METHOD FOR OPERATING A HYDROCARBON DEPOSIT BY INJECTION OF A GAS IN FOAM FORM

Abstract

Method for operating a hydrocarbon deposit by injection of gas in foam form, comprising a step of determination of a model of displacement of the foam, this model being a function of an optimal mobility reduction factor of the gas and of at least one interpolation function dependent on a parameter and constants to be calibrated.

Description

The present invention relates to the field of the operation of a fluid contained in an underground formation, more particularly the assisted recovery of a fluid, such as a hydrocarbon fluid, by foam injection.

The operation of an oil reservoir by primary recovery consists in extracting, via a so-called production well, the oil present in the reservoir through the effect of overpressure prevailing naturally in the reservoir. This primary recovery makes it possible to access only a small quantity of the oil contained in the reservoir, around 10 a 15% at the very most.

To make it possible to continue to extract the oil, secondary production methods are employed, when the pressure of the reservoir becomes insufficient to displace the oil still in place. In particular, a fluid is injected (re-injection of the water produced, diluted or not, injection of sea or river water, or even injection of gas, for example) into the hydrocarbon reservoir, in order to exert, in the reservoir, an overpressure specifically to drive the oil to the production well or wells. One standard technique in this context is water injection (also referred to as “waterflooding”), in which large volumes of water are injected under pressure into the reservoir via injection wells. The injected water drives a part of the oil that it encounters and pushes it to one or more production producing wells. The secondary production methods such as water injection however, make it possible to extract only a relatively small part of the hydrocarbons in place (typically around 30%). This partial scavenging is due in particular to the trapping of the oil by capillary forces, to the differences in viscosity and in density that exist between the injected fluid and the hydrocarbons in place, and to heterogeneity on microscopic or macroscopic scales (scale of the pores and also scale of the reservoir).

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, there are various so-called assisted recovery techniques (known by the acronym “EOR”, which stands for “Enhanced Oil Recovery”). Among these techniques, techniques can be cited that are similar to the abovementioned water injection, but that employ a water including additives such as, for example, soluble surface active agents in the water (this is called “surfactant flooding”). The use of such surface active agents induces in particular a reduction of the water/oil interfacial tension, which is specifically to ensure a more effective driving of the oil trapped in constrictions of pores.

Also known is assisted recovery by injection of gases, miscible or not (natural gas, nitrogen or CO₂). This technique makes it possible to maintain the pressure in the oil reservoir during its operation, but can also make it possible, in the case of miscible gases, to mobilize the hydrocarbons in place and thus improve the flow rate. A gas commonly used is carbon dioxide when it is available at low cost.

Also known are alternative techniques relying on an injection of foam into the oil reservoir. Because of its high apparent viscosity, the foam is considered as an alternative to gas as injection fluid in hydrocarbon reservoirs. The mobility of the foam is thus reduced relative to gas which, for its part, tends to segregate and perforate rapidly in the producing wells, notably in the heterogeneous and/or thick reservoirs. Assisted recovery by foam injection is particularly attractive because it requires the injection of smaller volumes than for other assisted recovery methods based on non-foaming fluids.

STATE OF THE ART

The following documents will be cited hereinbelow in the description:

• Ma, K., Lopez-Salinas, J. L., Puerto, M. C., Miller, C. A., Biswal, S. L., Hirasaki, G. J., 2013. Estimation of Parameters for the Simulation of Foam Flow through Porous Media. Part 1: The Dry-Out Effect. Energy & Fuels 27, 2363-2375 (ACS Publications).
• Farajzadeh, R., Lotfollahi, M., Eftekhari, A. A., Rossen, W. R. and Hirasaki, G. J., 2015. Effect of Permeability on Implicit-Texture Foam Model Parameters and the Limiting Capillary Pressure. Energy Fuels 29, 3011-3018 (ACS Publications).
• Kapetas, L., Vincent-Bonnieu, S., Farajzadeh, R., Eftekhari, A. A., Mohd-Shafian, S. R., Kamarul Bahrim, R. Z. and Rossen, W. R., 2015. Effect of Permeability on Foam-Model Parameters—An Integrated Approach from Coreflood Experiments through to Foam Diversion Calculations. 18th European Symposium on IOR, Dresden, 14-16 April.

Oil operation of a deposit consists in selecting the areas of the deposit that exhibit the best oil potential, in defining optimal operating schemes for these areas (notably using a numerical simulation of the flows in the deposit, in order to define the type of recovery, the number and positions of the operation wells, allowing for an optimal hydrocarbon recovery), in drilling operation wells and, in general, in putting in place the production infrastructures necessary to the development of the deposit.

Defining an operating scheme of an oil reservoir including a step of assisted recovery by foam injection may entail numerically simulating, in the most realistic way possible, the flows in the presence of foam in the reservoir considered. Such a simulation is performed using a flow simulator comprising a model of displacement of the foam.

Such a model may involve evaluating the performance levels of the foam in terms of mobility reduction. In general, this estimation involves conducting laboratory trials consisting in measuring the headlosses in displacements of foam on the one hand, of water and of non-foaming gas on the other hand in a sample of the oil reservoir. Then, this model of displacement of the foam, representative of the flows on a laboratory scale, is calibrated to the scale of the reservoir before carrying out the numerical simulations of the flows, in order to predict the benefit accrued by the injection of the foam in terms of improvement of the effectiveness of displacement of the fluids in place.

