A method for detecting the amount of water produced in the oxidation process of insulating oil

Abstract

The present application relates to a kind of detection methods of oxidation product water quantity in the oxidation deterioration process of insulating oil, belong to the technical field of testing and analyzing oil by determining the chemical properties of materials in measurement.The method first divides the pretreated insulating oil to be measured into multiple containers, then divides the multiple containers into two groups, the first group is connected with oxygen, the second group is connected with dry inert gas, and then aging is carried out;Then divide multiple time points, calculate the oxidation product water quantity estimate value and correction coefficient at each time point when aging is carried out to each time point;Finally, the oxidation product water quantity actual value of the insulating oil to be measured at the t time point is calculated according to the oxidation product water quantity estimate value and correction coefficient.The present application can accurately distinguish total water content and oxidation deterioration product water content, and accurately detect the oxidation product water content in the deterioration process of insulating oil.

Description

Technical Field

[0001] This invention relates to a method for detecting the amount of water produced during the oxidation and deterioration of insulating oil, belonging to the technical field of testing and analyzing oil by means of measuring the chemical properties of materials. Background Technology

[0002] Insulating oil is a crucial liquid insulation and cooling medium in power equipment such as oil-immersed transformers and reactors, and its performance directly affects the insulation reliability and operational safety of the equipment. Moisture is one of the important factors affecting its performance. Increased moisture content in the oil reduces the breakdown strength of the insulating oil, promotes the aging of insulating paper, and accelerates the deterioration of the insulation system. Therefore, understanding the changes in moisture content in insulating oil is essential for equipment operation and maintenance. Among these, accurately distinguishing and measuring the amount of water produced solely by the oxidation reaction of the insulating oil is a core prerequisite for understanding the changes in moisture content.

[0003] However, existing methods for detecting moisture in insulating oil typically focus on determining the total water content of the oil sample, such as the Karl Fischer method, humidity sensor method, or online micro-moisture monitoring method. While these methods can obtain the overall moisture content of the oil, in actual operation or accelerated aging processes, the moisture in the oil may originate from various factors, such as initial residual moisture, moisture absorption from the external air, moisture migration from the oil-paper insulation system, or even from the oxidative degradation reaction of the insulating oil itself. If only the total water content is measured, it is difficult to determine whether the increase in moisture is caused by the oxidative degradation process of the oil, and therefore difficult to accurately characterize the dynamic law of moisture formation during the oxidative degradation process of insulating oil. Furthermore, differences in oxygen content affect the oxidation reaction rate and degradation pathway of the insulating oil, thus affecting the formation, consumption, and migration processes of moisture, further increasing the evaluation error of the total water content index. Summary of the Invention

[0004] The technical problem to be solved by this invention is: how to accurately detect the amount of water produced by oxidation during the deterioration of insulating oil.

[0005] The technical solution proposed by this invention to solve the above-mentioned technical problems is: a method for detecting the amount of oxidation-induced water production during the oxidative deterioration process of insulating oil, comprising the following steps:

[0006] Step 1: Pre-treat the insulating oil to be tested. Divide the pre-treated insulating oil into N equal parts, each with an initial mass of m. Then, put the N parts of the insulating oil into N containers with a volume of V.

[0007] Step 2: Divide the N containers into two groups. Introduce oxygen at concentration y into the first group of containers, and introduce dry, inert gas into the second group of containers, at the aging temperature. Heat N containers to age them;

[0008] Set the interval x for detecting the water content in the container during the aging process. Detect the water content in the container once every interval x. Record the time of the first detection as time 1, the time of the second detection as time 2, and so on up to time T.

[0009] Step 3: When aging reaches time t, select one container from the first group of containers as the first oxygen-containing container, and select one container from the second group of containers as the first oxygen-free container; where t = 1, 2, ..., T;

[0010] Detect the first water content inside the first oxygen-containing container and the second water content in the first oxygen-free container. The first water content and the second moisture content The difference is used as an estimate of the oxidation water production of the insulating oil at time t. ;

[0011] Step 4: Calculate the correction coefficient at time t;

[0012] Step 5: Estimate the oxidation water production of the insulating oil at time t based on the correction coefficient at time t. After correction, the actual value of the oxidation water production of the insulating oil under test at time t is obtained. .

