WO2015044133A1

WO2015044133A1 - Calibration of a meter built into a line of a hydrocarbon production plant

Info

Publication number: WO2015044133A1
Application number: PCT/EP2014/070235
Authority: WO
Inventors: Cécile COUDROY; Jean Paul Couput; Renaud CAULIER
Original assignee: Total S.A.
Priority date: 2013-09-25
Filing date: 2014-09-23
Publication date: 2015-04-02
Also published as: AR097753A1

Abstract

A method for calibrating a meter built into a line of a hydrocarbon production plant, in which the meter is suitable for providing an estimate of the flow rate of a fluid in the line depending on a measurement of the fluid carried out by at least one sensor and on the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined condition of flow of the fluid in the line. The method includes at least one in situ measurement (S20) by the sensor under the predetermined condition, followed by the determination (S30) of the value of the calibration parameter depending on the result of the measurement. This offers a solution for the improved calibration of a meter.

Description

CALIBRATION OF A METRIC METER INTEGRATED WITH A LINE

HYDROCARBON PRODUCTION FACILITY

The present invention relates to the field of hydrocarbon production, and more specifically the calibration of a metric meter integrated in a line of a hydrocarbon production facility.

Hydrocarbon production facilities include lines in which fluids circulate. These may include production lines with a production well in which hydrocarbons circulate from a hydrocarbon reservoir to a wellhead. It may also be injection lines comprising an injection well and in which injection fluids circulate, that is to say fluids that it is desired to inject into the reservoir, for example polymers (in the context of assisted recovery, in particular).

It is now common to integrate in such a line metric meter adapted to provide an estimate of the flow rate of the fluid flowing in the line. Indeed, the estimate of the flow rate provided by the metric counter allows in known manner a better management of the line. This gives a better control or a better knowledge of the overall production of the installation, thanks to the estimate provided for each line integrating a metric counter, typically in real time and line by line.

Metric meters are devices adapted to measure the fluid flowing in the line, and to provide an estimate of the flow rate of the fluid in the line as a function of (at least) the measurement and the value of at least one parameter , commonly called "calibration parameter". In particular, a metric counter can perform one or more physical measurements (e.g., electrical, nuclear, and / or optical, for example, measurements of permittivity, conductivity, and / or gamma attenuation).

These metric meters need to be calibrated as correctly as possible. In other words, it is necessary that a value as correct as possible be attributed to the calibration parameters, otherwise the estimation of the flow rate of the fluid in the line provided by the metric counter would be too far from reality and by therefore unusable. This is especially true when considering flare gas meters or tax counters for which it is particularly critical that a near real-life estimate is made.

An existing solution for performing this calibration, with a counter intended to be on the surface, is to take a sample of the fluid flowing in the line and to determine the value of the calibration parameter in the laboratory with another counter that does not exactly have the same properties as the on-site counter. The parameter of Calibration provided is therefore not strictly "clean" on the on-site meter and may be far from the exact value. In addition, the sample taken is generally not representative of 100% of the fluid under the conditions of production: the calibration parameters deduced in laboratory from these samples will certainly be far from the exact values. Another method is to disconnect the counter to be calibrated from the production line and to successively pour the samples taken to calculate the calibration parameters in the latter case the difference to reality would always come from the representativeness of the samples and the fact that the calibration would be carried out at conditions far from those of the production.

When the meter is intended to be submarine, it is very complex to take representative samples of the production fluid because the line is then inaccessible. And it is also impossible to disconnect the meter from the production line to pour samples and calculate calibration parameters. Thus, for recalibration, an existing solution consists in calibrating the values of the calibration parameters with flow references provided by the production. This has the disadvantage of providing a value of the calibration parameter relatively far from the exact value, similarly or even more than in the case of counters intended to be on the surface.

The object of the present invention is to provide a calibration solution of a metric counter at least partially overcomes the aforementioned drawbacks. More particularly, the invention aims to provide a calibration solution of a metric counter that is simple to implement and that provides a value of the calibration parameter that is relatively close to its exact value.

To this end, the present invention proposes a method of calibrating a metric meter integrated in a line of a hydrocarbon production facility, in which the meter is adapted to provide an estimate of the flow rate of a fluid in the line. as a function of a measurement made by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined condition of flow of the fluid in the line. The method comprises at least one measurement by the sensor under the predetermined condition, in situ, then the determination of the value of the calibration parameter as a function of the result of the measurement.

