EP1468304A1

EP1468304A1 - Nuclear magnetic resonance method of detecting and monitoring the flocculation kinetics of heavy fractions of a complex fluid

Info

Publication number: EP1468304A1
Application number: EP03712250A
Authority: EP
Inventors: Jean-Pierre Cohen Addad; Marc Fleury
Original assignee: IFP Energies Nouvelles IFPEN
Current assignee: IFP Energies Nouvelles IFPEN
Priority date: 2002-01-16
Filing date: 2003-01-16
Publication date: 2004-10-20
Also published as: FR2834792A1; CA2472087A1; WO2003060539A1; AU2003216731A1; US20050083052A1; FR2834792B1; US7009393B2

Abstract

The invention relates to a nuclear magnetic resonance method of detecting and monitoring the flocculation kinetics of high molecular weight fractions of a complex fluid. The inventive method consists in applying the following to the fluid: a first static polarisation magnetic field and, subsequently, at least a second oscillating pulsed magnetic field which is intended to generate a nuclear magnetic resonance for the nuclei considered and the acquisition of relaxation signals from the nuclei in the fluid. Moreover, said method consists in: detecting, in the relaxation signals, a first part P1 which is representative of the relaxation of said flocculated fractions in the fluid and a second part P2 which is representative of the relaxation of the liquid fraction of the fluid; and determining the flocculation rate of said fractions by comparison with values Mx(t = 0) and Mx1(t = 0) which were extrapolated at the start of the acquisition times of said two parts. The invention is suitable, for example, for monitoring the flocculation kinetics of generally asphaltene polar fractions which are contained in the dissolved state and/or in the stable colloidal state in a liquid hydrocarbon fluid.

Description

METHOD OF DETECTION AND MONITORING BY NUCLEAR MAGNETIC RESONANCE OF THE KINETICS OF FLOCCULATION OF HEAVY FRACTIONS OF A COMPLEX FLUID.

The subject of the invention is a method of detection and monitoring by nuclear magnetic resonance of the flocculation kinetics of high molecular weight fractions of a complex fluid.

The method according to the invention notably has the application of monitoring the flocculation kinetics of generally asphaltenic polar fractions, which are contained in the dissolved state and / or in the stable colloidal state in a liquid hydrocarbon fluid.

The flocculation and deposition processes pose considerable problems in the petroleum industry. In particular for heavy oils, the very high molecular weight components (asphaltenes, resins) are often at the origin of these processes which can appear both in porous media during production and during transport. Flocculation is the formation of molecular aggregates of micron size resulting in sedimentation or deposition which can considerably modify the flow of the fluid, either by reducing the flow section or by increasing the viscosity. In addition, the intrinsic load of certain components (eg asphaltenes) generates a strong tendency to attach to charged surfaces.

The thermodynamic parameters which govern the flocculation processes are numerous (composition, pressure, temperature) and the complexity of the molecular structures involved make prediction and modeling very uncertain. Likewise, certain recovery processes (CO2 injection, acidification) can modify the fluid balances and cause these processes. Thus, it is necessary to carry out measurements but the techniques available do not make it possible to observe the first stages of flocculation and pose significant implementation problems in pressure and temperature or in situ in oil wells. State of the art

Many crude oils, in particular those called asphaltenic crude oils, are liquid hydrocarbon fluids which contain more or less large quantities of heavy fractions in the dissolved state and / or in the colloidal state stable under the conditions of pressure and temperature to which these are subject _. fluids. When these pressure and / or temperature conditions vary, especially when the pressure decreases, the heavy fractions contained in these fluids tend to flocculate and deposit in the formation around the wells, in the wells and in the production facilities. and transfer of said fluids. Thus, when one exploits a hydrocarbon deposit containing heavy fractions and generally before reaching the bubble point, the stability of these fractions decreases. When the saturation threshold is reached, the heavy fractions flocculate and deposit, which can cause clogging of porous media as well as the formation of plugs liable to cause. severe damage to production wells and surface facilities.

For petroleum producers, who have to extract and transport, via production wells and pipeline networks, liquid hydrocarbon fluids consisting of crude oil containing heavy fractions, for example asphaltic crude oil, from production fields, it is therefore important to have a precise knowledge of the pressure thresholds below which the heavy fractions are deposited, so as to implement the production and the transfer of said fluids under pressure and temperature conditions avoiding the deposits of heavy fractions in facilities or to provide appropriate treatment.

Various methods are known for determining the deposition threshold for heavy fractions, in particular asphaltenes, contained in liquid hydrocarbon fluids consisting of crude petroleum. These methods are most often optical methods by transmission or diffusion of light, conductimetric methods or even viscosimetric methods.

