JP2006250823A - System for evaluating corrosive deterioration of underground tank - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system for evaluating a corrosive deterioration of an underground tank, which eliminates the need for such a troublesome work that earth and sand for burying the periphery of the underground tank are removed when attaching an observation-use sensor to the underground tank in existence, and which can carry out an inspection under the operation of the underground tank without the need for an opened state (i.e., reserved liquid is once discharged in order to make it empty). <P>SOLUTION: The system comprises: a stage of setting an acoustic emission (AE) data sampling device which includes an AE sensor, on the underground tank, and then extracting AE data therefrom; a stage of processing the AE data and extracting an AE characteristic value; and a stage of analyzing the AE characteristic value and determining a corrosion rate. In the system, at least one portion of the AE sensor is in contact with the inner wall of the underground tank in the form of a shell through a waveguide rod. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an underground tank corrosion damage evaluation system, and more particularly to an underground tank corrosion damage evaluation system using an acoustic emission (hereinafter abbreviated as AE) method.

  Many tanks are used in petrochemical plants, various factories, and gas stations, and store liquids of organic or inorganic chemical substances such as oils and solvents. Therefore, damage / leakage (penetration cracks) of the tank is fatal, and it is essential to establish a maintenance technology that predicts this and takes preventive measures.

  In order to predict tank wall breakage / leakage, it is necessary to detect tank wall damage (pitting corrosion, etc.) due to local corrosion prior to tank wall breakage / leakage. There is a passive method and a passive method.

  In the active method, as represented by the ultrasonic flaw detection method, an elastic wave is incident on the tank wall and the reflected wave or transmitted wave is observed to detect damage inside the wall. In the method, as represented by the AE method, an elastic wave in a weak ultrasonic region is generated when damage in the wall occurs or grows (AE, also referred to as acoustic emission). Observe (sense) to detect damage inside the wall.

  The AE method can monitor progressive damage that leads to breakage / leakage continuously and without evacuating the tank while it is in operation, so as a preparatory step prior to the active method, suspected damage It is often used to get a comprehensive picture of where there is.

For example, Patent Document 1 discloses a technique for incorporating an AE signal as a part of failure monitoring means of an integrated operation support apparatus for plant equipment.
JP 2004-094631 A

  In the AE method, the technology for sensing weak waves in the ultrasonic region and the data processing technology for the sensed data have been improved and standardized both in Japan and overseas, and the noise components in the data are recognized and processed to produce signal patterns and actual corrosion. A database that is effective for collating with the state of damage is being developed.

  For example, the “AE User Group” consisting of major oil companies and major chemical companies established such a database for corrosion damage on the bottom surface of ground tanks in Europe. Currently, the AE method is applied to more than 1000 tanks every year. We carry out inspection by.

Also, in Japan, regarding the thickness reduction due to corrosion damage at the bottom of the tank, prior to the AE method, there is a legal accumulation of data on the bottom thickness measurement for the thickness reduction at the bottom by opening the tank. is there.
When this conventional legal plate thickness measurement data is statistically processed to obtain information on the corrosion amount / corrosion rate in consideration of aging deterioration, when the AE method and the plate thickness measurement are performed simultaneously, there is a certain correlation between the two. It is found that there is, and a database on these is being established.

The source of AE detected at the time of corrosion damage is the cracking of the hard corrosion product generated on the solid surface of the damaged part, rather than the deformation and the growth of minute defects (cracks) that occur locally inside the solid. It is said that it is in peeling.
Such a sudden change of the damaged portion (change amount is size μm to mm class) generates an elastic wave of 10 kHz to 1 MHz, that is, an elastic wave in the ultrasonic region.

When microgrowth is superimposed on the same corrosion damage, the damage also grows macroscopically and eventually penetrates the shell, leading to a serious situation of leakage of the stored liquid.
In terms of elastic waves due to rapid changes in defects (damaged parts) in solids, the spatial and temporal scales are different, and AE is equivalent to the occurrence of earthquakes. Can be applied in parallel.

  Since the sensor used in these AE methods observes the tank while it is in operation, it cannot be attached to the inner wall of the tank, so it must be attached exclusively to the outer wall of the tank. However, in the case of underground tanks, the outer wall is buried in the ground, and it is difficult to attach an observation sensor necessary for inspection. There was a problem that the AE method could not be applied to a huge number of underground tanks.

