FR2836227A1 - Method for evaluating physical parameters of oil well from rocky debris, comprises pressurizing the debris, measuring evolution of pressure or volume and aligning results with theoretical parameters - Google Patents

Abstract

Drilling fragments (F) are placed in a confining chamber (1) and pressure is applied by a water pump (2) and buffer reservoir (4) using a viscous liquid. The injection of the liquid is stopped at a threshold and the pressure evolution is measured by a manometer (7) and captured by a microcomputer (8). Physical parameters are found by aligning the measured data with analogous theoretical data An Independent claim is also included for equipment to contain rock debris, subject it to pressure by using a viscous fluid and measure the subsequent evolution of pressure or volume

Description

<Desc / Clms Page number 1>

The present invention relates to a method and experimental devices for evaluating the absolute permeability of an area of an underground hydrocarbon reservoir from rock samples taken in this area such as fragments obtained during drilling operations of well.

The current oil context leads operators to take an interest in new areas (deep offshore) but also in new types of deposits (marginal structures located near existing surface installations). Given the drilling costs associated with the difficult environment of these new discoveries or the limited size of certain structures, operators can no longer afford to drill additional appraisal wells without risking compromising the economic viability of the project. The development strategy set before the start of operations is therefore less strict in order to be able to adapt "in real time" to the nature of the information collected by drilling production wells. We are talking about appreciative development.

Petrophysical measurements play a key role in assessing the quality of a reservoir. However, the delays associated with this type of measure are often very long and therefore incompatible with the reactivity necessary for the success of appreciative developments. New, faster and less expensive means of evaluation are therefore sought as a support for decision-making.

Drilling debris (cuttings) brought up by mud has long been the subject of on-site examinations. They are carried out by the teams responsible for analyzing the sludge (known as Mud Logging) and are essentially used to complete the

<Desc / Clms Page number 2>

 description of the geological layers crossed during drilling carried out from logs.

STATE OF THE ART Work has already been carried out to try to evaluate petrophysical properties from drilling fragments. Acoustic properties have been measured for example with respect to S and P waves. Various parameters have also been studied such as the hardness and deformation of the rock fragments or their porosity and permeability.

According to a first known method for carrying out the permeability measurement, the piece of rock is previously coated in resin. A thin slice is cut from the coated rock and placed in a measuring cell. It includes means for injecting a pressurized fluid at a controlled rate therein and means for measuring the pressure drop created by the sample. As the resin is impermeable, the absolute permeability is deduced from the Darcy equation taking into account the real surface occupied by the rock fragments.

This method is described for example by: - Santarelli FJ, et al; Training assessment from logging on cuttings>>,
SPERE, June 1998; or - Marsala AF, et al; Transient Method Implemented under Unsteady State
Conditions for Low and Very Low Permeability Measurements on Cuttings>> SPE / ISRM n 47202, Trondheim, 8-10 July 1998.

This type of measurement is only obtained in the laboratory after long operations of conditioning the fragments.

Another known method is based on an NMR (Magnetic Resonance) measurement.
Nuclear) which is made directly on the drilling fragments after a preliminary washing followed by saturation in brine. This type of measurement provides a directly exploitable value of the porosity. The permeability K is determined by

<Desc / Clms Page number 3>

 through correlations of the same nature as those used in the context of NMR logs.

An illustration of this method is found in the following document: - Nigh E., et al; P-KTM: Wellsite Determination of Porosity and Permeability
Using Drilling Cuttings ", CWLS Journal, Vol 13, n'l, Dec 1984.