The invention also provides a computer program adapted to be recorded on a data recording memory, comprising instructions for executing the method.

The invention also proposes a computer program, adapted to be recorded on a data recording memory, comprising instructions for controlling the execution of the method by a metric counter, which metric counter is adapted to be integrated in a line of a hydrocarbon production facility, and to estimate the flow rate of a fluid in the line according to a measurement made by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined flow condition of the fluid in the line.

The invention also proposes a metric counter adapted to be integrated in a line of a hydrocarbon production facility, to estimate the flow rate of a fluid in the line as a function of a measurement made by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined flow condition of the fluid in the line, and comprising a memory having recorded the program comprising instructions for executing the method.

The invention also proposes a system adapted to communicate with a metric counter which is adapted to be integrated in a line of a hydrocarbon production facility, to estimate the flow of a fluid in the line according to a measurement. performed by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined condition of fluid flow in the line. The system includes a memory having recorded the program including instructions for controlling the execution of the method.

The invention also provides a hydrocarbon production facility comprising at least one metric counter above and / or the above system.

The invention also provides a method for producing hydrocarbons comprising the calibration method. In one example, the hydrocarbon production process comprises the initial calibration of the metric meter, the integration of the meter at a line of a hydrocarbon production facility, the use of the meter to estimate the flow rate of the fluid during a period of time. phase of production, then the new calibration of the metric meter, in situ, by the calibration process.

In preferred embodiments, the invention includes one or more of the following features:

the fluid is multiphase and the predetermined condition is the presence of a single phase of the fluid at the sensor;

the fluid comprises a water phase, a gas phase and an oil phase;

the measurement is slaved to a detection of a stop of the flow in the line; the line is a production line and the stop is a pit stop;

the line is a test line and the stop is a test stop; the method comprises, before the measurement, the segregation of the phases of the fluid, the measurement being then carried out on the phase at the sensor;

the metric counter is at the surface, the method comprising successive opening and closing of a fluid purge valve to repeat the measurement over several phases;

metric counter is underwater or underground, the method comprising successive opening and closing of a fluid production valve to repeat the measurement over several phases;

the measurement is iterated several times on the same phase, the determination of the value of the calibration parameter being a function of the result of the iterations of the measurement;

the iterations of the measurement are separated by an interval greater than one day, and / or less than six months;

the determination of the value of the calibration parameter comprises a process of reconciliation and data validation involving the result of the iterations of the measurement, the reconciliation being conditioned by an equality, at least substantially, between the result of the iterations of the measurement; and or

the process of reconciliation and data validation further implies at least one predetermined fluid hypothesis.

Other characteristics and advantages of the invention will appear on reading the following description of a preferred embodiment of the invention, given by way of example and with reference to the appended drawing.

Figure 1 shows a diagram of the calibration process.

Figures 2 to 4 illustrate the operation of an example metric counter.

Figure 5 shows a diagram of an example of a hydrocarbon production facility.

FIG. 6 shows an example of segregation with successive purges thus carried out during a test stop.

Figure 1 shows the method of calibration of a metric meter integrated in a line of a hydrocarbon production facility. The meter is adapted to provide an estimation of the flow rate of a fluid in the line as a function of a measurement made by at least one sensor on the fluid and the value of at least one calibration parameter. The calibration parameter corresponds (ie is related) to the measurement of the sensor under a predetermined condition of fluid flow in the line. The method comprises at least one S20 measurement by the sensor under the condition predetermined, in situ. The method then comprises the determination S30 of the value of the calibration parameter as a function of the result of the measurement. The method of FIG. 1 allows a calibration of the metric counter which is simple to implement and which provides a value of the calibration parameter relatively close to its exact value.

