JP2009222395A - Maximum corrosion speed estimating method of tank bottom plate - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a maximum corrosion speed estimating method of a tank bottom plate utilizing the plate thickness measuring data of a standard tank accumulated by past measurement. <P>SOLUTION: The plate thickness of the tank bottom plate is measured at 500 points or above at random in a lattice state or zigzag state, the respective corrosion depths 1a obtained by this measurement are divided by the use number of years of the tank, the obtained value is further divided by predetermined values within a range of 0.03-0.10 and 1 is added to the obtained values to calculate normalized corrosion speeds 1b. These normalized corrosion speeds 1b are arranged in the order of a decreasing speed and upper side cumulated probability 1c comprising the upper order of the normalized corrosion speeds 1b obtained by dividing the number of the plate thickness measuring data at each normalized corrosion speed 1b by the sum number of the plate thickness measuring data to calculate both logarithmic graphs 1d of the normalized corrosion speeds 1b and the upper side cumulated probability 1c. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a method for estimating the maximum corrosion rate of a bottom plate such as an oil storage tank.

  Conventionally, about 80% of domestic storage tanks are tanks that are used for a long period of 30 years or more after construction, and countermeasures against leakage due to corrosion thinning are an important facility management issue. In particular, it is one of the most difficult to manage the corrosion thinning from the back side (soil side) of the tank bottom plate, which is difficult to confirm the overall (exhaustive) corrosion thinning by visual inspection.

  Corrosion thinning management of the bottom plate of domestic storage tanks is regulated by the Fire Service Act, and open inspection is required within 5 to 13 years for specified outdoor tanks of 1000 kl (kiloliters) or more.

  In the conventional fire fighting method, the thickness is measured discretely by the ultrasonic method at intervals of 1000 mm. Ultrasonic plate thickness measurement of the tank bottom plate is made by entering ultrasonic waves from the inner surface of the tank, passing through the plate thickness and reflecting from the back side (sound is the same as light, and reflection and refraction occur at the interface) If the time of reflection and return is measured, the speed of sound is a material-specific value and can be converted to a plate thickness.

  Then, the portion below the reference plate thickness is repaired between the actually measured maximum corrosion rate (design thickness-minimum plate thickness) and the remaining thickness until the next release.

  In the tank bottom plate backside corrosion, the long-term use tank is locally corroded, and the above discrete measurement interval is much larger than the size of this local corrosion. The current situation is that the "actual maximum corrosion thinning" is not measured.

  In order to solve such a problem, Japanese Patent No. 3670525 (Patent Document 1) proposes a plate thickness measuring device for a cylindrical tank bottom plate capable of continuously measuring the plate thickness of the entire tank bottom plate.

Japanese Patent No. 3670525

  However, continuously measuring the plate thickness of the entire bottom plate of each tank takes a lot of work time and uses the standard tank plate thickness measurement data accumulated and measured in the past. There was a request that it would be efficient if the corrosion rate of the steel could be estimated.

  SUMMARY OF THE INVENTION The present invention solves the above-mentioned problems, and an object of the present invention is to provide a tank capable of estimating the maximum corrosion rate of the tank bottom plate using standard tank plate thickness measurement data accumulated in the past. It is intended to provide a method for estimating the maximum corrosion rate of the bottom plate.