By European patent application EP 1 167 948, a system is known for evaluating physical parameters such as their absolute permeability of porous rocks in an area of an underground deposit, from rock fragments taken from this area such as rock fragments brought up in the mud of a borehole. After immersing the fragments in a viscous fluid contained in an enclosure, fluid is injected under increasing pressure over time, up to a defined pressure threshold, so as to compress the gas trapped in the pores of the rock. This injection phase is followed by a relaxation phase with stopping the injection. The pressure variation during these two successive phases is recorded. The evolution of the pressure during the injection process having been modeled starting from initial values chosen for the physical parameters of the fragments, the calculator adjusts iteratively to make coincide as well as possible the modeled pressure curve with the pressure curve actually measured THE METHOD AND THE DEVICE ACCORDING TO THE INVENTION The purpose of the method according to the invention is to evaluate physical parameters such as the absolute permeability and the porosity of rocks in an area of an underground deposit from rock fragments ( drilling fragments, for example) taken from this area.

It involves the immersion of fragments contained in a confinement enclosure, in a viscous fluid, a step of placing the enclosure in communication with viscous fluid under pressure so as to compress the gas trapped in the pores of the rock, a step of measuring a quantity indicative of the evolution of the absorption of fluid by the rock, modeling the evolution of a physical quantity in the enclosure, from initial values chosen for

<Desc / Clms Page number 4>

 the physical parameters of the fragments, and a step of iterative adjustment of the values of the physical parameters of the rock fragments so that the modeled evolution fits as best as possible with the measured evolution of the physical parameter in the enclosure.

According to a first mode of implementation, the method is characterized in that: - the step of placing in communication comprises a brief period of putting in communication of the enclosure with a container containing fluid under a determined pressure, so to cause a rapid increase in the pressure in the enclosure and a compression of the gas trapped in the pores of the rock followed by a relaxation period after isolation of the enclosure, - the step of measuring the quantity includes measuring of the evolution of the pressure in the enclosure during the two periods, and - the modeling step relates to the modeling of the evolution of the pressure, from initial values chosen for the physical parameters of the fragments.

According to a second embodiment, the method is characterized in that: - the step of placing in communication also comprises putting the enclosure into communication with a container containing fluid under a determined pressure, so as to cause a rapid and prolonged increase in the pressure in the enclosure and a compression of the gas trapped in the pores of the rock, - the measurement step of the quantity includes the measurement of the evolution of the volume of fluid injected as a function of time , and - the modeling step relates to the modeling of the evolution of the volume of fluid injected, from initial values chosen for the physical parameters of the rock fragments.

<Desc / Clms Page number 5>

Depending on the case, the containment cell can be loaded either directly with drilling fragments invaded by drilling fluids, or with previously cleaned drilling fragments.

The device according to the invention makes it possible to evaluate physical parameters such as the absolute permeability and the porosity of rocks in an area of an underground deposit, from fragments of rock taken from this area. It comprises a confinement enclosure for fragments of porous rock, means for injecting a viscous fluid into the enclosure to fill the enclosure containing the rock fragments, firstly, and to carry out a cycle comprising a fluid injection phase in the enclosure of the means for measuring the evolution of a physical quantity in the enclosure and a processing system for modeling the evolution of said physical quantity from initial values chosen for the physical parameters of the rock fragments, and to iteratively adjust the values to be given to these physical parameters so that the modeled evolution of the physical quantity fits as best as possible with the measured evolution of the said quantity in the enclosure. It is characterized in that it comprises a container (a buffer bottle containing viscous oil and a gas cap under a predefined pressure) containing fluid under a determined pressure and means controlled by the treatment system, for controlling its communication with the enclosure containing the rock fragments.

According to a first embodiment, the device comprises means for measuring the evolution of the pressure in the enclosure, as a function of time, the processing system being adapted to model the evolution of the pressure from initial values chosen for the physical parameters of the rock fragments, and to iteratively adjust the values to be given to these physical parameters so that the modeled evolution of the pressure fits as best as possible with the measured evolution of the pressure in the enclosure.

According to a second embodiment, the device comprises means (such as a flow meter or a differential pressure sensor) for measuring the change in the volume of oil injected into the enclosure, as a function of time, the system of being adapted to model the evolution of the volume injected from

<Desc / Clms Page number 6>

 initial values chosen for the physical parameters of the rock fragments, and to iteratively adjust the values to be given to these physical parameters so that the modeled evolution of the injected volume fits as best as possible with the measured evolution of the volume of fluid in the pregnant.

