JPH05157609A - Liquid level measuring method - Google Patents

Abstract

PURPOSE:To detect the surface level of a liquid with accuracy lower than distance resolution by finding the correlation between a measured temperature value which is contained in temperature distribution data along an optical fiber in its length direction and obtained in the vicinity of the boundary between a liquid and gas and liquid level fluctuation. CONSTITUTION:An optical fiber sensor with a mark at an arbitrary location is dipped in water until the mark meets the surface of the water and temperatures are measured by sinking and raising the fiber by fixed lengths shorter than distance resolution. Constants A1 and A2 of an expression X=Lo+(A1.D2.DELTAl/Tp)+A2-L1 (where, X, Lo, L1, and D2 respectively represent the liquid level, set distance to a temperature changing point, distance from the starting edge to the reference liquid level position of an object to be measured, and feature amount of variation and DELTAl. Tp, and A1 and A2 respectively represent the pulse width, temperature difference between a gas and liquid, and constants representing amounts corresponding to the DELTAl and sampling time interval) are found in advance from the measured temperature values by using the method of least squares. During the course of actual measurement, the temperature distribution along the optical fiber in its length direction obtained when the fiber is dipped in the liquid to be measured is measured and the surface level of the liquid is found by substituting the measured value and found A1 and A2 values into the above- mentioned equation.

Description

Detailed Description of the Invention

The present invention relates to a liquid level measuring method using a distributed optical fiber temperature measuring system, and more particularly to a measuring method capable of detecting a liquid level with accuracy less than a distance resolution.

[0002]

2. Description of the Related Art A distributed optical fiber temperature measuring system (DTS: Distributed Temperature Sensor) is known as one of the techniques for continuously measuring a physical quantity along an optical fiber. The liquid level using this system is known. As for the measuring method, a patent application filed by the applicant earlier [April 1, 1991]
No. 0 application, title of invention; liquid level measuring method (hereinafter referred to as Document 1)] has been proposed.

This utilizes the fact that there is generally a difference between the air temperature and the liquid temperature. The optical fiber, which is the sensor part of the DTS, is immersed in the liquid to measure the temperature, and temperature distribution data in the longitudinal direction of the optical fiber. The temperature difference is expressed as a step on the temperature data, and the liquid level is detected based on the distance to the step.

[0004]

By the way, "distance resolution" is one of the performance characteristics of the distributed optical fiber temperature measuring system. This is an amount determined by the pulse width of the light source used by the device, the band of the light receiving circuit, the sampling interval, etc., and a distance resolution finer than the sampling interval cannot be obtained.

Here, for the sake of simplicity, only the sampling interval will be taken up as a factor that determines the distance resolution, and for example, a stepwise temperature change as shown in FIG. 8A will be sampled as shown in FIG. 8B. When measuring at intervals, the average temperature within each sampling interval is detected, so the temperature distribution data obtained is as shown in FIG. Therefore, if it is left as it is, it is known that the change point is in the section of n, but it is not possible to determine the finer position in it. In particular, even if the liquid level changes from this state, if the change is slight, it is impossible to judge how much the liquid level has changed and what the accurate distance to the liquid surface is after the change.

[0006]

The present invention has been made in order to solve the above problems, and first of all, it detects the change of the liquid surface within a certain range below the distance resolution. Secondly, the purpose is to detect the liquid level. Here, the liquid level refers to the distance from the liquid level reference position of the measurement object to the liquid surface, for example, if the measurement object is a well, the opening of the well is the liquid level reference position, and from here to the liquid surface. The distance becomes the liquid level.

[0007]

In order to achieve the above-mentioned object, the method of the present invention comprises: immersing an optical fiber, which is a sensor portion of a distributed optical fiber temperature measuring system, in a liquid to obtain a temperature distribution in the longitudinal direction of the optical fiber. The measured temperature near the interface between the liquid and gas indicated on the obtained temperature distribution data changes corresponding to the change in the liquid level. It is characterized in that the correlation between the measured values is obtained, and the liquid level is detected in units smaller than the distance resolution based on this correlation (first configuration).

Further, based on the concept of the above first construction,
It is configured so that the liquid level can be automatically determined using a computer or the like. This method is based on pre-measurement in which a slight change in liquid level and the corresponding change in temperature are obtained in advance in a laboratory, etc. for each measuring device, and the results obtained by this pre-measurement are used to actually This is a liquid level measuring method comprising two steps of main measurement for measuring the liquid level of a measurement target.

