FR3120087A1 - Method for selecting, for at least one underground formation comprising hydrocarbons, a technique for exploiting said underground formation - Google Patents

Abstract

The present invention relates to a method for selecting, for at least one underground formation comprising hydrocarbons, a technique for exploiting said underground formation, from among a predefined list of possible exploitation techniques, characterized in that it comprises the implementation by data processing means (21) of a client equipment (2), of steps of: Obtaining a value of at least one static parameter descriptive of physical properties of said underground formation; Calculation, for each mining technique of said predefined list, of an adequacy score of this mining technique with said underground formation as a function of the value obtained from at least one static parameter descriptive of the physical properties of said underground formation; Estimation, for each mining technique of said predefined list, of a value over time of at least one dynamic parameter descriptive of the mining of said underground formation when this mining technique is implemented; Calculation, for each mining technique of said predefined list, of an overall score for this mining technique as a function of the adequacy scores, of the values of said dynamic parameter descriptive of the mining of said underground formation, and of at least one ontology characterizing said exploitation techniques; Selection of at least said mining technique of said underground formation having the maximum overall score. Figure for the abstract: Fig. 2

Description

Method for selecting, for at least one underground formation comprising hydrocarbons, a technique for exploiting said underground formation

GENERAL TECHNICAL AREA

The present invention relates to the field of the exploitation of hydrocarbons contained in an underground formation, more particularly a method for selecting a technique for exploiting said underground formation.

STATE OF THE ART

Oil fields are usually exploited in three successive phases. Phase I: natural depletion / Phase II: waterflood (water injection) / Phase III: IOR or EOR (in French: “tertiary recovery”).

Phase I, called primary recovery, consists in extracting, via a so-called production well, the oil present in the reservoir by the effect of overpressure naturally prevailing within the reservoir. This is therefore a “natural depletion”. This primary recovery only allows access to a small quantity of the oil contained in the reservoir, around 10 to 15% at most.

To allow the extraction to continue when the pressure in the reservoir becomes insufficient to move the oil still in place, in phase II called secondary recovery, a fluid is injected (re-injection of the water produced, diluted or not, injection of sea or river water, for example) within the reservoir, with a view to exerting within the reservoir an overpressure capable of driving the oil towards the production well or wells. A common technique in this context is water injection (also referred to by the English term “waterflooding”), in which large volumes of water are injected under pressure into the reservoir via injection wells. The injected water drags some of the oil it encounters and pushes it towards one or more producing wells. Secondary production methods such as water injection, however, only extract a relatively small part of the hydrocarbons in place (typically around 30%). This partial sweeping is due in particular to the trapping of the oil by capillary forces, to the differences in viscosity and density existing between the injected fluid and the hydrocarbons in place, as well as to heterogeneities at micro- or macroscopic scales (scale pores and also reservoir scale).

To try to recover the rest of the oil, which remains in the underground formations after the implementation of the primary and secondary production methods, the so-called tertiary recovery phase III (known by the acronym "EOR", corresponding to " Enhanced Oil Recovery "), offers a set of techniques involving the injection of products intended to physically improve recovery. Among these products, mention may be made of additives such as surfactants or other surfactants, polymers such as hydrolyzed polyacrylamide, gas such as CO ₂ , etc.

The difficulty nevertheless consists in choosing the optimal EOR strategy among all those possible for a given subterranean formation.

Today, we generally rely on a process called "screening" which consists of a multi-criteria analysis to determine an optimal EOR strategy. However, these methods are often limited because they are based on analytical or simplified models (approximate numerical models, also called "proxy models"), aimed at approximating an additional recovery rate. Moreover, such a process ignores the effects due to the uncertainty introduced by other factors such as performance in an uncertain geology and heterogeneous environment, accessibility, the risk of injecting in areas that have been well swept by phase II, problems of injectivity, the price of the barrel and its evolution etc., due to a lack of data. If we consult a development plan usually used, we realize that many answers are likely to arrive too late, or that certain problems are only explored after costly efforts, for lack of anticipation, ignoring expertise or information, considered initially "fuzzy".

Thus, it would be desirable to have a new, more efficient and more universal selection tool, for at least one underground formation comprising hydrocarbons, of a technique for exploiting said underground formation.

PRESENTATION OF THE INVENTION

Thus, the present invention relates, according to a first aspect, to a method of selecting, for at least one underground formation comprising hydrocarbons, a technique for exploiting said underground formation, from among a predefined list of possible exploitation techniques, characterized in what it includes the implementation by means of data processing of customer equipment, of steps of:

1. Obtaining a value of at least one static parameter descriptive of physical properties of said underground formation from measurements supplied from at least one sensor;
2. Calculation, for each mining technique of said predefined list, of an adequacy score of this mining technique with said underground formation as a function of the value obtained from at least one static parameter descriptive of the physical properties of said underground formation;
3. Estimation, for each mining technique of said predefined list, of a value over time of at least one dynamic parameter descriptive of the mining of said underground formation when this mining technique is implemented;
4. Calculation, for each mining technique of said predefined list, of an overall score for this mining technique as a function of the adequacy scores, of the values of said dynamic parameter descriptive of the mining of said underground formation, and of at least one ontology characterizing the implementation of said exploitation techniques
5. Selection of at least said mining technique of said underground formation having the maximum overall score.

According to advantageous and non-limiting characteristics:

Said predefined list of possible operating techniques includes at least one miscible CO2 injection technique, one polymer injection technique, and one surfactant injection technique.

Said dynamic parameter descriptive of the exploitation of said underground formation is a recovery rate of said hydrocarbons from said formation.

Step (c) comprises, for each technique of exploitation of said predefined list, the construction according to the value obtained of the at least one static parameter descriptive of physical properties of said underground formation of a linear curve by pieces representing over time the rate of recovery of said hydrocarbons from said formation when this exploitation technique is implemented.