The foam displacement models used by the industry are relatively simple models which, subject to the conditions of existence of the foam, simulate only the effects of the foam in terms of mobility reduction and not the foam generation-destruction processes. Generally, the foam displacement models depend nonlinearly on numerous parameters (calibration constants). Determining the parameters of this model therefore involves solving a nonlinear inverse problem. However, the complexity of the displacement of a foam in a confined environment that any natural porous medium constitutes makes the modeling difficult because the great number of parameters influencing the foam can lead to indeterminacies (multiple solutions).

The approach proposed in the document (Ma et al., 2013) is known, which consists in simultaneously determining certain parameters of the foam displacement model by a graphic approach, completed by a numerical adjustment.

Also known is the technique proposed in the document (Farajzadeh et al., 2015) which proceeds with the determination of the unknown parameters (calibration constants) of the foam displacement model by an iterative least squares approach. However, since the problem posed is nonlinear with respect to these unknowns, there is no unique solution, or in other words, the parameters thus determined are one solution out of other possible ones (see for example Kapetas et al., 2015).

The method according to the invention aims to determine, pragmatically, the parameters of the foam displacement model. Unlike the existing methods, the method according to the invention consists, from experimental data, in a sequential adjustment of the parameters of the foam model, and not in an overall adjustment. Thus, the method according to the invention makes it possible to minimize the numeric adjustments, by trying to extract the maximum of information on the dynamic behavior of foam from experimental data.

THE METHOD ACCORDING TO THE INVENTION

Thus, the present invention relates to a method for operating an underground formation comprising hydrocarbons, by means of an injection of an aqueous solution comprising a gas in foam form and a flow simulator relying on a displacement model of said gas in foam form, said displacement model being a function of an optimal mobility reduction factor of said gas and of at least one interpolation function of said optimal mobility reduction factor, said interpolation function being a function of at least one parameter relating to at least one characteristic of the foam and of at least one constant. The method according to the invention is implemented from at least one sample of said formation, measurements of conventional relative permeabilities to said gas in non-foaming form and measurements of conventional relative permeabilities to said aqueous phase, and comprises at least the following steps:

• • A. said displacement model of said simulator is determined according to at least the following steps:
    • i. a plurality of values of said parameter is defined relative to at least one of said interpolation functions, an injection is performed into said sample of said gas in non-foaming form and of said gas in foam form according to said values of said parameter relative to said function, and a headloss with foam and a headloss without foam are measured respectively for each of said values of said parameter relative to said function;
    • ii. from said measurements of headloss relative to said interpolation function, an optimal value of said parameter relative to said function is determined, said optimal value making it possible to maximize a ratio between said headlosses without foam and said headlosses with foam measured for said function;
    • iii. from said measurements of headloss performed at said optimal value determined for at least said interpolation function, from said measurements of conventional relative permeabilities to said gas in non-foaming form and from said measurements of conventional relative permeabilities to said aqueous phase, said optimal mobility reduction factor is determined;
    • iv. for at least said interpolation function, from said optimal mobility reduction factor, from said measurements of headloss relative to said interpolation function, from said measurements of conventional relative permeabilities to said gas in non-foaming form and from said measurements of conventional relative permeabilities to said aqueous phase, said constants of said interpolation function are calibrated;
  • B—from said displacement model and from said flow simulator, an optimal operating scheme for said deposit is determined and said hydrocarbons are exploited.

According to one implementation of the invention, said displacement model of the foam can be expressed in the form:

k _rg ^FO(S_g)=FM k_rg(S_g),

in which k_rg^FO(S_g) is the relative permeability to said gas in foam form for a given gas saturation value S_g, k_rg(S_g) is the relative permeability to said non-foaming gas for said gas saturation value S_g, and FM is a functional expressed in the form:

$FM=\frac{1}{\left(M^{\mathrm{opt}}-1\right)*\prod _kF_k}$

in which M^opt is said optimal mobility reduction factor of said gas and F_k is one of said interpolation functions, with k≧1.

According to one embodiment of the invention, there can be four of said interpolation functions and said parameters of said functions can be a foaming agent concentration, a water saturation, an oil saturation, and a gas flow rate.

Advantageously, said interpolation function F_k of a parameter V_k can be written in the form:

$F_k\left(V_k\right)=\frac{M_k^{\mathrm{opt}}-1}{M^{\mathrm{opt}}-1}$

in which M^opt is said optimal mobility reduction factor and M_k^opt is an optimal mobility reduction factor for said parameter V_k.

According to one implementation of the invention, prior to the step iii), optimal conditions can be defined as corresponding to said optimal values determined for each of said interpolation functions, said gas in non-foaming form and said gas in foam form can be injected into said sample according to said optimal conditions, and a headloss with foam and a headloss without foam can be respectively measured.

Advantageously, said constants of at least one of said interpolation functions can be calibrated by a least squares method, such as an inverse method based on the iterative minimization of a functional.

Other features and advantages of the method according to the invention will become apparent on reading the following description of nonlimiting exemplary embodiments, with reference to the figure attached and described hereinbelow.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 presents an example of the trend of a mobility reduction factor R as a function of the gas flow rate Q.

DETAILED DESCRIPTION OF THE METHOD

In general, one of the objects of the invention relates to a method for operating an underground formation comprising hydrocarbons, by means of an injection of an aqueous solution comprising a gas in foam form, and in particular the determination of an optimal operating scheme for the underground formation studied. In particular, the method according to the invention targets the determination of the parameters of a model of displacement of the gas in foam form. Hereinbelow, foam describes a phase dispersed in another phase by the addition of a foaming agent in one of the two phases. One of the phases can be water and the other phase is a gas, such as natural gas, nitrogen or CO_2.