The method is satisfactory for rocks with very different permeabilities ranging from a few tenths of a milliDarcy to a few Darcy. Furthermore, a satisfactory estimate of the porosity is obtained. Given the small space occupied by the device, the ease of implementation but also the speed with which it is possible to carry out the measurements and the calibration between the theoretical data and the experimental data, the method lends itself particularly well. on site conditions. It is therefore quite possible to envisage a measurement and an interpretation directly on site within a very short time, consequently without common measure with those which are necessary to obtain equivalent results by laboratory methods. This opens up important perspectives in terms of characterization since we can take advantage of this new source of information as a support for the interpretation of electrical logs and refine the evaluation of a well in terms of production potential. SUMMARY PRESENTATION FIGURES Other characteristics and advantages of the method and of the device according to the invention will appear on reading the following description of nonlimiting exemplary embodiments, with reference to the appended drawings where: - FIG. 1 schematically shows the device; - Fig. 2 schematically shows the structure of a fragment or particle of porous rock in which the effects of the injection of a fluid with high viscosity such as oil are modeled; - Figs. 3a to 3c schematically show the pressure variation curves prevailing in the cell of the device of FIG. 1, during the injection and relaxation phases, for four different rocks with experimental procedure 1;

<Desc / Clms Page number 7>

 - Fig. 4 shows the good agreement between the permeabilities obtained for several types of rock, by a conventional method of measurement on cores and by the method according to the invention (procedure 1); - Figs. 5a to 5d show for the four preceding rocks, the precision which one obtains in the adjustment of the pressure curves modeled compared to the experimental curves in the case where one operates according to a first procedure which will be described later; - Fig. 6 shows experimental pressure developments within the framework of a second procedure; - Figs. 7a and b show two examples of adjustment of the experimental curves in the context of the second procedure; - Fig. 8 shows experimental changes in the volume of oil injected into the enclosure of the cuttings, within the framework of a third procedure; - Figs. 9a and b show two examples of adjustment of the experimental curves in the context of the third procedure; - Fig. 10 shows the good agreement between the permeabilities obtained according to the third procedure, with measurements carried out on cores; and - Fig. It highlights the good agreement obtained in the estimation of the porosity of the rock fragments tested.

DETAILED DESCRIPTION As we have seen, the determination of physical parameters of rocks such as their absolute permeability for example, essentially comprises three steps: 1) a step of acquiring experimental measurements of pressure variations (procedures 1 and 2) or evolution of the volume injected (procedure 3) from the drilling fragments, giving rise to experimental curves; It) a step of modeling the physical phenomena occurring within the drilling fragments during the same operating cycle, for arbitrary values

<Desc / Clms Page number 8>

 sought physical parameters (permeability K and porosity <1>) involved in the model, making it possible to establish similar theoretical curves; and ttt) an adjustment or calibration step in which the values to be given to the physical parameters involved in the model are determined so that the experimental and theoretical curves are best adjusted.

1) Acquisition of measurements The first step is carried out using a device such as that shown diagrammatically in FIG. 1. It comprises a confinement cell 1 into which the drilling fragments are initially introduced. A constant-flow water pump 2 communicates via a pipe 3 with the base of a buffer tank 4 containing oil having a high viscosity. The opposite end of the buffer tank 4 communicates by a valve V6 with a line L 1. A first end of the confinement cell 1 communicates with the line L 1 by means of two valves V1, V2 and with a purge line L2 via valve V1 and a purge valve V3. The opposite end of the confinement cell 1 communicates by means of an isolation valve V5 with a separator 6. A pressure gauge 7 is connected to the outlet of the cell 1.

The pressure variations measured by the pressure gauge 7 are acquired by a processor 8 such as a microcomputer. The valve V2 is controlled directly by the processor 8. The line L 1 also communicates via a valve V7 with a buffer bottle 9 containing viscous oil put under a pressure determined by a cap of gas under pressure. A flow meter or a differential pressure sensor 10 is placed if necessary on the line L 1 between the cell 1 and the bottle 9 to measure the flow rate of injected fluid.