First, in the pre-measurement, a mark is made on an arbitrary position of the optical fiber which is the sensor portion of the distributed optical fiber temperature measuring system, and the optical fiber is immersed in a tub filled with water to make the mark on the water surface. In addition, the optical fibers are sequentially immersed or pulled up by a fixed length less than the distance resolution to measure the temperature. This is one in which the actual ascending / descending state of the liquid surface is relatively created by the above-mentioned constant length dipping or pulling up. From the temperature data obtained by this measurement, for example, a constant A _{1 of} the following mathematical formula A is used by using the least square method, Obtain A ₂ in advance.

In the main measurement, the optical fiber is immersed in the liquid to be measured, the temperature distribution in the longitudinal direction of the optical fiber is measured, and this measurement result and A obtained in the previous measurement are measured.
_The liquid level is obtained with an accuracy equal to or less than the distance resolution by the formula obtained by substituting ₁ and A ₂ into the formula A (second configuration).

[0011]

[Formula 1] X = Lo + (A ₁ · D ₂ · Δl / Tp) + A ₂ −L ₁ (Formula A)

In this equation, X is the liquid level (the distance from the liquid level reference position of the object to be measured to the liquid surface), Lo is the set distance to the temperature change point, and L ₁ is the starting point of the optical fiber serving as the sensor from the object to be measured. To the liquid level reference position, D ₂ is the change feature amount, Δl is the pulse width, Tp is the temperature difference between the gas and the liquid, A
₁ and A ₂ are constants determined for each device used and are quantities depending on the pulse width and the sampling time width of the light receiving device.

By the way, what can be judged by the above-mentioned second structure is that the change of the liquid surface can be judged only within a constant width equal to or less than the distance resolution, and the actual liquid level after the change cannot be accurately detected. This is because the computer must determine the displacement reference point on the measurement data in order to automatically determine the liquid level, and the value calculated depends on how this reference point is set. is there. In the mathematical expression A, Lo is set as the reference point of displacement, and the liquid level is calculated by calculating how much displacement from this set distance is. Therefore, in order to accurately detect the liquid level, it is necessary to correct the error caused by the way of setting the reference point, and the method for this purpose is the third configuration described below. This consists of three steps of correction value calculation measurement for calculating an error (correction value) caused by how the reference point is set, in addition to the previous measurement and the main measurement.

In the third configuration, the procedure of the pre-measurement and the main measurement is the same as that of the second configuration, so the description thereof will be omitted. The procedure of the correction value calculation measurement and the handling of the correction value obtained thereby will be described below. .. This measurement can also be done in advance in a laboratory, etc. First, make a mark on the part where the length of the optical fiber that is the sensor part of the distributed optical fiber temperature measurement system is known, and from the optical fiber start end to that mark. The distance of is the measurement distance. Next, the optical fiber is immersed in water, the temperature is measured by aligning the mark with the water surface, and the measurement result and A ₁ and A ₂ obtained in the previous measurement are expressed by the mathematical formula A
The value calculated by the equation obtained by substituting into Then, the difference between the calculated distance and the measured distance is obtained and used as the correction value.

In this correction value calculation measurement, if there is no error in the way of setting the reference point (Lo in the above formula A), the liquid level (measured value) calculated by the second configuration and the actual liquid level Is supposed to agree. Then, the actual measurement is performed on an actual measurement object, and the measured value obtained as a result is corrected by the correction value to obtain an accurate liquid level. The correction value calculation measurement may be performed either between the previous measurement and the main measurement, or after the previous measurement and the main measurement.

First, the above-mentioned first structure will be specifically described with reference to FIG. In the figure, when the actual temperature distribution changes as shown in (A) along with the change in liquid level, the measured values at the sampling intervals shown in (B) are shown in (C). It is shown. In the description of the first configuration, for convenience, the temperature from the DTS device to the section n is substantially constant, and the distance from the DTS device to the liquid surface is the liquid level. Since the DTS displays the average temperature of each sampling interval, the temperature measurement value is extracted as the change from a to e in the section n in (C) corresponding to the change from a to e in (A).

Therefore, the change in liquid level and the sampling interval n
When the measured value in the interval n is g, the distance from the DTS device to the sampling interval n is L, the sampling interval is Δl, the average temperature to L is T ₁ , the sampling interval is Assuming that the average temperature from n + 1 to the end of the optical fiber is T ₂ , the liquid level X can be expressed by the following formula B in units of distance resolution or less.

