JPH07190879A - Estimation method of leak point of gaseous substance - Google Patents

Abstract

PURPOSE:To reduce an estimation error when wind is weak or when many obstacles exist by a method wherein lattice points inside a plant yard are supposed to be individual gas leak points, points are given to the supposed points whose estimated value by a case study exceeds an actually measured value and a supposed leak point whose point is maximum is regarded as the center of an estimated leak source. CONSTITUTION:For example, a plant yard 81 which handles a flammable gas is selected as an object, sampling modules 82 from M1 to M12 are arranged, and sensor modules 83 are connected respectively to them. The gas which has been gathered by the modules 82 is sucked by a pump, its concentration is measured by the sensor modules 83, and a measured value is sent to a processing device 86 together with data by an anemometer 84. When a gas leak is generated, its generation source is estimated by using a field method. That is to say, a supposed leak point and a wind direction are set, a list on gathered data is made, a gathered data set is selected, a leak amount is estimated roughly, and a threshold value is found. The instruction value, of the modules 82, which exceeds the threshold value is computed, points are given to the value which exceeds an actually measured value, and the supposed leak point whose point is largest is regarded as the center of an estimated leak.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for handling gaseous substances in a broad sense such as gas, suspended substances and aerosols leaked from plant equipment in a plant handling combustible substances or harmful substances, and particularly to a leak point. Regarding the estimation method.

[0002]

2. Description of the Related Art As a method of tracing a leakage point and a leakage amount of a gaseous substance using a conventional gas sensor group and a data processing device,
For example, (1) JP-A-1-140039 (New Cosmos Electric Co., Ltd.), (2) JP-A-2-47527 (JGC Corporation), (3) Japanese Patent Application No. 4-041486 (Mitsui Toatsu Kagaku). Co., Ltd. and (4) Japanese Patent Application No. 4-262084 (the inventors).

Among these, (1) and (2) are for tracing a leakage source by a discrete arrangement of one-point type gas sensors in a plant, and (1) is based on simultaneous detection conditions of three one-point type gas sensors, (2) proposes a method of tracking by using past wind direction variation data from the detection of one gas sensor. On the other hand, in (3) and (4), the concentration value of the leaked substance in the atmosphere and the wind direction and wind speed data collected by using the sampling module having a plurality of air samplers are analyzed in real time by the data processing device. It is a method of discovering the leakage of gaseous substances and estimating the leakage position. That is, by using the sampling module, spatial detection omission is prevented and at the same time spatial error is absorbed. Especially in (4), the position and quantity can be estimated with high reliability even under the condition that the measured concentration value and the measured wind direction and speed of the sensor module have a considerably large error and the atmospheric stability at the time of leakage cannot be measured. (Hereinafter, abbreviated as position quantity)
We propose a system to do.

[0004]

When the position quantity is estimated by using the diffusion equation as in the inventions of (3) and (4), the diffusion equation is effective and the leakage of the gaseous substance actually occurs. The difference from the condition becomes a problem. For example,
The relational expression based on the wind speed above a certain level and the open diffusion space may be difficult to apply when the wind speed is extremely low or there are many obstacles. In particular, when there are many obstacles, there are double obstacles that hinder the diffusion and complicate the local wind direction distribution. Also, depending on the plant, the place where gas leakage may occur is usually the explosion-proof area, and it is difficult to install a normal wind direction and anemometer.Therefore, it may happen that the wind direction information in the plant yard cannot be obtained at all. .

