Multi-phase metering device for oilfield applications

Abstract

This application is related to a system and methods for sampling fluids and gases using nuclear magnetic resonance (NMR) technology. Specifically the system is related to an improved metallic pipe design for use at oil and gas well heads that includes integral coils for transmitting an NMR pulse sequence and detecting NMR signals and can be used as a component of an NMR instrument. The methods are related to obtaining and analyzing NMR spectra in stationary and flowing states.

Claims

1. A method of measuring the relative quantity of components in a high temperature and pressure fluid using nuclear magnetic resonance (NMR) relaxometry in an NMR pressure tube having a magnetic field, comprising the steps of: a) calibrating the NMR pressure tube with a pure water sample by putting a pure water sample within the NMR pressure tube and applying an electromagnetic field to the pure water sample and recording a signal response using an NMR coil thereby providing a pure-water calibration curve; b) calibrating the NMR pressure tube with a pure oil sample by putting a pure oil sample within the NMR pressure tube and applying an electromagnetic field to the pure oil sample and recording a signal response using the NMR coil thereby providing a pure-oil calibration curve; c) introducing an at least two-component mixture into the NMR pressure tube; d) measuring relaxation spectra of the mixture in the NMR pressure tube by applying an electromagnetic field to the mixture and recording a response using the NMR coil; and e) calculating, using a processor, a quantity of oil and water in the at least two-component mixture by fitting the mixture relaxation spectra of step d) to a linear combination of the pure-water and pure-oil calibration curves of steps a) and b).

2. The method as in claim 1 wherein steps a) and b) include calculating total water amplitude and total oil amplitude respectively.

3. The method as in claim 2 further comprising repeating steps a) and b) over a selected temperature range.

4. The method as in claim 1 wherein during steps c) and d) the at least two-component mixture is stationary within the NMR pressure tube.

5. The method as in claim 4 wherein step e) includes averaging a component saturation value along a length of the NMR pressure tube.

6. The method as in claim 1 wherein during steps c) and d) the at least two-component mixture is flowing within the NMR pressure tube.

7. The method as in claim 1 wherein step e) includes calculating a volumetric component-cut and mass component-cut of one component of the mixture.

8. The method as in claim 1 wherein the at least two-component mixture includes different phases.

9. The method as in claim 8 wherein the different phases include a liquid phase and a gas phase.

10. The method as in claim 9 wherein the liquid phase is water.

11. The method as in claim 10 further comprising the step of calculating a gas-water ratio of the mixture.

12. The method as in claim 8 wherein the different phases include a hydrocarbon phase and a gas phase.

13. The method as in claim 12 further comprising the step of calculating a hydrocarbon-gas ratio of the mixture.

14. The method as in claim 1 wherein the at least two-component mixture includes a gas component, water component and hydrocarbon component.

15. The method as in claim 1 wherein the magnet and the applied electromagnetic fields are configured to obtain responses from hydrogen atoms from materials making up samples or mixtures within the NMR pressure tube.

16. An NMR apparatus for measuring the relative quantity of components in a high temperature and pressure fluid using nuclear magnetic resonance (NMR) relaxometry, the apparatus comprising: an NMR pressure tube; a magnet configured to apply a magnetic field within the NMR pressure tube; an NMR coil configured to apply an electromagnetic field within the NMR pressure tube; an electronic circuit for controlling the NMR coil to generate an electromagnetic field; a data acquisition board for receiving an NMR response of a sample from the NMR coil; a processor configured to process the NMR response to determine a quantity of oil and water in the sample by: a) storing a pure-water calibration curve obtained by calibrating the NMR pressure tube with a pure water sample by putting a pure water sample within the NMR pressure tube and applying an electromagnetic field to the pure water sample and recording a signal response using the NMR coil; b) storing a pure-oil calibration curve by calibrating the NMR pressure tube with a pure oil sample by putting a pure oil sample within the NMR pressure tube and applying an electromagnetic field to the pure oil sample and recording a signal response using the NMR coil; c) receiving a mixture relaxation spectra corresponding to an at least two-component mixture within the NMR pressure tube, the relaxation spectra of the mixture in the NMR pressure tube having been obtained by applying an electromagnetic field to the mixture and recording a response using the NMR coil; and d) calculating the quantity of oil and water in the at least two-component mixture by fitting the mixture relaxation spectra of step c) to a linear combination of the pure-water and pure-oil calibration curves of steps a) and b).