[0018]

X = {Δl (g−T ₂ ) / (T ₁ −T ₂ )} + L (Formula B)

In such a method, L can be easily discriminated by visually observing it from the graph of temperature data. However, when the liquid level is automatically discriminated by a computer or the like, a stable portion before and after the change point is detected. It is difficult to detect the temperature (T ₁ and T _{2 in the} above example) and discriminate the position of the L (sampling interval n). Therefore, next, the process of deriving the mathematical formula A will be described for the method (second configuration) for automatically determining the liquid level.

(Assumption of Preconditions) The optical pulse of DTS is a rectangular pulse having a width W (s) and a height P, and the head of the pulse is the incident end of the optical fiber (distance x =
0), time t = 0. The attenuation of light in the optical fiber is zero. The speed of light in the optical fiber is represented by C = Co / n (Co: speed of light in vacuum, n: refractive index). The temperature distribution does not change with time. (For the speed of light,
Relatively)

Under the above assumption, the light pulse is at the position of x = Ct at time t as shown in FIG. 2A, and the backscattered light from the distance x at this time is x at time 2t. -(C
It reaches the incident end together with the backscattered light generated from t to t + (W / 2) in the range of W / 2). Here, when the backscattered light intensity at x is f (x) and the instantaneous signal intensity of the backscattered light returning to the incident end is I (t), the signal intensity at time 2τ is shown in FIG. As described above, the signal is from the distance l = C · τ to l−Δl = C (τ−W / 2). Therefore, the instantaneous signal strength I (t) is represented by Expression C, and the actually observed signal S (t) is the integration of this I (t) during the sampling time width ΔT. Is represented as

[0022]

[Equation 3]

[0023]

[Equation 4]

(Assumption of scattered light intensity f (x)) f (x) represents the temperature θ (x) at the distance x. That is, f (x)
= Θ (x) (Equation E).

(Assumption of Temperature Distribution) The temperature distribution of the object to be measured is assumed to be stepwise as shown in FIG. 3 (A). In the temperature distribution of FIG. 3A, the distance to the change point is L, the high temperature portion is A, and the low temperature portion is B. In this case, the observation amount S (T) is calculated by the following formula F from the formulas D and E.
It is expressed as.

[0026]

[Equation 5]

Here, substituting l = Ct / 2, each condition is expressed as in Equation G, and the observed amount S (T) is in Equation H.
become that way. Further, when this is divided by T, it becomes as in Formula I. Then, when integration is performed for each of these, the results are arranged and standardized by Δl · ΔT, the finally obtained measured temperature S ′ (T) is represented by the mathematical expression J.

[0028]

[Equation 6]

[0029]

[Equation 7]

[0030]

[Equation 8]

[0031]

[Equation 9]

[0032]

[Equation 10]

(Study of Liquid Level Detection) How the measured temperature distribution changes when the change point L of the stepwise temperature distribution changes within a range smaller than the distance response determined by the sampling interval and the pulse width. Think Now, assuming that sampling is performed with a sampling time width ΔT from t = 0, the obtained measured amount is S ′ (iΔT), i is a discrete amount of 0, 1, 2, 3, ... In addition, since the temperature distribution changes stepwise, T + ΔT <2L / C, 2
In the range of (L + Δl) / C ≦ T, S ′ (T) has a constant value. Then, in the vicinity of the change point, a change appears in S ′ (T), but this aspect is as shown in FIG.
It can be classified into two patterns. CASE1: When at least one sampling interval is included in the width of change determined by the pulse width. CASE2: When one sampling interval extends inside and outside the width of change determined by the pulse width. CASE3: When one sampling interval is wider than the width of change determined by the pulse width and crosses the width of change.

(Calculation of Difference D ₁ ) Here, T ≦ 2L / C
When S ′ (T) that satisfies <T + ΔT is S ′ (To), the distribution of each measured value is as shown in FIG. Next, a change in the measured value distribution when the temperature change point L changes by α · ΔT (0 ≦ α <1) smaller than the sampling time width ΔT is derived for each CASE. As a method therefor, a difference D ₁ between adjacent measurement values is obtained in order to extract a change point which is a reference for a change in calculation, and a minute change time α is obtained from these.
・ Determine ΔT. Hereinafter, the difference D ₁ [D ₁ (T) = S ′ (T) −S ′ (T + ΔT)] is obtained for each CASE. The result of CASE 1 is shown in equation K, the result of CASE 2 is shown in equation L, the result of CASE 3 is shown in equation M, and the difference D ₁
Fig. 5 shows a graph of the distribution of.

