JP2007285728A - Stratum mise-a-la-masse method and its auxiliary equipment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To heighten accuracy in measuring the resistivities of strata by improving a stratum mise-a-la-masse method for loading a casing pipe of a bore hole as an electrode with a voltage and examining underground resistivity distributions of an object section in which electric potentials in the longitudinal direction of the casing pipe have been conventionally analyzed to be constant. <P>SOLUTION: In the mise-a-la-masse method for measuring resistivity distributions of strata, one or a plurality of sondes are suspended in one casing pipe sunk underground to bring their electrode pads into contact with the inner surface of the casing pipe and measure electric potential distributions along the longitudinal direction of the casing pipe when a hole opening of the casing pipe or the inside of a well is energized. A density distribution of a current flowing out from the casing pipe is evaluated on the basis of measured values and used to analyze resistivity of strata. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention is an electric exploration method for exploring underground structures and the like, in which a voltage is applied using a linear metal such as a casing pipe or a rod as a line current source, and the resistivity of the formation in other locations is investigated and measured. The present invention relates to an electrogalvanic potential method and an apparatus used therefor. More specifically, the present invention relates to a technique for improving the accuracy of the specific resistance analysis of the formation by accurately obtaining the distribution of the amount of current flowing out of the casing pipe along the longitudinal direction of the casing pipe in the implementation of the streaming potential search method. The strata targeted by the galvanostatic method are mainly deep geothermal exploration and petroleum exploration, but can also be applied to general strata and ground.

  The geoelectric potential method of the formation is an exploration method that has been developed as one of the fluid behavior monitoring methods in high-temperature rock power generation (geothermal power generation). In hot rock power generation, boreholes are drilled in the underground hot rocks, cracks are generated by hydraulic fracturing in the boreholes, water is injected into these cracks, and hot water storage tanks are installed underground by the heat from the hot rocks. Electricity is generated by rotating the turbine with water vapor formed and taken out from the steam. It is indispensable to monitor the movement of underground water in order to put injected water into the reservoir and take out hot water without resistance, and the geoelectric potential method of the formation has been used as a tool for that purpose.

Recently, in the oil field, in order to improve the oil recovery rate, water, water vapor, CO ₂ or the like is injected into the water from the borehole, and the moved oil is recovered in another production well. There is an oil production technology called Enhanced Oil Recovery. In this technology as well, the galvanostatic method is expected as a monitoring tool for oil storage tanks.

  In reservoir behavior monitoring in the geothermal and petroleum fields, it is often applied as a hybrid exploration method combining the natural potential method and the electrodynamic potential method, focusing on the streaming potential caused by fluid movement. Together they are called the fluid streaming potential exploration method.

  The galvanostatic method is classified as a kind of resistivity electrical exploration of the formation. Electric resistivity exploration is an exploration in which an alternating DC current is passed between two electrodes placed on the ground, and the electrical resistance between the other two electrodes is measured to determine the specific resistance between the electrodes. Is the method. Here, the specific resistance is a physical property value indicating difficulty in the flow of electricity. In electrical resistivity exploration, current-carrying electrodes are often provided on the ground surface, and one-dimensional and two-dimensional exploration have been put to practical use. Further, in order to extend this and increase the accuracy, research for practical use of three-dimensional exploration has been advanced, and an efficient construction method and analysis method have been devised (for example, see Patent Document 1).

  Furthermore, in order to improve the exploration accuracy in the deep part, in addition to the electrode installed on the ground surface, the same exploration is also carried out using the electrode buried in the deep underground and the electrode buried in the borehole, When measured in such a measurement arrangement, it is particularly called specific resistance tomography.

  The galvanic potential method is based on the same exploration principle as those of resistivity electrical exploration and resistivity tomography, but instead of the commonly used electrodes, a linear conductor structure such as a casing pipe of a borehole or a boring rod is used. There is a big feature in using as a current electrode.

As an application of the fluid streaming potential method. Application of the forced oil recovery method to monitoring technology is known (for example, see Non-Patent Document 2).
Application for Japanese Patent Application No. 2004-153539 Geophysical exploration: vol. 51, no. 6 (1998) p659-675 Ushijima et al .: “Oil Forced Recovery Method Monitoring by Fluid Flow Potential Method”

  The analysis of the galvanopotential method is performed according to the same flow as the electrical resistivity exploration of general geological formations. That is, first, a model indicating the specific resistivity of the area to be explored is created. This model is generally called the initial model. The initial model may be, for example, a horizontal structure model that assumes that a horizontal formation has been formed using logging data measured by boring, or a homogeneous resistivity model created from apparent resistivity data. Many.

  Using the initial model of the specific resistance thus created, the measured potential is calculated by numerical simulation.

  If the initial model is a good representation of the actual underground resistivity distribution, the calculated potential and the actually measured potential agree. However, since the two usually do not match, the initial model is corrected so that the difference is reduced. The simulation is further performed using the corrected model, and the same correction is repeated. In this way, correction is repeated until the discrepancy between the measurement data and the simulation data becomes sufficiently small, and a final specific resistance model is obtained.

  In the analysis of the galvanic potential method, the problem is how to accurately calculate the potential distribution when the casing pipe is used as an electrode. Generally, an approximate potential calculation is performed on the assumption that the current supplied to the casing pipe flows into the ground with a uniform distribution along the longitudinal direction of the casing pipe.