Indeed, the method starts from the observation that the metric counter estimates the flow rate of the fluid as a function, inter alia, of at least one calibration parameter which corresponds to the measurement of a meter sensor, under a predetermined flow condition fluid in the line (ie a predefined physical state of the fluid flowing in the line). In other words, the metric counter implements, to provide an estimation of the flow, a calculation model having among its inputs at least this calibration parameter. The method proposes, in order to improve the estimation provided by the counter, to determine at S30 the value of this parameter according to an operation carried out in situ, thus excluding any possible extraction of the counter. Furthermore, the method can exclude any hardware intervention on the line, the process operations can all be controlled remotely or be performed automatically. Thus, the method is simple to implement and allows recalibration at any time, even if the counter is underwater or underground and thus difficult to access. The operation performed in situ is a measurement (on the fluid) at S20 by the meter sensor under the predetermined condition (corresponding to the calibration parameter). Since the measurement S20 is performed under the predetermined condition, a value directly representative of the calibration parameter is obtained. Measurement S20 being performed in situ, a value corresponding to the production conditions is obtained. Effectively the measurement S20 is made on the same fluid that is produced or injected, and under the same conditions. This makes it possible to obtain a relatively accurate calibration, in that the estimation of the flow rate of the fluid in the line subsequently supplied by the meter is improved (closer to the actual flow rate).

Thus, the calibration process can be implemented within a more comprehensive hydrocarbon production process. As explained above, the integration of a metric meter with a line of a hydrocarbon production facility allows a better management of the line, with a better control or a better knowledge of the overall production of the installation and / or quantities of fluids injected, thanks to the estimate provided for each line incorporating a metric counter, eg in real time and / or line by line. The calibration method then makes it possible to improve the estimates provided by each counter on which the calibration process is implemented, which improves the overall management of the production. For example, the hydrocarbon production process may include an initial calibration of the metric counter. This initial calibration can be performed according to the method of Figure 1 or be performed according to the standard laboratory calibration technique, so ex situ. In this second case, the meter can then be integrated with a line of the hydrocarbon production facility and the meter used to estimate the flow rate of the fluid during a production phase in a conventional manner. Then, it is possible to perform a new calibration (a "re-calibration") of the metric counter, in situ, by the method of FIG. 1. This makes it possible to improve the calibration and / or to update it, and thus the subsequent estimates provided by the meter.

The process of Figure 1 can be repeated during production, with production phases during which the meter provides an estimate, and re-calibration phases, which can be either production phases during which the meter does not provide. no estimate or phases of injection stop and / or production stop. Indeed, a metric meter may be misaligned during use, even if it followed a proper initial calibration process. For example, the fluid flowing in the line may evolve over time so that the measurement of the meter sensor under the predetermined flow condition of the fluid in the line can be changed in theory. Similarly, the sensor itself may undergo material changes, for example wear or drift, so that its measurement under the predetermined condition of fluid flow in the line can be changed, even in the case where the fluid would be the same. The method of FIG. 1 allows in all these cases a re-calibration in order to take account of these temporal modifications and thus to adjust the value of the calibration parameter.

Moreover, the method of FIG. 1 can be repeated for several metric meters and / or for several calibration parameters, in the case where the counter considered provides an estimation of the flow rate of a fluid in the line as a function of several parameters of calibration corresponding to the sensor measurement under a predetermined flow condition of the fluid in the line.

The fluid flowing in the line may be monophasic, and in this case the estimation of the flow rate of the fluid in the line is simply to provide a value for the flow of the single phase of the fluid. This is typically the case of an injection line. A counter can make this estimate on the basis of any model using a measurement made on the fluid by at least one meter sensor. If the model furthermore involves the value of at least one calibration parameter which corresponds to the measurement of the sensor under a predetermined condition of fluid flow in the line (for example a flow rate or a velocity predetermined flow), then the method of Figure 1 can be used for the calibration of this counter with the same advantages as those described above.

Alternatively, the fluid flowing in the line can be multiphasic. This is typically the case of a production line, because the hydrocarbons produced usually comprise a water phase, a gas phase and an oil phase. In this case, the estimation of the flow rate of the fluid in the line may consist in providing the information making it possible to determine the flow rate of each phase. This information may directly include an estimate of the flow rate of each phase, and / or an estimate of the fraction (e.g., instantaneous) of each phase in the multiphase fluid in addition to the cumulative flow (e.g., instantaneous) of the multiphase mixture in the line. In the second case, it suffices to simply multiply between the fraction of a respective phase and the cumulative flow to obtain the flow rate of the respective phase.

In the case of multiphase fluid, we speak of multiphasic meter, commonly called MPFM (acronym from the English "multiphase flowmeter"). This may be for example a mass flow meter or an analyzer. Such a counter involves complex technologies, and the measurements given by the sensor (s) included in such a counter are particularly sensitive to the properties of the fluids. These meters being directly installed on the production lines, a periodic intervention and therefore a re-calibration is not possible. Therefore, it often happens that the meter does not use the correct calibration parameters and / or some of these sensors may drift faster than expected. The method of Figure 1 overcomes this problem.