It is also possible to use fragments available after cleaning from which all the fluids have been removed beforehand. In the case where the confinement cell 1 is loaded with cleaned fragments, helium from a bottle 5 is injected by opening a valve V4, so as to expel the air from the cell.

Cell 1 is then filled with an oil of high viscosity via the pump. The oil occupies the free space between the drilling fragments and it

<Desc / Clms Page number 9>

 also penetrates by spontaneous imbibition inside the rock. 1 / degassing occurs, the intensity and duration of which depends on the nature of the rock (mainly the porosity). This degassing only affects part of the gas. A certain residual volume remains trapped in the drilling fragments in the form of disconnected clusters.

Three procedures are possible for carrying out the experimental phase: Procedure 1 As it has already been described in detail in the European patent application already cited, procedure 1 essentially consists in injecting oil from the buffer tank 4 at a constant rate, increasing gradually the injection pressure by means of pump 2 (part C1 of the pressure curve). The amount of oil entering the pores of the rock is measured as the residual gas trapped in the pores is compressed. When the pressure reaches a certain PMI threshold, the oil injection is stopped. We then witness a relaxation. The fluids tend to rebalance in the drilling fragments and a slow rebalancing of the pressure is observed (part C2 of the pressure curve: Fig. 3).

Procedure 2 The second procedure essentially consists in putting cell 1 containing the rock fragments C in communication with the buffer bottle 9 containing viscous oil under pressure by opening the valve V2 controlled by the control computer 8.

Initially, the valve is closed. On the side of the buffer bottle 9, the pressure is equal to Pmax while on the side of cell 1, the pressure is equal to the ambient pressure. The valve V2 controlled by the processor 8 is then opened for a few tenths of a second to rapidly increase the pressure of the cell 1 up to the pressure Pmax, then this valve is closed and a relaxation of the pressure is observed which in this case also corresponds a return to equilibrium of the pressure in the rock fragments (Fig. 6). Compared to procedure 1, the pressure rise time is minimized, which increases the

<Desc / Clms Page number 10>

 sensitivity of the system in terms of permeability detection. On the other hand, this procedure is less precise in terms of volumetric balance (oil injected) compared to injection at constant flow.

Procedure 3 Compared to the two previous procedures for which a distinction was made between an injection period and a relaxation period, there is only one phase (injection). The preparation of the experiment is exactly the same as in the procedure 2. The flow meter or a differential pressure sensor 10 is added to the connection line between the cell 1 and the buffer bottle 9 (FIG. 1). Initially, we are in the same conditions as for procedure 2, (the flow meter or the differential pressure sensor are initially under pressure). The valve V2 is then simply opened and the evolution of the flow rate or of the differential pressure is recorded over time. When using the differential sensor, a calibration curve previously measured by means of pump 2 is used which gives the relationship between the viscous oil flow rate and the differential pressure. This then makes it possible to convert the experimental differential pressure measurements and to deduce the change in the volume of oil injected over time. The evolution of the volume injected is obtained directly if a flow meter is used.

Compared to previous procedures, this approach makes it possible to simplify the course of the experiment while keeping good control of the volume injected, which facilitates the acquisition of measurements and the interpretation of results with the simulator. On the other hand, as the parasitic part of the gas trapped outside the fragments or cuttings C (inter-cuttings space and compressibility of the system) is compressed during the first moments of the experiment, we observe a kind of natural decorrelation of the signal which allows to explore, at equal size, much larger permeability ranges compared to other methods but also to better measure the volume of gas actually trapped in the cuttings.

Figs. 3a to 3d show examples of evolution of the pressure signal observed for fragments of four different rocks for a flow rate of 480 cc / h

<Desc / Clms Page number 11>

 (procedure 1). Whatever the rock considered, we observe the same general evolution of the pressure. There is a gradual rise during the injection phase as the residual gas compresses. The time required to increase the pressure by 5 bars varies depending on the rocks from 15 to 40 seconds depending on the initial volume of trapped gas. As soon as the injection is stopped, the pressure decreases. If this decrease is significant for rocks 1 and 2, it remains more moderate for rocks 3 and 4. At long times, we observe a gradual stabilization of the signal.