  However, when applied in a field where the specific resistance value of the underground is small as in the oil field, the electrical resistance of the casing pipe itself cannot be ignored, and the current does not reach the depth sufficiently, so the current density along with the depth is Attenuates. For this reason, it is expected that the approximate calculation assuming that the current density is constant will cause a decrease in analysis accuracy, and it has been confirmed by a simulation experiment. Such a decrease in analysis accuracy can be prevented by separately measuring and evaluating the electrical resistance and current density distribution of the casing pipe.

  The present invention focuses on the above-described problems, and an object thereof is to provide an improved technique for obtaining a current density distribution flowing out along the longitudinal direction of a casing pipe and performing a simulation based thereon.

  The present invention has been made in order to solve the above-described problems, and is a geoelectric potential method for the formation, characterized by taking the following technical means. That is, the present invention relates to an electrodynamic potential method for measuring the resistivity distribution of an underground formation, in which a voltage is applied to one vertical casing pipe set in the ground and a sonde is suspended in the casing pipe. Then, a potential distribution along the longitudinal direction of the casing pipe is measured, a potential distribution along the longitudinal direction of the casing pipe is measured, and a specific resistance analysis of the formation is modified based on the measured potential distribution.

  In the above-described stratoelectric potential method, it is preferable to use a plurality of connected sondes as the sondes because measurement is easy.

  As a means for obtaining the outflow current density from the potential distribution, the relational expression between the potential and the current may be obtained by direct integration, or may be obtained by a resistance network.

  The apparatus used for carrying out the method of the present invention comprises one or more sondes that move up and down in the casing pipe and come into contact with the inner surface of the casing pipe, an elevating device that raises and lowers the sonde, and a data recording device. It is an auxiliary device for the galvanic potential method of the formation.

  According to the present invention, in the electrokinetic potential method for measuring the resistivity distribution of the underground formation, a voltage is applied to one vertical casing pipe set in the ground, and a sonde is suspended in the casing pipe. To measure the potential distribution along the longitudinal direction of the casing pipe, and to modify the resistivity analysis of the formation based on the measured value, the accuracy of the galvanostatic potential method can be improved, There is an excellent effect that the reliability of this method is remarkably increased.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is an explanatory diagram conceptually showing a method for exploring a galvanoelectric potential method, and shows a model view in which a geological formation as a target for carrying out the present invention is seen through, and an arrangement of an analysis apparatus according to the present invention. is there. FIG. 2 is a flowchart showing an outline of the method of the present invention.

  As shown in FIG. 1, a casing pipe 10 is sunk from the ground surface in a specific resistance analysis target area 100 of the formation, and a transmitter 11 is disposed on the head thereof. The transmitter 11 applies a voltage to the casing pipe 10 and analyzes the potential distribution in the vertical direction of the casing pipe 10 as described later.

  At each of a large number of measurement points 102 arranged in a desired arrangement on the ground surface 101, a potential difference between two measurement points or a potential difference between the far electrode P2 and each measurement point is measured. The measured potential difference between the two points is recorded in the recorder 20. This record is sent to a computer 21 which analyzes this data and calculates the resistivity of the formation. A simulation method is used for this calculation.

  FIG. 2 is a flowchart showing an outline of the analysis of the electrokinetic potential method exploration method. An analysis model is created 31 in advance. As the analysis model, a finite element mesh or a difference grid can be used.

  The measurement data 32 recorded in the recorder 20 is sent to the computer 21 through a data quality check 33 step. And it inputs into the computer 21 according to the said analysis model 31. FIG.

  Next, an initial specific resistance model is created and set 34. As the initial specific resistance model, a uniform specific resistance model, a horizontal multilayer structure model, or the like can be used.

  Next, theoretical potential calculation 35 is performed. Here, a finite element method, a difference method, or the like is used in accordance with the analysis model 31 set in advance.

  A residual (difference between the measured resistance value and the calculated value) 36 is calculated, and a convergence determination 37 is performed. If the residual does not converge, the specific resistance model is corrected 38, and the process returns to the theoretical potential calculation 35 again to perform recalculation. For the modification 38 of the resistivity model, a restricted least square method, a CG method (conjugate gradient method), or the like may be introduced.

  If convergence is recognized in the convergence determination 37, this is adopted and the output 39 of the specific resistance distribution is performed.

  As shown in FIG. 1, in the analysis of the galvanic potential method using the casing pipe 10 as a current source, it is generally assumed that the density of the current 12 flowing out of the casing pipe 10 is constant along the longitudinal direction of the casing pipe 10. Approximate calculation was performed.

  However, the actual density distribution of the current 12 on the electrode flowing out of the casing pipe 10 varies depending on the grounding resistance between the casing pipe and the ground and the electrical resistance of the casing pipe 10 itself. Therefore, the assumption of constant current density is not necessarily satisfied in all cases.

  This is an important issue to evaluate the influence of approximation in the analysis when considering the accuracy of the galvanostatic potential method. For this purpose, it flows out along the longitudinal direction of the casing pipe 10 by some means. It is necessary to measure the current density distribution.

  In addition, if analysis of the galvanostatic potential method can be performed using the measured current density distribution, the accuracy of the analysis can be expected to be improved.

  The grounding resistance between the casing pipe 10 and the natural ground, which is the first factor affecting the current density flowing out of the casing pipe 10, is the clearance between the casing pipe 10 and the natural ground, the corrosion around the casing pipe 10, This is related to various factors such as the influence of cement filled around the casing pipe 10, the diameter of the casing pipe 10, or the specific resistance of the surrounding ground.