In the case where the fluid is multiphasic and comprises in particular a water phase (typically salt water), a gas phase and an oil phase, the predetermined fluid flow condition in the line may consist of the presence of a single phase fluid at the sensor. The phase which is "at the level" of the sensor is, by definition, that which is the object of measurement according to the operation of the sensor, and in particular its hardware configuration. For example, the phase being measured can be next to the sensor. Thus, the metric counter may comprise at least one sensor, adapted to perform at least one type of measurement on the fluid. The counter is based, for the flow estimation, on a measurement of the sensor on the fluid and on predetermined values of calibration parameters, comprising precisely the (or) measurement (s) of the (or) sensor (s) ) on each of the three phases of the isolated fluid of the two others (for example at any rate, eg at zero flow rate). It should be noted that the calibration parameters may also include one or more predetermined fluid hypotheses, i.e. predetermined values of physico-chemical and / or thermodynamic variables relating to the fluid flowing in the line (for example for each of the phases in the case of a multiphase fluid).

The counter may in particular operate according to the following principle, described with reference to FIGS. 2-4.

The meter sensor may be a conventional gamma sensor, sending waves through the fluid at the sensor and measuring the mass attenuation of the wave. The sensor can be adapted to send several energy levels, so as to allow cross-checking of information sufficient for a multiphase fluid. We note below the linear attenuation in m ^"1 , μ the mass attenuation in m ² / kg, and p the density density in kg / m ^3. By definition, we have α = μ * p. mass μ depends on the compound and the gamma energy level.

For the hydrocarbon fluid of the above example (a water phase, an oil phase, and a gas phase), it is sufficient for the sensor to be able to make at least two measurements (eg corresponding to two different energy levels). ). We classically speak of low energy level and high energy level (or average), respectively corresponding to indices 1 and 2 in the notations below. The calibration parameters then comprise an attenuation value for each pure phase, by energy level, which can be noted μ _83Ζ _ ₁ , μ _{§ £ 18} _2, μωι _ε _ι, μι ™ ι _ε _2, μ _{ε £ ω} _ι and μ _β! ηι_2 · Thus, μ _ρ ί is the mass attenuation μ measured by the sensor for the energy level i, in the presence of the single phase ph.

It should be noted that if, for a given hydrocarbon production facility, the salinity of the wells is different from one well to another, then these parameters may vary. On the other hand, two wells with the same salinity should have the same water attenuation values for the same production facility. In addition, it should be noted that the sensor can actually measure a third level of energy, for example even higher than the other two. At such a level of energy, and in general at particularly high energy levels, the different compounds absorb substantially the same amount of energy (it could be translated by the fact that the mass attenuations are approximately equal to this level of energy. energy). However, this energy can be used to validate the estimation of the flow rate and / or water, oil and gas fractions provided based on the other two energy levels, preferably with high uncertainty. It is also possible to connect the measurement made for this third energy level to the density of the mixture by a line obtained after calibration. Thanks to the knowledge of the parameters μ _ΆΖ , μ ^ ι, μΐπύΐε ι, μΐπύΐε 2, μ _ε αιι i and liem_2, we can draw a triangle called "calibration triangle" on the basis of the following six values:

· Clgaz_2 μ az_2 * Pgaz

• Cthuile_l ^- μΐιυϋβ_1 Oil

• ¾uile_2 ^- μΐιυϋβ_2 Oil

• deau_l ^- μβ8υ_1 Oil

• C ^ water_2 μβ3υ_2 * Oil

The volume densities p are variables which depend on the pressure values P and temperature T of the line in which the counter operates and on the composition of the fluid, information given elsewhere (eg by a measurement made by dedicated sensors, for example, for example for P and T, or by a predetermined value, for example for the GOR - gas / oil ratio of the hydrocarbon to the exclusion of water, of the English "gas / oil ratio" - and / or the salinity which are predetermined fluid hypotheses) and therefore known to the metric counter which can therefore deduce the densities p. Thus, if the operating conditions P and T change and / or if the composition of the fluid (GOR, salinity) changes, the calibration triangle is also modified because the gas, oil and water densities vary. Thus linear attenuations also vary. In this sense, volume densities are calibration parameters that correspond in particular to predetermined fluid hypotheses (GOR and salinity).