[0035]

[Equation 11]

[0036]

[Equation 12]

[0037]

[Equation 13]

(Determination of change characteristic amount D ₂ ) Since a stepwise temperature distribution is assumed, at least one non-zero D ₁ (T) exists, as shown in FIG.
From this result, the change feature amount D ₂ is determined. D ₂ quantitatively indicates the change of the measured value S ′ (T), that is, the state of D ₁ (T), and there are several methods for deciding the method. If non-zero D ₁ (T) is one, when D ₁ is not a D ₁ = D ₂ 0 (T) is plural, and D ₂ the largest of the D ₁ (T). If there are multiple non-zero D ₁ (T), then D ₁ and T with the smallest T
Let D ₂ be the difference from the maximum D ₁ . If there are multiple non-zero D ₁ (T), then D ₁ with the smallest T is D
_Set to ₂ .

(Calculation of minute change amount d) The minute change amount d is obtained from D ₂ described above. CL / 2 as the reference point
Let us assume that the point at which = To is Lo (CLo / 2 = To), and the change point changes by Lo + d. In addition, d = α
CΔT / 2 (0 ≦ α <1). Here, according to the method in (Determination of change characteristic amount D ₂ ), the higher order term of d can be eliminated from the formula of D ₁ to obtain d, and only the first order term can be eliminated. That is, focusing on D ₁ of each CASE, the quadratic term of d, that is, To- (2Lo / C), is canceled by taking the difference between the one with the largest T and the one with the smallest T out of D ₁ that is not 0. I understand. Therefore, with this difference as D ₂ , d is obtained for each CASE. The calculation procedure of CASE 1 to 3 is shown in Equations N to P.

[0040]

[Equation 14]

[0041]

[Equation 15]

[0042]

[Equation 16]

(Summary) If the minute change amount d is obtained for each CASE and the distance from the optical fiber start end to the liquid level reference position of the object to be measured is L ₁ , the liquid level X is Lo.
It is represented by + d−L ₁ . By the way, when the temperature is actually measured by the DTS, the incident light pulse has a half width of about 1 or less.
The pulse width C is slightly different from the assumption described above due to the fact that the waveform has a skirt spread of 1 to 2 m in m.
W / 2 = Δl, temperature difference AB = Tp, and each CASE
Summarizing (numerical formulas N, O, and P), the mathematical formula A is approximately derived.

[0044]

X = Lo + (A ₁ · D ₂ · Δl / Tp) + A ₂ −L ₁ (Formula A)

In the actual measurement, first, the pre-measurement is performed to obtain the constants A ₁ and A _2, and then the A ₁ and A ₂ are substituted into the mathematical expression A to perform the main measurement. The liquid level of the object to be measured can be obtained by performing.

With such a method, it is possible to detect a change in liquid level in units of distance resolution or less. In addition, the above Lo
Has already described that there is an error in the way of setting itself, but means for correcting this (correction value calculation measurement) is as follows. A mark is made in advance at a location where the length of the optical fiber is known by a major measurement or the like (for example, 30 m from the device [optical fiber starting end]), and the optical fiber is dipped in a tub filled with water to make the above-mentioned mark on the water surface. Measure the points together. Here, if there is no error in the setting of Lo, the value obtained by the automatic determination should show 30 m. However, if there is an error due to the setting method, for example, if it is calculated as 30.5 m, the Lo
Since the setting error of 1 is known to be 0.5 m, it is possible to perform measurement with higher position detection accuracy by correcting 0.5 m after performing actual measurement (third configuration).

[0047]

EXAMPLES Examples of the second constitution of the present invention will be described below. The system configuration used for the measurement is, as shown in FIG. 6, a DTS device 1 (Sumithermo SUT-100 "product name"), a computer 2 for performing arithmetic processing, and a DTS device 1.
It is composed of an optical fiber 3 which is connected to the sensor and serves as a sensor, and an electric heating device 4 which heats the optical fiber. The prepared optical fiber 3 has a length of about 100 m and is inserted into the stainless steel tube 7 provided with an insulating coating layer as described in the above-mentioned document 1, and the optical fiber 3 is passed through the electric heating device 4 and the electric head 5. Is heated uniformly in the longitudinal direction. This is to detect the temperature difference between the liquid and the gas more clearly by utilizing the fact that the degree of cooling of the heated optical fiber differs between the liquid and the air. In addition, in the case of Sumithermo SUT-100,
Although W = 10 ns and ΔT = 8 ns, it is considered that W> 10 ns because the pulse is not substantially rectangular as in the above assumption and has a shape with widened tails, which is almost the same as CASE.
It seems to correspond to 1.