  Further, it has been clarified by simulation that the electrical resistance of the casing pipe 10 itself, which is the second factor, has a particularly large influence in an extremely low region where the specific resistance of the natural ground is 1 Ωm or less.

  In the FEM analysis program, the grounding resistance between the casing pipe 10 and the natural ground and the electrical resistance of the casing pipe 10 itself are approximated by a line element and a joint element, so that a potential simulation considering these effects to some extent can be performed. it can. However, for this purpose, it is necessary to obtain in advance the ground resistance between the casing pipe 10 and the natural ground and the electrical resistance of the casing pipe 10 itself with high accuracy.

  FIG. 3 shows a conceptual diagram of measurement of current flowing out from the casing pipe 10.

  When the electrical resistance per unit length of the casing pipe 10 is R and the potential along the longitudinal direction of the casing pipe 10 during energization is φ, the current j flowing along the longitudinal direction (z direction) of the casing pipe 10 is It can be written as follows:

  From the conservation law of the amount of current, the current j in the casing pipe 10 decreases with depth by an amount commensurate with the flowing current i. From this, the following equation holds.

Equation (2) indicates that if the potentials φ ₁ , φ ₂ , and φ ₃ are measured at least at three points in the casing pipe 10 during the measurement of the galvanic potential method, the amount of current flowing out of the casing pipe 10 between the three points is calculated. It shows that it can be estimated.

  When creating a measurement system for the purpose of electrologging, it is necessary to set the electrode interval (the above-mentioned three points interval) to at least several tens of cm or less. However, if the potential is measured in a small section in the casing pipe 10 having a small electrical resistance, the measured potential difference becomes a very small value and is easily affected by noise.

  On the other hand, there is no need to measure current density with the same resolution as the above-mentioned electric logging in the galvanic potential method, so it is possible to minimize the influence of noise instead of sacrificing resolution by increasing the electrode spacing. It is good to make it.

Hereinafter, a method of measuring the outflow current density by measuring a potential or a potential difference at a relatively large electrode interval and performing inverse analysis on the data will be described. As this method,
(A) A method for obtaining the relationship between the potential distribution and the outflow current density by directly integrating an electrical relational expression (direct integration method)
(B) Method of obtaining by resistance network (c) Method of analyzing potential potential measured in casing pipe together with data measured on ground surface in analysis of streaming potential search method using FEM.
There are three.

  First, the direct integration method will be described.

  The relationship between the current density on the casing pipe and the potential q can be written as follows.

Where R: casing pipe electrical resistivity (electrical resistance per unit length)
q: Potential on casing pipe FIG. 4 is a graph showing the relationship between the depth of the casing pipe and the outflow current density. Assuming that the electrical resistance R of the casing pipe is constant, the current density is also divided into a plurality of small sections D as shown in FIG. 4, and the current density q ₁ , q is divided into each section. Given ₂ , q ₃ ,..., Q _m , equation (3) can be integrated as follows.

  Here, the second term RI on the right side of Equation (4) can be evaluated by a value obtained by multiplying the amount of current flowing in the casing pipe in the vertical direction at the well opening by the resistivity of the casing pipe. Eventually, the integrated function Q (z) + RI becomes a function of the form shown in equation (4).

When both sides of the equation (4) are further integrated, the potential φ on the casing pipe, which is a measured value, is obtained. Incidentally, the potential φ (z) at the length (depth) z along the longitudinal direction of the casing pipe is the area of the region shown by hatching in FIG. 5 (from the depth 0 to Z surrounded by the dφ / dz curve). It is a value obtained by adding the potential φ ₀ at the well opening to the area).

Since the expression (4) can be expressed as a linear function of q _j (j = 1 to m), the potential evaluation expression is finally derived in the form of the following expression.

  When the second and third terms on the right side of Equation (5) are moved to the left side, the following linear equation is obtained.

Here, it is f (z) = φ (z ) -RIz one phi _0.

  Equation (6) assumes a measurement corresponding to the bipolar method in electrical exploration.

  FIGS. 6 to 8 show schematic diagrams of the potential measurement of the casing pipe using one, two or three potential electrodes. In FIG. 6, one measuring electrode sonde 40 is inserted into the casing pipe 10, a voltage V is applied to the one sonde 40, and the top of the casing pipe 10 is expressed by the equation (6) while descending the casing pipe 10. The current A is measured at (Pile well opening 43), and the outflow current density distribution in the longitudinal direction of the casing pipe 10 is obtained. In the figure, 41 is a voltmeter and 42 is an ammeter.

  In FIG. 7, two measurement electrode sondes 40 a and 40 b are suspended in the casing pipe 10 at regular intervals, a voltage is applied, and the outflow current from the casing pipe 10 is measured at the well opening 43. The voltage difference between the two measurement electrode sondes 40a and 40b is ΔV. At this time, the calculation is performed using the above equation (7).

FIG. 8 shows three measurement electrode sondes 40 a, 40 b, and 40 c suspended in the casing pipe 10 at intervals, and brought into contact with the inner surface of the casing pipe 10. The voltage difference between the sondes 40a, 40b, and 40c is ΔV ₁ and ΔV ₂ . The calculation formula at this time is formula (8).

  In the case of FIGS. 6, 7, and 8, the outflow current density can be estimated using Equation (6), Equation (7), or Equation (8).