[0013] Furthermore, after step 5, step 6 is executed: the actual value of the oxidation water production of the insulating oil at time T is obtained. Substituting into equation (6), the limiting oxidation water production of the insulating oil under oxygen concentration y is calculated. C max and oxidation water production rate constant k4

[0014] (6).

[0015] Furthermore, in step 2, the oxygen concentration y is adjusted to other oxygen concentrations to simulate different leakage conditions, and steps 1-5 are repeated to obtain the actual value of the oxidation water production of the insulating oil under different leakage conditions at time t.

[0016] Furthermore, the pretreatment in step 1 is vacuum drying and degassing, with the vacuum gauge pressure set to ≤−0.1 MPa, the drying temperature to 35-65℃, and the drying time to 24-72 h.

[0017] Furthermore, in step 1, during the pretreatment process, the moisture content of the insulating oil to be tested is detected every time interval L. When the following formula (7) is satisfied, the pretreatment is stopped.

[0018] (7);

[0019] In formula (7), S1 and S2 are the moisture content of the insulating oil to be tested obtained from two adjacent tests, and S is the average moisture content of the insulating oil to be tested obtained from the two adjacent tests.

[0020] Furthermore, the dry inert gas in step 2 is nitrogen.

[0021] Furthermore, in step 3, when t=1, one container from the first group of containers is randomly selected as the first oxygen-containing container, and one container from the second group of containers is randomly selected as the first oxygen-free container; when t>1, the first oxygen-containing container and the first oxygen-free container selected at time t are controlled to be different from the containers selected at all times before time t.

[0022] Furthermore, the correction coefficients in step 4 include a first correction coefficient f1, a second correction coefficient f2, a third correction coefficient f3, and a fourth correction coefficient f4, which are specifically calculated through the following steps:

[0023] Step 4.1: Calculate the first correction coefficient f1 according to the following formula (1).

[0024] (1);

[0025] In equation (1), K is the moisture vapor-liquid partition correction coefficient; K ref It is the first water vapor-liquid partition coefficient at a reference thermodynamic temperature, which is 298.15 K; R is the molar enthalpy of solubility of water in the insulating oil to be tested in the first oxygen-containing container; R is the universal gas constant, taken as 8.314 J / (mol・K); It is the actual thermodynamic temperature inside the first oxygen-containing container; ρ is the reference thermodynamic temperature; ρ is the density of the insulating oil to be tested in the first oxygen-containing container; r1 is the ratio of the remaining mass of the insulating oil to be tested in the first container to the initial mass m.

[0026] Step 4.2: Calculate the second correction coefficient f2 according to the following formula (2).

[0027] (2);

[0028] In equation (2), k1 is the oxygen consumption rate constant; It is the apparent rate constant of oxygen consumption at the reference temperature; is the apparent activation energy of oxygen consumption in the oxidation reaction of the insulating oil to be tested; k2 is the apparent decay rate constant of dissolved oxygen; r2 is the volume ratio of the gas phase to the oil phase in the first oxygen-containing container;

[0029] Step 4.3: Calculate the third correction coefficient f3 according to the following formula (3).

[0030] (3);

[0031] In equation (3), E is the apparent activation energy of the oxidation reaction of the insulating oil to be tested; It is the average actual thermodynamic temperature inside the first oxygen-containing container; and These are the aging temperatures. The second moisture gas-liquid partition coefficient and the first oxygen gas-liquid partition coefficient; and These are the actual thermodynamic temperatures. The third water vapor-liquid partition coefficient and the second oxygen vapor-liquid partition coefficient; γ is a kinetic influencing factor of oxygen partitioning;

[0032] Step 4.4: Calculate the fourth correction factor f4 using the following formula (4).

[0033] (4);

[0034] In equation (4), k3 is the evaporation rate correction constant; It is the evaporation rate constant at the reference temperature; It is the apparent activation energy of evaporation of the light component of the insulating oil to be tested.

[0035] Furthermore, in step 5, the estimated value of the oxidation water production is obtained by formula (5). After correction, the actual value of the oxidation water production of the insulating oil in the first oxygen-containing container at time t is obtained.

[0036] (5);

[0037] In equation (5), F is the coupling correction coefficient; α is the first influence coefficient, 0.15≤α≤3; β is the second influence coefficient, 0.08≤β≤0.2.