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention is described with reference to the accompanying figures in which:

(2) FIG. 1 is a schematic view of an NMR apparatus for use in an NMR instrument in accordance with one embodiment of the invention;

(3) FIG. 2 is a cross-sectional view of an NMR compatible pipe in accordance with one embodiment of the invention;

(4) FIG. 2A is a schematic end view of an NMR pipe in accordance one embodiment of the invention;

(5) FIG. 3 is a front sectional view of an NMR apparatus in accordance with one embodiment of the invention;

(6) FIG. 4 is a graph of a relative quality factor (Q factor) of a resonator versus the ratio of the resonator diameter (D.sub.r) and the ratio of an outer sleeve diameter (D.sub.sl) of an NMR pipe in accordance with one embodiment of the invention;

(7) FIG. 5 is a graph of a relaxation spectrum of an oil and water mixture in accordance with one embodiment of the invention; and

(8) FIG. 6 is a graph of total water amplitude versus total oil amplitude for field measurements performed in accordance with one embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

(9) With reference to the figures, an NMR-compatible system 10 for use with an NMR instrument 12 and methods for determining the content of a fluid at oil and gas well heads using the NMR-compatible system and NMR instrument are described.

(10) NMR-Compatible Apparatus

(11) As shown in FIG. 3, the NMR-compatible system 10 consists of an NMR pipe 40 that is connected to conventional well-head piping (not shown). The NMR pipe 40 has connection devices 40b, 40c at each end to secure the pipe to the conventional pipe. A portion of or all fluid from the conventional pipe diverted through the NMR pipe 40 in a one-way direction, as shown in FIG. 3, wherein the fluid within the NMR pipe is subjected to NMR measurement by the NMR instrument 12 to determine the content of the fluid sample.

(12) FIGS. 1, 2 and 2A illustrate the pipe 40 as having a resonator 4 located inside the NMR pipe for the transmission of a pulse sequence and NMR signal detection, and several layers including an outer layer 44, a shielding layer 46, a gap layer 48 and a core layer 50. An appropriate NMR magnet 30 surrounds the pipe 40 such as a permanent low field magnet (1-5 MHz).

(13) As shown in FIGS. 1 and 3, to enable field-use and deployment, the apparatus 10 is secured to a skid 18b and further includes an electronic circuit 20 for generating and detecting the NMR signals. The electronic circuit preferably comprises a matching circuit 1, a preamplifier 2, a transceiver switch 3, an amplifier and filter 4, a data acquisition board 5, a computer/processor 6, a peripheral interface 7, a pulse forming board 8 and a transmitter 9. Preferably, various components of the electronic circuit 20 are enclosed in an explosion proof cabinet 18, shown in FIG. 1 and FIG. 3, in which an inside positive pressure is maintained preferably through the use of compressed air. A temperature control and purge unit 18c is located on the outside of the cabinet 18, along with a port 7b for the peripheral interface 7. First and second resistive thermal devices (RTDs) 54a, 54b are attached to the input and output ends of the pipe 40 to measure temperature.

(14) As is known to those skilled in the art, the magnet 30 creates a strong, homogenous magnetic field that causes certain nuclei within the fluid sample 52 to line up within the magnetic field. The pulse forming board 8 provides pulsations of radiofrequency (RF) energy in a CPMG (Carr. Purcell, Meiboom and Gill) sequence that are transmitted to the resonator 42. The RF signal excites aligned molecules within the sample that then cause certain atomic nuclei to resonate. When the RF signal is turned off, the nuclei “relax” and produce a weak RF signal which induces a small current in the resonator coil that is received by the data acquisition board. The current is processed and analyzed by the processor to create NMR spectra for the sample using a standard NNLS (non-negative least-squares) algorithm. As atoms of different substances relax at different rates, it is possible to determine the relative amounts of particular atoms in the sample using NMR relaxometry analysis, of which methods are described below. In the preferred embodiment of the invention, hydrogen atoms are excited and hydrogen bearing moles riles are detected.