              Here, Δd is an electrode interval.

Equation (6) to (8) are both first-order equation of _q j (j = 1 to m), the known value left side obtained from the measured value _{φ (z), q j (} j = 1~ The coefficient of m) is an amount calculated by the aforementioned integration. Therefore, if data is acquired at a large number of depths, equations (6) to (8) are all multi-dimensional simultaneous equations with q _j (j = 1 to m) as unknowns. The outflow current density q _j (j = 1 to m) of the section can be calculated.

  Since the direct integration method described above is formulated on the assumption that the hole opening is energized, it is necessary to measure the potential and the potential difference in the entire section from the hole opening to the hole bottom with the energization point as the upper end of the casing pipe. When the depth of the casing pipe is large, the potential data of the deep part cannot be acquired with sufficient accuracy unless the current source is lowered to the deep part.

  Next, a method using a resistance network will be described.

As shown in FIG. 9, the method using the resistance network approximates the casing and the surrounding ground by an equivalent resistance network. In FIG. 9, ρ _{1 to} ρ _m are electrical resistances of the casing pipe itself, and R _{1 to} R _{m + 1} are ground resistances of the casing pipe and the formation. The resistance value of the equivalent circuit is obtained by inverse analysis from the potential in the casing pipe or the potential difference distribution when the casing 13 is energized from the power source 13. Next, the potential with respect to an arbitrary energization point is calculated, and the current flowing out from the casing pipe is obtained using the potential difference between both ends of the ground resistances R _{1 to} R _{m + 1} .

The electrical resistances ρ _{1 to} ρ _m of the casing pipe itself are measured in advance by the Wenner method or similar electrode arrangement. In this case, there are two ways to handle the electrical resistance of the casing pipe itself in the analysis of the resistance network. One is a method in which the electrical resistance value obtained in advance is treated as a known parameter and fixed in the inverse analysis. The other is a method of simultaneously analyzing the electrical resistances ρ _{1 to} ρ _{m of} the casing pipe itself as unknown parameters by adding the data of the Wenner method arrangement to the inverse analysis data.

  The method using the resistance network shown in FIG. 9 can analyze not only the data energized in the hole (the upper end of the casing pipe) but also the data energized in the hole (in the casing pipe). It is possible to easily cope with measurement by internal energization.

Next, the simulation was examined according to the flow shown in FIG. A calculation model shown in FIG. 11 was created as a simulation model. The length L of the casing pipe 10 was two types of 500 m and 1000 m. The case where the background specific resistance of the casing pipe 10 was 1, 10, 100, 1000 Ωm was examined. The casing pipe 10 has an inner diameter of 20 cm, a wall thickness of 6 mm, and a conductivity of 10 ⁴ S / m (10 ⁴ Ωm). In addition, a model in which a casing pipe having a length of 50 m was in a horizontal layer structure was also examined.

  The flow shown in FIG. 10 will be described. First, potential data 52 and current density distribution 53 are obtained by potential simulation 51 using FEM. Data generation 54 for current density calculation is performed from the potential data 52. In this current density calculation, setting of the electrode interval, creation of 1 to 3 pole data, and addition of noise assuming on-site conditions are performed. Next, inverse analysis 55 is performed. By the inverse analysis 55, an estimated value 56 of the current density distribution is obtained. A comparative study 57 is performed between the estimated value 56 of the current density and the current density distribution data 53 obtained in advance, and the simulation is evaluated.

  In the calculation model shown in FIG. 11, a casing pipe 10 having a length L is sunk in the vertical direction (depth z direction) in the target area 100, and a surface orthogonal to the casing pipe 10 is a ground surface 101. A number of measurement points are set by taking p.

  A simulation was performed for the purpose of showing that the above current density distribution measurement method is effective.

  The potential simulation was performed by 2.5D FEM, and a 2D FEM model as shown in FIG. 12 was created. The 2.5-dimensional FEM is a method for calculating a three-dimensional field in a two-dimensional resistivity structure (the structure does not change in a direction orthogonal to the cross section). Compared to the three-dimensional FEM, the mesh can be finely divided even when the memory capacity is small, and the discretization error due to the roughness of the division can be reduced. Further, a structure such as a horizontal multilayer structure can be easily provided.

  In FIG. 12, the horizontal axis indicates the horizontal separation distance (−10,000 m to 10,000 m) from the casing pipe 10, and the vertical axis indicates the depth (0 m to 12,000 m). FIG. 13 shows a portion of the target region up to a depth of 600 m and a horizontal separation distance of ± 500 m, and shows an FEM model when the casing pipe length (depth) is 500 m. The casing pipe 10 was modeled using a line element having a length of 5 m of FEM, and this was inserted into a section having a depth of 0 m to 500 m. The 2.5-dimensional element adjacent to the line element is a square element having a side of 5 m.

When the casing pipe length was 1000 m, a line element was given to a section having a depth of 0 m to 1000 m in the same model as FIGS. 12 and 13.

In this case, the resistance of the line element (resistance per unit length) is calculated as 0.000273469 Ω / m from the diameter, thickness and specific resistance of the casing pipe 10 shown in FIG. It was set as a physical property value.

  Next, potential simulation results for showing that this idea is effective will be described. First, the case of a casing pipe length of 500 m was examined.