  The first configuration of the tank bottom plate maximum corrosion rate estimation method according to the present invention for achieving the above object is to measure the thickness of the tank bottom plate in a random, grid or zigzag manner at 500 points or more and measure Divide each corrosion depth obtained by the years of use of the tank, obtain the normalized corrosion rate from the value, arrange the normalized corrosion rates in descending order, and plate thickness for each normalized corrosion rate The upper cumulative probability of the normalized corrosion rate divided by the plate thickness measurement data total number is obtained in an upper order, and a log-log graph of the normalized corrosion rate and the upper cumulative probability is obtained. The logarithmic graph is linearly approximated, the absolute value D ′ of the slope of the approximate straight line 1e is obtained, and the penetration of the 3D lattice block assumed in the tank bottom plate from one point on the surface is stopped at a predetermined penetration probability by penetration simulation. Infiltrate this Performing the infiltration simulation up to a fixed number of blocks, obtaining the maximum penetration depth of each three-dimensional lattice block, obtaining the normalized penetration depth from the maximum penetration depth, arranging the normalized penetration depth in descending order, An upper cumulative probability of the normalized penetration depth in descending order is calculated by dividing the number of three-dimensional lattice blocks for each normalized penetration depth by the total number of penetration simulation blocks, and the normalized penetration depth and the upper cumulative The logarithmic graph with probability is obtained, the logarithmic graph is linearly approximated, the absolute value D ″ of the slope of the approximated straight line is obtained, this operation is performed with various penetration probabilities, and D ′ and D ″ Are obtained under the same conditions, and the number obtained by dividing the total area of the tank bottom plate by the predetermined measurement area is obtained under this condition. 15 or less range It is divided by a predetermined value, and obtains the maximum corrosion rate is multiplied by a predetermined value of 0.03 or more and 0.10 or less.

  When obtaining the normalized corrosion rate, each corrosion depth obtained by measurement is divided by the age of use of the tank, and the value is further in the range of 0.03 or more and 0.10 or less. It is preferable to obtain a normalized corrosion rate consisting of a value obtained by dividing by a predetermined value and adding 1, and among them, each corrosion depth obtained by measurement is divided by the years of use of the tank, and the value is calculated. It is preferable to obtain a normalized corrosion rate consisting of a value obtained by dividing by 0.05 and adding 1.

  Further, when obtaining the normalized penetration depth by dividing the maximum penetration depth, the maximum penetration depth is divided by a predetermined value in the range of 3 or more and 15 or less. It is preferable to obtain the normalized penetration depth, and among them, it is preferred to obtain the normalized penetration depth consisting of a value obtained by dividing the maximum penetration depth by 7 and adding 1.

  Further, the predetermined value in the range of 3 to 15 excluding the maximum corrosion rate is preferably 7, and is similarly in the range of 0.03 to 0.10 multiplied by the maximum corrosion rate. The predetermined value is preferably 0.05.

  Further, when obtaining the number obtained by dividing the total area of the tank bottom plate by a predetermined measurement area, it is preferable to obtain the number obtained by dividing the total area of the tank bottom plate by the area of the probe of the predetermined ultrasonic measurement device.

Further, the second configuration of the method for estimating the maximum corrosion rate of the tank bottom plate according to the present invention includes a number N of three-dimensional lattice blocks for performing the permeation simulation, and a case where the permeation simulation is performed in the first configuration. The relationship between the maximum penetration depth F of each three-dimensional lattice block is
[Equation 1]
F = a × ln (N)
a = α × exp (β × D ′)
In this case, α is a predetermined value in the range of 5.5 to 7.7 and β is a predetermined value in the range of −0.16 to −0.14.

  The plate thickness measurement data for measuring the plate thickness of the tank bottom plate and obtaining D ′ is preferably 1000 points or more. Further, the number of blocks for obtaining D ″ in the penetration simulation is preferably a predetermined number of 500 points or more, and more preferably the same number of blocks as the number of plate thickness measurement data. The maximum penetration depth of each three-dimensional lattice block may be the number of lattices.

  According to the present invention, the maximum corrosion rate of the tank bottom plate can be estimated by using standard tank plate thickness measurement data accumulated in the past.