Said statistical parameter descriptive of physical properties of said underground formation is chosen from among depth, temperature, porosity, permeability, heterogeneity, residual hydrocarbon saturation, hydrocarbon zone thickness, hydrocarbon viscosity, clay rate, salinity, of said underground formation.

Step (e) comprises the exploitation of said hydrocarbons of said formation by implementing infrastructures relating to said selected exploitation technique.

Said problem of selecting an exploitation technique in a hierarchical system is previously decomposed by Hierarchical Process Analysis, AHP, by means of said ontology; step (d) comprising aggregating, to calculate the overall score of a mining technique, the suitability scores and the values of said dynamic parameter descriptive of the mining of said subterranean formation in accordance with said hierarchical system using a multi-criteria decision-making method, such as Technique for Order of Preference by Similarity to Ideal Solution, TOPSIS.

Said hierarchical system comprises a plurality of factors, and optionally sub-factors of said factors, of the problem of selecting an exploitation technique including at least one factor of physical properties of said underground formation and an exploitation factor of said formation subterranean, step (d) comprising the construction of at least one judgment matrix comparing two by two the factors or the sub-factors of a factor, then the calculation from the judgment matrix or matrices of weightings to be applied to each suitability score and each dynamic parameter descriptive of the exploitation of said underground formation.

The method comprises a preliminary step (a0) of building said ontology characterizing said operating techniques by processing means of a server connected to said client equipment.

The method is also a method for selecting at least one underground formation from among a predefined set of underground formations comprising hydrocarbons, comprising the implementation of steps (a) to (c) for each underground formation of said predefined set, the step (d) comprising the calculation, for each pair of an underground formation of said predefined set and of a technique for exploiting said predefined list, of an overall score for this pair as a function of the suitability scores, of the values said dynamic parameter descriptive of the exploitation of each underground formation, and of at least one ontology characterizing said exploitation techniques; and step (e) comprising selecting at least said pair having the maximum overall score.

According to a second and a third aspect, the invention relates to a computer program product comprising code instructions for the execution of a method according to the first selection aspect, for at least one underground formation comprising hydrocarbons, of a technique for exploiting said underground formation; and computer-equipment readable storage means on which a computer program product includes code instructions for performing a method according to the first selection aspect, for at least one subterranean formation comprising hydrocarbons, of a technique for exploiting said underground formation.

PRESENTATION OF FIGURES

Other characteristics, objects and advantages of the invention will emerge from the description which follows, which is purely illustrative and not limiting, and which must be read in conjunction with the appended drawings in which:

the represents an architecture for the implementation of the method according to the invention;

the illustrates the steps of a preferred embodiment of a method for selecting a technique for exploiting an underground formation comprising hydrocarbons according to the invention;

the represents a preferred embodiment of step (b) of a method for selecting a technique for exploiting an underground formation comprising hydrocarbons according to the invention;

the represents a preferred embodiment of step (c) of a method for selecting a technique for exploiting an underground formation comprising hydrocarbons according to the invention;

the represents an example of a continuous curve per piece of a dynamic parameter descriptive of the exploitation of said underground formation when an exploitation technique is implemented;

the represents a preferred embodiment of a first part of step (d) of a method for selecting a technique for exploiting an underground formation comprising hydrocarbons according to the invention;

the represents a preferred embodiment of a second part of step (d) of a method for selecting a technique for exploiting an underground formation comprising hydrocarbons according to the invention;

the represents a preferred embodiment of a preliminary step (a0) of a method for selecting a technique for exploiting an underground formation comprising hydrocarbons according to the invention.

DETAILED DESCRIPTION

Principle of the invention

With reference to the , the invention proposes a method for selecting, for at least one underground formation F1 comprising hydrocarbons, a technique for exploiting said underground formation, from among a predefined list of possible exploitation techniques.

Preferably, there is even a predefined set of underground formations F1, F2, F3 comprising hydrocarbons, and the method is also a method for selecting at least one underground formation F1, F2, F3 from among said predefined set of underground formations F1, F2, F3 containing hydrocarbons.

In summary, it will be understood that the present method is a method for selecting an underground formation F1, F2, F3 comprising hydrocarbons and/or a technique for exploiting said underground formation F1, F2, F3.

In the rest of this description, we will take the example of a single underground formation F1, and we will then explain how to extend to the case of a predefined set of underground formations F1, F2, F3.

As explained, it is assumed that said underground formation F1 has undergone a prior phase of exploitation of said hydrocarbons from said formation in accordance with a water injection technique (Phase II – itself after an exploitation phase by natural depletion – Phase I), and the ultimate objective is to further exploit said hydrocarbons from said formation by implementing infrastructures relating to said selected exploitation technique.

Preferably, said predefined list of possible exploitation techniques includes so-called EOR techniques, i.e. at least one miscible CO2 injection technique, one polymer injection technique, and one surfactant injection technique. The present method does not in itself propose any new exploitation technique and those skilled in the art may use any known EOR technique.

There can of course be several techniques of each type, for example several polymer injection techniques, differing by the polymer and/or the injection methods. In general, we will assume that we have K candidate exploitation techniques in the predefined list.

The invention can be summed up as the implementation of a methodology which revolves around several cognitive steps, allowing the quantification of criteria for choosing a particular exploitation technique.

The main output is a score assigned to the techniques in relation to various "decision factors", previously prioritized. The number of factors considered can vary from one to several dozen, depending on whether a focused or, on the contrary, integrated assessment is desired.

As represented by , it is assumed that there is at least one underground formation F1, F2, F3 comprising hydrocarbons, equipped with at least one sensor 10 for acquiring data relating to said underground formation F1.

The method is implemented using a decision-making system, coupled to this or these underground formation(s) F1, F2, F3 (and more precisely the sensors 10).