  A simulation corresponding to the potential simulation 51 of the flow of FIG. 10 was performed by applying a current as a point source to the nodes of X = 0 and Z = 0, with the energization current amount being 1A. FIG. 14 shows the distribution of the outflow current density 53 (see FIG. 10) along the longitudinal direction of the casing pipe obtained by the potential simulation when the depth of the casing pipe is 500 m and the background resistance is 1, 10, 100, and 1000Ω, respectively. This is shown in (a) to FIG. 14 (d), respectively. 14 (a) to 14 (d) correspond to the potential distribution diagrams shown in FIGS. 15 (a) to 15 (d), respectively.

  As shown in FIGS. 14C and 14D, the outflow current density has a substantially constant value regardless of the depth when the background specific resistance is large. On the other hand, as shown in FIGS. 14 (a) and 14 (b), when the background specific resistance is reduced, the outflow current density is abruptly attenuated with the depth and the current does not reach the deep part. As shown in FIGS. 15A to 15D, it can be seen that the potential distribution approaches the potential distribution for the point current source as the background specific resistance decreases.

  Next, similarly, the current density distribution and the potential distribution when the casing pipe length is 1000 m are shown in FIGS. 16 (a) to 16 (d) and FIGS. 17 (a) to 17 (d). The trend of the current density distribution is the same as that in the case where the casing pipe length shown in FIGS. 15 and 16 is 500 m. However, as the casing pipe length becomes longer, it is strongly affected by the attenuation, and the outflow current density near the well bottom Is even smaller than 500 m.

  Next, potential simulation results when a horizontal multilayer structure model is used as a study model will be described.

  A horizontal multilayer structure model for potential simulation is shown in FIG. FIG. 18 is an example in which it is assumed that horizontal layers with a constant specific resistance distribution are layered. For example, the layers from the surface to −500 m are 100, 10, 1, 10, 1, 10 Ωm in the depth direction. This is assumed to be a horizontal layer having a certain specific resistance. As these specific resistance values, for example, values obtained in the sinking process of the casing pipe 10 are used.

  FIG. 19 shows the potential distribution obtained by the simulation and the outflow current density distribution on the casing pipe. When the resistance of the casing pipe is infinitely small, it is known that the magnitude of the outflow current of each layer in the horizontal multilayer structure is inversely proportional to the formation resistivity. 18 and 19, since the specific resistance of the casing pipe is not infinitely small, the outflow current density distribution is almost inversely proportional to the formation specific resistance. In addition, the specific resistance of the casing pipe is also added. There was also a tendency for the amount of outflow current to decay with depth due to the effect of.

  In addition, for the purpose of examining the influence of noise mixed in the acquired data assuming on-site measurement, there is also a data set in which normal noises with standard deviations of 0.01 mV and 0.1 mV are given to the data set created from the simulation results. In addition, data creation 54 was performed (see FIG. 10).

  FIG. 20 shows a case where the background specific resistance is uniformly 1 Ωm and the casing pipe length is 500 m. FIG. 21 shows a case where the background specific resistance is uniformly 1 Ωm and the casing pipe length is 1000 m. 7 shows the seven potential data created by the potential simulation when there is no noise. 20 and 21, the vertical axis represents the depth (m), and the horizontal axis represents the potential difference (mV) (logarithmic scale). 1 pole measurement with 1 sonde, 2 pole measurement with 2 sondes (each with electrode spacing 5, 25, 50 m) and 3 pole measurement with 3 sondes (electrode spacing 5, 25, respectively) 7 curves showing 50 m) are shown.

  From FIG. 20 and FIG. 21, it is understood that the potential decreases as the depth increases, and the potential is larger in the unipolar measurement, and the potential is smaller in the order of the second and third pole measurements. In the 2- and 3-pole measurement, the smaller the electrode interval, the smaller the potential. Further, the casing pipe length of 1000 m has a lower measurement potential level, and the accuracy near the bottom of the well is markedly lower than that of 500 m.

  FIG. 22 shows data when the casing pipe length is 500 m in the case of the horizontal multilayer structure model. In the case of the horizontal multilayer structure model, the measurement data of the three-pole measurement at the minimum electrode interval of 5 m is similar to the actual outflow current density distribution, and the relative distribution of outflow current density can be grasped only by looking at this data. be able to. However, as the electrode spacing is set larger, the curve becomes smoother as if a low-pass filter was applied, and high-frequency components are lacking. Furthermore, since the bipolar measurement data has a form in which the outflow current density distribution is integrated once, and the single pole measurement data has a form in which the outflow current density distribution is integrated twice, the high frequency component disappears with the number of integrations. I can see it going. From these facts, in the ideal case where there is no noise, it was confirmed that the data of the tripolar measurement carried out with a small electrode interval contained the highest amount of high frequency components and is suitable for high resolution measurement. It was.

  On the other hand, when the measurement data contains noise, the measurement value below the noise level has no meaning, and the level of the measurement potential level is a large factor that governs the accuracy of the analysis result. From FIG. 20 to FIG. 22 from this point of view, the magnitude of the potential and the measured potential difference is smaller in the 2-pole measurement than in the 1-pole measurement, and smaller in the 3-pole measurement than in the 2-pole measurement. As the electrode spacing decreases, the potential difference tends to decrease, and the above measurement arrangement giving priority to high resolution can be said to be the measurement arrangement most susceptible to noise.

  In actual exploration, it is necessary to find an optimal arrangement that satisfies both accuracy and resolution requirements. Therefore, we conducted reverse analysis for several measurement cases with different noise levels, and examined the optimum measurement specifications for designing an actual measuring instrument.