[0005]

[Means for Solving the Problems] Even when the wind direction is unknown, the leakage amount of the gaseous substance and the wind speed are known,
If the indicated value of the sampling module becomes larger than the set threshold value, it means that the leakage source exists within a limited distance from the sampling module. For example, in FIG. 1, the area 11
Is a threshold value that sets the indicated value of the sensor module 13 connected to the sampling module 12 when a wind of wm / s is blowing and there is a leakage of Qm ³ / hr from any point inside. It represents a range in which there is at least one wind direction. That is, points A to Q in area 11
When there is a leak of m ³ / hr and the wind direction is θ, the indicated value of the sensor module 13 exceeds the set threshold value Sppm, but when there is a leak of Qm ³ / hr from the point B outside the area 11, In any wind direction, the sensor module 12
Does not exceed the set threshold value Sppm. At this time, it is considered that the point A is more likely to be a leakage source candidate than the point B. By using the field method utilizing this inference, when the wind direction information cannot be obtained at all or accurate wind direction information is obtained. It is possible to estimate the source of leakage even when it is unknown.

The field method is performed in the following procedure. [FIG. 2] is a diagram showing a flow of the procedure. In a plant yard as shown in "Procedure 1" [Fig. 3], a horizontal plane 31 that is substantially centered in the vertical direction is taken out, and is divided by evenly spaced meshes as shown in [Fig. 4]. Each grid point is set as a virtual leak point. A point is given to each set virtual leak point, but all of them are set to zero. As shown in “Procedure 2” [FIG. 5], the wind direction in the plant yard is set by equally dividing it into all directions. The wind direction 51 is an example of the set wind direction. Alternatively, when the accuracy of ± θ degrees can be expected in the information of the wind vane, as shown in FIG.
Is evenly divided. The wind direction 63 is an example of the set wind direction. "Procedure 3" Wind speed data w (m
/ S) and the concentration data Ci (ppm, i is the sensor number) of the gaseous substance detected by each sensor module are recorded for a predetermined time and averaged, and a combination is stored as a collection data set. This is repeated each time new averaging is performed over time, and a list (S) of many collected data sets is created as shown in FIG. One of the collected data sets (Si) is selected from the list ([FIG. 7]) created in “Procedure 4” and “Procedure 3”. With respect to the collected data sets selected in “Procedure 5” and “Procedure 4”, the leakage amount of the gaseous substance is calculated using [Equation 1].

[0007]

[Equation 1] Q = (wVCave / L) * 10 ^{-6 In} [Equation 1], V is the volume (m3) of the plant yard obtained by appropriately determining the range in the height direction, for example, [Fig. 3]. In case of, plant width 32: W and plant depth 3
Calculated as the product of 3: D and plant height 34: H. C
Since ave is the average (ppm) of the sensor readings of all sampling modules, it is calculated by [Equation 2].

[0008]

[Equation 2] Cave = (C1 + ... + Cn) / n
n: number of sensors L is the gas leaked from one point inside the plant yard
A plurality of determination methods can be considered for the average distance (m) that travels until reaching the outside of the plant yard, but if the wind direction is horizontal in a planar plant, it is calculated by [Equation 3].

[0009]

[Formula 3] L = (W × D / π) ^1/2 “Procedure 6” For the collected data set selected in “Procedure 4”, the maximum sensor reading value of the sampling module was searched for and set to it. The density value multiplied by the ratio ρ is calculated as a threshold value. One from the wind direction set in "Procedure 7" and "Procedure 2" (di)
Is selected. “Procedure 8” One (Pi) is selected from the virtual leak points set in “Procedure 1”. “Procedure 9” One sampling module (Mi) showing a sensor indication value exceeding the threshold value calculated in “Procedure 6” is selected. "Procedure 4" for the wind direction selected in "Procedure 10" and "Procedure 7"
From the virtual leak point selected in “Procedure 8” when the wind at the wind speed included in the collected data set selected in
The value Ces that the sensor of the module selected in "Procedure 9" should show when the amount of leakage calculated in "Procedure 5" occurs.
Calculate t using the gas diffusion equation. The module sensor indication value Cest calculated in “procedure 11” and “procedure 10” is used as the measured sensor indication value Ci of the module selected in “procedure 9” in the collected data set selected in “procedure 4”. In comparison, when Cest is larger than Ci, 1 is added to the virtual leak point selected in "procedure 8". If any of the sampling modules showing the sensor indication value exceeding the threshold value calculated in “procedure 12” and “procedure 6” are not yet selected, the procedure returns to “procedure 9”. Then, the last (Mi) is reached. Of the virtual leak points set in “Procedure 13” and “Procedure 1”,
If there is any item that has not been selected, the procedure returns to "Procedure 8". Then, the last (Pi) is reached. If there is an unselected wind direction among the wind directions set in "procedure 14" and "procedure 2", the procedure returns to "procedure 7".
Then, the last (di) is reached. If any of the collected data sets set in “Procedure 15” and “Procedure 3” is not yet selected, the procedure returns to “Procedure 4”. Then, the last (Si) is reached. “Procedure 16” Among the virtual leak points, a certain range around the point having the maximum point or the center of gravity of the point group having the maximum point is expected to be the leak source of the gaseous substance.