The system essentially consists of at least one client equipment 2 comprising data processing means 21 such as a processor, data storage means 22 such as a memory (for example a hard disk) storing a database reference data to which we will return later. The equipment 2 is typically a workstation, in particular provided with an input and output interface to restore the results of the process (and if necessary define the technique of exploitation of said underground formation to be implemented). The client equipment 2 is connected directly or indirectly to the sensors 10, for example by a network of the Internet type.

In addition, the decision-making system also advantageously comprises a server 1 itself connected to said client equipment 2, again by a network of the Internet type. As will be seen later, the server 1 also comprises data processing means 11 such as a processor and data storage means 12, and allows the construction of an ontology characterizing said exploitation techniques.

Process

With reference to the , the present method is implemented by the data processing means 21 of the client equipment 2 and begins with a step (a) (called "scraping") for obtaining a value of at least one static parameter description of physical properties of said underground formation, and preferably a plurality (it will be assumed that there are N static parameters considered).

By static parameter descriptive of physical properties of said underground formation, we mean physical quantities called "reservoir data" which are substantially invariant over time and make it possible to characterize in a unique way the nature of an underground formation and to compare them with each other.

In a non-limiting manner, each static parameter descriptive of physical properties of said underground formation is chosen from among a depth, a width, a temperature, a porosity, a permeability, a heterogeneity, a residual hydrocarbon saturation, a thickness of the hydrocarbon zone , hydrocarbon viscosity, clay content, lithology (sandstone, carbonates, etc.), salinity, of said underground formation, etc.

Each parameter is obtained from measurements provided by means of said at least one sensor 10, in particular the value of the parameter can be directly measured by the sensor 10 (this is the case of the depth for example) or else calculated at from measurements made by the sensor 10 (this is the case of the residual saturation for example).

If the measurements are slightly variable over time, an average value can be taken from a time series (which is the case for temperature in particular).

In practice, there is generally a database of values of static parameters descriptive of physical properties of said set of underground formations stored by the data storage means 22, supplemented gradually from the measurements provided from the sensors 10, and step (a) consists in extracting the N values of interest from said base.

In a step (b), illustrated in more detail by the , a score of adequacy of this mining technique with said underground formation is calculated, for each mining technique of said predefined list, as a function of the value obtained from the at least one static parameter describing the physical properties of said underground formation.

Note that step (b) does not require the use of every static parameter obtained in step (a), some can be set aside and used only for steps (c) or (d).

Step (b) typically comprises a comparison of each value of a static parameter considered with at least one threshold value (preferably two: a low threshold and a high threshold) associated with each exploitation technique, then the weighting of the results of all the comparisons so as to obtain said adequacy score. Each threshold is thus an “applicability criterion” of said exploitation technique.

Said threshold values are advantageously those observed in several existing cases of EOR operations having or have not been successful, and can also be entered in dedicated databases (one per possible operating technique), also stored by the storage means 22 Those skilled in the art may also take applicability criteria from the literature (see for example the documents: Taber, JJ, Martin, FD, and Seright, RS 1997a. EOR Screening Criteria Revisited—Part 1: “ Introduction to Screening Criteria and Enhanced Recovery Field Projects ", SPE Res Eval & Eng 12 (3): 189–198. SPE-35385-PA; Taber, JJ, Martin, FD, Seright, RS, 1996. " EOR screening criteria revisited . Proceedings of the SPE/DOE Tenth Symposium on Improved Oil Recovery”, April 21–24, 1996, held at Tulsa, Oklahoma, USA (SPE 35385); Kang PS, Jong-Se Lim JS, and Chun Huh C., “ Screening Criteria and Considerations of Offshore Enhanced Oil Recovery ” MDPI – Energy; and Akanji L., Sa ndrea R. “ A Neuro-Fuzzy Approach to Screening Reservoir Candidates for EOR ”, Advances in Petroleum Exploration and Development Vol. 12, No. 1, 2016, p. 1-14 DOI:10.3968/8743).

In the example of the , the RPi with 1≤i≤N are the values of the N static parameters descriptive of physical properties of said underground formation. We see that on a loop we test each parameter (i ranging from 1 to N), by comparing its value with the two threshold values CRmin and CRmax. We understand here that we have as many thresholds (and therefore comparisons) as there are possible exploitation techniques. More precisely, there are in practice thresholds CRmin j and CRmax j with 1≤j≤K, where K is the number of techniques for using said predefined list.

If, for a technique, the value of the static parameter RP _i is between these min. and max, the possibility of application is indicated by an elementary score S _i of 1, otherwise 0 (ie S _i has binary values).

The adequacy score of this exploitation technique is then calculated as the linear combination of the elementary scores S _i by weights W _i (between 0 and 1), with the formula ∑ _i (S _i *W _i ). As explained before, we have an adequacy score by exploitation technique.

It should be noted that the weights Wi are “confidence factors” translating the certainty into the value of the static parameter descriptive of the physical properties of said corresponding underground formation (a parameter lacking a solid statistical basis is considered to be uncertain). They are thus specific to static parameters and therefore common to all exploitation techniques (unlike thresholds and therefore elementary scores). These confidence factors can be predetermined (and stored again in a database), but also calculated on the fly by analyzing the database of static parameter values, or even be directly entered by the user (hence the "input" in the ). By default, the value of the weights Wi can be set to 1 (maximum certainty), and they can not be used (ie directly sum the elementary scores Si).

Each fit score can ultimately be normalized, to arrive at a value between 0 and 1.

In order for an eligibility of the EOR technique to be validated, the adequacy score of the technique must possibly be higher than a floor such as 0.5. In addition, the corresponding scores can (if necessary for the eligible techniques) be classified and therefore a priori offer a criterion to the decision maker to choose the methods for which a more in-depth study will take place.

Note that we will not be limited to the embodiment of the (comparison with thresholds), and it is possible for example to have an evaluation of the criteria by maximizing the distances with reference values, or even the calculation of a polynomial whose coefficients are predetermined.