FIG. 2 schematically shows a calibration triangle 20 corresponding to such a calibration, with coordinate vertices (a _gas , a _gas _ ₂ ), (cthuii, cihuiie _ ₂ ), and (a _water i, ct _ea u 2 ), corresponding to the six values listed above.

FIG. 3 shows an example of the counter: the counter 30. In order to perform the calibration, the counter 30 can receive a sample 34 of each of the phases, typically in the laboratory in the context of the prior art but in situ as part of the process. of FIG. 1. Then, the counter 30 measures at S20, thanks to a sensor 32 provided for this purpose, the mass attenuations of each of the phases, which makes it possible to determine S30 the calibration parameters (ie the linear attenuations, and / or or the mass attenuations themselves, depending on the chosen point of view, knowing that one or the other of these data can be recorded by the counter as a calibration parameter, leaving then to perform the appropriate calculation).

In use, the counter 30 measures a mass attenuation of the fluid (which is then a multiphase mixture). The sensor sensor 32 measures and records N'counts counts for each energy level (at least the "low" and "high" levels, but possibly an even higher level as well). With the following equations:

NOounts = No * decay * exp [- (x water

Skin + Xgaz Pgaz +

X oil *

* Phui) x Dihroa]

for i = 1 and 2,

and

Xeau Xgaz ^~ ^ ^~ X oil 1>

with x _water , gas and X uiie which denote the fractions of each of the phases, the densities p which are predetermined and given by the composition and thermodynamics, and N ₀ , decay and Dthroat which are known,

the counter 30 can determine the instantaneous fractions of each of the phases.

Figure 4 illustrates a real-world example of a possible problem with poor calibration. In this example, if the counter was calibrated all the operating points 15 would be inside the calibration triangle 10 of vertices 1 1, 12 and 13 (drawn according to the principle of FIG. 2). This is not the case and we obtain negative gas fractions. The method of FIG. 1 makes it possible to update this calibration triangle.

A computer program may be provided for the metric counter to execute the calibration process. This program is known in computer form adapted to be recorded on a data recording memory and may include instructions for executing the method. Thus, the metric counter may include a memory having recorded the program.

Alternatively, a computer program may be provided to control (remotely) the execution of the method by one (or more) metric counter (s). This program is known in computer science adapted to be recorded on a data recording memory and may include instructions. Thus, the computer program is not recorded by a memory of the metric counter, but by a memory of a system adapted to communicate with the metric counter. The system may typically include a processor and a transceiver for communicating with the metric counter and controlling the execution of the method. The metric counter can then include instructions for performing the calibration process in the form of a communication protocol with the system. The two computer programs mentioned above may comprise a preliminary step of detecting a stoppage of the flow in the line, for example by periodic tests, the subsequent execution of the method being slaved to this detection. This is detailed below.

The two computer programs mentioned above may include any type of known computer instructions. These instructions may be lines of code, written in any computer language, for example object-oriented, possibly in the form of source code, compiled code or precompiled code. They can also be installation programs (ie making a metric counter or a system suitable for executing the process or controlling its execution.) The programs can be tangibly recorded on a storage memory adapted to this. be a volatile or non-volatile memory, for example EPROM, EEPROM, flash memory, or CD-ROM disks.

FIG. 5 represents an example of a hydrocarbon production installation comprising several metric meters, in which the method can be implemented.

The plant 60 comprises a hydrocarbon production line, consisting of several pipes in which the fluid flows. The production line comprises in particular several fluid production wells 78 opening out from the hydrocarbon reservoir 66. Three production wells 78 are shown in the figure, but any number suitable for optimal coverage of the reservoir 66 is conceivable to take into account the geological complexity of the reservoir, the quality of the fluids present in the reservoir, the geographical location of the reservoir (on land, at sea, in deep sea), and inherent constraints. The production line comprises a fluid line 74 supplied with fluid by a manifold 40. The production fluid line 74 is a main line receiving the fluid from all the production wells 78 drilled in the tank 66 via a manifold (here the manifold 40), serving as a convener. The production fluid line 74 is located on a seabed 64, and feeds a "riser" 72 (ie a substantially vertical pipe) leading to a main station, eg in this case a floating unit for production, storage and unloading 68 (known under the acronym FPSO, for "Floating Production Storage and Offloading") located at the marine surface 62. The facility 60 also comprises an injection line comprising several injection wells 79, and operating according to a principle symmetrical to that of the production line.