  An embodiment of a method for estimating the maximum corrosion rate of a tank bottom plate according to the present invention will be specifically described with reference to the drawings. FIG. 1 is a flowchart for explaining a method for estimating the maximum corrosion rate of a tank bottom plate according to the present invention, FIG. 2 is a flowchart for explaining an infiltration simulation, and FIG. 3 (a) is a normal one based on the thickness measurement data of the tank bottom plate obtained by actual measurement. FIG. 3 (b) is a diagram showing how to determine the chemical corrosion rate and the upper cumulative probability, and FIG. 3 (b) is a diagram illustrating how the normalized penetration depth and the upper cumulative probability are obtained from the penetration depth of each block obtained by the simulation. Is a diagram showing a simulation under typical infiltration conditions, a typical actual tank risk risk curve and a normalized corrosion risk curve, FIG. 5A is a diagram for explaining an infiltration model of the infiltration simulation, and FIG. FIG. 6 is a conceptual diagram of the infiltration simulation model, FIG. 6 is a diagram showing the relationship between the slope D ′ of the normalized corrosion risk curve and the infiltration probability in the depth direction, and FIG. FIG. 8 is a diagram for explaining the estimation method from the slope D ′ of the normalized corrosion risk curve of a (a = 5.9exp (−0.15 × D ′)) while showing the relationship with the number of thickness measurement data. , A = α × exp (β × D ′) is a diagram showing the relationship between the coefficients α, β and the number of data, FIG. 9 is a diagram showing a state in which the permeation simulation results are verified by comparison with the actual measurement results is there.

In step S ₁ of FIG. 1, for example, the thickness of the tank bottom plate by the present applicant such as a plate thickness measuring device of Patent Document 1, developed random lattice shape or in a zigzag pattern, measured over 500. In this embodiment, 1000 or more points of plate thickness measurement data for actually measuring the plate thickness of the tank bottom plate are obtained.

Then, each corrosion depth obtained by the measurement (see the corrosion depth column 1a in FIG. 3A) is divided by the service life y of the tank, and the value is further 0.05 (mm / y). ) To obtain the normalized corrosion rate consisting of the value obtained by adding “1” (see the normalized corrosion rate column 1b in FIG. 3A), and arrange these normalized corrosion rates in descending order. The upper cumulative probability (refer to the upper cumulative probability column 1c in FIG. 3 (a)) obtained from the higher order of the normalized corrosion rate obtained by dividing the number of plate thickness measurement data for each corrosion rate by the total number of plate thickness measurement data, The logarithmic graph 1d of the normalized corrosion rate corresponding to the corrosion depth of 0.1 mm or more and the upper cumulative probability is obtained (see FIG. 3B), and the logarithmic graph 1d is linearly approximated, The absolute value D ′ of the inclination of the approximate straight line 1e is obtained (step S ₂ ).

  Here, the “0.05” divided when obtaining the normalized corrosion rate (see the normalized corrosion rate column 1b in FIG. 3A) is the unit [0.05 mm / year]. This is a positional parameter for normalizing the velocity, and means a value on the horizontal axis of the bending point of the corrosion risk curve in FIG. 4C, and the practically usable range is shown in FIG. ) In the range of 0.03 or more and 0.10 or less shown on the horizontal axis, and the most desirable value is 0.05.

  In the linear approximation, the data range to be approximated is not limited to the corrosion depth of 0.1 mm or more, but may be approximated within a range where the linear relationship is substantially established.

On the other hand, as shown in FIG. 5, a three-dimensional lattice block assumed in the tank bottom plate is infiltrated from one point on the surface until the permeation stops at a predetermined permeation probability by the permeation simulation, and this permeation simulation is performed up to a predetermined number of blocks. Implement (step S ₃ ).

  Here, as shown in the schematic diagram of FIG. 5A, the penetration (percolation) model gives a probability Pr to the penetration site 3 with a random number, and the magnitude of the preset penetration probability Ps. It is a probabilistic infiltration model that determines whether to infiltrate in a relationship.

  For example, if Pr ≧ Ps, it does not penetrate, and if Pr <Ps, it penetrates. Here, the penetration probability Ps is a parameter corresponding to the corrosion rate.

  In the infiltration simulation of the tank bottom plate, as shown in FIG. 5 (b), a lattice-like infiltration block 2 that becomes a three-dimensional lattice block is considered, and a model in which corrosion progresses and expands from one point in the center. It is.