The method according to the invention requires the availability of:

• • a sample of the underground formation studied, taken by in situ coring for example;
  • a flow simulator relying on a model of displacement of the gas in foam form (see below);
  • measurements of conventional relative permeabilities to the gas in non-foaming form and measurements of conventional relative permeabilities to the aqueous phase: these can be measurements performed expressly for the requirements of the method according to the invention (the expert has perfect knowledge of how to conduct such laboratory experiments), but they can also be analytical functions calibrated from correlations well known to the expert.

Thus, the method according to the invention requires the availability of a flow simulator comprising a model of displacement of the foam. According to the invention, the model of displacement of the foam relies on the hypothesis that the gas present in foam form has its mobility reduced by a given factor in set conditions of formation and of flow of the foam. The formulation of such a model, used by many flow simulators, consists in modifying the relative permeabilities to the gas when the gas is present in foam form which, for a given gas saturation S_g, is expressed according to a formula of the type:

k _rg ^FO(S_g)=FM k_rg(S_g)  (1)

in which k_rg^FO(S_g) is the relative permeability to the gas in foam form, that is expressed as the product of a function FM by the relative permeability to the non-foaming gas k_rg(S_g) for the same gas saturation value S_g (later denoted S_rg^FO). One assumption underpinning the current foam models is that the relative permeability to water (or to liquid by extension) is assumed unchanged, whether the gas is present in continuous phase form or in foam form. Given this assumption, the gas mobility reduction functional, hereinafter denoted FM, is expressed according to a formula of the type:

$\begin{array}{cc}FM=\frac{1}{1+\left(M_{\mathrm{mod}}^{\mathrm{opt}}-1\right)*\prod _kF_k\left(V_k\right)}& \left(2\right)\end{array}$

in which:

• • M_mod^opt is the optimal mobility reduction factor, that is to say the ratio of the relative permeabilities to the gas (k_rg) and to the foam (k_rg^FO) in optimal conditions for reducing the mobility of the gas, that is to say the conditions in which the terms F_k(V_k) defined hereinbelow have the value 1, i.e.:

$\begin{array}{cc}{M}_{\mathrm{mod}}^{\mathrm{opt}}=\frac{k_{\mathrm{rg}}\left(S_{g,\mathrm{opt}}^{\mathrm{FO}}\right)}{k_{\mathrm{rg}}^{\mathrm{FO}}\left(S_{g,\mathrm{opt}}^{\mathrm{FO}}\right)}=\frac{1}{FM_{\mathrm{opt}}}& \left(3\right)\end{array}$

• • the terms F_k(V_k) (with k equal to or greater than 1) are the values of the interpolation functions F_k of the mobility reduction factor between the value M_mod^opt and 1, which each depend on a parameter V_k relative to at least one characteristic of the foam, and which involve a certain number of calibration constants to be calibrated as explained hereinbelow.

In order to provide a model of displacement of the foam to the simulator that is representative of the reality, the method according to the invention aims to determine, reliably from representative displacement measurements, the following modelling data:

• • the optimal mobility reduction factor of the gas M_mod^opt as defined according to the equation (3);
  • the calibration constants of each of the functions F_k considered in the definition of the model of displacement of the foam according to the equations (1) and (2).

According to one implementation of the invention, the parameter V_k can notably be the foaming agent concentration C_s^w, the water saturation S_w, the oil saturation S_o, or even the gas flow rate u_g.

According to one implementation of the invention, the gas mobility reduction functional, denoted FM, comprises four interpolation functions F_k(V_k) and each of these functions comprises two constants to be calibrated from experimental data. According to an implementation of the invention in which the gas mobility reduction functional comprises four interpolation functions F_k(V_k), the following are defined:

• • the interpolation function F₁ relative to the parameter V₁=C_s^w (foaming agent concentration C_s^w) by a formula of the type:

$\begin{array}{cc}{F}_1={\left(\frac{\mathrm{Min}\left(C_s^w,C_s^{w\text{-}\mathrm{ref}}\right)}{C_s^{w\text{-}\mathrm{ref}}}\right)}^{e_s}& \left(4\right)\end{array}$

• • and for which the constants to be calibrated are the exponent e_s, and the constant C_s^w-ref which corresponds to the foaming agent concentration in reference optimal conditions;
  • the interpolation function F₂ relative to the parameter V₂=s_w (water saturation), by a formula of the type:

$\begin{array}{cc}{F}_2=\left[0.5+\frac{\arctan \left[f_w\left(S_w-S_w^{*}\right)\right]}{\pi }\right]& \left(5\right)\end{array}$

• • and for which the constants to be determined are the constant f_w which governs the transition (according to the water saturation) between the foaming and non-foaming states and the constant S_w* which represents the transition water saturation between stable and unstable foaming states;
  • the interpolation function F₃ relative to the parameter V₃=S_o (oil saturation) by a formula of the type:

$\begin{array}{cc}{F}_3={\left(\frac{\mathrm{Max}\left[0;S_o^{*}-S_o\right]}{S_o^{*}}\right)}^{e_0}& \left(6\right)\end{array}$

• • in which S_o* is the oil saturation beyond which the foam loses all its faculties to reduce the mobility of the gas, and the exponent e_o is a constant to be determined;
  • the interpolation function F₄ relative to the parameter V₄=u_g (gas flow rate) by a formula of the type:

$\begin{array}{cc}{F}_4={\left(\frac{N_c^{*}}{\mathrm{Max}\left(N_c,N_c^{*}\right)}\right)}^{e_c}\mathrm{with}N_c=\frac{{\mu }_gu_g}{{\mathrm{\phi \sigma }}_{\mathrm{vg}}\left(C_s^w\right)}& \left(7\right)\end{array}$

• • in which N_c* is the reference value of the capillary number N_c calculated for the reference optimal flow rate. The variables involved in the calculation of N_c are the velocity of the gas u_g, the porosity φ of the formation considered, the water-gas interfacial tension σ_gw (which is dependent on the foaming agent concentration C_s^w), and the viscosity of the gas μ_g. The exponent e_c is also a constant to be calibrated.