Fig. 6 shows examples of changes in pressure in the context of procedure 2. As in the context of procedure 1, there are significant variations in the relaxation curves depending on the nature of the rocks tested. The lower the permeability of the rocks, the more marked pressure relaxation is observed.

Fig. 8 shows examples of changes in pressure in the context of procedure 3. We observe significant variations in the filling curves depending on the nature of the rocks tested. The lower the permeability of the rocks, the more we observe a slow filling kinetics. The higher the porosity of the rocks, the greater the volume of cumulative oil injected.

 H) Modeling The aim of this modeling is to obtain an estimate of the permeability K from pressure measurements or the volume of oil injected.

It is considered that the drilling fragments are of homogeneous size and spherical in shape and that the gas is assumed to be perfect. The viscous pressure drop of the gas is neglected compared to that of the oil, given the difference between the viscosities. The residual gas trapped in the rock fragments after the spontaneous imbibition of the oil is in the form of disconnected clusters distributed homogeneously. It is also considered that the capillary pressure is negligible.

Given the spherical shape of the fragments, we will reason on a cap of thickness dr (Fig. 2). and calculate the evolution of the pressure at the boundary of the rock particle when an oil flow q is injected.

<Desc / Clms Page number 12>

p pression P : Sg = S 0 Po (Po est la pression de l'huile). Dans la calotte, on fait un g g p

bilan matière sur l'huile. L'accumulation est égale à la différence entre ce qui rentre et ce qui sort. De là, on déduit :

We consider that the N rock particles share the total flow rate Q of the injected fluid in an equitable manner, and that each receives the flow rate q = Q. The law of perfect N gases makes it possible to deduce the local saturation in gas Sg as soon as we know the

p pressure P: Sg = S 0 Po (Po is the oil pressure). In the cap, we make a ggp

material balance on oil. Accumulation is equal to the difference between what comes in and what comes out. From there, we deduce:

p Comme = (1- S gO PO), on en déduit que : P

p As = (1- S gO PO), we deduce that: P

- KV--griidp Comme par ailleurs, = p^ 0 et que la pression capillaire peut être

considérée comme négligeable ce qui fait donc que Po = Pgaz = P, l'équation précédente s'écrit :

- KV - griidp As elsewhere, = p ^ 0 and that the capillary pressure can be

considered negligible, which means that Po = Pgaz = P, the previous equation is written:

c.

vs.

On obtient donc la forme classique d'une équation de type diffusion avec toutefois un terme en 1/p2 facteur de l'accumulation qui provient de la nature compressible du gaz.

As a result, K p2 at

We therefore obtain the classic form of a diffusion type equation with, however, a term in 1 / p2 accumulation factor which comes from the compressible nature of the gas.

l'équation à résoudre s'écrit : JLÂf In spherical coordinates, the Laplacian is equal to. 2 '. Finally, r Dr ar

the equation to be solved is written:

<Desc / Clms Page number 13>

/ avec a (2) K Lors de sa mise en place, l'huile chasse l'air dans l'espace libre entre les fragments de forage et pénètre dans la roche par imbibition spontanée. Malgré certaines précautions, il est possible qu'il reste un certain volume de gaz retenu à l'extérieur du fait de la forme non régulière des fragments de forage. Ce volume piégé (Vp) joue un rôle direct sur la forme générale de la réponse en pression et doit être pris en compte dans la résolution.

/ with a (2) K When it is put in place, the oil expels air into the free space between the drilling fragments and penetrates into the rock by spontaneous imbibition. Despite certain precautions, it is possible that a certain volume of gas remains outside due to the irregular shape of the drilling fragments. This trapped volume (Vp) plays a direct role on the general shape of the pressure response and must be taken into account in the resolution.