Here, the permeation simulation flow is shown in FIG. 2, first, in step S _11, as in FIG. 5 (b), to ensure the sequence of the infiltration block 2 is a 3-dimensional lattice block composed of a plurality of sites, the particular surface of the three-dimensional lattice The penetration start point “1” is input at the center (step S ₁₂ ). Then enter the penetration probability Ps in step S _13, in step S _14, to search for all the penetration site substitutes "1" in the previous attempt. Then search all empty site 3 adjacent to all of the penetration site in step S _15.

Next, in step S ₁₆ , random numbers “0 to 1” are generated in all the empty sites 3, and in step S ₁₇ , when the random number generated in step S ₁₆ is smaller than the penetration probability Ps. Goes to step S ₁₈ and assigns “1” to the site. Next, in step S _19, to count the number of penetration sites that put a "1". In step S ₁₇ , no value can be entered for the empty site 3 in which the random number generated in step S ₁₆ is greater than the penetration probability Ps.

Next, in step S ₂₀ , if there is no penetration site 3 in which “1” is entered in this trial, the process proceeds to step S ₂₁ , where the total number of penetration sites 3 in which “1” is entered, the penetration depth, Display vertical and horizontal spread. Incidentally, in the step S _20, when the penetration site 3 containing the "1" in this trial there are one or more repeats the steps S ₁₄ to S ₂₀ proceeds to step S _14.

  Here, as shown in FIG. 5 (b), one corroded portion of the tank bottom plate is regarded as one permeation block (cluster) 2 and is made to correspond to one ultrasonic measurement data of the actual tank.

  How far it penetrates can be controlled by the penetration probability Ps, but when it exceeds a certain value, it penetrates infinitely (critical probability). Here, the permeation simulation shown in FIG. 2 is performed until the permeation stops under the condition that the permeation is stopped within the critical probability (a condition where the permeation is terminated somewhere).

  The number of permeation blocks 2 acquired by simulation with the same permeation probability Ps (which can be regarded as the same tank data) is preferably 500 points or more, and more preferably, the number of measurement data of the target real tank. The maximum penetration depth of each penetration block 2 obtained (see the penetration depth column 4a in FIG. 3C) is divided by “7” and “1” is added (FIG. 3C). Normalized penetration depth column 4b) is calculated, the normalized penetration depths are arranged in descending order, and the upper cumulative value obtained by dividing the number of penetration blocks 2 for each normalized penetration depth by the number of total penetration simulation blocks 2 A probability (see the upper cumulative probability column 4c in FIG. 3C) is obtained, and a log-log graph 4d of the normalized penetration depth and the upper cumulative probability is obtained (see FIG. 3D). Further, by performing linear approximation with all the data of FIG. 3D, the slope D ″ of the normalized risk curve in the penetration simulation can be obtained from the slope D ″ of the approximate line 4e.

  Both the slope D ″ of the normalized risk curve in this simulation and the slope D ′ of the normalized corrosion risk curve of the actual tank shown in FIG. 3B have no physical dimension and should be directly compared. Can do.

  In the linear approximation, the data range to be approximated is not limited to all data, but may be approximated within a range where the linear relationship is substantially established.

  Here, “7” divided when obtaining the normalized corrosion rate in simulation (see the normalized corrosion rate column 1b in FIG. 3A) is a positional parameter for normalizing the penetration depth. 4 is a bending point of the corrosion risk curve in the simulation of FIG. 4A, which means a value on the horizontal axis, and a practically usable range is “3” shown on the horizontal axis of FIG. 4A. ”And not more than“ 15 ”, and the most desirable value is“ 7 ”.