  The object of the present invention, which has been made in order to cope with the above-mentioned problems, is that when attaching an observation sensor to an existing underground tank, it requires a troublesome work of removing earth and sand buried in the outer periphery of the underground tank. In addition, an underground tank that can be inspected while it is in operation without opening the underground tank for the purpose of inspection only (dispensing the stored liquid (discharging it once) and emptying it) It is to provide a corrosion damage evaluation system.

  Another object of the present invention is to provide a highly reliable and economical underground tank corrosion damage evaluation system by applying a database formed for the above-mentioned ground tank.

  An underground tank corrosion damage evaluation system according to the present invention, which has been made to solve the above-mentioned problems, includes an AE data collection device including an acoustic emission (hereinafter referred to as AE) sensor installed in an underground tank. A stage for collecting AE data, a stage for processing the AE data to extract an AE characteristic value, and a stage for analyzing the AE characteristic value to determine a corrosion degree, and at least one of the AE sensors. The portion is in contact with the inner wall of the shell of the underground tank via a waveguide rod.

  Further, as shown in claim 2, in the system according to claim 1, the AE characteristic value extracted in the stage for extracting the AE characteristic value is at least “total number of hits” and “maximum amplitude for each hit”. The stage for judging the corrosion degree by analyzing the AE characteristic value includes (a) a correlation database of the measured shell thickness and the corrosion degree regarding the past underground tank, and (b) the correlation between the total number of hits and the corrosion degree. And (a) a database that obtains a corrosion risk parameter (CRP) from a correlation analysis of the shell thickness reduction amount and its cumulative frequency, and (a2) a correlation analysis of CRP and the total number of hits. A database for obtaining a calibration curve of CRP, and from the “total number of hits” in the step (b), the CRP of the underground tank based on the databases (a1) and (a2) It estimates a limit to predict corrosion rates and life of the underground tank, characterized in that.

  In the underground tank corrosion damage evaluation system according to the present invention, one or more waveguide rods are simply submerged in the liquid stored from the upper part of the operating underground tank so that one end of the waveguide rod is in contact with the inner wall of the underground tank. The AE method can be observed through a sensor connected to the other end of the waveguide rod.

  Therefore, when attaching an observation sensor to an existing underground tank, there is no need for the troublesome work of removing the earth and sand that embeds the outer periphery of the underground tank. Inspection can be performed without stopping the operation. That is, economical inspection becomes possible.

  Further, the corrosion damage evaluation system for underground tanks according to the present invention can be used with an inspection database related to ground tanks established by domestic and foreign AE methods, and is therefore highly reliable and economical.

  In addition to this, maintenance of underground tanks is much more expensive than ground tanks, so it is important to determine the right or wrong timing of maintenance work, but according to the present invention, it depends on "total number of hits" etc. Empirical maintenance timing decisions can be reinforced by direct data: estimation of corrosion damage rate of shells in underground tanks, residual shell thickness, and hence penetration life of damaged parts.

An underground tank corrosion damage evaluation system according to the present invention will be described with reference to the drawings.
1A and 1B are diagrams showing an underground tank corrosion damage evaluation system according to the present invention, in which FIG. 1A is an overall view and FIG. 1B is a schematic diagram showing a corrosion damage portion which is a source of AE.
FIG. 2 is a diagram illustrating various characteristics of an ultrasonic waveform by AE.
FIG. 3 is a flow diagram of the AE method according to the present invention.
In FIG. 4, (A) is a table used for primary determination of the corrosion degree, and (B) is a table showing specific determination results of the corrosion degree in (A).
FIG. 5 is a diagram showing portions according to the second embodiment in the flow chart of the AE method according to the present invention.
FIG. 6 is an actual measurement diagram showing the correlation of parameters related to the degree of corrosion used in the second embodiment.

  Referring to FIG. 1, (A) is an overall view of a corrosion damage evaluation system for an underground tank according to the present invention. The underground tank is surrounded by a shell 10 and buried in the ground 30, and only its upper surface is above the ground. It is exposed so that it can be accessed from. In particular, in the case of an existing underground tank, the outer surface of the shell is underground and buried in the soil, so that it is difficult to access by ordinary means.