  From the data set created as described above, the outflow density distribution was estimated by solving Equations (6) to (8) (55 in FIG. 10).

The above formulas (6) to (8) are unstable due to the problem of improperness or the influence of noise, and a significant solution cannot be obtained as they are. In order to avoid these problems, the following treatments (a) to (d) were performed.
(A) A flattening constraint was given as an a priori a priori condition for solution stabilization.
(B) Since it is known in advance that the current density may change by two digits or more depending on the location and does not take a negative value, the unknown g _j (j = 1 to m) is log-transformed. Inverse analysis was performed.
(C) Although the above equations (6) to (8) are linear equations, they become non-linear with the logarithmic transformation of unknowns. Therefore, a solution was obtained by a method in which a uniform current density was started as an initial model and was sequentially corrected by iteration.
(D) Since the total amount of current is known, this was added as a constraint.

  The estimated value 56 (see FIG. 10) of the outflow current density distribution obtained by the inverse analysis as described above is superimposed on the true outflow current density distribution directly obtained by the potential simulation and shown in FIGS. .

  FIG. 23 shows the analysis result of the current density distribution when the casing pipe length (depth) is 500 m, the background specific resistance is 1 Ωm, and there is no noise. In the figure, black circles indicate the true current density distribution, and solid lines indicate the analysis results. Both agree well. In addition, (a)-(g) in the figure has shown 1 pole measurement, 2 pole measurement ((DELTA) 5m, (DELTA) 25m, (DELTA) 50m), and 3 music measurement ((DELTA) 5m, (DELTA) 25m, (DELTA) 50m) similarly to FIG.

  24 and 25 are the same as FIG. 23, except that the noise is 0.01 mV and 0.1 mV, respectively. 24 and 25, there is a slight deviation between the true current density distribution (black circle in the figure) and the analysis result (solid line in the figure).

  26 to 28 show the analysis results of the current density distribution when the casing pipe length (depth) is 1000 m, the background specific resistance is 1 Ωm, there is no noise, and 0.01 mV and 0.1 mV, respectively. Other displays are the same as those shown in FIGS.

  FIGS. 29 to 31 show analysis results of the current density distribution when the casing pipe length (depth) is 500 m and the horizontal multilayer structure model is set to have no noise, noise of 0.01 mV, and noise of 0.1 mV, respectively. Other displays are the same as those shown in FIGS. In FIGS. 24 to 25, FIGS. 27 to 28, and FIGS. 30 to 31, portions that become regions below the noise level are indicated by broken lines with arrows.

  In addition to the above results, the same analysis was performed for cases where the background specific resistance was 1Ωm, 10Ωm, 100Ωm, and 1000Ωm. From these analyses, differences in analysis error due to casing pipe length, measurement method, and noise conditions are shown in FIGS.

  FIGS. 32 to 35 are uniform resistivity models, for cases where the background resistivity is 1 Ωm, 10 Ωm, 100 Ωm, and 1000 Ωm, and for analysis errors in the case of a casing pipe length of 500 m, one pole measurement and two pole measurement, respectively. For each of (electrode spacing Δ5m, Δ25m, Δ50m) and tripolar measurement (electrode spacing Δ5m, Δ25m, Δ50m), no noise (noise free), noise is 0.01 mV, noise is 0.1 mV The case is shown.

  FIGS. 36 to 39 are uniform resistivity models, and analysis errors in the case where the background resistivity is 1Ωm, 10Ωm, 100Ωm, and 1000Ωm and the casing pipe length is 1000 m are the same as those in FIGS. The measurement method and noise conditions are shown.

  FIG. 40 shows the analysis error when the casing pipe length is 500 m in the horizontal multilayer structure model, in the case of the measurement method and noise conditions similar to those of FIGS.

  The results of the inverse analysis shown in FIGS. 32 to 40 are summarized as follows (A) and (B).

(A) Uniform resistivity model In the absence of noise, the outflow current density distribution is accurately reproduced in any case. As the noise increases, the reproducibility becomes worse and the accuracy of the inverse analysis gradually decreases. True outflow current density distribution is generally reproduced in the section where the measurement data of potential or potential difference is larger than the noise level, but the reproducibility decreases in the deep section where the measurement data is below the noise level. Is recognized.

  In the case of the casing pipe length (depth) of 1000 m where the measured potential level is small, the accuracy drop near the bottom of the well is remarkable compared to the case of the casing pipe length (depth) of 500 m.

  The influence of noise increases as the background specific resistance decreases. This is also considered to be because the measurement potential becomes small.

  In measuring the potential along the longitudinal direction of the casing pipe, the influence of noise deterioration due to noise is strongly recognized as the number of poles increases in one-pole measurement, two-pole measurement, and three-pole measurement. This is presumably because the measured value of the potential difference is smaller as the number of poles is larger. Further, in the two-pole measurement and the three-pole measurement, a decrease in accuracy due to the influence of noise is strongly recognized as the electrode interval decreases. This is also because the potential difference measured by reducing the electrode spacing is reduced.

(B) In the case of the horizontal multilayer structure model The horizontal multilayer structure model has poor reproducibility of the outflow current density distribution along the longitudinal direction of the casing pipe as compared with the uniform specific resistance model. This is because the portion where the outflow current density changes abruptly cannot be accurately reproduced by the inverse analysis, and the actual outflow current density distribution is expressed as a gentle change distribution. This tendency becomes more prominent as the noise level becomes higher. This is presumably because high-frequency components cannot be accurately obtained due to lack of data or the influence of noise.