First, the pre-measurement is performed by such a system. The mark 6 is attached to an arbitrary position of the optical fiber 3 in advance, the optical fiber 3 is left in the air, and when the temperature becomes stable, the optical fiber 3 is immersed in water as shown in FIG. Adjust 6 to the liquid level. Then, the optical fiber 3
Immerse each at 0 cm and measure the temperature at each liquid level.
Sampling interval Δl: 0.8m, Air temperature: 55 ° C,
The temperature in water: 28 ° C, and the temperature difference Tp was 27 ° C. still,
For convenience of explanation, the value of L ₁ is 0 this time, but the temperature data obtained in the actual measurement is represented as shown in FIG. 7. Therefore, to perform automatic determination of the liquid level, to release the data in the corresponding data and L ₃ since not relevant to the measurement of up to L ₁ in the liquid level reference position from processing, pre-L ₁ and L ₃ Just measure the distance to. In the actual main measurement,
Of course, L ₃ can be the end of the optical fiber.
The measured value when the mark was positioned on the water surface was set to 0, and thereafter 10 cm each was immersed in water in order 10, 20,
Table 1 shows the measurement results for each sampling interval as 30 ...

[0049]

[Table 1]

The difference D ₁ is obtained from this measurement result, but the actual measured value is about ± 1 due to the influence of noise and the like.
Since the error of 0 ° C. is included, 2 ° C. is set as the error width E and the temperature difference beyond this is extracted as a significant difference. The error width E is one of the factors that determine the measurement accuracy, and the ratio of the temperature difference Tp and the error width E determines the measurement accuracy. For example, if the temperature difference is 20 ° C. and the error width is 2 ° C. with a measuring instrument having a distance resolution of 1 m, the liquid level change can be measured with an accuracy of 1/10 of the distance resolution, that is, in units of 10 cm. Therefore, in order to improve the measurement accuracy, the temperature difference Tp may be made larger by the heating means described in Document 1. Of the plurality of D ₁ (x) extracted from the results in Table 1, the smallest one of x is the change feature amount D _2, and x at that time is Lo, d, D
₂ shows each data of Lo and Lo.

[0051]

[Table 2]

Here, in the formula A, Δl = 0.8, Tp
= 27 was substituted, and the constants A ₁ and A ₂ were obtained from this and the results in Table 2 by the method of least squares. As a result, A ₁ =-
It was found that 0.067 and A ₂ = −0.19 should be set. Originally, the actual measurement is performed using the formula obtained by substituting A ₁ and A ₂ into the formula, but this time, for simplicity, this formula was used to calculate the liquid level at each immersion stage It was confirmed whether or not the measurement could be performed in units of 10 cm. When this measurement is performed, the temperature difference Tp is determined by averaging each data within the error width E from L ₁ toward the end of the optical fiber and each data within the error width E from L ₃ toward the optical fiber start end. This is done by finding the difference from the average value. Liquid level X is L
It is represented by o + d−L ₁ , but as described above, L ₁ = 0.
And the liquid level X calculated from this equation is shown in the bottom row of Table 2 above.

As can be seen from the result of this calculation, there is a portion (data with an asterisk) in which there is a slight error because the decimal point is taken as one digit, but it is finer than the distance resolution of 0.8 m of this device and 10 cm. It was confirmed that the change in liquid level could be measured in units. The data when d = 0, 74.1 will be described as an example. The liquid level X obtained here coincides with the actual liquid level (in this case, from the optical fiber start end to the marked position). Not necessarily. This is because Lo, which is the reference of displacement, is set in the first place, and the distance from the optical fiber start end to the marked position is 74.1.
The difference from m is the correction value.

[0054]

As described above, according to the method of the present invention, it is possible to detect the liquid level in a unit finer than the distance resolution of the conventional DTS device. Therefore,
For example, it can be effectively used for liquid level detection in wells, boring holes, etc., which require measurement at intervals of about 10 cm.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of the basic principle of the method of the present invention, (A) shows an actual temperature distribution, (B) shows a sampling interval, and (C) shows a distribution of measured values.