Generally, it can be shown that any interpolation function F_k of the parameter V_k can be written in the form:

$\begin{array}{cc}{F}_k\left(V_k\right)=\frac{\frac{1}{FM}-1}{\frac{1}{FM_{\mathrm{opt}}}-1}=\frac{M_{\mathrm{mod}}\left(V_k\right)-1}{M_{\mathrm{mod}}^{\mathrm{opt}}-1}& \left(8\right)\end{array}$

in which M_mod(V_k) is the reduction of mobility for a value V_k of the parameter k affecting the foam (and for optimal values of the other parameters V_j, j being different from k) and in which M_mod^opt=M_mod(V_k^opt) is the reduction of mobility obtained for the optimal value V_k^opt of the parameter V_k. The method according to the invention thus consists, for each parameter V_k affecting the foam, in determining the factors M_mod(V_k) for various values of this parameter, and M_mod^opt, then in determining, from these factors, the constants of the interpolation function F_k considered.

According to an implementation of the invention in which the functional FM defined in the equation (2) involves the interpolation functions F₁, F₂, F₃ and F₄ defined in the equations (4) to (7), the determination of the model of displacement of the foam entails calibrating the 8 constants: C_s^w-ref, e_s, f_w, S_w*, S_o*, e_o, N_c^ref, e_c.

According to the invention, the determination of the constants of the interpolation functions F_k involved in the equation (2) is performed via a calibration, interpolation function by interpolation function (and not globally, for all the functions), from experimental measurements relative to each of the interpolation functions, performed in the optimal conditions established for the other interpolation functions.

The method according to the invention comprises at least the following steps, the step 1 being able to be repeated for each of the interpolation functions of the model of displacement of the foam, and the step 2 being able to be optional:

1. Laboratory measurements relative to an interpolation function

• • 1.1. Definition of values of the parameter relative to the interpolation function
  • 1.2. Injections with/without foam and measurements of headlosses
  • 1.3. Determination of an optimal parameter value

2. Laboratory measurements according to optimal conditions

3. Determination of the foam displacement model

• • 3.1. Determination of the optimal mobility reduction factor
  • 3.2. Calibration of the constants of the interpolation functions

4. Operation of the hydrocarbons of the formation

The various steps of the method according to the invention are detailed hereinbelow.

1. Laboratory Measurements Relative to an Interpolation Function

During this step, laboratory experiments are performed relative to a given interpolation function F_k of the model of displacement of the foam defined according to the equations (1) and (2). According to one implementation of the invention, this step is repeated for each of the interpolation functions involved in the model of displacement of the foam defined according to the equations (1) and (2). It should be noted that the model of displacement of the foam can however comprise only a single interpolation function (case for which k=1). Optimal values are adopted for the other parameters affecting the foam in such a way that the other interpolation functions F_j, j different from k, have the value 1 or are invariant in these experiments relative to the interpolation function F_k.

During this step, applied to each interpolation function independently of one another, a plurality of values of the parameter is defined relative to the interpolation function considered, then an injection into said sample of the gas in non-foaming form and of the gas in foam form is performed according to the values of the parameter relative to the interpolation function considered, and a headloss with foam and a headloss without foam are measured respectively for each of the values of the parameter relative to this function. This step is detailed hereinbelow for a given interpolation function F_k.

1.1. Definition of Parameter Values Relative to the Interpolation Function

During this substep, the aim is to define a plurality of values V_k,i (with i lying between 1 and I, and I>1) of the characteristic parameter V_k of the interpolation function F_k considered.

According to one implementation of the invention, the aim is to define a range of values of this parameter and a sampling step for this range.

According to one implementation of the invention, the plurality of values of the parameter V_k relative to the interpolation function F_k considered are defined from the possible or realistic values of the parameter considered (for example, a mass concentration of foaming agent less than 1% in all cases) and so as to sample in an ad hoc manner the curve representative of the interpolation function considered (an interpolation function that has a linear behavior does not need a high number of measurements, unlike other types of function). The expert in foam injection-assisted recovery has a perfect knowledge of how to define a plurality of ad hoc values of the parameters of each of the interpolation functions F_k. According to an implementation of the invention in which the interpolation function considered relates to the fluid flow rate (parameter V₄ of the function F₄ of the equation (7)), an injection flow rate on coring of between 10 and 40 cm³/h is for example chosen, with a step of 10 cm³/h.