  When there is no noise, in one-pole measurement, two-pole measurement, and three-pole measurement, an analysis result with higher resolution is obtained and the reproducibility is superior when the number of poles is large. Moreover, if the electrode distribution is the same, an analysis result with higher resolution can be obtained as the electrode interval is reduced, and the reproducibility is excellent.

  -On the other hand, when there is noise, a decrease in accuracy due to the influence of noise is strongly recognized as the electrode interval is reduced. This is considered to be due to the fact that the measured potential difference becomes smaller and is more susceptible to noise by reducing the electrode spacing.

  In the case of tripolar measurement (electrode spacing 5 m) and a noise of 0.1 mV, the uniform outflow current density distribution given as the initial model was not improved at all and was directly output as the analysis result. This is presumably because the measured potential difference data was below the noise level in all sections, and there was a lack of significant information for determining leakage current leakage.

  Similarly, even in the case of tripolar measurement (electrode spacing 5 m) and a noise level of 0.01 mV, reproducibility was inferior compared to monopolar measurement, bipolar measurement, or tripolar measurement with a large electrode interval. This is presumably because the potential level of the measurement data was below the noise level in the interval of about half.

The results of the above simulation studies are summarized as follows (a) to (e).
(A) It was confirmed that the high frequency component contained in the measurement data decreased in the order of tripolar measurement, bipolar measurement, and monopolar measurement. Therefore, in order to obtain an analysis result with high resolution, it is advantageous that the number of electrodes is large.
(B) In the interval where the level of the measured potential (potential difference) is less than or equal to the noise level, the accuracy of reverse analysis is significantly reduced. Therefore, in order to suppress the influence of noise, it is necessary to select an arrangement that increases the value of the measurement data.
(C) Although the one-pole measurement has a high measurement potential level, it is assumed that the noise level also increases by picking up noise mixed in the far electrode cable. The accuracy is improved by setting the electrode interval as large as possible in the two-pole measurement or the three-pole measurement.
(D) If the casing pipe length is 1000 m and the background specific resistance is 1 Ωm, and the noise level is 0.01 mV (10 nV) or less, the measurement electrode arrangement and electrode spacing By making an appropriate selection, it is possible to determine an approximate outflow current density distribution. In this case, the case where the electrode interval was 50 m and the measurement arrangement was bipolar measurement was effective.
(E) It was shown that the accuracy is improved by using data energized in the casing pipe.

From the above, the measuring device needs to satisfy the following requirements (i) to (iii).
(I) When the measured potential level is 0.01 mV or higher, the noise level must be suppressed to 0.01 mV or lower.
(Ii) Considering that the measurement potential level depends on geological conditions, it is desirable to design the measurement potential level so that it can be easily changed in the field. Both 2-pole measurement and 3-pole measurement are measured in parallel. It is preferable that an extension cable for adjusting the electrode interval can be inserted between the electrode probes so that the electrode interval can be easily changed.
(Iii) It is necessary to determine detailed specifications based on the examination of the measured potential level and noise level based on the actual measurement data.

  Next, when measuring the outflow current of the casing pipe by energization in the well, when the depth exceeds 1000 m or the background specific resistance is less than 1 Ωm, it is difficult to obtain the current density distribution in the deep part by energizing the well opening. Therefore, in order to compensate for this, it is necessary to use a means for flowing current in the well.

When the casing pipe length (depth) exceeds 1000 m, or when the background specific resistance is small, the outflow current is measured by the ammeter 42 from the uppermost end (well opening) 43 of the casing pipe 10 as shown in FIG. Instead of the system, as shown in FIG. 41, the detection end 44 is inserted into the casing pipe 10 and measurement is performed. FIG. 41 is a schematic diagram showing an auxiliary device according to the embodiment, in which voltages are applied to three sondes 40a, 40b, and 40c inserted into the casing pipe 10 to generate potential differences ΔV ₁ and ΔV ₂ to detect the detection end. 44 shows an apparatus for measuring the outflow current.

  In order to obtain the outflow current density distribution at the time of energization of the well opening from the data obtained by energization in the well, a resistance network model comprising the electrical resistance of the casing pipe 10 and the ground resistance between the casing pipe 10 and the surrounding ground is used. After creating and determining model parameters (resistance values), it is necessary to calculate the current density distribution. In the case of a three-dimensional resistivity reverse analysis using a line element, a method of analyzing these wellhead measurement data together with resistivity data can be considered.

  Here, the results of a simulation study on an analysis method using a resistor network are shown. The examination follows the procedure shown in the flow of FIG. 10, and after obtaining the resistance value of the network using the method using the resistance network as the method of the inverse analysis 55 of FIG. 10, the estimated value 56 of the current density distribution is obtained. It was.

  In creating simulation data, the following measurements were assumed. That is, as shown in FIG. 42, the measurement sonde is lowered into the casing pipe 10, and the potential difference is measured between the two electrodes P1 and P2 while energizing the current electrode C1. When this was completed, the sound was moved 5 m and the same measurement was repeated, and similar data was acquired at each depth in the casing pipe. The distance between the current electrode C1 and the electrode P1 was 5 m, and three types of simulation data of 5 m, 25 m, and 50 m were created for the distance between P1 and P2. In addition, the potential simulation used 2.5-dimensional FEM similarly to the above-mentioned simulation.