In a step (c), for each mining technique of said predefined list, a value over time of at least one dynamic parameter descriptive of the mining of said underground formation is estimated when this mining technique is Implementation.

In other words, an additional criterion this time dynamic can be used, in order to better characterize the choices and the final evaluation of the candidate extraction techniques. Indeed, the EOR technique to be chosen must not only be in line with the nature of the underground formation, but also prove to be relevant over time.

Said dynamic parameter descriptive of the exploitation of said underground formation is therefore preferably a rate of recovery of said hydrocarbons from said formation (as a function of time). In addition, other dynamic parameters can be used, in particular a breakthrough time (time at which the first traces of the injected fluid, such as water, tracers, etc., are detected at the level of the producing wells).

It is understood that where the static parameters are “exogenous” parameters to the exploitation of the underground formation, that is to say specific to the formation, and undergone, as opposed to the dynamic parameters which are “endogenous”, i.e. whose value is a consequence of exploitation. Thus the value of "exogenous" static parameters is a "cause" of the state of the operation, i.e. influencing the value of the endogenous parameter(s), hence the idea of the present process to refine the result of the step (b) with the dynamic parameter(s).

The relations allowing an estimation of the value of the dynamic parameter are typically multiple regressions (response surface) obtained from a series of simulations, the input data of which cover a large statistic of the reservoir properties as well as the operating conditions. injection of these products in field applications. The regressions are expressed in terms of similarity criteria, estimated or calculated, from the values of the static parameters descriptive of physical properties of said underground formation. An example of the calculation procedure is schematized in the .

As can be seen, said values of the static parameters descriptive of physical properties of said underground formation are used again, in particular extracted from the ad hoc database.

This time, work is advantageously done on “dimensionless numbers” denoted DM derived from the values of the static parameters descriptive of physical properties of said underground formation.

In general, the dynamic parameters can thus be derived from a response surface (regression) given by the equation: parameter = a ₁ *(DM ₁ ) + a ₂ *(DM ₂ ) +…. a _n *(DM _n ). The number of “dimensionless numbers” depends on the mining technique. And it is the coefficients a ₁ , a ₂ , a _n which are obtained by regression following a series of simulations (note that they can be predefined within the framework of the present method, but we will show how to calculate them).

In a known way, an adimensional number, or criterion of similarity, makes it possible to compare scales, or to apply results of the response surface (ex. recovery rate) to an underground formation which has a similar adimensional number.

The person skilled in the art may refer to the documents

- Wood, Derek: “ A Screening Model for CO2 Flooding and Storage in Dipping Gulf Coast Reservoirs Based on Dimensionless Groups ” The University of Texas at Austin, MS thesis, 2006,

- Chang, HL, Zhang, ZQ, Wang, QM et al. 2006. " Advances in Polymer Flooding and Alkaline/Surfactant/Polymer Processes as Developed and Applied in the People's Republic of China" . J Pet Technol 58 (2): 84–89. SPE-89175-MS

- and Rai, K, Russell, TJ, Lake, LW, Delshad, M., “ Oil-Recovery Predictions for Surfactant Polymer Flooding ”, SPE ATCE, New Orleans, Louisiana USA, 4-7 October 2009, SPE 124001.

Advantageously, dimensionless numbers are generated (dimensional analysis or inspectional analysis of process equations) for each mining technique and for each underground formation.

Each dimensionless number is constructed from a combination of static parameters. For example, the dimensionless number called "Aspect Ratio" is given by the formula L/A(Kz/Kx) ^1/2 ) where L=length A=width Kz = vertical permeability and Kx = horizontal permeability of the formation, the dimensionless number “dip ratio” is given by the formula L/H*tan(α), where α is the dip, and the “dimensionless time” to which we will return later. Note that some static parameters can be dimensionless numbers directly (like saturation).

It is assumed that one has on the one hand "reference" dimensionless numbers for a plurality of known underground formations (these reference dimensionless numbers can also be predetermined) and on the other hand "candidate" dimensionless numbers for the underground formation on which the present method is implemented. For dimensionless reference numbers, one can calculate min, max, average values, and normalize all values (normalization is typically calculated by subtracting from the value of the dimensionless number the difference between the min. and max. value of the base and dividing the whole by half of the difference between the min and max value of the base).

From there, one can make reservoir simulations (advantageously up to more than 600 simulations) by means of a reservoir simulator on the basis of the reference values, assuming that one knows or can model the associated changes over time of the dynamic parameter (for example, “reference” recovery rates are known). The idea is to estimate on the basis of all the reference values the coefficients a ₁ , a ₂ , a _n mentioned before which make it possible to correctly link the dimensionless reference numbers and the static reference parameters.

If the calculation of the "candidate" values of DM is impossible due to the lack of reservoir data (the static parameters), the possibility is offered of either choosing from a database the value of the dimensionless number (between a min and a max) , or let the user enter it.

Preferably, step (c) then comprises (always for each exploitation technique of said predefined list) the construction of a linear curve by pieces representing over time the value of said dynamic parameter descriptive of the exploitation of said underground formation when this mining technique is implemented. Said curve can be defined according to the coefficients a ₁ , a ₂ , a _n mentioned before and the candidate dimensionless numbers, it suffices that we have? dimensionless parameters either as a function of time, and in particular either a dimensionless time, see below.

For example, in case the dynamic parameter is a recovery rate, a piecewise linear recovery curve is generated, see .

The generation of this curve advantageously requires 4 points for each reservoir, from the estimated coefficients. These points are preferentially:

• the coordinate point (t _D0 ; 0) where t _D0 is the breakthrough time of the fluid to be injected into said exploitation technique (CO2, polymer or surfactant), it is understood that the recovery is nil as long as one does not has not started injecting.
• the coordinate points (t _D1 ; R _D1 ), (t _D2 ; R _D2 ) and (t _D3 ; R _D3 ), where the times t _D1 , t _D2 and t _{D 3} are arbitrary dimensionless times such that 0.8, 1 and 1.2, and RD ₁ , RD ₂ and RD ₃ are the corresponding estimated recovery rates.

