21|

---

WO2011157764A4|2012-03-29|

---

DK2583121T3|2015-04-07|

---

US20130073269A1|2013-03-21|

---

MX2012014727A|2013-05-30|

---

US20110313744A1|2011-12-22|

---

EA201291174A1|2013-06-28|

---

JP2013534002A|2013-08-29|

---

AU2011267039A1|2013-01-10|

---

CN103003718A|2013-03-27|

---

EA030299B1|2018-07-31|

---

US9031821B2|2015-05-12|

---

BR112012032159A2|2016-11-16|

---

MY164433A|2017-12-15|

---

EP2583121A2|2013-04-24|

---

US8412501B2|2013-04-02|

---

CN103003718B|2016-08-03|

---

CA2801804C|2019-02-12|

---

WO2011157764A3|2012-02-09|

---

AU2011267039B2|2014-05-22|

---

ES2532766T3|2015-03-31|

---

WO2011157764A2|2011-12-22|

---

JP5797262B2|2015-10-21|

---

PL2583121T3|2015-06-30|

---

CO6640278A2|2013-03-22|

---

CA2801804A1|2011-12-22|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

US5305209A|1991-01-31|1994-04-19|Amoco Corporation|Method for characterizing subterranean reservoirs|

---

US6980940B1|2000-02-22|2005-12-27|Schlumberger Technology Corp.|Intergrated reservoir optimization|

---

US6775578B2|2000-09-01|2004-08-10|Schlumberger Technology Corporation|Optimization of oil well production with deference to reservoir and financial uncertainty|

---

US6810332B2|2003-01-31|2004-10-26|Chevron U.S.A. Inc.|Method for computing complexity, confidence and technical maturity indices for reservoir evaluations|

---

US7835893B2|2003-04-30|2010-11-16|Landmark Graphics Corporation|Method and system for scenario and case decision management|

---

FR2855631A1|2003-06-02|2004-12-03|Inst Francais Du Petrole|METHOD FOR OPTIMIZING THE PRODUCTION OF AN OIL DEPOSIT IN THE PRESENCE OF UNCERTAINTIES|

---

WO2005020044A1|2003-08-26|2005-03-03|The Trustees Of Columbia University In The City Of New York|Innervated stochastic controller for real time business decision-making support|

---

US20070016389A1|2005-06-24|2007-01-18|Cetin Ozgen|Method and system for accelerating and improving the history matching of a reservoir simulation model|

---

CN100429528C|2005-08-10|2008-10-29|大庆油田有限责任公司|Method of deposition phase control for casting sandstone oil reservoir attribute|

---

US9043188B2|2006-09-01|2015-05-26|Chevron U.S.A. Inc.|System and method for forecasting production from a hydrocarbon reservoir|

---

US7774184B2|2006-10-17|2010-08-10|Schlumberger Technology Corporation|Brownfield workflow and production forecast tool|

---

US8024123B2|2007-11-07|2011-09-20|Schlumberger Technology Corporation|Subterranean formation properties prediction|

---

CN101661514B|2008-05-21|2014-04-16|中国石化股份胜利油田分公司地质科学研究院|Oil deposit black oil model numerical simulation system|

---

US8793111B2|2009-01-20|2014-07-29|Schlumberger Technology Corporation|Automated field development planning|

---

CN101644783B|2009-07-27|2011-12-07|中国石化集团胜利石油管理局|Reservoir description method based on condition recursion|

---

US8532968B2|2010-06-16|2013-09-10|Foroil|Method of improving the production of a mature gas or oil field|CA2783787A1|2010-02-12|2011-08-18|Exxonmobil Upstream Research Company|Method and system for creating history-matched simulation models|

---

EP2564309A4|2010-04-30|2017-12-20|Exxonmobil Upstream Research Company|Method and system for finite volume simulation of flow|

---

AU2011283190A1|2010-07-29|2013-02-07|Exxonmobil Upstream Research Company|Methods and systems for machine-learning based simulation of flow|