It is also necessary to take into account a certain compressibility due to the experimental device. It comes from the cell, the lines and the properties of the oil. The equivalent compressibility observed is of the order of 0.0005 bar '.

As the oil used is saturated with gas at atmospheric pressure, dissolution phenomena occur when the pressure increases during the measurement. These aspects are taken into account by introducing a diffusion parameter reflecting the exchange of molecules at the gas / oil interfaces.

et en espace P (R, t) =Pet- (0, ) =0. Lorsque)'on simu) e des expériences à ar pression imposée, la pression Pext est connue et l'équation se résout de manière explicite. Lorsque l'on simule des expériences à débit imposé, la valeur de Pext est calculée par l'intermédiaire d'une boucle de convergence dont le test repose sur une comparaison entre la saturation de gaz restant dans la particule de roche et la valeur obtenue par bilan volumique à partir de la quantité d'huile injectée. The diffusion equation is solved by the finite difference method with an explicit scheme and by imposing the boundary conditions in time P (r, O) = Patm

and in space P (R, t) = Pet- (0,) = 0. When we simulate experiments at imposed pressure, the pressure Pext is known and the equation is solved explicitly. When simulating imposed flow experiments, the value of Pext is calculated by means of a convergence loop whose test is based on a comparison between the saturation of gas remaining in the rock particle and the value obtained by volume balance based on the quantity of oil injected.

The resolution of the diffusion equation during the relaxation period (procedures 1 and 2) is identical and is based on the same convergence loop. Only the test condition changes since stopping the injection causes the volume of gas in the rock particle to be maintained.

<Desc / Clms Page number 14>

III) Adjustment of the model to the experimental results The model is implemented in a calculator such as computer 8 (cf. Fig. 1) in the form of software and inserted in an iterative optimization loop. We rotate the model with permeability K and porosity values chosen a priori, we compare the resulting simulated pressure curve with the experimental curve and by successive iterations by changing the values of K and porosity in the model, we find the one (s) that allow you to best adjust the theoretical and experimental curves. As it is the product <M which counts for the calibrations, the value of is taken equal to 25% to estimate the value of (D.

Figs. 5a to 5d show the good agreement which is quickly obtained by successive iterations, between the theoretical and experimental curves for four rock fragments tested by means of procedure 1. As also shown in FIG. 4, the results obtained by applying the method are entirely comparable with those obtained in the laboratory after long periods of conditioning by conventional methods for several rocks of different permeabilities.

Figs. 7a and b show two examples of adjustment by successive iterations on experimental results using procedure 2.

Figs. 9a and b show two examples of adjustment by successive iterations on experimental results using procedure 3.

In all cases, there is a very good agreement between the simulations and the experimental results. The results obtained by application of the method are completely comparable with those obtained in the laboratory after long periods of conditioning by conventional methods for several rocks of different permeabilities as shown in FIG. 4 (procedure 1) and FIG. 10 (procedure 3). Furthermore, FIG. 11 shows that a satisfactory estimate of the porosity of the rocks tested is also obtained.

<Desc / CRUD Page number 15>

 The programming of this modeling within a code makes it possible to calibrate the experiments by trial and error and therefore to deduce the corresponding value of K and to give a correct estimate of the porosity.

Claims (9)