Between these points the recovery is considered to be linear (approximations obtained from the curves generated by simulation during the establishment of the multiple regressions).

The dimensionless time here is understood in relation to a time to reach "1 HCPV", i.e. an injection of a volume of fluid equal to 100% of the volume of the reservoir (HCPV means "hydrocarbon pore volume" ). This explains why we can have dimensionless times greater than 1, i.e. we injected more than 1 HCPV of EOR fluid.

This approach is considered realistic since most industrial projects rarely exceed 1 HCPV.

From there, in a main step (d), on which the final decision-making is essentially based, an overall score of this exploitation technique is calculated, still for each exploitation technique of said predefined list, according to suitability scores, values of said dynamic parameter descriptive of the exploitation of said underground formation, and of at least one ontology characterizing said exploitation techniques.

In general, step (d) is preferably carried out in two sub-steps (d1) and (d2): in step (d1), represented by the , said problem of selecting an exploitation technique is previously decomposed into a hierarchical system by Hierarchical Process Analysis, AHP, by means of said ontology, of scores to be calculated, of elementary final scores, and in step (d2 ), represented by the , we have the aggregation, to calculate the overall score of an exploitation technique, of the elementary final scores using a multi-criteria decision-making aid method, such as the TOPSIS method (Technique for Order of Preference by Similarity to IdealSolution).

We will begin by defining this notion of ontology characterizing said exploitation techniques. By “ontology”, we mean the structured set of terms and concepts representing the meaning of a field of information, in this case that of said exploitation techniques.

More precisely, it is a question of being able to represent the preliminary knowledge on the exploitation of the reservoirs by one of the said EOR techniques by an ontological cartography.

Thus, the ontology lists and defines in particular a plurality of factors, and possibly sub-factors of said factors, of the problem of selecting an exploitation technique.

Advantageously, there is at least one factor of physical properties of said underground formation and one factor of exploitation of said underground formation (corresponding directly to the static and dynamic parameters), but more can be had as represented on the (with their sub-factors):

• economic factor
  • geological reserves
  • payback time
  • costs
  • hydrocarbon prices
  • depreciation
• operating factor (production dynamics)
  • production performance
  • injection performance
  • water production ratio (water cut)
  • hydrocarbon prices
  • maintaining pressure
• physical properties factor (formation.fluid properties)
  • the different static parameters
• numerical simulation factor
  • hydrocarbon production rate
  • hydrocarbon recovery
  • water production performance
  • pressure performance
• risk factor
  • underground formation geological analogy
  • hydrocarbon price stability
  • structural damage on formation
  • frequency of infrastructure renewal
  • uncertainty on simulation input data
• fuzzy factor
  • availability of fluid injection
  • HSE
  • Legislation and regulations
  • Public acceptance
  • Impact on employment

The invention will of course not be limited to this list of factors and sub-factors. Note that the user may have the ability to edit these factors/sub-factors.

The AHP (Hierarchical Process Analysis) method is a structured technique for organizing and analyzing complex decisions, based on mathematics and psychology.

It has in particular applications in decision-making, and proves to be particularly suitable in the context of the present method, once one has the various adequacy scores, and values of said dynamic parameter descriptive of the exploitation of said formation. underground.

Rather than prescribing an “exact” decision, AHP helps find the one that best fits the understanding of the problem. It provides a comprehensive and rational framework for structuring decision-making, representation and evaluation of the elements, in order to relate them to the objectives and to evaluate the alternatives.

AHP starts by decomposing the problem leading to a decision into a hierarchy of sub-problems that are more and more easily restricted in order to be able to be analyzed independently, which is possible thanks to the ontology based on the said factors or sub-problems. factors. The elements of the hierarchy can relate to an aspect of the decision, to a material or immaterial problem, carefully measured or estimated, well or badly understood, while relating to a decision-making. Once the hierarchy is constructed, its various elements can be systematically evaluated by comparing them to each other, two at a time, with respect to their impact on an element above them in the hierarchy.

Thus, in the first sub-step (d1), the problem of selecting an exploitation technique is resumed and reworked according to the ontology using AHP.

As such, said hierarchical system comprises as explained a plurality of said factors, and possibly sub-factors of said factors, of the problem of selecting an exploitation technique, and step (d1) comprises the construction of at least a judgment matrix comparing two by two the factors or sub-factors of a factor, then the calculation from the weighting judgment matrix(es) to be applied to each suitability score and each dynamic parameter descriptive of the exploitation of said underground formation, so as to obtain said elementary final scores.

The idea is to evaluate the different hierarchical elements with regard to their impact on the factors higher in the hierarchy.

As we see on the , step (d1) begins with a formulation of the goals of the decision-making process, in particular: “What is the best exploitation technique from said predefined list to be selected for said underground formation, all considerations combined?”.

The judgment of the relevance of such and such a factor is made by filling in a “judgment matrix” (known as the Saaty table), according to a scoring system directly derived from the AHP analysis method. This notation is a system ranging from 1 to 9. The meaning of this notation is as follows:

• 1: Both activities (factors or sub-factors) contribute equally to the given objective.
• 3: One activity is slightly favored over the other based on judgment and experience
• 5: One activity is strongly favored over the other based on judgment and experience
• 7: One activity is very strongly favored over the other (demonstrated by a few application cases)
• 9: One activity obviously favored over the other (demonstrated by numerous application cases)
• 2,4,6,8: Intermediate values
• 1/2, 1/3, ¼, etc. : The inverse of the previous statements

The “judgment matrix” compares the factors/sub-factors in pairs by forming all the possible combinations (it is a square matrix). Then weighting coefficients of the different factors/sub-factors can be obtained from this matrix.