[0038] The beneficial effects of this invention are as follows: By forming a control group with oxygen-containing and oxygen-free containers, the initial moisture content, background moisture from thermal effects, and system interferences are subtracted, thereby obtaining the true oxidation water production amount related to the oxidation degradation reaction. Simultaneously, for four types of systematic errors—gas-liquid distribution, dynamic oxygen consumption, temperature fluctuations, and oil phase evaporation—a partial correction formula and a second-order all-factor coupled correction model are established. The first step uses four partial coefficients to complete the independent benchmark correction for each individual error source. The second step introduces a multi-factor interactive coupling correction coefficient F to specifically correct the nonlinear interactive deviations caused by global temperature control and the evaporation-gas-liquid distribution-oxygen mass transfer chain reaction, ultimately obtaining the accurate actual value of oxidation water production during the oxidative degradation of insulating oil. Compared with existing insulating oil moisture detection methods, this invention can accurately distinguish between total moisture content and oxidation water production, and accurately detect the oxidation water production amount during the insulating oil degradation process. Detailed Implementation

[0039] Example 1

[0040] This embodiment of a method for detecting the amount of water produced during the oxidative deterioration of insulating oil includes the following steps:

[0041] Step 1: Pre-treat the insulating oil to be tested. Divide the pre-treated insulating oil into N equal parts, each with an initial mass of m. Place the N parts of the insulating oil into N sealed containers, each with a volume of V; N is a positive even number.

[0042] Pretreatment consisted of vacuum drying and degassing to reduce the initial moisture content of the insulating oil being tested, ensuring that all subsequent experimental groups were in the same or essentially the same initial state. The vacuum gauge pressure was set to ≤−0.1 MPa, the drying temperature to 60℃, and the drying time to 60 h.

[0043] During the pretreatment process, the moisture content of the insulating oil to be tested is checked every L time interval. When the following formula (7) is satisfied, the pretreatment is judged to be completed and the pretreatment is stopped.

[0044] (7);

[0045] In equation (7), S1 and S2 are the moisture content of the insulating oil to be tested obtained from two consecutive tests, and S is the average moisture content of the insulating oil to be tested obtained from two consecutive tests.

[0046] Step 2: Divide the N containers into two groups. Introduce oxygen at a concentration of y into the first group of containers, and introduce a dry, inert gas, nitrogen, into the second group of containers. This will characterize the effects of heat, initial moisture release, container background, and moisture changes caused by non-oxidizing factors. The aging temperature will be used as the aging temperature. Heat N containers to age them;

[0047] Set the interval x for detecting the water content in the container during the aging process. Detect the water content in the container once every interval x. Record the time of the first detection as time 1, the time of the second detection as time 2, and so on up to time T.

[0048] Step 3: When aging reaches time t, select one container from the first group as the first oxygen-containing container, and select one container from the second group as the first oxygen-free container; t=1,2,...,T;

[0049] Detect the first water content in the first oxygen-containing container and the second water content in the first anaerobic container The first water content Second moisture content The difference is used as an estimate of the oxidation water production of the insulating oil at time t. That is, by subtracting the water content of the first oxygen-containing container at time t from the water content of the first oxygen-free container under the same aging time, the amount of additional water generated by the oxidative degradation reaction of the insulating oil under the oxygen-containing conditions is obtained.

[0050] In this embodiment, a one-time selection method after parallel aging is adopted, meaning that each container in the aging process is selected only once at time t and is not used for subsequent aging. This method avoids problems such as changes in oil sample quality, gas phase space, oxygen content, introduction of external moisture, and changes in the oil-gas ratio caused by repeated sampling of the same container. A sealed sampling valve is preferably used during the moisture content detection process to reduce the contact between the insulating oil under test and the outside air. Specifically, when t=1, one container from the first group can be randomly selected as the first oxygen-containing container, and one container from the second group can be randomly selected as the first oxygen-free container; when t>1, the first oxygen-containing container and the first oxygen-free container selected at time t are controlled to be different from the containers selected at all times before time t.