A known method called stain test, consists in depositing a little mixture on a filter paper, and to observe the stain which forms. The flocculation aggregates which form in a mixture diffuse less quickly than the surrounding liquid. Thus, if the stain is not uniform, it is an indication that it contains flocculating particles. The aforementioned methods call for the detection of the variation of a physical quantity, for example absorption coefficient or absorbance of the light rays in the visible or in the infrared domain, electrical conductivity or viscosity, which results from the change of structure of the fluid as a result of flocculation and deposition of heavy fractions.

A major drawback of such methods is that they are not very selective in the sense that it is not always easy to relate the variation of the physical quantity measured to the flocculation and the deposition of heavy fractions and that they do not are not always sensitive to the deposition of a small amount of such fractions. Certain methods, such as the measurement of absorbance in the infrared, are very sensitive, but difficult to implement under deposit conditions.

In addition, as these methods are often implemented in the laboratory, there is also the question of the representativeness of the samples, on which the measurements of the physical quantity are carried out. In fact, for a sample to be representative of the fluid sampled, it is necessary to maintain this sample under the pressure and temperature conditions prevailing for the fluid sampled, for example reservoir fluid, throughout the sampling operations, transport and storage of the sample preceding the measurement.

By patent FR 2,818,753 of the applicant, a method is known for determining the threshold for depositing heavy fractions, which are contained in the dissolved state and / or in the stable colloidal state in a liquid hydrocarbon fluid. The invention provides a method for determining the threshold for depositing heavy fractions, in particular asphaltenes, contained in the dissolved state and / or in the stable colloidal state in a liquid hydrocarbon fluid, which is based on the formation of a increasing pressure drop due to the flow, at increasing flow, of a sample of the fluid through a capillary passage.

The fluid sample being in the dissolved state and / or in the stable colloidal state, at the entry of a capillary passage a pressure drop is generated between inlet and outlet at least equal to the difference between the pressure of the fluid sample and the bubble pressure of said sample. A significant change in the variation as a function of time of ΔP is detected (difference between the fluid pressure at the inlet of the capillary and the pressure at its outlet) and by a magnitude D representative of the flow rate of liquid flowing through the capillary passage, which makes it possible to characterize the threshold for depositing heavy fractions of the fluid. Furthermore, it is known that (NMR) can be used in particular for measuring certain physical characteristics of mixtures of fluids such as hydrocarbons, in particular viscosity or their gas-liquid ratio or GOR. The viscosity of the mixture and its coefficient GOR are obtained from the NMR measurement of a diffusion coefficient D and from the measurement of longitudinal relaxation time T1, or transverse T2. Such an application is described for example in patents WO 0142817 or US 5696448 or in the following document:

- Prammer, MG et al, (2001); "the Dowhole NMR Fluid Analyzer"; SPWLA 42 ^nd Annual Logging Symposium.

It should be noted that these measurements are generally carried out in a time interval after the excitation which turns out to be unsuitable for the detection and interpretation of the flocculation phenomena.

It is also known to use NMR apparatuses to detect and follow a very different phenomenon which is the crystallization of particles in fluids.

The method according to the invention

The inventors have observed that the NMR type methods applied to the detection of solid particles could also surprisingly be applied to non-solid particles of high molecular weight in slow rotation which gradually aggregate together, and verified that the this type of method could determine their rate of flocculation.

The method of detection and monitoring by nuclear magnetic resonance of the kinetics of flocculation of non-solid aggregates of high molecular mass of a complex fluid comprises the application to the fluid of a first static magnetic magnetic field of polarization then of at least a second pulsed magnetic field oscillating perpendicular to the first created by windings connected to an excitation generator to carry out the nuclear magnetic resonance of the nuclei considered and the acquisition of the relaxation signals of the nuclei in the fluid.

It comprises the detection on the relaxation signals, of a first part representative of the relaxation of these aggregates in the fluid and of a second part representative of the relaxation of the liquid fraction of the fluid, and the determination of the rate flocculation (Tf) of the aggregates by comparison of the extrapolated values at the origin respectively of the acquisition times of the first part and of the second part.

We can determine the flocculation rate by the relation:

Tf = (M _x (t = 0) - M _xl (t = 0)) / M _x (t = 0) or

(M _x (t = 0) and M _xl (t = 0) are the extrapolated values at the origin respectively of the acquisition times of the first part and the second part.

One can obtain for example the flocculation threshold of the fluid by modeling the relaxation signals actually obtained by means of a combination of exponential functions depending on an adjustment parameter and one searches for the threshold corresponding to a maximum value of said adjustment parameter ..