The installation 60 incorporates (i.e. includes), for each production well 78 and each injection well 79, a metric counter 50, as previously described. The metric counter 50 is shown in the figure upstream of the wellheads 44, each time on a pipe line, but it can be at any other suitable place according to the appreciation of the skilled person. In addition, the installation 60 could include a quantity of metric meters 50 smaller or greater than that shown in the figure, depending on the needs of the operation. In any case, the installation 60 is adapted to the execution of the calibration method of FIG. 1 of at least a part of the metric meters 50, or by the metric meters 50 themselves which then have a program comprising instructions for this, or by a system adapted to the control of the metric meters 50 (the metric meters 50 then being adapted to be controlled in this sense, for example by recording programs for communication with the system according to a predetermined protocol) . The installation 60 allows a better hydrocarbon production, thanks to a simple calibration (thus easily repeatable) of the metric meters 50, providing correct calibration parameters, thus ensuring a good subsequent estimation of the flow rates by the meters 50.

Various options of the calibration method of Figure 1 are now described.

As mentioned above, the measurement S20 can be slaved to a detection of a stoppage of flow in the line. In other words, the method comprises detecting a stoppage of the flow in the line, by any detector provided for this, and systematically following this detection performs the measurement S20 (therefore at substantially zero rate). This systematization can typically be provided by the instructions included in one of the programs mentioned above. The installation may include valves adapted to block the passage of fluid in the lines of the line. Stopping the flow can be achieved by closing such a valve. For this purpose, alarms or events can be identified by rules, for example as closing a valve and it can further be verified that the delta P (ie pressure differential) in the meter is less than a minimum value. equivalent to the maximum static P delta. Performing the S20 measurement during a flow stop in the line makes it possible to perform a calibration that does not impede the production, and thus to optimize the overall operation of the tank using the metric meter estimate at any time. where the fluid flows in the line with significant flow.

In the case where the line is a production line, the detected stop may be a pit stop. In other words, for whatever reason, desired or unwanted, the production of the well in the line is (substantially) stopped, and this judgment is taken advantage of to perform a (re) calibration. The line can also be a test line, that is to say a line being used only during test phases, to test the values provided by the metric meters (production) of the production lines. For example, one or more pipes are derived incorporating a metric counter to locally measure the flow (with a metric test counter) and cross-check with the information provided by the metric meters (production) in question. In this case, the detected stop may be a test stop, that is, a phase where the test output is (substantially) stopped. This advantage is then taken advantage of to perform a (re) calibration, for example of the metric test counter.

When the measurement S20 is slaved to a detection of a stoppage of the flow in the line, the method may comprise, before the measurement S20, a segregation of the phases of the fluid. The measurement is then performed on the phase at the sensor. In other words, when the flow is stopped in the line, the flow rate of the fluid in the pipe to which the meter is integrated becomes substantially zero. When this fluid is multiphasic, the process can then comprise a segregation (ie a separation) of these different phases. This can be done naturally (eg by decantation), for a great simplicity. The method then consists in waiting for a given period. This period can be determined by prior tests. It is also possible to carry out the measurement S20, and depending on the value provided, it is possible, thanks to predefined knowledge, to determine the end or otherwise of the segregation (for example, it is known that the water attenuates more than the oil, and the oil attenuates more than gas). One can also repeat the measurement S20 and detect if it stabilizes, which is characteristic of an end of segregation. It can also be detected if the pressure stabilizes (for example, a standard deviation of less than 1 bara pressure is detected to signal an event, such as the end of the segregation). The time criteria for optimum segregation in the meter depend in particular on the type of fluid produced and the possibility or not of emulsion. This solution makes it possible to reach the predetermined condition of fluid flow (ie in the presence of a single phase of the fluid at the level of the sensor) in a simple manner.

When the metric counter is at the surface, the line where the meter is integrated is also on the surface and typically includes a manually accessible fluid purge valve. The process may then comprise successive opening and closing of the fluid purge valve to repeat the measurement S20 over several phases. In other words, following the segregation, the different phases of the multiphasic fluid are laminated vertically, with one of them at the meter sensor. The first measurement S20 is made on this phase at the sensor. But then it is possible by partial and successive purges, suitably adjusted according to the quantity of fluid of each of the phases, to repeat the measurement S20 on all the other phases above the first phase, until all the phases have been visited. This makes it possible to optimize the process, then performing the S30 determination for all the respective parameters at each of the phases on which the measurement S20 could thus be made.