[0051] Step 4: Calculate the correction coefficient at time t;

[0052] The correction factors include a first correction factor f1, a second correction factor f2, a third correction factor f3, and a fourth correction factor f4, which are calculated through the following steps:

[0053] Step 4.1: In a sealed container, the water molecules generated by oxidation will reach thermodynamic equilibrium between the oil phase and the gas phase. Detecting only the water content in the oil phase will ignore the water distributed in the gas phase space, leading to an underestimation of the total water production. Conventional correction methods usually treat the water vapor-liquid partition coefficient and the oil-gas ratio as fixed constants, without considering the regulatory effect of temperature on the water dissolution equilibrium and the dynamic change in the proportion of the gas phase space due to oil evaporation. This easily introduces additional errors in variable temperature and long-term aging scenarios. In order to reflect the exponential change of the water vapor-liquid partition coefficient (oil phase equilibrium concentration / gas phase equilibrium concentration) in insulating oil with thermodynamic temperature, the first correction coefficient f1 is calculated according to the following formula (1).

[0054] (1);

[0055] In equation (1), K is the moisture vapor-liquid partition correction coefficient; K ref It is the first water vapor-liquid partition coefficient of the insulating oil to be tested in the first oxygen-containing container at the reference thermodynamic temperature. It belongs to the inherent physical property parameter of the oil and can be obtained by combining the static gas-liquid equilibrium method with the Karl Fischer moisture determination experiment. The reference thermodynamic temperature is taken as 298.15 K. It is the molar enthalpy of water solubility in the insulating oil to be tested in the first oxygen-containing container, with the unit J / mol. It characterizes the degree of influence of temperature on the water dissolution equilibrium and can be obtained by fitting the experiment of multi-temperature point distribution coefficient determination; R is the universal gas constant, which is taken as 8.314 J / (mol・K). It is the actual thermodynamic temperature inside the first oxygen-containing container; The reference thermodynamic temperature is 298.15 K; ρ is the density of the insulating oil to be tested in the first oxygen-containing container; r1 is the ratio of the remaining mass of the insulating oil to be tested in the first container to the initial mass m.

[0056] The first correction coefficient f1 is essentially a gas-liquid two-phase distribution correction coefficient. In equation (1), the value 1 corresponds to the total water content of the oil phase itself, and the second term is the conversion ratio of the water content distributed in the gas phase space relative to the oil phase. Together, they constitute the conversion coefficient between the total water content and the water content of the oil phase, which can convert the measured water content of the oil phase into the total water content in the sealed container. By coupling the temperature effect and dynamic volume change, dynamic correction of water gas-liquid distribution under different temperatures and different degrees of evaporation is achieved.

[0057] Step 4.2: The oxidation reaction continuously consumes oxygen in the container, and the oxygen concentration decreases exponentially with aging time, rather than being a constant initial value. The decrease in oxygen concentration will lead to a gradual decrease in the oxidation water production rate, and the apparent water production will be lower than the theoretical water production under constant oxygen partial pressure. The dissolved oxygen concentration is jointly regulated by the oxygen gas-liquid distribution and the oil-gas ratio, and the rate constant itself has a significant temperature dependence. Therefore, the second correction coefficient f2 is calculated according to the following formula (2).

[0058] (2);

[0059] In equation (2), k1 is the oxygen consumption rate constant; It is the apparent rate constant of oxygen consumption at the reference temperature, in weeks. -1 It can be calibrated through oxygen concentration decay experiments at different temperatures; is the apparent activation energy of oxygen consumption in the oxidation reaction of the insulating oil under test, expressed in J / mol, characterizing the sensitivity of the oxygen consumption rate to temperature, and can be obtained through fitting experiments with multiple temperature gradients; k2 is the apparent decay rate constant of dissolved oxygen, expressed in cycles. -1 r2 is the volume ratio of the gas phase to the oil phase in the first oxygen-containing container;

[0060] The second correction coefficient f2 is essentially a dynamic correction coefficient for oxygen consumption. Equation (2) reflects the mass transfer-reaction coupling law that the larger the oil-gas ratio, the smaller the oxygen distribution coefficient, the lower the dissolved oxygen ratio, and the slower the apparent reaction rate. Then, by performing integral averaging correction on the exponentially decaying oxygen concentration, the measured water production under the dynamically decaying oxygen concentration is converted into the standard water production under the equivalent initial constant oxygen partial pressure, so as to achieve horizontal comparability of data at different time points and under different operating conditions.