(15) Pipe Layers

(16) The different layers of the pipe 40 are designed to maximize the signal-to-noise (SNR) ratio of the NMR instrument by maximizing the sample volume of the fluid 52 for a given diameter of pipe. In NMR, SNR is proportional to the square root of the quality factor (Q) of the resonator 42 and to the sample volume. It is preferable that the Q factor is optimized for SNR and for ringing time constant, which is proportional to Q. At some point of Q, ringing time (recovery time or dead time) is minimal in order to maximize SNR. As known to those skilled in the art, there are practical methods, such as active damping, that can be used to improve SNR while keeping recovery time minimal.

(17) The non-magnetic outer layer 44 has mechanical characteristics designed to withstand the high temperatures and pressures that pipes used in oil and gas operations typically encounter. Suitable materials include stainless steel, beryllium, copper, and titanium. Preferably titanium (Grade 2 or Grade 5) is used, as a lesser thickness of titanium is required in comparison to beryllium, copper, and stainless steel to provide the necessary mechanical characteristics. The smaller wall thickness translates into a larger available volume inside the pipe for the sample fluid, which effectively increases the SNR of the instrument.

(18) Located interior to the outer layer 44 is the shielding layer 46 that is designed to shield the resonator 42 from outside noise. The shielding layer is preferably made from the same material as the resonator 42, such as copper. Alternatively the shielding layer is manufactured from a non-magnetic material with a higher conductivity than the material of the resonator in order to maximize the Q factor of resonator 42. Table 1 below illustrates the ratio of the resonator diameter (D.sub.r) to shielding layer diameter (D.sub.sl) to maximize the Q factor of the resonator for a given shielding layer material.

(19) TABLE-US-00001 TABLE 1 Ratio Of Resonator Coil Diameter (D.sub.r) To Shielding Layer Diameter (D.sub.sl) To Maximize The Q Factor For A Given Shielding Layer Material Relative high frequency Best ratio of D.sub.r/D.sub.sl to Shielding Layer effective resistance to maximize Q factor of Material copper resonator Super Conductor 0 0.659 Silver 0.98 0.552 Copper 1 0.55 Titanium 5.06 0.369

(20) As shown in FIG. 4, when the shielding layer 46 and the resonator 42 are made of the same material (e.g. copper), the ratio of the diameter of the resonator coil (D.sub.R) and the diameter of the shielding layer (D.sub.SL) are preferably optimized at about 0.55, which causes the resonator coil to have the highest Q factor for the limited volume available within the outer layer 44. Furthermore, it can also be seen in FIG. 4 that an increase in the resonator coil diameter in order to fit it into the available inner diameter of the outside conducting pipe, i.e. where the D.sub.R/D.sub.SL ratio approaches 1, leads to a drastic drop in the Q value.

(21) Interior of the shielding layer 46 is the gap or insulating layer 48 that creates a non-conductive space between the shielding layer and the innermost core layer 50 for optimum Q as shown in FIG. 4. The gap layer is filled with material in order to prevent flow of fluid on the outside of the resonator coil and to prevent the coil from mechanical wear. With this configuration, the gap layer also transfers the high pressure forces of a fluid sample within the pipe to the outer layer 44. Resins, as known to those skilled in the art, are suitable materials for the gap layer, as they are non-conductive and have sufficient mechanical strength. Another suitable material is polyetheretherketone (PEEK). The thickness of the gap layer 48 is defined by (D.sub.r-D.sub.sl)/2.