1.2. Injections with/without Foam and Headloss Measurements

During this substep, at least two series of experiments are carried out on at least one sample of the underground formation for the interpolation function F_k considered:

• • injection of gas in non-foaming form (more specifically, a co-injection of water and of gas in non-foaming form) into the sample considered for each of the values V_k,i of the parameter V_k relative to the function F_k considered. The flow rates of gas and of water adopted for each of these co-injections are the same as the flow rates of gas and of water injected in foam form in the tests which follow these co-injections. For example, in the case of the interpolation function F₄ of the equation (7), the flow rate only is made to vary in the sample considered, the parameters of the other interpolation functions F₁, F₂, F₃ (for example, the foaming agent concentration, the quality of the foam and the oil saturation) being fixed. During each of the experiments of this first series, a headloss (that is to say a pressure difference) is measured, that is denoted ΔP_k,i^NOFO, for each value V_k,i;
  • injection of foam: the same experiment is repeated, for the same values of the parameter considered (for example the flow rate for the interpolation function F₄ according to the equation (7)), but by this time injecting the water and the gas in foam form. During each of the experiments of this second series, a headloss (that is to say a pressure difference) is measured, that is denoted ΔP_k,i^FO, pr for each value V_k,i;

According to one implementation of the invention, the injections of gas in non-foaming form and in foam form are performed on samples of the formation initially saturated with a liquid phase (such as water and/or oil), the latter being able to be mobile or residual depending on the history of the coring and the measurement objectives (checking mobility of the gas in secondary or tertiary injection, after injection of water). The displacements studied are then draining processes in which the saturation of the gas phase increases in all cases.

According to a variant implementation of the invention, it is possible to measure, in addition to the headlosses, the productions of liquid phase (water and/or oil) and of gas, and, possibly, the gas saturation profiles during the transitional period of the displacement and in the steady state. These optional measurements make it possible to validate the model once the interpolation functions F_k are calibrated.

1.3. Determination of an Optimal Parameter Value

During this substep, the aim is to determine the value V_k^opt, that will hereinafter be called optimal value, maximizing the ratio between the headlosses without foam ΔP_k,i^NOFO and the headlosses with foam ΔP_k,i^FO relative to the interpolation function F_k considered and measured during the preceding substep. Thus, if M_lab^k,i is used to denote the ratio of the headlosses measured in the presence and in the absence of foam for the value V_k,i of the parameter V_k, i.e.

$M_{\mathrm{lab}}^{k,i}=\frac{\Delta P_{k,t}^{\mathrm{FO}}}{\Delta P_{k,i}^{\mathrm{NOFO}}}=\frac{k_{\mathrm{rg}}\left(S_{g\left(k,i\right)}^{\mathrm{NOFO}}\right)}{k_{\mathrm{rg}}^{\mathrm{FO}}\left(S_{g\left(k,i\right)}^{\mathrm{FO}}\right)},$

it is then possible to define the optimal value V_k,iopt as the value V_k,i which maximizes M_lab^k,i whose value is then denoted as follows:

$\begin{array}{cc}{M}_{\mathrm{lab}}^{k,\mathrm{iopt}}=M_{\mathrm{lab}}^{\mathrm{kopt}}=\underset{i}{\mathrm{Max}}M_{\mathrm{lab}}^{k,i}& \left(9\right)\end{array}$

According to a preferred implementation of the invention, the step 1 as described hereinabove is repeated for each of the parameters V_k relative to each of the interpolation functions F₁, taken into consideration for the implementation of the method according to the invention. Thus, at the end of such a repetition, an optimal value V_k^opt is obtained for each parameter V_k.

Subsequently, “optimal conditions” is the term used to denote the set of the values V_k^opt determined on completion of the step 1, the latter if necessary being repeated for each of the interpolation functions taken into consideration for the implementation of the method according to the invention.

2. Laboratory Measurements According to Optimal Conditions

During this step, the aim is to perform two types of experiments on at least one sample of the underground formation, by injecting gas in non-foaming form, and gas in foam form, similarly to the substep 1.2, but this time in the optimal conditions determined on completion of the substep 1.3, this substep being repeated if necessary for each of the interpolation functions taken into consideration for the definition of the model of displacement of the foam according to the equations (1) and (2)). In other words, the following measures are carried out:

• • injection of gas in non-foaming form (more specifically, a co-injection of water and of gas in non-foaming form) into the sample considered, this injection being performed in the optimal conditions (defined by the set of the optimal values V_k^opt determined for each parameter V_k) determined on completion of the step 1. During this first experiment, a headloss (that is to say a pressure difference) is measured, that is then denoted ΔP_opt^NOFO;
  • injection of foam (that is to say an injection of gas and of water, with an addition of a foaming agent into one of the water or gas phases) into the sample considered, this injection being performed in the optimal conditions (defined by the set of the optimal values V_k^opt determined for each parameter V_k) determined on completion of the step 1. During this second experiment, a headloss (that is to say a pressure difference) is measured, that is then denoted ΔP_opt^FO.

Subsequently, M_lab^opt will be used to denote the optimal mobility reduction factor relative to the laboratory measurements, defined by a formula of the type:

$\begin{array}{cc}{M}_{\mathrm{lab}}^{\mathrm{opt}}=\frac{\Delta P_{\mathrm{opt}}^{\mathrm{FO}}}{\Delta P_{\mathrm{opt}}^{\mathrm{NOFO}}}=\frac{k_{\mathrm{rg}}\left(S_{g,\mathrm{opt}}^{\mathrm{NOFO}}\right)}{k_{\mathrm{rg}}^{\mathrm{FO}}\left(S_{g,\mathrm{opt}}^{\mathrm{FO}}\right)}& \left(10\right)\end{array}$

This step is not necessary in practice if the precaution to perform the experiments of the step 1 relative to each of the parameters V_k by adopting optimal values V_j,j≠k^opt of the other parameters V_j affecting the foam has indeed been taken. This step nevertheless makes it possible to refine the value of M_lab^opt, if the assumed optimal conditions of the parameters V_j,j≠k were not perfectly satisfied.