[0061] Step 4.3: The oxidation reaction rate is highly sensitive to temperature. Even a small deviation between the actual oil temperature and the aging temperature inside the container can lead to an exponential systematic deviation in the water production. Conventional temperature correction only adjusts the oxidation reaction rate and does not consider the indirect impact of temperature on the quantitative water production by altering the moisture and oxygen vapor-liquid distribution balances, thus limiting the scope of correction. Therefore, this embodiment integrates the triple regulatory effects of temperature on the oxidation reaction rate, moisture vapor-liquid distribution, and oxygen vapor-liquid distribution, constructing a third correction coefficient f3 with multiple coupled effects.

[0062] The third correction factor f3 is calculated using the following formula (3).

[0063] (3);

[0064] In equation (3), E is the apparent activation energy of the oxidation reaction of the insulating oil to be tested, and the unit is J / mol; It is the average actual thermodynamic temperature inside the first oxygen-containing container, which can be obtained in real time by pre-embedding temperature sensors inside the container; and These are the aging temperatures. The second moisture gas-liquid partition coefficient and the first oxygen gas-liquid partition coefficient; and These are the actual thermodynamic temperatures. The third water vapor-liquid partition coefficient and the second oxygen vapor-liquid partition coefficient; γ is the kinetic influencing factor of oxygen distribution, which characterizes the order of influence of oxygen distribution balance on the oxidation reaction rate. For the first-order free radical oxidation reaction of the insulating oil to be tested, the water production rate is proportional to the dissolved oxygen concentration, so γ is taken as 1. It can be verified by multi-temperature and multi-oil-gas ratio experiments.

[0065] The third correction factor f3 is essentially a temperature uniformity correction factor. By fully correcting the water production at the measured temperature to the baseline water production at the aging temperature, it completely eliminates the multi-dimensional systematic errors caused by temperature fluctuations between batches.

[0066] Step 4.4: During the aging process, a small amount of low-boiling-point light components in the insulating oil to be tested will evaporate into the gas phase space, resulting in a decrease in the total mass of the oil phase. Since the water content is the mass concentration, the decrease in the oil phase mass will lead to an overestimation of the apparent concentration. Therefore, the fourth correction coefficient f4 is calculated according to the following formula (4).

[0067] (4);

[0068] In equation (4), k3 is the evaporation rate correction constant; It is the evaporation rate constant at the reference temperature, with units of cycles. -1 It can be calibrated through evaporation weight loss experiments at different temperatures; It is the apparent activation energy of evaporation of the light component of the insulating oil to be tested, expressed in J / mol. It characterizes the sensitivity of the evaporation rate to temperature and can be obtained by fitting through multi-temperature gradient experiments.

[0069] The fourth correction factor f4 is essentially the correction factor for oil phase evaporation loss. As the core input of the dynamic oil-gas ratio, the fourth correction factor f4 is embedded in the calculation process of the gas-liquid two-phase distribution correction factor, forming a closed-loop coupling relationship between evaporation and gas-liquid distribution.

[0070] Step 5: Based on the four correction coefficients at time t, estimate the oxidation water production of the insulating oil at time t using the following formula (5). After correction, the actual value of the oxidation water production of the insulating oil under test at time t is obtained.

[0071] (5);

[0072] In equation (5), F is the coupling correction coefficient; in the formula for the coupling correction coefficient F, the first term, constant 1, corresponds to the baseline state without interaction effect; the second term is the temperature-evaporation interaction correction term, where α is the first influence coefficient of the temperature-evaporation interaction, characterizing the degree of influence of the synergistic effect of temperature deviation and evaporation loss on the quantitative result of water production, that is, the greater the temperature deviation from the set value, the more significant the evaporation loss, the greater the dynamic deviation of the oil-gas ratio from the preset value of single-factor correction, and the larger the correction amount of this term, which is used to correct the additional error caused by the chain reaction of temperature and evaporation; the third term is the gas-liquid distribution-oxygen consumption interaction correction term, where, β is the second influence coefficient of the gas-liquid distribution-oxygen consumption interaction, which characterizes the degree of mass transfer coupling influence between water vapor-liquid distribution and oxygen consumption kinetics. That is, the larger the gas-liquid distribution correction and oxygen kinetics correction, the further the oil-gas ratio of the container deviates from the standard state, and the more significant the coupling deviation of the two-phase mass transfer process of water and oxygen. The larger this correction is, the more it is used to correct the additional error caused by the coupling of the mass transfer process. α and β are both inherent physical property correlation parameters of the corresponding insulating oil, which can be obtained by fitting through orthogonal calibration experiments with multiple temperatures and multiple oil-gas ratios. The value range of α is 0.15≤α≤3, and the value range of β is 0.08≤β≤0.2.

