RU2762597C1 - Method for diagnosing oil product leakage from a coil during fire heating in a pipe furnace - Google Patents

Abstract

FIELD: diagnostic methods.

SUBSTANCE: invention relates to diagnostic methods. A method for diagnosing oil leakage from a coil during firing in a tubular furnace is described, which consists in the fact that according to several measured and corresponding predicted values ​​of the wall temperatures of the radiant and/or convective zones of the furnace, averaged over some time intervals τ, the values ​​of the residuals between the calculated and the measured values ​​and when these residuals exceed the boundary values, the fact and place of the leak is diagnosed, while the localization of the leakage area is determined at the place of installation of temperature sensors for which the residual is maximum, for which the wall temperatures of the radiant and/or convective zones of the furnace are measured with at least three temperature sensors in each zone, and the readings are transmitted to the computing device, in which, based on of the measured temperature values, the average temperature values ​​for a given period are calculated, according to which the values ​​of the same temperatures are calculated using the models of virtual sensors, comparing the calculated values ​​with the measured ones.

EFFECT: diagnostics of oil product leakage from the coil.

1 cl, 2 dwg, 1 ex

Description

The invention relates to the field of oil,

in the oil refining, petrochemical, gas industry, in particular, to methods for diagnosing oil product leaks from the coils of tube furnaces when fired with gas or liquid fuel, and can be used for prompt diagnosis of the fact of a leak, localization of the place, the amount of oil product leakage in normal operation on the basis of computational procedures for evaluating the heat balance of the heating process, analyzing the temperature profile of the radiant and / or convective zones of the furnace.

It is well known that tube furnaces (TP) are one of the most explosive and fire hazardous devices with high rates of accident probability. The most dangerous type of accidents, the so-called burnout of the coils, is associated with the depressurization of the tubular elements through which the heated oil product is pumped. At present, by order of Rostekhnadzor dated December 15, 2020 N 533 "On the approval of federal norms and rules in the field of industrial safety" General explosion safety rules for explosive and fire hazardous chemical, petrochemical and oil refineries ", a list of controlled parameters indicating the burnout of the coil, the procedure for their use for diagnostics are not clearly defined, and diagnostics based on them is possible only with large leaks. The practice of operating tube furnaces suggests determining burnout according to the following criteria:

- pressure drop in the coil at the TPP outlet;

- increasing the content of carbon monoxide in flue gases;

- reduction of oxygen content in flue gases;

- large (in comparison with the normal mode) increase in the temperature of the flue gases at the pass of the furnace.

These indicators for determining burnout give an adequate diagnosis only for very large volumes of oil product leakage due to depressurization of the coils. The decrease in pressure is noticeable when the raw material leakage rate is at least 2-3%. If the load on the raw material is, for example, 50 t / h, this means that there will be a noticeable pressure drop at leakage of the order of 500 kg / h. This figure is comparable to the mass fuel consumption. The content of carbon monoxide or oxygen in flue gases depends on a number of reasons and can be considered as a stable diagnostic sign only when there are significant changes in comparison with the average values. In addition, the measurement of these parameters encounters significant technical difficulties. The temperature of flue gases also fluctuates within significant limits (tens of degrees} and changes noticeably, as a rule, only when a fire has begun in combustion chambers. ensure production safety.

Known methods for diagnosing the state of furnace coils, for example, from https://nova78.ru/nerazrushayuschiy-kontrol-metody-nerazrushayuschego-kontrolya/. The known method is based on the methods of hardware-software non-destructive testing of the coil wall thickness in an idle furnace. As a result of the implementation of the method, existing or potentially dangerous places for the burnout of the coils are determined.

Known methods for diagnosing coil burnouts (Verevkin A.P., Matveev D.S., Khusniyarov M.Kh., Chikurov A.V. Construction of a mathematical model of a pyrolysis tube furnace for the purpose of optimizing modes and diagnosing coil burnouts // Oil and Gas Business. 2010. T 8, No. 1. P. 70-73; Khusniyarov M.Kh., Verevkin A.P., Kuzeev I.R., Tlyasheva PP, Matveev D.S., Gaevskaya O.I., Chikurov A.V. , Kharisov PM, Naumkin E.A., Simarchuk AS. Edited by Khusniyarov M.Kh. on the analysis of changes in some signs of furnace operation and changes in technological parameters, such as uneven heating of streams in multi-flow furnaces, a local increase in the temperature of the coils, a change in the pressure drop across the coil, the rate of change in the temperature difference of the product at the inlet and outlet of the heating furnace.