3. Determination of the Foam Displacement Model

3.1. Determination of the Optimal Mobility Reduction Factor

During this substep, the aim is, from the headloss measurements performed in the optimal conditions, from measurements of conventional relative permeabilities to the gas in non-foaming form and from measurements of conventional relative permeabilities to the aqueous phase, to determine an optimal mobility reduction factor, that is to say the factor of reduction of the relative permeabilities to the gas when, present at a given saturation within the porous medium, it circulates in foam form or in continuous phase form (in the presence of water).

According to one implementation of the invention, the optimal mobility reduction factor is determined according to at least the following steps:

• • from the conventional relative permeabilities to the gas k_rg and to the aqueous phase k_rw, the gas saturation is calculated in steady state conditions of flow of non-foaming gas and water S_g^NOFO according to a formula of the type:

$\begin{array}{cc}{S}_g^{\mathrm{NOFO}}={\left(\frac{k_{\mathrm{rg}}}{k_{\mathrm{rw}}}\right)}^{-1}\left(\frac{f_g}{1-f_g}\frac{{\mu }_g}{{\mu }_w}\right)& \left(11\right)\end{array}$

• • in which f_g is the fractional gas flow rate (ratio of the gas flow rate to the total flow rate), μ_g and μ_w are, respectively, the viscosity of the gas and of the water;
  • from the ratio of headlosses measured in the optimal conditions as defined on completion of the step 1 (the step 1 being able to be repeated if necessary for each of the interpolation functions F_k considered), from the gas saturation in steady state conditions of flow of non-foaming gas and water S_g^NOFO, the gas saturation in the presence of foam S_g^FO is calculated according to a formula of the type:

$\begin{array}{cc}{S}_{g,\mathrm{opt}}^{\mathrm{FO}}=1-{\left(k_{\mathrm{rw}}\right)}^{-1}\left\{\frac{k_{\mathrm{rw}}\left(S_w^{\mathrm{NOFO}}=1-S_g^{\mathrm{NOFO}}\right)}{M_{\mathrm{lab}}^{\mathrm{opt}}}\right\}& \left(12\right)\end{array}$

• • This relationship devolves from the known assumption of invariance of the functions of relative permeability to water flowing in the form of foam films or in conventional continuous form.
  • from the gas saturation in steady state conditions of flow of non-foaming gas and water S_g^NOFO from the gas saturation in the presence of foam S_g^FO, in the optimal conditions, from the factor M_lab^opt determined in the optimal conditions (see step 2), the mobility reduction factor M_mod^opt is determined according to a formula of the type:

$\begin{array}{cc}{M}_{\mathrm{mod}}^{\mathrm{opt}}=M_{\mathrm{lab}}^{\mathrm{opt}}\frac{k_{\mathrm{rg}}\left(S_{g,\mathrm{opt}}^{\mathrm{FO}}\right)}{k_{\mathrm{rg}}\left(S_{g,\mathrm{opt}}^{\mathrm{NOFO}}\right)}& \left(13\right)\end{array}$

3.2. Calibration of the Constants of the Interpolation Functions

During this substep, the constants of each of the interpolation functions F_k considered are calibrated, from the optimal mobility reduction factor M_mod^opt, from the headloss measurements relative to the interpolation function considered, from the measurements of conventional relative permeabilities to the gas in non-foaming form and from the measurements of conventional relative permeabilities to the aqueous phase.

According to one implementation of the invention, the procedure described in the substep 3.1 is applied beforehand to the ratios of the headlosses M_lab^k,i measured in the presence and in the absence of foam for the different values V_k,i of the parameter V_k. Thus, the mobility reduction factors M_mod^k,i relative to the values V_k,i of the parameter V_k are determined according to a formula of the type:

$\begin{array}{cc}{M}_{\mathrm{mod}}^{k,i}=M_{\mathrm{lab}}^{k,i}\frac{k_{\mathrm{rg}}\left(S_{g,\left(k,i\right)}^{\mathrm{FO}}\right)}{k_{\mathrm{rg}}\left(S_{g\left(k,i\right)}^{\mathrm{NOFO}}\right)}& \left(14\right)\end{array}$

in which the gas saturation in the presence of foam S_g(k,i)^FO for the values V_k,i of the parameter V_k is obtained according to a formula of the type:

$\begin{array}{cc}{S}_{g\left(k,i\right)}^{\mathrm{FO}}=1-{\left(k_{\mathrm{rw}}\right)}^{-1}\left\{\frac{k_{\mathrm{rw}}\left(S_{w\left(k,i\right)}^{\mathrm{NOFO}}=1-S_{g,\left(k,i\right)}^{\mathrm{NOFO}}\right)}{M_{\mathrm{lab}}^{k,i}}\right\}& \left(15\right)\end{array}$

Advantageously, this operation is repeated for each of the interpolation functions F_k. Then, the constants of each of the interpolation functions F_k considered are calibrated, from the optimal mobility reduction factor M_mod^opt and from the values of the mobility reduction factors M_mod^k,i relative to each interpolation function determined as described above. In the case of the function F₄ for example, a value of the exponent e_c is determined which most closely adjusts the values of M_mod^4,i corresponding to the values V_4,i of the parameter studied (flow rate in this example), which is formulated as follows:

$F_4\left(V_{4,i}\right)={\left(\frac{N_c^{*}}{\mathrm{Max}\left(N_{c,i},N_c^{*}\right)}\right)}^{e_c}=\frac{M_{\mathrm{mod}}^{4,i}-1}{M_{\mathrm{mod}}^{\mathrm{opt}}-1}$

According to one implementation of the invention, this calibration, interpolation function by interpolation function, can be performed by a least squares method, such as for example an inverse method based on the iterative minimization of a functional. The expert has a perfect knowledge of such methods. Advantageously, the implementation of a least squares method and in particular the iterative minimization of a functional, is performed by means of a computer.