[0073] Understandably, the four influencing factors—gas-liquid distribution, oxygen consumption, temperature deviation, and evaporation loss—are not entirely independent. Temperature, as a global control variable, simultaneously alters the moisture-gas-liquid distribution balance, the oxygen-gas-liquid distribution balance, the oxidation reaction rate, and the oil phase evaporation rate. Oil phase evaporation changes the oil-gas ratio within the container, thus simultaneously affecting the two-phase distribution and mass transfer process of moisture and oxygen. When multiple factors act together, the total error is not a linear summation of the errors of each individual factor. If a conventional simple multiplication method is used, the nonlinear interactive coupling effect will be ignored, introducing an additional 3% to 5% systematic error in scenarios such as high-precision quantification, high-temperature aging, and cross-batch comparisons.

[0074] Therefore, based on the four correction coefficients mentioned above, this embodiment constructs a second-order all-factor coupled correction model that combines single-factor independent benchmark correction with multi-factor interactive deviation correction. The first step involves independent benchmark correction of each individual error source using the four sub-term coefficients. The second step introduces a multi-factor interactive coupling correction coefficient F to specifically correct the nonlinear interactive deviations caused by global temperature control and the evaporation-gas-liquid distribution-oxygen mass transfer chain reaction, ultimately obtaining the actual value of the oxidation water production during the oxidative degradation process of the insulating oil.

[0075] Step 6: Calculate the actual value of water production from the oxidation of the insulating oil at time T. Substituting into equation (6), the limiting oxidation water production of the insulating oil under oxygen concentration y is calculated. C max and oxidation water production rate constant k4

[0076] (6);

[0077] Water production through limiting oxidation The oxidation water production rate constant k4 is used to quantitatively characterize the rate of oil oxidation and deterioration.

[0078] Example 2

[0079] The method for detecting the amount of water produced by oxidation during the oxidation and deterioration process of insulating oil in this embodiment is basically the same as that in embodiment 1. The difference is that the oxygen concentration introduced into the first group of containers is different. By adjusting the oxygen concentration y in step 2 to other oxygen concentrations, the oxidation and deterioration process under different leakage conditions can be simulated. Then, by repeating steps 1-5, the actual value of the amount of water produced by oxidation of the insulating oil under different leakage conditions at time t can be obtained.

[0080] The above description is only a preferred embodiment of the present invention, but the present invention is not limited thereto. All equivalent substitutions or modifications made to the concept and technical solutions of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for detecting the amount of water produced during the oxidative deterioration of insulating oil, characterized in that... The steps include the following: Step 1: Pre-treat the insulating oil to be tested. Divide the pre-treated insulating oil into N equal parts, each with an initial mass of m. Then, put the N parts of the insulating oil into N containers with a volume of V. Step 2: Divide the N containers into two groups. Introduce oxygen at a concentration of y into the first group of containers, and introduce dry, inert gas into the second group of containers, at the aging temperature. Heat N containers to age them; Set the interval x for detecting the water content in the container during the aging process. Detect the water content in the container once every interval x. Record the time of the first detection as time 1, the time of the second detection as time 2, and so on up to time T. Step 3: When aging reaches time t, select one container from the first group of containers as the first oxygen-containing container, and select one container from the second group of containers as the first oxygen-free container; where t = 1, 2, ..., T; Detect the first water content inside the first oxygen-containing container and the second water content in the first oxygen-free container. The first water content and the second moisture content The difference is used as an estimate of the oxidation water production of the insulating oil at time t. ; Step 4: Calculate the correction coefficient at time t; Step 5: Estimate the oxidation water production of the insulating oil at time t based on the correction coefficient at time t. After correction, the actual value of the oxidation water production of the insulating oil under test at time t is obtained. .