  FIG. 43 shows a comparison between the current density distribution analyzed using the resistance network (distribution when energized at a point of depth 0) and the current density distribution 53 (see FIG. 10) based on the original FEM simulation. 43 (a) shows the result of analysis using data with an interval between the electrodes P1 and P2 of 5 m, and FIG. 43 (b) shows the result of analysis using data with an interval between the electrodes P1 and P2 of 5 m and 25 m. FIG. 43 (c) shows the result of analysis using all three types of data in which the distance between the electrodes P1 and P2 is 5 m, 25 m, and 50 m. It was confirmed that the accuracy was higher than that of FIGS. 29, 30, and 31 analyzed from the data of energization of the hole opening, and the accuracy improved as the types of data used increased.

The measurement physical object in the present invention is a minute voltage (minimum input conversion voltage resolution 10 μv) and current in the casing pipe well. Thereby, the following two measurements are performed.
(1) Electrical resistance of the casing pipe in the specified casing pipe section (2) Outflow current to the formation in the specified casing pipe section When calculating the electrical resistance of the casing pipe 10, the casing pipe is closed circuit from the sonde electrode in the well In order to pass a current to 10 points, a current conducting electrode and a voltage measuring electrode are attached to a pad of one tool, and a current is passed through a casing pipe between tools in a well. A well current drive unit tool may be inserted into the well separately, but this is not recommended in terms of realistic operation, safety, and cost. In addition, it is necessary to consider the validity from the amount of current injection and other physical relationships.

  The electrode spacing can be arbitrarily determined by changing the length of the relay connection cable. In this case, the measurement system can measure a minute voltage stably.

  The borehole system is a combination of basic tool units, and the uppermost electrode and the lowermost electrode of the upper and lower tool electrode pads serve as current electrodes. In this case, it is preferable that this electrode can be used not only as a current electrode but also in other cases so that the system can have general versatility so that the current path can be changed.

It is a conceptual explanatory view showing an embodiment of an electric exploration method according to the present invention. 4 is a flowchart of the process of the present invention. It is a measurement conceptual diagram of the electric current which flows out from a casing. It is a graph which shows the relationship between the depth of a casing, and an outflow current density. It is a graph which shows the electric potential on a casing pipe. It is explanatory drawing which shows arrangement | positioning of a measurement sonde. It is explanatory drawing which shows arrangement | positioning of a measurement sonde. It is an arrangement plan of a measurement sonde. It is explanatory drawing of a resistance network. It is a flowchart of simulation. It is a figure which shows the calculation model of simulation. It is a figure which shows the two-dimensional FEM model of simulation. It is a figure which shows the two-dimensional FEM model of simulation. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a schematic cross section which shows formation electric potential distribution. It is a schematic cross section which shows formation electric potential distribution. It is a schematic cross section which shows formation electric potential distribution. It is a schematic cross section which shows formation electric potential distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a schematic cross section which shows formation electric potential distribution. It is a schematic cross section which shows formation electric potential distribution. It is a schematic cross section which shows formation electric potential distribution. It is a schematic cross section which shows formation electric potential distribution. It is explanatory drawing of a horizontal multilayer structure model. It is a current density distribution diagram. It is a graph which shows the electric potential data by electric potential simulation. It is a graph which shows the electric potential data of a horizontal multilayer structure model. It is a graph which shows the electric potential data of a horizontal multilayer structure model. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows current density distribution. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a graph which shows an analysis result. It is a schematic diagram which shows the auxiliary | assistant apparatus of an Example. It is a schematic diagram which shows the measurement electrode arrangement | positioning assumed by simulation. It is a graph which shows current density distribution.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Casing pipe 11 Transmitter 12 Current 13 Power supply 20 Recorder 21 Computer 31 Analysis model 32 Measurement data 33 Quality check 34 Initial specific resistance model creation and setting 35 Theoretical potential calculation 36 Residual 37 Convergence determination 38 Resistivity model correction 39 Specific resistance Distribution output 40, 40a, 40b, 40c Sonde 41 Voltmeter 42 Ammeter 43 Wellhead 44 Detection end 45 Voltmeter 51 Potential simulation 52 Potential data 53 Current density distribution 54 Data creation 55 Reverse analysis 56 Estimated current density distribution 57 Comparison

Claims (5)

  In the electrokinetic potential method for measuring the resistivity distribution of the underground formation, a voltage is applied to one casing pipe set in the ground, and a sonde is suspended in the casing pipe so as to contact the casing pipe. A stratoelectric potential method for measuring a potential distribution along a longitudinal direction of a casing pipe, obtaining an outflow current density based on the measured potential distribution, and modifying a specific resistance analysis of the stratum.   2. The geoelectric potential method of the formation according to claim 1, wherein a single sonde or a plurality of sondes connected at an arbitrary interval in the longitudinal direction of the casing pipe is used as the sonde.   2. The geoelectric potential method for a formation according to claim 1, wherein the relationship between the potential distribution and the outflow current density is obtained by direct integration of a relational expression.   2. The method of claim 1, wherein the relationship between the potential distribution and the outflow current density is obtained by a resistance network.   An auxiliary device for a geoelectric potential method of a formation, comprising: one or a plurality of sondes that move up and down in the casing pipe and contact the inner surface of the casing pipe; a lifting device that raises and lowers the sonde; and a data recording device.

Images