When the metric counter is underwater or underground, stopping the flow in the line is the closing of a production valve. The process may then comprise successive opening and closing of the fluid production valve to repeat the measurement S20 over several phases. Since the pressure at the meter is high, the opening of the valve restarts the production (at least partially). Thus, as in the case of the purge presented above, it is possible to alternate between opening and closing (eg by a remote control) of the valve to put the vertically stratified phases of the multiphase fluid one after the other at the sensor . Conversely, in the case of the purge, the measurements S20 subsequent to the first measurement S20 are then performed on the phases below the first phase. This allows, as in the case of the purge, to optimize the process.

FIG. 6 shows an example of segregation with successive purges thus carried out during a test stop. Curve 176 represents the temperature which decreases naturally and gradually in the chamber of the meter. Curve 170 represents the pressure which remains constant. The curves 170, 172 and 174 represent nuclear measurements. We see in the figure a well test stop indicator 160 (and thus the beginning of the segregation) in the form of an end of differential pressure differential (curve 178). Indicators 162 of the purges are also shown in the form of step changes in the measurements. Indicator 164 of a new well test is also seen. When the curve 170 is around 3000 counts, this is representative of a 100% water point. When the curve 170 is around 8000 counts, this is representative of a 100% oil point. When the curve 170 is around 16000 counts, this is representative of a 100% gas point. Thus, as previously mentioned, once the segregation is complete, the measurement interval indicates whether it is in the presence of water, oil, gas or emulsion. If the data are exploitable (phase and stable conditions), the measurements are used to check the composition (eg the density, and / or the amount of sulfur) in the presence of gas or oil and the parameters the calculator and the fluid parameters necessary for the measurement processes are updated, and the salinity if it is in the presence of water. This example thus shows that the method can well perform the segregation and discriminate the different phases according to the measurements made.

It sometimes happens that the raw measurement S20 itself can be erroneous (for example derives from the gamma source, and / or bias on the temperature). One solution for identifying these errors is to compare and reconcile data from a 100% point of a given phase to several given dates. In other words, the measurement S20 can be iterated several times on the same phase (the operation can itself be reproduced for several phases). The determination S30 of the value of the calibration parameter is then a function of the result of the iterations of the measurement S20 (i.e. the result of several measurements S20 is taken to determine the value of the calibration parameter). The method can exclude any intervention between two iterations. This allows a temporal redundancy of data further improving the accuracy of the calibration.

For example, the iterations of the measurement S20 (on the same phase) are separated by an interval greater than one day, and / or less than six months, typically of the order of ten days. One can typically do between two and ten iterations, for example three or four iterations. This allows a moderate difficulty in gathering the information, while ensuring a good accuracy of the calibration. The S30 determination of the value of the calibration parameter can for this purpose include a Data Reconciliation and Validation (DVR) process involving the result of the iterations of the measurement. The reconciliation is conditioned by an equality (eg at least substantial, ie with a difference lower than a predetermined error, or even zero) between the result of the iterations of the measurement. In other words, the DVR is made on the basic assumption that measurements made at different iterations for the same phase should be equal. The data reconciliation and validation process may furthermore involve at least one predetermined fluid hypothesis (which may be assumed to be constant along the different iterations), thereby possibly correcting the assumption and keeping account of this correction during the determination S30 to better refine the result. According to tests (some of which are presented briefly below), the DVR provides a particularly accurate calibration, quickly and efficiently.

The DVR is a known process for providing input values (ie the data involved and ultimately reconciled by the process) and modifying these values according to predetermined uncertainty intervals and predetermined constraints. The DVR generally consists of the resolution of the following minimization program:

with y; the data involved in the process (i.e. the data to be reconciled),

σ; uncertainty the the measure i,

yi * the reconciled value of measure i,

σ; * the reconciled uncertainty of measure i,

Xj the unmeasured variables and thus for example calculated, and the term

under certain constraints linking the variables.

In this case, the method can execute a DVR to "equalize" the result of the S20 measurements made at different time iterations on the same phase, these being theoretically supposed to give the same result for the determination. in S30 of the same value of the calibration parameter. Thus the sum of the penalties is minimized by respecting, at least substantially, an equality constraint between the yi (which include the measurements made at the iterations of the S20 for each phase, and possibly one or more predetermined fluid hypothesis (s). ).