2. The detection method according to claim 1, characterized in that: Step 6 is executed after step 5: The actual value of the oxidation water production of the insulating oil at time T is calculated. Substituting into equation (6), the limiting oxidation water production of the insulating oil under oxygen concentration y is calculated. C max and oxidation water production rate constant k4 （6）。 3. The detection method according to claim 1 or 2, characterized in that: In step 2, the oxygen concentration y is adjusted to other oxygen concentrations to simulate different leakage conditions, and steps 1-5 are repeated to obtain the actual value of the oxidation water production of the insulating oil under different leakage conditions at time t.

4. The detection method according to claim 1, characterized in that: In step 1, the pretreatment is vacuum drying and degassing, with the vacuum gauge pressure set to ≤−0.1 MPa, the drying temperature to 35-65℃, and the drying time to 24-72 h.

5. The detection method according to claim 1, characterized in that: In step 1, during the pretreatment process, the moisture content of the insulating oil to be tested is detected every L time interval. When the following formula (7) is satisfied, the pretreatment is stopped. （7）； In formula (7), S1 and S2 are the moisture content of the insulating oil to be tested obtained from two adjacent tests, and S is the average moisture content of the insulating oil to be tested obtained from the two adjacent tests.

6. The detection method according to claim 1, characterized in that: The dry inert gas in step 2 is nitrogen.

7. The detection method according to claim 1, characterized in that: In step 3, when t=1, one container from the first group of containers is randomly selected as the first oxygen-containing container, and one container from the second group of containers is randomly selected as the first oxygen-free container; when t>1, the first oxygen-containing container and the first oxygen-free container selected at time t are controlled to be different from the containers selected at all times before time t.

8. The detection method according to claim 1, characterized in that: The correction coefficients in step 4 include a first correction coefficient f1, a second correction coefficient f2, a third correction coefficient f3, and a fourth correction coefficient f4, which are calculated through the following steps: Step 4.1: Calculate the first correction coefficient f1 according to the following formula (1). （1）； In equation (1), K is the moisture vapor-liquid partition correction coefficient; K ref It is the first water vapor-liquid partition coefficient at a reference thermodynamic temperature, which is 298.15 K; R is the molar enthalpy of solubility of water in the insulating oil to be tested within the first oxygen-containing container; R is the universal gas constant, taken as 8.314 J / (mol・K); It is the actual thermodynamic temperature inside the first oxygen-containing container; ρ is the reference thermodynamic temperature; r1 is the density of the insulating oil to be tested in the first oxygen-containing container; and r1 is the ratio of the remaining mass of the insulating oil to be tested in the first container to the initial mass m. Step 4.2: Calculate the second correction coefficient f2 according to the following formula (2). （2）； In equation (2), k1 is the oxygen consumption rate constant; It is the apparent rate constant of oxygen consumption at the reference temperature; is the apparent activation energy of oxygen consumption in the oxidation reaction of the insulating oil to be tested; k2 is the apparent decay rate constant of dissolved oxygen; r2 is the volume ratio of the gas phase to the oil phase in the first oxygen-containing container; Step 4.3: Calculate the third correction coefficient f3 according to the following formula (3). （3）； In equation (3), E is the apparent activation energy of the oxidation reaction of the insulating oil to be tested; It is the average actual thermodynamic temperature inside the first oxygen-containing container; and These are the aging temperatures. The second moisture gas-liquid partition coefficient and the first oxygen gas-liquid partition coefficient; and These are the actual thermodynamic temperatures. The third water vapor-liquid partition coefficient and the second oxygen vapor-liquid partition coefficient; γ is a kinetic influencing factor of oxygen partitioning; Step 4.4: Calculate the fourth correction factor f4 using the following formula (4). （4）； In equation (4), k3 is the evaporation rate correction constant; It is the evaporation rate constant at the reference temperature; It is the apparent activation energy of evaporation of the light component of the insulating oil to be tested.

9. The detection method according to claim 1, characterized in that: Step 5 estimates the oxidation water production using the following formula (5). After correction, the actual value of the oxidation water production of the insulating oil in the first oxygen-containing container at time t is obtained. （5）； In equation (5), F is the coupling correction coefficient; α is the first influence coefficient, 0.15≤α≤3; β is the second influence coefficient, 0.08≤β≤0.2.

Images