[0010]

In the field method, as seen in the above-mentioned "Procedure 11", points are given to virtual points when the sensor indication value under the selected condition is larger than the actual measurement value. That is, the significance of this point is that the virtual leak point is sufficiently close to the sampling module of the sensor that shows a large indication value that exceeds the threshold value of the “procedure 6”. Therefore, since the leakage source estimated by the field method is inevitably located around the sampling module showing a high indication value, the estimation result according to the gas diffusion formula is far from the sampling module showing a high indication, especially when the wind is weak. It has the effect of significantly reducing the number of cases where it becomes an open place.

[0011]

Embodiments of the present invention will now be described with reference to the drawings. The present invention is not limited to the examples described below.

In this embodiment, 80 m × 80 as shown in FIG. 8 of one section of a plant that handles combustible gas.
The plant yard 81 of m is targeted. In the plant yard 81, a total of 12 sampling modules 82 from M1 to M12 are arranged and connected to the sensor modules 83, respectively. The range in the height direction is 5 m, and the sampling module 82 is installed at a height of 2.5 m above the ground. The gas collected by the sampling module 82 is sucked by the pump and its concentration is measured by the sensor module 83. The signal measured by the sensor module 83 is transmitted to the data processing device 86 in the control room via the communication cable 85 together with the signal measured by the anemometer 84.

With these systems, the sensor indication values and the wind direction and wind speed information of 12 modules are constantly collected, and when gas actually leaks, for example, [Fig. 9].
The state becomes as shown in.

After that, the source of leakage is estimated as follows in accordance with the procedure.

In accordance with "Procedure 1", the plant yard 81
A virtual leak point 87 is set inside the.

According to "Procedure 2", the wind direction around the entire circumference is set as shown in FIG.

According to "Procedure 3", the wind speed data and the sensor indication value data as shown in FIG. 9 are collected,
A collection data set as shown in FIG. 10 is created.

According to "Procedure 4", one collection data set is selected. The case where the collection data set 101 illustrated in FIGS. 9 and 10 is selected is taken as an example.

In accordance with "Procedure 5", the amount of gas leakage is calculated. Because it is an 80m x 80m plant yard,
From [Equation 3], L = (80x80 / 3.1416) ^1/2 = 45.14 m is calculated. Also since the height direction 5m, V = 80x80x5 = 32000 from m ³ collected data sets 101, w = 1.21m / s, Cave = 4.
Since it is 01ppm (calculation result by [Equation 2]), the leakage amount is calculated by [Equation 1] as Q = (1.21x32000x4.01 / 45.14) x10 ^-6 ＝ 3.44m ³ / s = 12.4m ³ / hr .

Collected data set 1 according to "Procedure 6"
From 01, specify the maximum indicated value of 14.43 ppm (M7) and determine the threshold value. If the ratio ρ is 0.5, the threshold value is 7.22 ppm.