FIG. 2 is an explanatory diagram showing a traveling state of an optical pulse.

FIG. 3A is a graph showing a stepwise temperature distribution,
FIG. 6B is an explanatory diagram of case classification caused by the relationship between the pulse width and the sampling interval.

FIG. 4 is a graph showing a temperature distribution in each case of FIG.

FIG. 5 is a graph showing a difference distribution.

FIG. 6 is an explanatory diagram of an example of the present invention.

FIG. 7 is a graph showing temperature data obtained by measurement in Examples.

FIG. 8 is an explanatory diagram showing a measurement state in a conventional DTS device.

[Explanation of symbols]

 1 DTS device 2 Computer 3 Optical fiber 4 Electric heating device 5 Electric head 6 Mark 7 Stainless tube

Claims (3)

[Claims] 1. A liquid represented on the temperature distribution data obtained by immersing an optical fiber, which is a sensor portion of a distributed optical fiber temperature measuring system, in a liquid, measuring the temperature distribution in the longitudinal direction of the optical fiber. Since the temperature measurement value near the boundary surface of the gas and the gas changes corresponding to the change in the liquid level, the correlation between the liquid level change and the temperature measurement value near the boundary surface is obtained, and the liquid level is determined based on this correlation. A liquid level measuring method characterized by detecting. 2. As a pre-measurement, a mark is made on an arbitrary position of an optical fiber which is a sensor part of a distributed optical fiber temperature measuring system, the optical fiber is immersed in water to align the mark with the water surface, and a distance resolution is sequentially obtained. The temperature is measured by immersing or pulling up each of the following fixed lengths, and the constants A ₁ and A ₂ of the following equations are obtained in advance from this temperature data. As the main measurement, the optical fiber is immersed in the liquid of the measurement object. A liquid level measuring method characterized by immersing and measuring the temperature distribution in the longitudinal direction of the optical fiber, and obtaining the liquid level with accuracy less than the distance resolution by the measurement result and the formula obtained in the previous measurement. X = Lo + (A ₁ · D ₂ · Δl / Tp) + A ₂ −L _{1 In} this formula, X is the liquid level (distance from the liquid level reference position of the measurement object to the liquid surface), Lo is the temperature change point. Set distance, L ₁ is the distance from the optical fiber starting end serving as a sensor to the liquid level reference position of the object to be measured, D ₂ is the change feature amount, Δl is the pulse width, Tp is the temperature difference between gas and liquid, A ₁ , A ₂ is a constant that is determined for each device used and is a quantity that depends on the pulse width and the sampling time width of the light receiving device. 3. As a pre-measurement, a mark is made on an arbitrary position of the optical fiber which is a sensor part of the distributed optical fiber temperature measuring system, the optical fiber is immersed in water to align the mark with the water surface, and the distance resolution is sequentially obtained. The temperature is measured by immersing or pulling up each of the following constant lengths, and the constants A ₁ and A ₂ of the following equations are obtained in advance from this temperature data, and the length of the optical fiber is known as a correction value calculation measurement. Mark the location where the mark is, and if the length from the optical fiber start end to the mark is the measurement distance, immerse the optical fiber in water, measure the temperature by aligning the mark with the water surface, and The value calculated by the formula obtained by substituting A ₁ and A ₂ obtained in the previous measurement into the following equation is used as the calculation distance, and the difference between the calculation distance and the measurement distance is corrected, and the main measurement is performed. , The optical fiber to be measured Immersed in the liquid, by measuring the temperature distribution in the longitudinal direction of the optical fiber, the measurement obtained by the measurement result and equation of A _1, A ₂ obtained was obtained by substituting the following equation in the previous measurement A liquid level measuring method, characterized in that the liquid level is detected by correcting the value by the correction value. X = Lo + (A ₁ · D ₂ · Δl / Tp) + A ₂ −L _{1 In} this formula, X is the liquid level (distance from the liquid level reference position of the measurement object to the liquid surface), and Lo is the temperature change point. Set distance, L ₁ is the distance from the optical fiber starting end serving as a sensor to the liquid level reference position of the object to be measured, D ₂ is the change feature amount, Δl is the pulse width, Tp is the temperature difference between gas and liquid, A ₁ , A ₂ is a constant that is determined for each device used and is a quantity that depends on the pulse width and the sampling time width of the light receiving device.