For example, suppose we take only the following 6 factors

A1 – Economic factor

A2 - Physical properties factor

A3 - operating factor

A4 – Simulation factor

A5 – Risk factor

A6 - Fuzziness Factor

We can for example obtain the following matrix: A1 A2 A3 A4 AT 5 A6 A1 1 3 2 5 4 7 A2 0.33333333 1 1 2 2 4 A3 0.5 1 1 4 5 6 A4 0.2 0.5 0.25 1 3 4 AT 5 0.25 0.5 0.2 0.33333333 1 4 A6 0.14285714 0.25 0.16666667 0.25 0.25 1

Naturally, each diagonal element is worth 1, and for example the notation A1/A6 is worth 7 because it is obvious that the economic factor is very strongly favored compared to the fuzzy factor. On the other hand, it is considered that the factors of physical properties and exploitation contribute in equal parts to the given objective, hence the rating A2/A3 which is worth 1.

It is repeated that this matrix and the values of these boxes can, once the choice of the factors and sub-factors has been completed, be deduced from the ontology in a manner known to those skilled in the art. may define as part of the AHP a rating scale that measures the importance or difference of one item over another.

The first calculations of the AHP method will concern the calculation of the vector of judgment coefficients from the values of the judgment matrix, i.e. the relative weights of each factor (in the AHP method, we also speak of " vector of priorities”). The formula applied is as follows:

Judgment coefficient (Factor weight) = sum of normalized lines / number of factors

With the same formula, the vector of the judgment coefficients of the sub-factors is calculated for each factor. This vector is then weighted by the weight of the factor, and so on recursively. We thus obtain the weights of each factor and of the sub-factors which will intervene in the weighting of the vectors of the judgment coefficients of the alternatives calculated with the same mathematical formula.

An option in the output results is provided to assess data consistency. This test is used to check the entry of values in the comparison tables. Indeed if the factor A1 is for example evaluated 2 times more important than the factor A2, and A2 is considered 3 times more important than A3, and A3 4 times more important than A1, then the test will make it possible to say that there is an inconsistency in the data. It is measured with 2 parameters: the coherence index (CI) and the coherence ratio (CR).

To obtain the average consistency, we first multiply the comparison matrix with its vector of coefficients which thus gives us a new vector, which should be a vector of principal components (eigenvalues) if the consistency were perfect, i.e. be proportional to the vector of coefficients. Thus, if we divide the product vector obtained by the coefficient of the vector of the element of the same line, the average of this normalized vector gives the average consistency, which corresponds to the eigenvalue λ.

The calculation formula for the consistency index is: CI = (mean consistency) - number of items / (number of items -1), where number of items is the number of columns or rows of the matrix.

The formula for calculating the coherence ratio is given by: RC = Coherence index / random coherence. We note that the random coherences depend only on the number of elements and are known, and we have for example 1.12 if N=6 as in our example.

If the coherence ratio is less than or equal to 0.1 then the evaluation is considered coherent.

If we take the previous example, the normalized matrix is: A1 A2 A3 A4 AT 5 A6 A1 0.41216879 0.48 0.433213 0.39735099 0.26229508 0.26923077 A2 0.1373896 0.16 0.2166065 0.1589404 0.13114754 0.15384615 A3 0.2060844 0.16 0.2166065 0.31788079 0.32786885 0.23076923 A4 0.08243376 0.08 0.05415162 0.0794702 0.19672131 0.15384615 AT 5 0.1030422 0.08 0.0433213 0.02649007 0.06557377 0.15384615 A6 0.05888126 0.04 0.03610108 0.01986755 0.01639344 0.03846154

The vector of judgment coefficients is then: 0.37570977 0.15965503 0.24320163 0.10777051 0.07871225 0.03495081

The product of the matrix and the vector is then 2.43943534 1.04086201 1.62505969 0.69948038 0.47683428 0.21569169

And we get the average consistency 6.40232613.

Finally, IC= 8.05% and RC=7.18%<0.1 so the consistency is acceptable.

Without going into detail, we can do exactly the same work on the scale of the sub-factors of a factor and obtain a "sub-sector" of judgment coefficients, which can be multiplied by the judgment coefficient of the factor whose they are the sub-factors (for example 0.16 for the sub-factors of the physical properties factor).

Until then, we only use the ontology in practice (and not the values of scores/parameters obtained in steps (b) and (c)), so that unless the user modifies the list of factors /sub-factors, the judgment matrix(es) can be pre-computed. However, steps (b) and (c) can also independently influence the analysis.

We will now take into account these judgment matrices to weight each adequacy score and each dynamic parameter descriptive of the exploitation of said underground formation.

It is indeed possible to judge the various sub-factors more precisely according to a grid of specific relevance (noting for example from 1 to 5 each sub-factor with regard to its impact with respect to the exploitation technique concerned ), associated with each mining technique and allowing them to be compared, and depending on one or more parameters. Again, one or more particular thresholds can also be defined in dedicated databases.

An "elementary" score (because it is specific to a factor, and not yet global) is obtained by summing the products between the factor's judgment coefficient, the judgment coefficient of each sub-factor, and the score assigned for each sub-factor. -factor.

Thus, for example, if we take the physical properties factor and its 10 sub-factors, the three examples of exploitation techniques taken into account “receive” an equal elementary score of 3.2683. In detail, we have: Factor coefficient Subfactor coefficient CO2 polymer surfactant A2
Physical properties 0.16 Depth 0.017 2 2 2 Temperature 0.262 4 4 4 Porosity 0.16 3 3 3 Permeability 0.089 3 3 3 Residual saturation 0.061 3 3 3 Heterogeneity 0.149 2 2 2 Thickness 0.0001 3 3 3 API density 0.016 3 3 3 Viscosity 0.037 2 2 2 clay rate 0.05 4 4 4 Salinity 0.159 4 4 4 Score 3.2683 3.2683 3.2683

We find the value 0.16 mentioned before. The coefficients of the sub-factors are obtained by the own judgment matrix, and for each technique there is a score grid. Note that we could recursively proceed in this way with possible “sub-sub-factors”.