In accordance with "Procedure 7", the wind direction is selected from the one shown in FIG. Here, it is assumed that the wind direction 111 of [FIG. 11] is selected.

According to "Procedure 8", one of the virtual leak points 87 is selected. Here, it is assumed that the point 112 in FIG. 11 is selected.

Collected data set 1 according to "Procedure 9"
For 01, module M exceeding the threshold of 7.22ppm
4, M5, M7 are selected (Module 1 of [Fig. 11]
13, 114, 115).

In accordance with "procedure 10", the module 11
Calculate the sensor readings of 3, 114, 115. The gas diffusion formula used is already Sutton, Crammer, Pasquill,
The relational expression reported by Sakagami et al. May be used, but here, the α model derived by the inventors based on the proof result based on the Gaussian model is used. In the normal Gaussian model, the gas continuously discharged from the point source with height z ₀ (m) at Q (m ³ / hr) is the wind speed w (m / s).
When diffused by the wind, the wind direction axis is the X axis (downward is positive),
A space rectangular coordinate point (X, with the leak point as the origin of the XY plane)
The gas concentration C (m ³ / m ³ ) of Y, Z) is represented by [Equation 4].

[0025]

[Equation 4] Of this formula

[0026]

[Outer 1]

[0027]

[Outside 2] Are the standard deviations of the concentration distributions in the Y and Z directions, respectively, which are generally a function of the atmospheric stability and the leeward distance x.

[0028]

[Outside 3] It is expressed by using [Equation 5].

[0029]

[Equation 5] Here α ₁ , α ₂ , β ₁ , and β ₂ are experimentally established constants. here,

[0030]

[Outside 4] However, since the purpose is to measure atmospheric stability, an anemometer that is located slightly away from the explosion-proof area of the plant yard may be used. When the sensor indication values of the modules 113, 114, and 115 are calculated using the α model, the sensor indication value 11 of FIG.
6, 117, 118.

According to "Procedure 11", the calculated value of the sensor indication value of M4 and the measured value are compared. The calculated value is 0.00ppm from [Fig. 11] and the measured value is 7.26ppm from [Fig. 9].
12 points will not increase.

According to the “procedure 12”, the comparison of the “procedure 11” is performed for M5 and M7. The calculated values from [Fig. 11] are 36.2 ppm and 17.8 ppm, respectively, and the measured values from [Fig. 9] are 11.84 ppm and 14.43 ppm, respectively. Both of them have larger calculated values, so the point 112 has two points. Increase.

According to "procedure 13", another virtual leak point 8
The same operation is performed for 7 as well, and the point increase of each virtual leak point 87 is obtained as shown in FIG.

In accordance with "Procedure 14", the same operation is performed for the wind directions of other [Fig. 5], and the point increase of each virtual leak point 87 is obtained as shown in [Fig. 13].

In accordance with "Procedure 15", the same operation is performed for the other collected data sets of [Fig. 10], and the final virtual leak point is obtained as shown in [Fig. 14].

According to "Procedure 16", a region 142 centered on the point 141 having the highest point in FIG. 14 is predicted as a leakage source.

[0037]

EFFECT OF THE INVENTION A system for estimating the position and amount of a leakage source of a gaseous substance by using a gas sensor of a multipoint sampling module type arranged in a plant and a wind direction and wind speed measuring device is a diffusion equation as a means for tracking a position amount. When the method of optimizing is used, the error tends to increase when the wind is weak or there are many obstacles, and even if the wind direction information is not available, it is impossible to configure the system.

In the present invention, in order to overcome the above problems, the inside of the plant yard is divided into meshes at equal intervals, and each of the lattice points is assumed to be a gas leak point. In many cases, a large number of wind direction case studies have been conducted over all directions or a highly reliable wind direction range, and the concentration value of the gaseous substance predicted by the diffusion formula exceeds the actually measured concentration value of the gaseous substance. By selecting the point, the effect of significantly increasing the estimation error of the same gas leak source estimation system is significantly reduced.