Since the underground tank is in operation, the liquid 20 is stored inside, and the portion of the inner surface of the underground tank that is in contact with the liquid 20 (hereinafter referred to as the liquid phase portion, Of these, the portion other than the liquid phase portion is referred to as the gas phase portion) with the tip 42 of the waveguide rod 40.
The waveguide rod 40 is a stainless steel rod having a size corresponding to the depth of the underground tank, for example, a length of 2.5 m and a diameter of 25 mm, and when inserted into the liquid 20 from the sharply sharpened tip 42 side, Strong contact with the bottom shell by its own weight.

The entire waveguide rod 40 serves as a sensor, that is, not only senses the AE wave propagating through the bottom shell at the tip 42 but also senses the AE wave propagating through the liquid 20 on the side surface of the waveguide rod. Therefore, the sensitivity can be increased, and these AE waves are propagated to the upper end of the waveguide rod without being substantially attenuated.
In this way, it is possible to easily access the AE generation source in the liquid phase portion of the shell, which is difficult to access by ordinary means.

  In this embodiment, two waveguide rods 40, 40 are installed so that the tip 42 is in contact with the inner surface of the bottom of the shell through the opening 12 provided in the upper portion of the shell 10, and the other end of the waveguide rod is provided. A sensor, for example, a piezoelectric sensor 44 is installed, and the piezoelectric sensor is connected to the personal computer 90 by a wired signal line 91 or wirelessly.

  As shown in FIG. 1 (B), corrosion at some point in the liquid phase part of the shell proceeds, that is, as the damaged part 50 grows, it occurs on the solid surface of the damaged part at random time intervals. A plurality of abrupt changes 55 of the damaged part, such as cracking and peeling of the hard corrosion product, and each abrupt change becomes an elastic wave in the ultrasonic region (hereinafter simply referred to as “ultrasonic wave”), When propagating through the shell and liquid while being attenuated and reaching the waveguide tip 42 and the side surface of the waveguide rod 40, respectively, they propagate through the waveguide rod without substantial attenuation and reach the piezoelectric sensor. The signal is converted into a voltage signal and transmitted to the personal computer 90 via the signal line 91.

Next, referring to FIG. 2, an ultrasonic waveform by AE (converted into a voltage waveform by a piezoelectric sensor) is illustrated, and various characteristics thereof are illustrated.
Actually, every time a corrosion product breaks or peels off and a sudden change occurs in the damaged part, a plurality of such ultrasonic waveforms are generated and transmitted superimposed on the signal line. A block of ultrasonic waveforms corresponding to a sudden change is called a hit.

  Since various noises are superimposed on the signal line, a certain threshold Ath is set, and a portion where the amplitude of the waveform continuously exceeds Ath is regarded as one hit. For each hit, characteristics such as maximum amplitude value Am, rise time Tr, duration Td, and count number n are defined.

  Next, referring to FIG. 3, the flowchart of the AE method according to the present invention is roughly divided into a first stage S10 for collecting AE data, a second stage S20 for processing AE data, and AE characteristics. It comprises a third stage S30 for analyzing values.

In step S11 of the first stage S10, AE sensors are installed at various locations on the shell of the underground tank.
Since the amount of attenuation of ultrasonic waves depends on the positional relationship between the damaged part and the waveguide rod, as in the case of estimation of the epicenter of an earthquake, a plurality of waveguide rods, for example, a minimum of two waveguide rods are appropriately used. When installed in, it is possible to estimate the approximate position of damage (growing and deforming) on the shell by analyzing the difference in the intensity of the ultrasonic waves received by each waveguide rod.

If it is desired to accurately evaluate damage to the gas phase portion of the shell, a waveguide rod may be installed on the inner surface of the shell and processed together with data from a sensor directly installed on the upper surface of the shell.
The installation quantity and installation location of the waveguide rod and the AE sensor are appropriately selected based on the past, particularly the database of the ground tank, according to the size and shape of the tank and the target evaluation accuracy.

  In step S12, the AE sensor, the signal line, and the personal computer with the built-in AE measurement board connected thereto are operated for a predetermined measurement time, for example, 1 hour, and the AE data is time-sequentially for each AE sensor. To be recorded.

Next, in the second stage S20, the time series data obtained in this way is subjected to a preparation process before analysis.
In step 21 of the second stage S20, noise processing is performed.
In other words, various noises are superimposed on the time series data, and some of them exceed the predetermined threshold Ath, and are removed. The causes of such noise include wind noise, rain noise, rain water flow noise, rain / condensation water drop falling sound, vibration sound of piping connected to the underground tank, and / or other noise peculiar to the underground tank (external to the tank). Various vibration sounds transmitted through the ground).