According to one mode of implementation, the method comprises the application to the fluid of a sequence of two 90 ° pulses called pseudo solid echoes in which a focalization pulse of the magnetization at 180 ° is interposed between two successive applications of the pulses at 90 °, with time intervals τ / 2 between the different pulses, and the maximum amplitude of the relaxation signals is measured. neighborhood of time t = 2τ for different values of τ in the sequence.

The proposed method has many advantages. It offers the possibility of continuous monitoring, useful for the analysis of flocculation depending on the chemical composition and the solvent. The entire volume is analyzed. Unlike optical methods, even non-transparent samples can be analyzed. The characterization of the kinetics is easy. Likewise, it is easy to assess the level of aggregates, which can lead to a rough estimate of the molecular weight.

Summary presentation of the figures

Features and _. advantages of the method according to the invention will appear more clearly on reading the description below of non-limiting example (s) of embodiment, with reference to the appended drawings where:

Figure 1 shows the decrease in transverse magnetization measured on a flocculated mixture; - Figure 2 shows the decay curves of the transverse magnetization measured for a mixture not initially flocculated, then flocculated;

- Figure 3 schematically shows an NMR type analysis device;

- Figure 4 shows the relaxation signal of an asphaltene / toluene / heptane solution; and - Figure 5 shows the evolution of the coefficient A of a function modeling the relaxation signal of Figure 4, which identifies the value of the flocculation threshold.

detailed description

The NMR analysis technique essentially consists, as it is recalled, of applying to a test object a first static magnetic magnetic field Bo of polarization intended to align the nuclei of the hydrogen protons initially oriented at random, in the direction of the field and then to a second pulsed magnetic field oscillating at the Larmor frequency perpendicular to the first created by windings excited by a control signal to carry out a nuclear magnetic resonance experiment. When this impulse field ceases, the return of the nuclei to their initial state or relaxation, generates electromagnetic signals (echoes) which are detected and analyzed. From the amplitude characteristics of these signals, the presence of this or that substance and some of its physical parameters are determined.

In the case where the analysis relates to a fluid mixture, we conventionally pass (Fig. 3) the conduit 1 where it circulates (or possibly the container which contains it) first in the air gap of a permanent magnet 2 where there is a static field B ₀ (z) polarized in a direction z, then in a winding 3 connected on the one hand to a generator 4 delivering on command a current pulse at a frequency in the range of radio frequencies creating a transverse pulse field oscillating Bι (t) in a direction x perpendicular to the direction z, and on the other hand to a circuit 5 for detecting relaxation signals.

Within the framework of the method according to the invention, the analysis of the relaxation of the transverse magnetization M _x (t) of the protons of a liquid considered, observed at short times, is applied both to the detection of flocculation, the determination of the rate of large structures formed in suspension in a liquid and the characterization of the kinetics governing this formation process. One can proceed conventionally by using a so-called Hahn sequence or even following a so-called pseudosolid echo sequence. After polarization by the static magnetic field Bo (z) of the order of 0.5 T for example, a pulse field Bι (t) is applied by means of the windings 3 which has for example a falling edge and a time of receiver desaturation less than a few microseconds. A series of at least two so-called 90 ° pulses separated by a time interval τ will be applied, with an interposed 180 ° pulse intended to focus the magnetization and thus to overcome the diamagnetic heterogeneities of the material. The instants of application of the pulses are separated by time intervals of duration τ / 2. The sequence can be diagrammed in time as follows:

(90 °) x - τ / 2 - (180 °) y - τ / 2 - (90 °) x - echo acquisition

By analyzing the echoes by means of a device 6, their maximum amplitude M _x (t) is determined in the vicinity of the time t = 2τ for different values of τ in the above sequence.

For a liquid in which certain fractions have flocculated, the relaxation signal has two very distinct parts corresponding to the appearance of large structures in the liquid: a first part which decreases very rapidly over twenty microseconds and whose importance increases regularly with the formation of flocculated structures, and a second part which decreases very slowly over several hundred milliseconds and corresponds to the liquid part of the fluid (Fig. 2). The flocculated structures produce a rapid decrease in magnetization because they can be considered as frozen

(Quasi-solid). They are thus clearly distinguished from the magnetic properties of the surrounding liquid. Furthermore, the flocculated structures are also subjected to the Brownian movement generated by the thermal agitation of the liquid, but this random movement is too long to induce a magnetic relaxation process. This approach applies as soon as the average size of the structures exceeds a few nanometers.

The rate of flocculated aggregates Tf can be obtained by extrapolating the signal from the liquid part M _xl at time t = 0.

A graphical method can be used when, as in the example in Fig.l, the distribution of the points relating to the liquid part of the fluid analyzed is substantially linear. One can also search for example a polynomial of sufficient degree or if necessary a sum of exponential functions modeling the distribution of the points of the liquid part, and to deduce from it the extrapolated value at time t = 0.