[Brief description of drawings]

FIG. 1 is a diagram showing a range in which a leakage source may exist with respect to a sensor instruction value of a sampling module when a wind speed and a leakage amount are known.

FIG. 2 is a diagram showing a flow of a field method.

FIG. 3 is a diagram simulating a plant into a rectangular parallelepiped.

FIG. 4 is a diagram showing setting of a virtual leak point.

FIG. 5 is a diagram showing a wind direction setting when the wind direction information is completely untrustworthy.

FIG. 6 is a diagram showing a wind direction setting when a part of the wind direction information can be trusted.

FIG. 7 is a diagram showing a list of collected data sets.

FIG. 8 is a configuration diagram of the present invention used in an embodiment.

FIG. 9 is a diagram showing elements constituting one collection data set.

FIG. 10 Same as [FIG. 7]. However, specific numerical values are used in the examples.

FIG. 11 shows selection of a sampling module for point calculation in the field method embodiment,
It is a figure which shows calculating a sensor module instruction value from each selected condition simultaneously.

FIG. 12 is a diagram showing a point increase of each virtual leak point in one selected collection data set and one selected wind direction in an example of the field method.

FIG. 13 is a diagram showing a point increase of each virtual leak point in one selected collected data set in the example of the field method.

FIG. 14 is a diagram showing a total point of each virtual leak point and at the same time a final leak source estimation result in the embodiment of the field method.

[Explanation of symbols]

M1 to M12 module 11 Leakage source existence range 12 Sampling module 13 Sensor module 14 Specific wind direction 31 Horizontal section taken from the plant yard to set a virtual leak point 32 Plant width 33 Plant depth 34 Plant height 51 Example of wind direction selection 61 Measured wind direction 62 Range of reliable wind direction 63 Wind direction selection example 81 Plant yard 82 Sampling module 83 Sensor module 84 Wind direction anemometer 85 Communication cable 86 Data processing device 87 Virtual leak point example 101 Collected data set example 111 Wind direction selection Example 141 Highest virtual leak point of 141 points 142 Estimated leak source

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Saburo Takahashi 1-6 Takasago, Takaishi, Osaka Prefecture Mitsui Toatsu Chemical Co., Ltd. (72) Inventor Masahiko Tsuchiya 2-8-1, Akanehama, Narashino, Chiba Toyo Engineering Ring Co., Ltd. (72) Inventor Ippanta Kishiguchi 2-8-1 Akanehama, Narashino-shi, Chiba Toyo Engineering Co., Ltd. (72) Inventor Yoshio Kawauchi 2-8-1 Akanehama, Narashino, Chiba Toyo Inside Engineering Co., Ltd.

Claims (2)

[Claims] 1. When a gaseous substance leaks from equipment in a plant that handles combustible substances or harmful substances, a plurality of sampling modules arranged in the plant and a sensor for detecting the concentration of the leaked substance are built-in. The sensor module, the wind direction and wind speed measurement device that measures the local wind direction and local wind speed in real time, the wind direction and wind speed data obtained from the wind direction and wind speed measurement device, and the data collected by each sampling module are collected by the sensor module. With the system configured by the data processing device that analyzes the detected concentration value of the leaking gaseous substance in the atmosphere in real time,
A method for estimating leakage points of gaseous substances, characterized by using the field method under the condition that wind direction information is not available or has low reliability. 2. The field method divides the inside of a plant yard into meshes at equal intervals, and assumes those lattice points as individual gas leak points, and selects a plurality of omnidirectional or highly reliable wind direction ranges within the plant yard. According to the wind direction case study, a point is given to the grid point where the concentration value of the gaseous substance predicted by the diffusion formula exceeds the measured concentration value of the gaseous substance, and the virtual leak point with the maximum point is estimated as the center of the leak source. The method for estimating a leakage point of a gaseous substance according to claim 1, wherein