  Most of these exhibit correlated waveforms peculiar to the time-series data of a plurality of sensors, so that discrimination can be eliminated. However, for example, in case of strong wind, heavy rain, liquid inflow / discharge of the liquid in the tank, etc., it becomes impossible to discriminate, so the AE method must be avoided.

In step S22, since most of these noises are removed, individual hit waveforms are recognized. In S25, the total number of hits N is calculated. In S26, the characteristic value for each hit, that is, the maximum amplitude value Am described above. Count number n, rise time Tr, duration Td, energy E, etc. are calculated and recorded.
Here, the energy E is defined as an integral value of an absolute value of a hit waveform or an integral value of a square value.

  In the third stage S30, the number of hits obtained in this way and the characteristic value for each hit are analyzed, and the progress of damage to the target underground tank, that is, the degree of corrosion is determined.

  For example, first, in step S31, the distribution map of the maximum amplitude value Am is analyzed. The maximum amplitude value Am is arranged in descending order, and the cumulative number of hits is calculated (the final value, that is, the cumulative number of hits with respect to the minimum maximum amplitude value becomes the total number of hits N). Plot.

  Usually, when damage is growing at one place or a plurality of places in an underground tank, that is, corrosion is progressing, a sudden change accompanying the damage is called a fractal, and its scale shows a power distribution. That is, the logarithmic plot result is a straight line with a negative slope, and the magnitude of the slope is one of the indices related to the corrosion mode of the underground tank.

  If the log-log plot does not become a straight line, based on the database, for example, it is assumed that the small amplitude region and the large amplitude region are different types of noise, and the intermediate region (total hit count becomes N ′ (<N). ) Can be taken out and analyzed.

  In step S32, the total hit number N 'and energy E' are graded according to the magnitude based on the characteristic data for each hit corresponding to the true damage thus obtained. At that time, the energy E ′ is the sum of the energy E of the individual hits.

  Referring now to FIG. 4, the total number of hits N ′ and energy E ′ are graded into 5 stages, A to E and a to e, respectively, as shown in FIG. As shown in FIG. 4 (B), five stages of judgments I to V are assigned to the intersections of the completed matrix, so it is primary judgment as to which of the corrosion degrees of the underground tank corresponds to I to V. it can.

If necessary, in steps S33 and S34 based on the legally measured shell thickness and corrosion degree correlation database S35, a correlation analysis of the total hit number N ′, energy E ′ and corrosion degree is performed to corrode the underground tank. The degree can be estimated in more detail.
In step S36, the overall determination of the corrosion degree is performed based on the above, that is, the maintenance work timing corresponding to the risk of damage is determined.

  Further, if necessary, in S37, the hit data can be decomposed into hit data from a plurality of sensors, and the position of the AE generation source regarded as the damaged portion can be analyzed to estimate the position of the damaged portion.

Referring to FIG. 5, it is a flowchart of the AE method according to the second embodiment, and among the stage S <b> 30 in FIG. 3, the steps according to the present embodiment are directly indicated by solid lines, and the steps not directly related are indicated by broken lines.
That is, this embodiment is a specific example S351, S352 of the correlation database S35 of shell thickness measurement and corrosion degree, and a specific example S331 of the correlation analysis S33 of the total hit number N ′ and the corrosion degree based on the example S351. According to this, the life prediction of the damaged part of the underground tank can be performed.

  In Japan, in the case of ground tanks, it is obliged to measure the thickness of the bottom plate thickness selected at random for each tank, and hence the thickness reduction of the bottom plate thickness. As a result of the logarithmic plot of the cumulative frequency, a straight line with a negative slope is obtained, and there is a close correlation between the slope (D, absolute value of the slope) and the corrosion rate, and a variable proportional to the reciprocal of the slope. Is defined as a “corrosion risk parameter” (hereinafter abbreviated as CRP), the corrosion rate of the underground tank, and thus the life to penetration, can be predicted (probabilistically) corresponding to the CRP value. it can. The first knowledge is obtained.

  Further, when the CRP obtained in this way and the total hit number N ′ obtained by simultaneously performing the above AE method are plotted, the upper limit of the CRP is given by a calibration curve proportional to the total hit number N ′. The second knowledge is obtained.