A test of a multi-period reconciliation on a 100% oil calibration point was carried out. Three sets of measurements were obtained:

1 / P = 35.5 Bara

T = 42 ° C

NcountsJ = 8050

NcountsJ = 11200

2 / P = 37.2 Bara

T = 43.5 ° C

NcountsJ = 8075

Ncounts_2 = 11150

3 / P = 36.3 Bara

T = 42.5 ° C

NcountsJ = 8041

Ncounts_2 = 11172

After reconciliation, we obtained

= 0.0131 and GOR = 51 Sm3 / Sm3.

Claims

1. Method for calibrating a metric meter integrated into a line of a hydrocarbon production facility, in which the meter is adapted to provide an estimate of the flow rate of a fluid in the line as a function of a measurement carried out by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined flow condition of the fluid in the line, the method comprising:

At least one measurement (S20) by the sensor under the predetermined condition, in situ, then

The determination (S30) of the value of the calibration parameter as a function of the result of the measurement.

2. The method of claim 1, wherein the fluid is multiphasic and the predetermined condition is the presence of a single phase of the fluid at the sensor.

3. The method of claim 2, wherein the fluid comprises a water phase, a gas phase and an oil phase.

4. The method of claim 2 or 3, wherein the measurement (S20) is slaved to a detection of a stop flow in the line.

The method of claim 4, wherein the line is a production line and the shutdown is a pit stop.

The method of claim 5, wherein the line is a test line and the stop is a test stop.

7. Method according to one of claims 4-6, wherein the method comprises, before the measurement (S20), the segregation of the phases of the fluid, the measurement is then performed on the phase at the sensor.

8. The method of claim 7, wherein the metric counter is at the surface, the method comprising a successive opening and closing of a fluid purge valve to repeat the measurement (S20) on several phases.

9. The method of claim 7, wherein the metric counter is underwater or underground, the method comprising a successive opening and closing of a fluid production valve to repeat the measurement (S20) over several phases.

10. Method according to one of claims 2-9, wherein the measurement (S20) is iterated several times on the same phase, the determination (S30) of the value of the calibration parameter being a function of the result of the iterations of the measurement. (S20).

11. The method of claim 10, wherein the iterations of the measurement (S20) are separated by an interval greater than one day, and / or less than six months.

The method according to claim 10 or 11, wherein the determination of the value of the calibration parameter comprises a data reconciliation and validation process involving the result of the iterations of the measurement, the reconciliation being conditioned by equality, at least substantially , between the result of the iterations of the measure.

The method of claim 12, wherein the reconciliation and data validation process further includes at least one predetermined fluid assumption.

14. A process for producing hydrocarbons comprising the calibration method according to one of claims 1-13.

The process for producing hydrocarbons according to claim 14, comprising:

• the initial calibration of the metric counter,

• integration of the meter with a line of a hydrocarbon production facility,

• the use of the meter to estimate the flow rate of the fluid during a production phase, then

The new calibration of the metric counter, in situ, by the method according to one of claims 1-13.

Computer program adapted for recording on a data recording memory, comprising instructions for executing the method according to one of claims 1-13.

17. A computer program adapted to be recorded on a data recording memory, comprising instructions for controlling the execution of the method according to one of claims 1-13 by a metric counter adapted to be integrated in a line. of a hydrocarbon production facility, and to estimate the flow rate of a fluid in the line as a function of a measurement made by at least one sensor on the fluid and the value of at least one corresponding calibration parameter measuring the sensor under a predetermined fluid flow condition in the line.

18. Metric counter adapted to be integrated in a line of a hydrocarbon production facility, to estimate the flow rate of a fluid in the line as a function of a measurement made by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined flow condition of the fluid in the line, and comprising a memory having recorded the program according to claim 16.

19. System adapted to communicate with a metric counter adapted to be integrated in a line of a hydrocarbon production facility, to estimate the flow rate of a fluid in the line as a function of a measurement made by at least one sensor on the fluid and the value of at least one calibration parameter corresponding to the measurement of the sensor under a predetermined flow condition of the fluid in the line, the system comprising a memory having recorded the program according to claim 17.

20. A hydrocarbon production installation comprising at least one metric meter according to claim 18 and / or the system according to claim 19.