According to another implementation of the invention, such a calibration is carried out, interpolation function by interpolation function, graphically. The expert has a perfect knowledge of such methods for calibrating constants of a function from a series of values of said function.

Thus, on completion of this step, there is a model of displacement of the foam that is calibrated and suitable for use by an ad hoc flow simulator.

4. Operation of the Hydrocarbons

During this step, the aim is to define at least an optimal scheme for operating the fluid contained in the formation, that is to say an operating scheme that allows for an optimal operation of a fluid considered according to technical and economic criteria predefined by the expert. It can be a scenario offering a high rate of recovery of the fluid, over a long period of operation, and requiring a limited number of wells. Then, according to the invention, the fluid of the formation studied is operated according to this optimal operation scheme.

According to the invention, the determination of said operational scheme is performed using a flow simulation that makes use of the foam displacement model established during the preceding steps. One example of flow simulator that makes it possible to take account of a foam displacement model is the PumaFlow software (IFP Energies nouvelles, France).

According to the invention, at any instant t of the simulation, the flow simulator solves all the flow equations specific to each mesh and delivers solution values of the unknowns (saturations, pressures, concentrations, temperature, etc.) predicted at that instant t. The knowledge of the quantities of oil produced and of the state of the deposit (distribution of the pressures, saturations, etc.) at the instant considered devolves from this resolution. According to one implementation of the invention, different schemes for operating the fluid of the formation studied are defined and the flow simulator incorporating the foam displacement model determined on completion of the step 3 is used to estimate the quantity of hydrocarbons produced according to each of the different operating schemes.

An operating scheme relative to a foam injection-assisted recovery can notably be defined by a type of gas injected into the formation studied and/or by the type of foaming agent added to this gas, by the quantity of foaming agent, etc. An operating scheme is also defined by a number, a geometry and a layout (position and spacing) of the injecting and producing wells in order to best take account of the impact of the fractures on the progression of the fluids in the reservoir. In order to define an optimal operating scheme, various tests of different production scenarios can be performed using a flow simulator. The operating scheme that offers the best fluid recovery rate for the lowest cost will for example be preferred. By selecting various scenarios, characterized for example by various respective layouts of the injecting and producing wells, and by simulating the fluid production for each of them, it is possible to select the scenario that makes it possible to optimize the production of the formation considered according to the technical and economic criteria predefined by the expert. The operating scheme offering the best fluid recovery rate for the lowest cost will for example be considered as the optimal operating scheme.

The experts then operate the fluid of the formation considered according to the scenario that makes it possible to optimize the production from the deposit, notably by drilling injecting and producing wells defined by said optimal operating scheme, and to produce the fluid according to the recovery method defined by said optimal operating scheme.

Production Example

The features and advantages of the method according to the invention will become more clearly apparent on reading about the following exemplary application.

More specifically, the present invention has been applied to an underground formation in which the reservoir rock consists of sandstone, of the Berea sandstone type. An assisted recovery of the hydrocarbons contained in the reservoir based on an injection of foaming CO2 is trialed.

For this example, a functional FM of the foam displacement model is used according to the equation (2) defined by the four interpolation functions according to the equations (4) to (7). As prescribed in the method according to the invention, the calibration of the constants of the interpolation functions is carried out interpolation function by interpolation function. Only the calibration of the interpolation function F₄ (see equation (7)) is detailed hereinbelow, but the same principle can be applied to the other interpolation functions.

According to the step 1.2 described above, a series of co-injections of gas and of water and of injections of foam were carried out in the laboratory, on a sample of the reservoir rock originating from the formation studied. The characteristics of this sample and the measurement conditions are presented in Table 1. A non-dense gaseous mixture consisting of 62% CO2 and 38% methane at a temperature of 100° C. and a pressure of 100 bar was injected. These displacements were performed with fixed fractional gas flow rate (equal to 0.8) and for different successive total flow rates (10, 20, 30 and 40 cm³/h). The oil is absent for this series of tests and the headlosses in steady state conditions of flow of water and of gas on the one hand, of foam on the other hand, were measured in the same conditions.

The conventional relative permeabilities required to resolve the equations (11) and (12) are analytical functions defined as power functions (called Corey functions) with exponents equal to approximately 2.5 for the gas and 3.9 for the water with an irreducible drainage water saturation equal to 0.15, and limit points equal to 0.2 for the gas and 1 for the water, i.e.:

$k_{\mathrm{rg}}\left(S_g^{\mathrm{NOFO}}\right)=0.2{\left(\frac{S_g^{\mathrm{NOFO}}}{0.85}\right)}^{2.5}$ $k_{\mathrm{rw}}\left(S_w^{\mathrm{NOFO}}\right)={\left(\frac{S_w^{\mathrm{NOFO}}-0.15}{0.85}\right)}^{3.9}$

These curves of relative permeabilities were estimated beforehand from literature data and checked afterwards by comparison of the headloss values calculated and measured during the co-injections of gas and of water. In this way the model of relative permeability to the foam does indeed return the gas mobility reductions, i.e. the ratios of relative permeability in the absence and in the presence of foam, but without necessarily well reproducing the real diphasic behavior (transient states in particular).