FIG. 6 shows an actual example showing the correlation of parameters related to the degree of corrosion.
That is, the CRP value obtained from the first finding is taken on the horizontal axis, and the AE hit number N ′ is plotted on the vertical axis.
However, the CRP value on the horizontal axis is inversely proportional to the absolute value D of the gradient, and its proportionality constant is h _0.001 / y.

Here, h _0.001 is a thinning amount corresponding to the cumulative frequency of 0.001 (= initial plate thickness−actual plate thickness) in the negative slope line of the logarithmic plot of the above-described thinning amount vs. cumulative frequency. Yes, y is the years of service. The vertical axis is actually normalized by dividing the number of AE hits N ′ by the number of sense channels (the number of waveguide rods) ch and the measurement time (1 hour in the case of the first embodiment). is there.

  In FIG. 6, with respect to the plot of CRP and N ′, a correlation line passing through the origin with a high CRP value data (that is, data corresponding to a case where corrosion is progressing) as a center is set as a calibration curve CL. In the figure, the plot at the upper left of the calibration curve CL is considered to have a large N ′ value due to the influence of noise or the like at the time of AE measurement, so if this calibration curve is used for a new measured N ′ value. The CRP value is estimated on the safe side (larger eye).

  Referring to FIG. 5 again, the present embodiment is based on these findings. First, a database S351 for obtaining CRP from a correlation analysis of a logarithmic distribution of an actual measurement value of the plate (shell) thickness reduction and its cumulative frequency. Is used.

Second, a database S352 that provides a CRP upper-limit estimated calibration curve from a correlation analysis between CRP and the total number of hits N ′ is used.
At this time, in the case of the present embodiment, the calibration curve is corrected in consideration of the use of the underground tank and the use of the waveguide rod.

  In step S331, the total hit number N ′ obtained in step S31 of the first embodiment is applied to the databases S351 and S352 to obtain the corresponding CRP value on the calibration curve, that is, the upper limit estimated value of CRP. Corresponding to the value, the corrosion rate of the damaged portion of the underground tank, and hence the life until penetration corrosion is predicted (probabilistically), and the empirical result obtained in step S32 described in the first embodiment is used. It is possible to reinforce the estimation of the degree of corrosion related to the maintenance timing and to obtain step S36.

It is a figure which shows the corrosion damage evaluation system of the underground tank by this invention, (A) is a general view, (B) is a schematic diagram which shows the corrosion damage part which is a generation source of AE. It is a figure which illustrates the ultrasonic waveform by AE, and shows the various characteristics. It is a flowchart of the AE method by this invention. (A) is a table | surface used for the primary determination of a corrosion degree, (B) is a table | surface which shows the specific determination result of the corrosion degree in (A). In the flowchart of AE method by this invention, it is a figure which shows the part which concerns on a 2nd Example. It is an actual measurement figure which shows the correlation of the parameter regarding the corrosion degree used in a 2nd Example.

Explanation of symbols

10 Shell of underground tank 12 Opening 20 Liquid 30 Underground (Soil)
40 Waveguide rod 42 One end of waveguide rod 44 Piezoelectric sensor 50 Damaged portion 55 Rapid change of damaged portion 90 PC (built-in AE board)
91 Signal line

Claims (2)

A stage for collecting AE data by installing an AE data collection device including an acoustic emission (hereinafter referred to as AE) sensor in an underground tank, a stage for processing the AE data and extracting an AE characteristic value, and the AE characteristic value Including a stage for analyzing the corrosion level to determine the degree of corrosion,
At least a portion of the AE sensor is in contact with the inner wall of the shell of the underground tank via a waveguide rod;
A corrosion damage evaluation system for underground tanks. The AE characteristic value extracted in the stage of extracting the AE characteristic value includes at least “total number of hits” and “maximum amplitude value for each hit”,
The stage for analyzing the AE characteristic value to determine the corrosion degree includes (a) a correlation database between the measured shell thickness and the corrosion degree regarding the past underground tank, and (b) a correlation analysis step of the total hit number and the corrosion degree. ,
Further, (a) (a1) obtains a corrosion risk parameter (CRP) from the correlation analysis of the shell thickness reduction amount and its cumulative frequency, and (a2) obtains a CRP calibration curve from the correlation analysis of CRP and the total number of hits. Database and
From the “total number of hits” in the step (b), the upper limit of the CRP of the underground tank is estimated based on the databases (a1) and (a2) to predict the corrosion rate and life of the underground tank. The corrosion damage evaluation system for an underground tank according to claim 1, wherein

Images