(22) The inner core layer 50 is a hollow cylinder for containing the fluid sample 52 within the inner core volume such that the fluid sample is in contact with the inner surfaces of the inner core layer. The inner core layer also provides support for the resonator 42 that is contained within the insulating layer. A suitable material for the core layer is polyetheretherketone (PEEK) which is non-metallic and has a high resistance to corrosion caused by a typical chemical environment of the fluid sample. Another suitable material for the inner core layer is Teflon®. It is preferable that the inner core layer 50 be made as thin as possible in order to maximize the sample volume; however the thinness of the inner core layer is restricted by factors including the abrasiveness of the fluids.

(23) The resonator 42 is preferably a standard solenoid coil wrapped around the core layer that is immersed and contained within the gap layer 48. Preferably, the length L of the coil along the tube is at least twice the diameter of the coil which increases the homogenous radiofrequency (RF) field area inside the coil. It is preferable to use multiple wires connected in parallel which increases both the RF field homogeneity and the Q value of the coil.

(24) Method for Determining Fluid Content

(25) Methods for determining the properties of fluids, including the oil, water, solvent and gas content, at oil and gas well heads using NMR relaxometry are described. The measurements are taken in either stationary or flowing modes for the fluid.

(26) Measurement of Oil and Water Content in a Stationary Fluid

(27) To determine the oil and water content of a stationary fluid in a pipe running through an NMR meter, a heavy oil (bitumen) and water signal are separated in the NMR T2 relaxation spectrum. The two measurements can be taken independently of each other. The graph in FIG. 5 illustrates a typical observed separation of oil and water peaks.

(28) Assuming that the pipe is totally and uniformly filled with a mixture of oil and water (or radial sensitivity of NMR is uniform), the signal from water is proportional to the amount of water in the mixture in the following sense:
A.sub.w(T,P)=∫AI.sub.w(T,P,{right arrow over (r)})ρ.sub.w(T,P)S.sub.w({right arrow over (r)})d.sup.3{right arrow over (r)}  (1)
where T is temperature, P is pressure, {right arrow over (r)} is a vector representing integration element position, A.sub.w(T, P) is total water amplitude, AI.sub.w(T, P, {right arrow over (r)}) is water (mass) amplitude index, ρ.sub.w(T, P) is water density, and S.sub.w({right arrow over (r)}) is current water saturation (portion of the fluid volume element occupied by water). In the case of the uniform fluids distribution only AI.sub.w(T, P, {right arrow over (r)}) is spatially dependent, then:
A.sub.w(T,P)=ρ.sub.w(T,P)S.sub.w∫AI.sub.w(T,P,{right arrow over (r)})d.sup.3{right arrow over (r)}  (2)

(29) The amount of oil in the fluid can be determined by replacing water with oil in the above formulae.

(30) Calibration of the system is performed with the pipe filled with water only based on the following:
A.sub.w,100%(T,P)=ρ.sub.w(T,P)∫AI.sub.w(T,P,{right arrow over (r)})d.sup.3{right arrow over (r)}  (3)
Water cut (volumetric) (S.sub.w) within a cross-section of the pipe inside the magnetic field can be obtained according to the following relation:

(31) $\begin{array}{cc}{S}_w=\frac{A_w\left(T,P\right)}{A_{w,100\%}\left(T,P\right)}& \left(4\right)\end{array}$

(32) As S.sub.w+S.sub.o=1, then S.sub.o=1−S.sub.w.