Table 1 presents the headlosses (pressure gradient) with and without foam for four values of the parameter V₄=u_g of the equation (7). From these values, the value of

$M_{\mathrm{lab}}^{\mathrm{kopt}}=\underset{i}{\mathrm{Max}}M_{\mathrm{lab}}^{k,i}$

is deduced therefrom. This value (M_lab^4opt in this example), equal to 83, was obtained for a flow rate V₄^opt equal to 20 cm³/h (see substep 1.3). These laboratory experiments were repeated for the other parameters of the other interpolation functions F₁, F₂, and F₃. The optimal conditions are then determined for all of the interpolation functions.

According to the step 2, measurements are performed with and without foam in the duly determined optimal conditions. The optimal mobility reduction factor M_mod^opt is then determined, in accordance with the step 3.1, as are the values of the mobility reduction factors M_mod^k,i relative to the sampled values V_k,i of the parameter V_k, in accordance with the step 3.2. The calibration of each of the interpolation functions is then carried out. In particular, the constant e_c of the function F₄ was calibrated and a value close to 0.6 was determined.

FIG. 1 shows by a solid line the trend of the mobility reduction factor R as a function of the parameter V₄ of the function F₄ (flow rate Q) deduced from the method according to the invention. The comparison with the mobility reduction factor deriving from the flow simulation (dotted line curve) shows a good consistency with the foam displacement model of according to the invention.

Thus, the method according to the invention allows for a reliable determination of the foam displacement model from experimental data produced and processed according to a sequential and systematic approach, parameter by parameter, and not by the overall adjustment of a set of measurements varying one or more parameters simultaneously. Moreover, given the parametric complexity of the behavior of the foams, the experiments according to the method according to the invention are carried out in conditions as close as possible to the reservoir conditions.

TABLE 1
Berea
sandstone	Total flow rate [cm3/h]	10	20	30	40
P = 100 bar	ΔPk,iFO	5.2	10.3	11.9	13.6
T = 100° C.	ΔPk,iNOFO	0.07	0.124	0.18	0.235
L = 15 cm
A = 12.56 cm2
φ = 0.19
Kw = 120 mD
μw = 0.28 cp
μg = 0.02 cp
ρg = 0.125 g/cm3	$M_{\mathrm{lab}}^{k,i}=\frac{\Delta P_{k,i}^{\mathrm{NOFO}}}{\Delta P_{k,i}^{\mathrm{FO}}}$	74	83	66	58

Claims

1. Method for operating an underground formation comprising hydrocarbons, by means of an injection of an aqueous solution comprising a gas in foam form and a flow simulator relying on a displacement model of said gas in foam form, said displacement model being a function of an optimal mobility reduction factor of said gas and of at least one interpolation function of said optimal mobility reduction factor, said interpolation function being a function of at least one parameter relating to at least one characteristic of the foam and of at least one constant, characterized in that, from at least one sample of said formation, measurements of conventional relative permeabilities to said gas in non-foaming form and measurements of conventional relative permeabilities to said aqueous phase: A. said displacement model of said simulator is determined according to at least the following steps:i. a plurality of values of said parameter is defined relative to at least one of said interpolation functions, an injection is performed into said sample of said gas in non-foaming form and of said gas in foam form according to said values of said parameter relative to said function, and a headloss with foam and a headloss without foam are measured respectively for each of said values of said parameter relative to said function;ii. from said measurements of headloss relative to said interpolation function, an optimal value of said parameter relative to said function is determined, said optimal value making it possible to maximize a ratio between said headlosses without foam and said headlosses with foam measured for said function;iii. from said measurements of headloss performed with said optimal value determined for at least said interpolation function, from said measurements of conventional relative permeabilities to said gas in non-foaming form and from said measurements of conventional relative permeabilities to said aqueous phase, said optimal mobility reduction factor is determined;iv. for at least said interpolation function, from said optimal mobility reduction factor, from said measurements of headloss relative to said interpolation function, from said measurements of conventional relative permeabilities to said gas in non-foaming form and from said measurements of conventional relative permeabilities to said aqueous phase, said constants of said interpolation function are calibrated;B—from said displacement model and from said flow simulator, an optimal operation scheme for said deposit is determined and said hydrocarbons are exploited.

2. Method according to claim 1, in which said displacement model of the foam is expressed in the form: krgFO(Sg)=FM krg(Sg)in which krgFO(Sg) is the relative permeability to said gas in foam form for a given gas saturation value Sg, krg(Sg) is the relative permeability to said non-foaming gas for said gas saturation value Sg, and FM is a functional expressed in the form:

3. Method according to claim 1, in which there are four of said interpolation functions and said parameters of said functions are a foaming agent concentration, a water saturation, an oil saturation and a gas flow rate.

4. Method according to claim 1, in which said interpolation function Fk of a parameter Vk is written in the form:

5. Method according to claim 1, in which, prior to the step iii), optimal conditions are defined as corresponding to said optimal values determined for each of said interpolation functions, said gas in non-foaming form and said gas in foam form are injected into said sample according to said optimal conditions, and a headloss with foam and a headloss without foam are respectively measured.

6. Method according to claim 1, in which said constants of at least one of said interpolation functions are calibrated by a least squares method, such as an inverse method based on the iterative minimization of a functional.