It is thus possible to obtain, for each factor and for each technique of said predefined list of exploitation techniques, said elementary score, and put them in the form of a “global” matrix such as: A1 A2 A3 A4 A4 A6 CO2 2,846 3.26 2.56 2,792 3.7 3,429 Polymer 2,923 3.26 3.16 3.125 3.45 3.19 Surfactant 2,436 3.26 3.24 3,792 3.6 2.71

The second part (step d2) of the final decision consists in aggregating said “single-factor” elementary scores into an overall “multi-factor” score.

It uses the technique of preference by comparison with an ideal solution (TOPSIS - “Technique for Order of Preference by Similarity to Ideal Solution”) which is a multi-criteria decision analysis. This method compares a set of solutions by identifying weights for each criterion (i.e. each factor), normalizing scores for each criterion, and calculating the geometric distance between each alternative and the ideal alternative, which is the best score for each criterion.

A strong assumption of the TOPSIS method is that the criteria are assigned monotonically increasing or decreasing. The advantage of this method is that it allows compromises between criteria, where a bad result in one criterion can be minimized by a good result in another criterion. This gives a more realistic form of modeling than non-compensatory methods, which include or exclude alternative solutions.

The TOPSIS method can be directly implemented on the global matrix presented above, but alternatively step (d2) is preferably in accordance with the , and itself advantageously comprises an AHP analysis comparing the techniques for using the list two by two on the basis of the elementary scores.

In other words, a new judgment matrix is reconstructed for each factor, so as to obtain for each technique and each factor a judgment coefficient (or rather a vector of judgment coefficients for each factor, which can be aggregated into a global matrix), in the same way as in the previous step. For each judgment matrix, ratings 1/9 to 9 are assigned based on a comparison of the elementary scores of the different techniques for the factor associated with the matrix.

For example, we see that for the factor A2 of physical properties of the formation we have the same score of 3.26 for the three techniques. The judgment matrix of this factor therefore has only 1s (and thus the vector of judgment coefficients all have values also at 0.333). If, on the other hand, we take the fuzzy factor A6, the scores are much more disparate (3.429, 3.19 and 2.71) so that we will have a matrix in which the notations CO2/polymer and polymer/surfactant are worth 2 because the first has a score slightly higher than the second for this factor, and the CO2/surfactant rating is worth 3 because the CO2 technique has a much higher score than the surfactant technique for this factor, hence a judgment matrix: CO2 Polymer Surfactant CO2 1 2 3 Polymer 0.5 1 2 Surfactant 0.3333333 0.5 1

The vector of judgment coefficients is then: 0.54 0.3 0.16

It is then possible to constitute from all the vectors of judgment coefficients a new global matrix and to normalize it.

Thus, in the example of 6 factors in columns and 3 techniques in rows (we understand that it is just tilted in relation to the global matrix at the end of step (d1)), we have the following values: A1 A2 A3 A4 A4 A6 CO2 0.47029356 0.57735027 0.25073566 0.25073566 0.84623284 0.84623284 Polymer 0.84623284 0.57735027 0.47029356 0.47029356 0.25073566 0.47029356 Surfactant 0.25073566 0.57735027 0.84623284 0.84623284 0.47029356 0.25073566

The idea is then to multiply each value of the normalized matrix by the judgment coefficient of the corresponding factor as obtained in step (d1) (for example 0.16 for the factor A2 of physical properties), so as to obtain a so-called “weighted normalized” matrix: A1 A2 A3 A4 A4 A6 CO2 0.17871155 0.09237604 0.06017656 0.02758092 0.06769863 0.02538699 Polymer 0.32156848 0.09237604 0.11287045 0.05173229 0.02005885 0.01410881 Surfactant 0.09527955 0.09237604 0.20309588 0.09308561 0.03762348 0.00752207

The decision then makes use of the maximum and minimum values of the normalized weighted matrix (or else of the simple global matrix, if one does not wish to make the second AHP) in accordance with the principle of TOPSIS. We can then, for each technique, calculate the Euclidean distances S_I ⁺and S_I ^-to extreme solutions (ideal and non-ideal), with formulas S_I ⁺= [∑_I(V_{I I}-V_I ⁺)²]^0.5and S_I ^-= [∑_I(V_{I I}-V_I ^-)²]^0.5, as well as the performance index, P_I= S_I ^-/ (S_I ⁺+ S_I ^-) as an overall score. Indeed, the higher the PI, the greater the performance of the method. In our example, we obtain the following overall scores: If+ Whether- Pi Rank CO2 0.21246819 0.0976688 0.31792148 3 polymer 0.11070767 0.23367694 0.67853479 1 surfactant 0.22897847 0.15819245 0.40858558 2

So that we can select the second exploitation technique.

As such, the method ends with a step of selecting at least said technique for exploiting said underground formation having the maximum overall score.

Then, step (e) advantageously comprises the exploitation of said hydrocarbons of said formation by implementing infrastructures relating to said selected exploitation technique, in a manner known to those skilled in the art.

Underground multi-formations

Preferably, as explained, there is a predefined set of underground formations comprising hydrocarbons, and a double selection can be made.

The method also becomes a method for selecting at least one underground formation from a predefined set of underground formations comprising hydrocarbons, i.e. a method for selecting at least a couple of underground formations from a predefined set of underground formations comprising hydrocarbons and a technique for exploiting said underground formation (from the predefined list).

The idea is to implement steps (a) to (c) for each underground formation of said predefined set, and thus step (d) can include the calculation, for each couple of an underground formation of said predefined set and of a technique for exploiting said predefined list, of an overall score of this pair as a function of the adequacy scores, of the values of said dynamic parameter descriptive of the exploitation of each underground formation, and of at least one ontology characterizing said exploitation techniques.