Then, this operation is performed with various penetration probabilities, and the condition that the D ′ and the D ″ are substantially the same is obtained (step S _{5 in} FIG. 1). Under this condition, from the total area of the tank bottom plate, The number N divided by the area of 25 mm x 5 mm, which is the section for continuous measurement, is obtained, the number of penetration blocks 2 is increased to this number N, and the assumed maximum penetration depth dmax is obtained from the penetration block 2 with the maximum depth ( In step S ₆ ) in FIG. 1 and in step S ₉ in FIG. 1, this value dmax is divided by “7” which is the penetration depth in the penetration simulation, and becomes 0 as a positional parameter for normalization in the actual tank data. .05 (mm / y) to find the maximum corrosion rate (step S _{10 in} FIG. 1)
). In step _{S 5,} when the D'and said D'' is not a substantially same repeats the steps _S 3 to S ₅ returns to the step _{S 3.}

  Here, the area of 25 mm × 5 mm, which is a section for continuous measurement, is not limited to this, and when estimating the maximum corrosion depth when surface inspection is performed with a normal ultrasonic probe, the measurement is performed. The area of the vessel can also be used. In addition, the values of “0.05” and “7” employed at the bending points of the actual tank and the simulation risk curve should be substantially the predetermined values within the above-mentioned range. it can.

  By the way, when the above penetration simulation is performed with various penetration probabilities, there is a penetration block that penetrates deeper when the number of penetration blocks 2 increases in any penetration probability. The relationship between the number of data and the penetration depth can be approximated by the maximum penetration depth F = a × ln (the number of corrosion data consisting of the number of penetration blocks 2) at each penetration probability.

  Then, there is a relationship shown in FIG. 7 between D ″ and a obtained with various penetration probabilities, and the value of a for any D ″ can be obtained.

Here, as shown in FIG. 7, D ″ changes depending on the number (the number of data) of the permeation blocks 2 for obtaining D ″. For this reason, the regression equation a = α × exp (β × D ′ The values of α and β when ') are changed as shown in FIG. That is, when performing the penetration simulation, the relationship between the number N of three-dimensional lattice blocks and the maximum penetration depth F of each three-dimensional lattice block when the penetration simulation is performed is:
[Equation 2]
F = a × ln (N)
a = α × exp (β × D ″)
Α is a predetermined value in the range of 5.5 to 7.7, and β is a predetermined value in the range of −0.16 to −0.14. The coefficient a is obtained (step S _{7 in} FIG. 1), and the maximum penetration depth F of each three-dimensional lattice block is obtained (step S _{8 in} FIG. 1).

Here, since the direct D ″ and the D ′ obtained from the actual tank data are substantially the same, the correspondence with the actual tank is “instead of“ D ″ ”in the above equation (2). D '"
Is substituted, and the maximum penetration depth F corresponding to the actual tank condition is obtained. That is, it is obtained by the following equation (1).

[Equation 1]
F = a × ln (N)
a = α × exp (β × D ′)

Then, a number N obtained by dividing the total area of the tank bottom plate by an area of 25 mm × 5 mm, which is a section for continuous measurement, is obtained, and this number N is obtained as the assumed maximum penetration depth dmax from the above equation 1, and the steps of FIG. in S _9, the value dmax, is divided by "7" as the penetration depth in penetration simulation, the maximum corrosion is multiplied by 0.05 a position parameter for normalization (mm / y) in the actual tank data The speed is obtained (step S _{10 in} FIG. 1).

  Here, the area of 25 mm x 5 mm, which is the section for continuous measurement, is not limited. When estimating the maximum corrosion depth when surface inspection is performed with a normal ultrasonic probe, the area of the measuring instrument is used. Can also be used. In addition, the values of “0.05” and “7” employed at the bending points of the actual tank and the simulation risk curve should be substantially the predetermined values within the above-mentioned range. it can.

  The maximum corrosion rate estimated above and the actual maximum corrosion rate actually measured have a correlation as shown in FIG. 8, and it was verified that a good estimation result was given.

  As an application example of the present invention, it can be applied to a method for estimating the maximum corrosion rate of a bottom plate of a petroleum storage tank or the like.