From this value M _xl (t = 0) and the corresponding value of the first part of the distribution close to the origin of times, similarly extrapolated to time t = 0 or M _x (t = 0), we obtains the rate Tf. For example, we can define Tf by the relation:

Tf = [M _x (t = 0) - M _xl (t = 0)] / M _x (t = 0) (1)

where M _x ι = M _x (t> tc), te being the time separating the two regions of decrease of M _x (t). It is implicitly assumed that the total mass analyzed does not change.

The detection of the first part of the relaxation distribution (Fig. 1) supposes that the relaxation signals are acquired practically from the origin of the times, that is to say within an interval much less than 1 ms, which is not the case with conventional methods using NMR analysis apparatus.

The observation of the kinetics of flocculation as a function of time is illustrated in FIG. 2 for a viscous liquid (100 centipoises). At f = 5 days, there is no flocculation because the signal does not show rapid decay. At t = 29 days, certain flocculated structures appear and become more numerous at t = 36 days. The acquisition time of a curve M _x (t) being a few minutes, faster kinetics can easily be observed.

The example of application of the method below, which relates to alphaltene / toluene / heptane solutions, clearly shows a method of detecting the flocculation threshold.

It is known that when the proportion of heptane increases, this type of flocculated solution. Figure 4 shows an example of the transverse magnetization signals (Mx (t)) obtained on these solutions. The parts PI and P2 previously described are connected to it by an intermediate part P3. In general, the signals are correctly described by a function which is a sum of exponentials. In the following function:

Mx (t) = exp [- (t / τ) ^α ] + A exp (-t / T _2S ) (2) the second term A exp (-t / T _2S ) is characteristic of the relaxation time of the heptane / toluene mixture. Equation 2 is useful for separating the high molecular weight part characterized by very short relaxation times (order of magnitude τ), from that of the heptane / toluene mixture (order of magnitude T ₂ s). The first term describes the PI part as well as the intermediate part P3. For the mixture considered, the high molecular weight buildings are visible before flocculation.

By determining the coefficients of relation 2 so that the modeled curve coincides with the curve in Figure 4, we obtain the value of A.

The aggregation rate of this mixture is then calculated from the following relationship: Tf = (l-A) / A (3)

This way of calculating Tf is equivalent to that which one obtains by applying the relation 1. The follow-up of the evolution of the coefficient A, or in an equivalent way of Tf, shows that the flocculation threshold corresponds to the maximum of this coefficient, in accordance with standard techniques.

In FIG. 5, we see the evolution of the coefficient A as a function of the T / H proportion of toluene / heptane, when the asphaltene concentration is respectively 10.7, 8.8 and 7.6%. We see that the flocculation threshold which is at the heptane / toluene mass proportion 35/65%. by the arrow, corresponds to the maximum of the coefficient A.

The NMR measuring device used for implementing the method can be miniaturized for the purpose of being installed in an oil well in conjunction with a fluid sampler of a type known in any other surface installation for monitoring. flocculation phenomena. Whatever the analysis conditions, approximately 50 mg of substance is sufficient in practice.

Claims

1) Method for detection and monitoring by nuclear magnetic resonance of the kinetics of flocculation of non-solid aggregates of high molecular mass of a complex fluid comprising the application to the fluid of a first static magnetic field of polarization Bo (z) then at least one second pulsating magnetic field oscillating Bι (t) perpendicular to the first created by windings (3) connected to an excitation generator (4) to carry out nuclear magnetic resonance of the nuclei considered and the acquisition of signals of nucleus relaxation in the fluid, characterized in that it comprises the detection on the relaxation signals, of a first part representative of the relaxation of said aggregates in the fluid and of a second part representative of relaxation of the liquid fraction of the fluid, and the determination of the flocculation rate (Tf) of said fractions by comparison of the values (M _x (t = 0) and M _xl (t = 0) ext originally collected respectively the acquisition times of said first part and said second part.

2) Method according to claim 1, characterized in that in that the flocculation rate is determined by the relation:

Tf = (M _x (t = 0) - M _xl (t = 0)) / M _x (t = 0)

3) Method according to claim 1, characterized in that the flocculation threshold of the fluid is obtained by modeling the relaxation signals actually obtained by means of a combination of exponential functions dependent on an adjustment parameter (A) and the threshold corresponding to a maximum value of said adjustment parameter is sought.

4) Method according to one of the preceding claims, characterized in that it comprises the application to the fluid of a sequence of two 90 ° pulses called pseudo solid echoes in which a focusing pulse of the magnetization at 180 °, between two successive applications of the pulses at 90 °, with time intervals τ / 2 between the different pulses, and the maximum amplitude of the relaxation signals is measured in the vicinity of the time t = 2τ for different values of τ in the sequence.