---

EP2599023B1|2010-07-29|2019-10-23|Exxonmobil Upstream Research Company|Methods and systems for machine-learning based simulation of flow|

---

US9058445B2|2010-07-29|2015-06-16|Exxonmobil Upstream Research Company|Method and system for reservoir modeling|

---

US9058446B2|2010-09-20|2015-06-16|Exxonmobil Upstream Research Company|Flexible and adaptive formulations for complex reservoir simulations|

---

WO2013039606A1|2011-09-15|2013-03-21|Exxonmobil Upstream Research Company|Optimized matrix and vector operations in instruction limited algorithms that perform eos calculations|

---

CA2879063A1|2012-07-27|2014-01-30|Landmark Graphics Corporation|Systems and methods for estimating opportunity in a reservoir system|

---

EP2901363A4|2012-09-28|2016-06-01|Exxonmobil Upstream Res Co|Fault removal in geological models|

---

WO2014186545A1|2013-05-16|2014-11-20|Exa Corporation|Mass exchange model for relative permeability simulation|

---

EP2981866B1|2013-05-24|2020-03-11|Halliburton Energy Services Inc.|Methods and systems for reservoir history matching for improved estimation of reservoir performance|

---

CA2909170C|2013-06-10|2020-02-18|Exxonmobil Upstream Research Company|Determining well parameters for optimization of well performance|

---

WO2015020650A1|2013-08-07|2015-02-12|Landmark Graphics Corporation|Static earth model calibration methods and systems using permeability testing|

---

RU2016105167A|2013-09-16|2017-08-22|Лэндмарк Графикс Корпорейшн|PSEVDOFAZOVOE SIMULATION OF MINING METHOD FOR SIGNAL PROCESSING FOR EVALUATION quasi-multiphase flows with extraction by a model-driven sequence of analog step function relative permeability modeling of flow into the reservoir to RANKING SETS PETROPHYSICAL realization|

---

RU2016131935A|2014-01-14|2018-02-16|Бейкер Хьюз Инкорпорейтед|METHOD FOR DETECTING DISPLACEMENT OF THE MOBILE FLUID FRONT FRONT BY THE COMBINATION OF ELECTRIC AND GRAVIMETRIC MEASUREMENTS IN WELLS|

---

AU2015298233B2|2014-07-30|2018-02-22|Exxonmobil Upstream Research Company|Method for volumetric grid generation in a domain with heterogeneous material properties|

---

EP3213126A1|2014-10-31|2017-09-06|Exxonmobil Upstream Research Company|Handling domain discontinuity in a subsurface grid model with the help of grid optimization techniques|

---

CN104933246A|2015-06-15|2015-09-23|中国海洋石油总公司|Method for quantitatively representing contribution of gas field reserves parameters to reserves change|

---

KR101766917B1|2015-12-23|2017-08-11|한국지질자원연구원|Apparatus and method for evaluating economics of bitumen|

---

RU2692369C1|2018-12-26|2019-06-24|Публичное акционерное общество "Газпром нефть"|Method of selecting deposit development system|

法律状态:
2018-12-26| B06F| Objections, documents and/or translations needed after an examination request according art. 34 industrial property law|

---

2019-11-05| B06U| Preliminary requirement: requests with searches performed by other patent offices: suspension of the patent application procedure|

---

2020-06-16| B09A| Decision: intention to grant|

---

2020-11-17| B16A| Patent or certificate of addition of invention granted|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 15/06/2011, OBSERVADAS AS CONDICOES LEGAIS. |

优先权:

---

申请号 | 申请日 | 专利标题

---

US12/816,940|2010-06-16|

---

US12/816,940|US8412501B2|2010-06-16|2010-06-16|Production simulator for simulating a mature hydrocarbon field|

---

PCT/EP2011/059968|WO2011157764A2|2010-06-16|2011-06-15|Production simulator for simulating a mature hydrocarbon field|

---