(33) Volumetric ter cut can be converted into the mass water cut (WC.sub.m) by the following:

(34) $\begin{array}{cc}{\mathrm{WC}}_m=\frac{S_w{\rho }_w\left(T,P\right)}{S_w{\rho }_w\left(T,P\right)+S_o{\rho }_o\left(T,P\right)}& \left(5\right)\end{array}$

(35) Radial variations of AI.sub.w(T, P, {right arrow over (r)}) in a properly designed NMR relaxometer can be as low as 1% and even less. However, within the length of the measured volume of the pipe inhomogeneities of the magnetic field will exist. There may also be variations in water saturation along the length of the pipe if the system is flowing. To account for these variations, the above formulae become:
A.sub.w(T,P)=∫AI.sub.w(T,P,z)ρ.sub.w(T,P)(∫S.sub.w(x,y,z)dxdy)dz  (6)
or
A.sub.w(T,P)=∫AI.sub.wT,P,z)ρ.sub.w(T,P)S.sub.w(z)dz  (7)
where S.sub.w(z)=(∫S.sub.w(x, y, z)dxdy) is water saturation averaged over the pipe cross-section. With the flow the pipe stationary and settled this value does not depend on z and again:

(36) $\begin{array}{cc}{A}_w\left(T,P\right)={\rho }_w\left(T,P\right){\stackrel{\_}{S}}_w\int {\mathrm{AI}}_w\left(T,P,z\right)dz& \left(8\right)\\ A_{w,100\%}\left(T,P\right)={\rho }_w\left(T,P\right)\int {\mathrm{AI}}_w\left(T,P,z\right)dz& \left(9\right)\\ {\stackrel{\_}{S}}_w=\frac{A_w\left(T,P\right)}{A_{w,100\%}\left(T,P\right)}& \left(10\right)\end{array}$
Measurement of Water and Gas Content in a Stationary Fluid

(37) The last set of formulae can be applied to any two phase system present in the pipe If it is known that the only phases present are gas and water then the above formulae still give the volumetric water saturation S.sub.w and volumetric gas saturation S.sub.g=1−S.sub.w. The mass gas-water ratio (GWR) can be established based on the equation of state of gas at the known pressure and temperature:

(38) $\begin{array}{cc}\mathrm{GWR}=\frac{S_g{\rho }_g\left(T,P\right)}{S_w{\rho }_w\left(T,P\right)}& \left(11\right)\end{array}$
Measurement of Oil and Gas Content in a Stationary Fluid

(39) The above water-gas measurement procedure is directly transferable to oil-gas flows.

(40) $A_o\left(T,P\right)={\rho }_o\left(T,P\right){\stackrel{\_}{S}}_o\int {\mathrm{AI}}_o\left(T,P,z\right)dz$$A_{0,100\%}\left(T,P\right)={\rho }_o\left(T,P\right)\int {\mathrm{AI}}_o\left(T,P,z\right)dz$${\stackrel{\_}{S}}_o=\frac{A_o\left(T,P\right)}{A_{o,100\%}\left(T,P\right)}$

(41) It should be understood that the measurement of the value:
A.sub.o,100%(T,P)=ρ.sub.o(T,P)∫AI.sub.o(T,P,z)dz
for oil will require a sufficient amount of oil in order to perform a calibration procedure. As oil properties are subject to more variation than water properties, calibration procedures must occur more frequently.
Measurement of Oil, Water and Gas Content in a Stationary Fluid

(42) In order to determine the oil, water and as content of the flow, the measurements of oil and water signals in an appropriate range of relaxation times can be applied as follows:
A.sub.w(T,P)=ρ.sub.w(T,P)S.sub.w∫AI.sub.w(T,P,z)dz(Water)
A.sub.o(T,P)=ρ.sub.o(T,P)S.sub.o∫AI.sub.o(T,P,z)dz(Oil)
A.sub.o,100%(T,P)=ρ.sub.o(T,P)∫AI.sub.o(T,P,z)dz(Pure Oil)
A.sub.w,100%(T,P)=ρ.sub.w(T,P)∫AI.sub.w(T,P,z)dz(Pure Water)

(43) Integration above is performed over the oil or water peak accordingly (see FIG. 5). Typically, the entire oil spectrum is below 300 ms and the water spectrum is above this threshold.

(44) $\begin{array}{cc}{S}_w=\frac{A_w\left(T,P\right)}{A_{w,100\%}\left(T,P\right)}& \left(12\right)\\ S_o=\frac{A_o\left(T,P\right)}{A_{o,100\%}\left(T,P\right)}& \left(13\right)\\ S_g=1-S_w-S_o& \left(14\right)\end{array}$

(45) Conversion of volume fractions into mass fractions can be performed as above with the use of PVT properties of each phase.