It is a flowchart explaining the maximum corrosion rate estimation method of the tank bottom plate which concerns on this invention. It is a flowchart explaining an infiltration simulation. (A) is a figure which shows a mode that a normalization corrosion rate and an upper accumulation probability are calculated | required from the thickness measurement data of the tank bottom plate obtained by actual measurement, (b) is normalized from the penetration depth of each block obtained by simulation It is a figure which shows a mode that a penetration depth and an upper accumulation probability are calculated | required. It is a figure which shows the simulation under typical penetration | invasion conditions, and the corrosion risk curve and typical corrosion risk curve of a typical actual tank. (A) is a figure explaining the penetration model of a penetration simulation, (b) is a conceptual diagram of a penetration simulation model. It is a figure which shows the relationship between inclination D 'of a normalized corrosion risk curve, and the penetration probability of a depth direction. The figure which shows the relationship between the penetration depth and the number of plate thickness measurement data, and the estimation method from inclination D 'of the normalized corrosion risk curve of a (a = 5.9exp (-0.15 * D')) It is. It is the figure which showed the relationship with the data number of each coefficient (alpha) and (beta) of a = (alpha) * exp ((beta) * D '). It is a figure which shows a mode that the penetration simulation result was verified in comparison with the measurement result.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1a ... Corrosion depth column 1b ... Normalization corrosion rate column 1c ... Upper accumulation probability column 1d ... Double logarithm graph 1e ... Approximate straight line 2 ... Penetration block 3 ... Penetration site 4a ... Penetration depth column 4b ... Normalization penetration depth column 4c: Upper cumulative probability column 4d: Log-log graph 4e: Approximate straight line dmax: Assumed maximum penetration depth F ... Maximum penetration depth Pr ... Probability Ps ... Penetration probability y ... Year of use

Claims (2)

Measure the thickness of the tank bottom plate at random over 500 points in a random, grid, or staggered pattern, divide the corrosion depth obtained by the measurement by the age of the tank, and use the normalized corrosion from that value. Obtain the speed, arrange the normalized corrosion rates in descending order, and the upper cumulative probability consisting of the upper order of the normalized corrosion rate obtained by dividing the number of plate thickness measurement data for each normalized corrosion rate by the total number of plate thickness measurement data A logarithmic graph of the normalized corrosion rate and the upper cumulative probability, a linear approximation of the logarithmic graph, and an absolute value D ′ of the slope of the approximate line,
The three-dimensional lattice block assumed in the tank bottom plate is infiltrated from one point on the surface by infiltration simulation until the infiltration stops at a predetermined infiltration probability, and this infiltration simulation is performed up to a predetermined number of blocks, and each three-dimensional lattice block The maximum penetration depth is calculated, the normalized penetration depth is obtained from the maximum penetration depth, the normalized penetration depths are arranged in descending order, and the number of three-dimensional lattice blocks for each normalized penetration depth is total penetration. The upper cumulative probability consisting of the higher order of the normalized penetration depth divided by the number of simulation blocks is obtained, the logarithmic graph of the normalized penetration depth and the upper cumulative probability is obtained, and the logarithmic graph is linearly approximated Then, the absolute value D ″ of the inclination of the approximate straight line is obtained,
This operation is performed with various penetration probabilities, and the condition that the D ′ and the D ″ are substantially the same is obtained. Under this condition, the number obtained by dividing the total area of the tank bottom plate by the predetermined measurement area is obtained. Determine the value of the penetration simulation up to this number, divide this value by a predetermined value in the range of 3 to 15 and multiply by the predetermined value in the range of 0.03 to 0.10 to obtain the maximum corrosion rate. A method for estimating the maximum corrosion rate of a tank bottom plate, characterized in that The relationship between the number N of three-dimensional lattice blocks that perform the penetration simulation and the maximum penetration depth F of each three-dimensional lattice block when the penetration simulation is performed is as follows:
[Equation 1]
F = a × ln (N)
a = α × exp (β × D ′)
Wherein α is a predetermined value in a range of 5.5 to 7.7 and β is a predetermined value in a range of −0.16 to −0.14. The method for estimating the maximum corrosion rate of the tank bottom plate as described in 1.

Images