CLAIMS 1) Method for evaluating physical parameters such as the absolute permeability and the porosity of rocks in an area of an underground deposit from rock fragments (F) taken from this area, comprising the immersion of fragments contained in a confinement enclosure (1) in a viscous fluid, a step of placing the enclosure containing the fragments in communication with a source of pressurized fluid so as to compress the gas trapped in the pores of the rock, a measurement step of a quantity indicative of the evolution of the absorption of fluid by the rock, a modeling of the evolution of a physical quantity in the enclosure, from initial values chosen for the physical parameters of the fragments (F) , and a step of iterative adjustment of the values of the physical parameters of the rock fragments so that the modeled evolution fits as best as possible with the measured evolution of the physical parameter in the enclosure , characterized in that: - the step of placing in communication comprises a brief period of putting in communication of the enclosure with a container containing fluid under a determined pressure, so as to cause a rapid increase in the pressure in the enclosure and a compression of the gas trapped in the pores of the rock followed by a period of relaxation after isolation of the enclosure, - the step of measuring the quantity includes measuring the evolution of the pressure in the enclosure during the two periods, and - the modeling step relates to the modeling of the evolution of the pressure, from initial values chosen for the physical parameters of the fragments. 2) Method for evaluating physical parameters such as the absolute permeability and the porosity of rocks in an area of an underground deposit from rock fragments (F) taken from this area, comprising the immersion of fragments in a fluid viscous contained in a confinement enclosure (1), a step of placing the enclosure containing the fragments in communication with a source of pressurized fluid, so as to compress the trapped gas <Desc / Clms Page number 17>  in the pores of the rock, a step of measuring a quantity indicative of the evolution of the absorption of fluid by the rock, a modeling of the evolution of a physical quantity in the enclosure, from values initials chosen for the physical parameters of the fragments (F), and a step of iterative adjustment of the values of the physical parameters of the rock fragments so that the modeled evolution fits as best as possible with the measured evolution of the physical parameter in the enclosure, characterized in that: the step of placing in communication also comprises putting the enclosure into communication with a container containing fluid under a determined pressure, so as to cause a rapid and prolonged increase in the pressure in the enclosure and a compression of the gas trapped in the pores of the rock, - the step of measuring the quantity includes measuring the evolution of the volume of fluid injected as a function of time, and - the step of modeling relates to the modeling of the evolution of the volume of injected fluid, starting from initial values chosen for the physical parameters of the fragments. 3) Method according to claim 1 or 2, characterized in that the confinement cell is loaded with drilling fragments invaded by drilling fluids. 4) Method according to claim 1 or 2, characterized in that the confinement cell is loaded with previously cleaned drill fragments. 5) Device for evaluating physical parameters such as the absolute permeability and the porosity of rocks in an area of an underground deposit, from rock fragments taken from this area, comprising a confinement enclosure (1) for fragments of porous rock (F), means for injecting a viscous fluid into the enclosure in order to fill the enclosure containing the rock fragments, firstly, and to carry out a cycle comprising a fluid injection phase in the enclosure, means (7, 10) for measuring the evolution of a physical quantity in the enclosure and a processing system (8) for modeling the evolution of said physical quantity from <Desc / Clms Page number 18>  initial values chosen for the physical parameters of the rock fragments, and to iteratively adjust the values to be given to these physical parameters so that the modeled evolution of the physical quantity fits as best as possible with the measured evolution of the said quantity in l enclosure, characterized in that it comprises a container (9) containing fluid under a determined pressure and means (V2) controlled by the treatment system to control the placing in communication of the container (9) with the enclosure ( 1) containing the rock fragments. 6) Device according to claim 5, characterized in that said container (9) is a buffer bottle containing viscous oil and a gas cap under a predefined pressure. 7) Device according to claim 5 or 6, characterized in that it comprises means (7) for measuring the evolution of the pressure in the enclosure, as a function of time, the treatment system (8) being adapted to model the evolution of the pressure from initial values chosen for the physical parameters of the rock fragments, and to iteratively adjust the values to give to these physical parameters so that the modeled evolution of the pressure fits as best as possible the measured evolution of the pressure in the enclosure. 8) Device according to claim 5 or 6, characterized in that it comprises means (10) for measuring the change in the volume of oil injected into the enclosure, as a function of time, the treatment system (9 ) being adapted to model the evolution of the injected volume from initial values chosen for the physical parameters of the rock fragments, and to iteratively adjust the values to be given to these physical parameters so that the modeled evolution of the injected volume adjusts at best with the measured evolution of the volume of fluid in the enclosure. 9) Device according to claim 8, characterized in that the means (10) for measuring the change in the volume of fluid injected comprise a flow meter or differential pressure sensor (10).