(46) The instrument must be calibrated by filling the pipe with water (equations (3)/(9)) or oil (for oil equivalent of equation (9)). A.sub.w,100%(T, P), A.sub.o,100%(T,P) for the full range of operating temperatures and pressures is done prior to installation. If the produced water and oil do not chemically change during production, then this calibration is sufficient. However, in order to account for noise and changes in production fluid properties, biannual calibrations are preferable. If the instrument is move to a different production location recalibration is preferable.

(47) Calibration Procedure for Oil

(48) In order to minimize the frequency of performing the on calibration procedure, the following alternate oil calibration procedure can be performed.

(49) For three phase measurements, the A.sub.o,100%(T,P) may be difficult to obtain. If the system can be operated in two-phase mode without gas then the following calibration can be made. The system is run in two-phase mode (no gas) and measurements are taken. For the flow without a gas phase S.sub.o+S.sub.w=1. With the use of previous relations this can be represented as

(50) $\begin{array}{cc}\frac{A_w\left(T,P\right)}{A_{w,100\%}\left(T,P\right)}+\frac{A_o\left(T,P\right)}{A_{o,100\%}\left(T,P\right)}=1& \left(15\right)\end{array}$
A.sub.w,100%(T,P) is a relatively simple function to measure in the laboratory. Then the following equation will be applicable:

(51) $\begin{array}{cc}{A}_{o,100\%}\left(T,P\right)=A_{w,100\%}\left(T,P\right)\frac{A_o\left(T,P\right)}{A_{w,100\%}\left(T,P\right)-A_w\left(T,P\right)}& \left(16\right)\end{array}$
The above relation allows for the extraction of the unknown function A.sub.o,100%(T,P) that can be used in three phase measurements later.
Measurement of Oil and Water Content in a Continuous Flow

(52) The above water-oil measurement procedure can be adapted for continuous oil-water Row if only the oil component is tracked in velocities that allow collection of the oil relaxation signal without counting for the water relaxation signal.

(53) $A_o\left(T,P\right)={\rho }_o\left(T,P\right){\stackrel{\_}{S}}_o\int {\mathrm{AI}}_o\left(T,P,z\right)dz$$A_{0,100\%}\left(T,P\right)={\rho }_o\left(T,P\right)\int {\mathrm{AI}}_o\left(T,P,z\right)dz$$S_o=\frac{A_o\left(T,P\right)}{A_{o,100\%}\left(T,P\right)}$
It should be understood that the measurement of the value:
A.sub.o,100%(T,P)=ρ.sub.o(T,P)∫AI.sub.o(T,P,z)dz
for oil will require a sufficient amount of oil in order to perform calibration procedure. As oil properties are subject to more variation than water properties this will mean more often calibration procedures. The water signal calculated as:S.sub.w=1−S.sub.o
Field Trials/Examples

(54) The graph in FIG. 6 presents field trial measurements for the NMR apparatus and method. As seen, the data fail closely on a straight line according to the relation as shown in Equation 16.

(55) Table 2 shows a comparison of the NMR field data compared to Dean-Stark (lab) measurements for the same samples. The sample was split into two samples for the Dean-Stark measurements.

(56) TABLE-US-00002 TABLE 2 A Comparison of NMR and Dean-Stark (DS) Water Cut (WC) Measurements For Four Samples. WC from DS WC from NMR Sample # (%) (%) 1a 82.85 82.4 1b 82.77 2a 78.95 80.2 2b 80.29

(57) Accordingly, the results show good correlation between the field measured and laboratory analysis samples.

(58) Although the present invention has been described and illustrated with respect to preferred embodiments and preferred uses thereof, it is not to be so limited since modifications and changes can be made therein which are within the fun, intended scope of the invention as understood by those skilled in the art.