Finally, step (e) comprises the selection of at least said pair having the maximum overall score.

Note that in this embodiment it is possible that said predefined list only contains an EOR mining technique (which would make no sense if there was only one underground formation), the goal is thus to select the best formation where to apply this technique.

Construction of the ontology

Preferably, the method comprises a preliminary step (a0) of building said ontology characterizing said exploitation techniques by processing means 11 of the server 1 connected to said client equipment 2 (and its transmission to the client equipment 2 for implementation of the process).

This construction step can be much earlier than the selection process itself.

We repeat that by “ontology”, we mean the structured set of terms and concepts representing the meaning of a field of information, in this case that of said exploitation techniques.

This step (a0) can be implemented in any way known to those skilled in the art, but advantageously, with reference to the , the construction of the ontology involves:

• load a pre-established methodological “model” regarding EOR aspects from a list of relevant tasks (e.g. laboratory, simulation, surface facilities, optimization, uncertainties, etc.) with the respective glossary of concepts and “operational language” . In this, the operation can be assimilated to an “influence diagram” in the broad sense; Where
• load a general glossary of concepts and “operational language”.

computer program product

According to a second and a third aspect, the invention relates to a computer program product comprising code instructions for the execution of a method according to the first selection aspect, for at least one underground formation comprising hydrocarbons, of a technique for exploiting said subterranean formation, when said program is executed on a computer; and a storage medium readable by computer equipment on which this computer program product is located.

Claims (12)

Method for selecting, for at least one underground formation comprising hydrocarbons, a technique for exploiting said underground formation, from among a predefined list of possible exploitation techniques, characterized in that it comprises the implementation by data processing means (21) of a client equipment (2), steps of:

1. Obtaining a value of at least one static parameter descriptive of physical properties of said underground formation from measurements supplied from at least one sensor (10);
2. Calculation, for each mining technique of said predefined list, of an adequacy score of this mining technique with said underground formation as a function of the value obtained from at least one static parameter descriptive of the physical properties of said underground formation;
3. Estimation, for each exploitation technique of said predefined list, of a value over time of at least one dynamic parameter descriptive of the exploitation of said underground formation when this exploitation technique is implemented;
4. Calculation, for each mining technique of said predefined list, of an overall score for this mining technique as a function of said adequacy scores, of the values of said dynamic parameter descriptive of the mining of said underground formation, and of at least one ontology characterizing the implementation of said exploitation techniques;
5. Selection of at least said mining technique of said underground formation having the maximum overall score.

A method according to claim 1, wherein said predefined list of possible operating techniques includes at least one miscible CO2 injection technique, one polymer injection technique, and one surfactant injection technique. Method according to one of Claims 1 and 2, in which the said dynamic parameter descriptive of the exploitation of the said underground formation is a rate of recovery of the said hydrocarbons from the said formation. Method according to claim 3, in which step (c) comprises, for each exploitation technique of said predefined list, the construction according to the value obtained of at least one static parameter describing physical properties of said formation underground of a piecewise linear curve representing over time the rate of recovery of said hydrocarbons from said formation when this mining technique is implemented. Method according to one of Claims 1 to 4, in which the said statistical parameter descriptive of the physical properties of the said underground formation is chosen from among a depth, a temperature, a porosity, a permeability, a heterogeneity, a residual saturation in hydrocarbons, a thickness hydrocarbon zone, hydrocarbon viscosity, clay content, lithology, salinity, of said underground formation. Process according to one of Claims 1 to 5, in which step (e) comprises the exploitation of said hydrocarbons from said formation by implementing infrastructures relating to said selected exploitation technique. Method according to one of claims 1 to 6, in which said problem of selecting an exploitation technique in a hierarchical system is previously decomposed by Hierarchical Process Analysis, AHP, by means of said ontology; step (d) comprising aggregating, to calculate the overall score of a mining technique, the suitability scores and the values of said dynamic parameter descriptive of the mining of said subterranean formation in accordance with said hierarchical system using a multi-criteria decision-making method, such as Technique for Order of Preference by Similarity to Ideal Solution, TOPSIS. A method according to claim 7, wherein said hierarchical system comprises a plurality of factors, and optionally sub-factors of said factors, of the mining technique selection problem including at least one factor of physical properties of said subterranean formation and an exploitation factor of said underground formation, step (d) comprising the construction of at least one judgment matrix comparing two by two the factors or the sub-factors of a factor, then the calculation from of the weighting judgment matrix(es) to be applied to each adequacy score and each dynamic parameter descriptive of the exploitation of said underground formation. Method according to one of Claims 1 to 8, comprising a preliminary step (a0) of constructing said ontology characterizing said exploitation techniques by processing means (11) of a server (1) connected to said client equipment (2 ). Method according to one of Claims 1 to 9, also for selecting at least one underground formation from among a predefined set of underground formations comprising hydrocarbons, comprising the implementation of steps (a) to (c) for each underground formation of said predefined set, step (d) comprising the calculation, for each pair of an underground formation of said predefined set and of a technique for exploiting said predefined list, of an overall score for this pair as a function of the scores of adequacy, of the values of said dynamic parameter descriptive of the exploitation of each underground formation, and of at least one ontology characterizing said exploitation techniques; and step (e) comprising selecting at least said pair having the maximum overall score. Computer program product comprising code instructions for executing a method according to one of Claims 1 to 10 for selecting, for at least one underground formation comprising hydrocarbons, a technique for exploiting said formation underground, when said program is executed on a computer. Storage means readable by computer equipment on which a computer program product comprises code instructions for the execution of a method according to one of Claims 1 to 10 of selection, for at least one underground formation comprising hydrocarbons , of a technique for exploiting said underground formation.