A known method for detecting burnout of a cooled heating unit, which includes measuring the temperature difference of the coolant at the inlet and outlet of the heating unit and the rate of change of this difference. The values of the temperature difference between the inlet and outlet flows of the cooling unit of the refrigerant and the rate of change of this difference are determined in advance, which correspond to the moment of burnout of the thermal unit. Burnout of the wall of the thermal unit is recorded while simultaneously exceeding predetermined maximum permissible values of the temperature difference between the inlet and outlet flows of the cooling unit of the refrigerant and the rate of its change (patent RU2243265, IPC C21V 7/10, publ. 27.12.2004).

There is a known method for diagnosing a burnout of a coil, which can be used for prompt diagnosis of the fact of a leak (Matveev D.S. P., Kuzeev I.R., Tlyasheva PP, Matveev D.S, Gaevskaya O.I., Chikurov A.V., Kharisov PM, Naumkin E.A., Simarchuk A.S., edited by Khusniyarov M.Kh Technogenic risk and industrial safety management of oil refineries: textbook, Ufa, "Oil and Gas Business", 2012., 324 p.). The method involves the use of diagnostic variables, which are determined by the differences in the values of the measured parameters of the technological object and the parameters of the dynamic model ("digital twin") of the automated technological complex of the furnace and allow, on the basis of production rules (rules for the formation of a diagnosis), to establish causal relationships between the facts of exceeding the diagnostic limits variables and possible reasons for their appearance, including malfunctions of the measuring instruments and burnout of the coil. For the formation of diagnostic indicators, the following are used: the temperature of heating the product at the outlet of the furnace, the temperature at the pass of the furnace, the consumption of fuel gas, the magnitude of the discrepancy between the actual heat balance and the theoretical one.

The disadvantage of this method is that the influence of random factors on the parameters of the furnace is not taken into account, such as the ambient temperature, fluctuations in the calorific value of the fuel and the heat capacity of the raw materials, the errors of the measuring transducers, and the location is not localized and the amount of oil product leakage is not determined. Therefore, this method for diagnosing coil burnout does not provide full diagnostic functionality and is not sufficiently informative in terms of diagnostic accuracy.

The technical problem solved by the invention is to increase the safety of operation of fired tube furnaces by diagnosing early burnouts with the establishment of the fact of oil product leakage from the coil and localization of the place and size of the leak.

EFFECT: determination of the fact of oil product leakage from the coil with localization of the leak and the amount of leakage, taking into account the influence of random factors on the parameters of the furnace.

в соответствии с выражением:The problem is solved, and the technical result is achieved by a method for diagnosing oil leakage from a coil during fired heating in a tubular furnace, which consists in the fact that according to several measured and corresponding predicted values of the temperatures of the walls of the radiant and / or convective zones of the furnace, they are calculated averaged over some time periods. at intervals t, the values of the residuals between the calculated and measured values and when these residuals exceed the boundary values, the fact and location of the leak is diagnosed, while the localization of the leakage area is determined at the location of the temperature sensors, for which the residual is maximum, for which the wall temperatures of the radiant and / or convective zones are measured ovens with at least three temperature sensors in each zone, and the readings are transmitted to a computing device, in which, on the basis of the measured temperature values, the average temperatures over a given period are calculated, according to which the values of the same temperatures are calculated according to the mode lam virtual sensors

according to the expression:

- среднее за период τ расчетное по модели

значение температуры в радиантной и/или конвективной зонах установки датчика T_i;where

is the average over the period τ calculated according to the model

the temperature value in the radiant and / or convective zones of the sensor installation T _i ;

моделей получают регрессионным анализом статистических данных;parameters

models are obtained by regression analysis of statistical data;

- среднее за период τ измеренное значение температуры в радиантной и/или конвективной зонах датчиком T_j, который также установлен в радиантной и/или конвективной зонах соответственно, при этом i,j=1,2,3, i≠j, причем средние значения температур стенки радиантной и/или конвективной зон печи рассчитывают как

где n - количество измерений на периоде τ, а также вычисляют невязки моделей, сравнивая расчетные значения с измеренными:

- the average over the period τ measured temperature value in the radiant and / or convective zones by the sensor T _j , which is also installed in the radiant and / or convective zones, respectively, with i, j = 1,2,3, i ≠ j, and the average values wall temperatures of the radiant and / or convective zones of the furnace are calculated as

where n is the number of measurements over the period τ, and the residuals of the models are also calculated by comparing the calculated values with the measured ones:

then the discrepancy of the heat balance of the furnace ΔВ is determined as

where F _t , F _c - mass flow rates of fuel and oil product, respectively;

q is the calorific value of the fuel;

n - efficiency ovens;

С _с - heat capacity of the oil product,

t _out , t _in are the temperatures of the oil product of the raw material at the outlet and inlet to the furnace, respectively,

and, based on the values of the residuals of the thermal imbalance, the calculated and measured values, a diagnostic conclusion is made about the state of the measuring instruments and the furnace coil, while the burnout of the coil in the zone of the temperature sensor Ti is diagnosed if the condition is met

виртуального датчика; i=1,2,3…N, j=i-1, k=i+1, при этом для j<1 принять j=N, для k>N принять k=1;where ε ^add. - permissible error of the model

virtual sensor; i = 1,2,3… N, j = i-1, k = i + 1, while for j <1 take j = N, for k> N take k = 1;

N is the number of temperature sensors participating in the diagnostic algorithm;

- допустимый тепловой дебаланс, невязка, после чего рассчитывают величину утечки нефтепродукта как F_y=ΔВ/(q_y*k),

- permissible thermal imbalance, discrepancy, after which the amount of oil product leakage is calculated as F _y = ΔВ / (q _y * k),

where q _y is the calorific value of the petroleum product;

k - coefficient of effective use of the heat of combustion of a petroleum product, is assigned empirically.

The technical result is achieved by the stated set of essential features due to the fact that the following are used for diagnostics:

- parameter values averaged over time intervals;

- models of virtual sensors;

- models for calculating the heat balance, taking into account the averaged values of the parameters and the technical condition of the measuring instruments (temperature and flow sensors).

FIG. 1 illustrates the claimed method for the case of installing three temperature sensors in the radiant zone of the furnace with a schematic diagram of an automated technological complex for diagnosing oil leakage from the coil of a fired tube furnace.

FIG. 2 shows a diagram for determining the relationship of models of virtual temperature sensors of the radiant zone of the furnace (for three sensors).

The figures indicate:

1 - diagnostic object (fired heating furnace);

TI 1, TI 2, TI 3 - temperature sensors of the wall of the radiant zone of the furnace;

TI 4, TI 5 - oil product temperature sensors at the inlet and outlet of the furnace;

FT 1, FT 2 - oil and fuel flow sensors, respectively;

η - efficiency ovens;

C _c - heat capacity of the oil product;

q is the calorific value of the fuel;

k is the coefficient of effective use of the heat of combustion of a petroleum product;

и значения невязок ε_ij, которые используются для формирования диагностического сообщения, а также рассчитывается объем утечки;2 - a computing device in which, on the basis of the measured values of temperatures TI 1, TI 2, TI 3, the values of temperatures are calculated using models of virtual sensors

and the values of the residuals ε _ij , which are used to generate a diagnostic message, and the amount of leakage is calculated;

I - oil product at the entrance to the heating furnace;

II - oil product at the exit from the heating furnace;

III - fuel for the burners of the furnace.

T _i - the value of the "wall temperature" parameter, measured by the sensor i, i = l ... 3;

M _ij is a model of a virtual sensor for calculating the value of T _i from the readings of the sensor T _j , j = il and j = i + 1 (for j <1, take j = 3, for j> 3, take j = 1).

Diagnostics of the leakage of oil from the coil during fired heating in tubular furnaces is carried out as follows.

по которым рассчитывают значения тех же температур по моделям виртуальных датчиков

и значения невязок ε_ij, которые используются для формирования диагностического сообщения, а также рассчитывается объем утечки нефтепродукта из змеевика в случае диагностирования прогара на основе данных измерения FT 1, FT 2, TI 4, TI 5 и множества А, задаваемых в (2) в качестве постоянных значений,

The oil product (raw material) enters the coil of the furnace 1, where it is heated with an open flame from the combustion of liquid or gaseous fuel in the burners. The combustion process in the furnace is controlled by wall temperature sensors in the radiant zone of the furnace. The number of sensors must be at least 3. The signals about the wall temperature values from the sensors TI 1, TI 2, TI 3 are sent to the computing device 2. In the device 2, on the basis of the measured values of the temperatures TI 1, TI 2, TI 3, the average values over a given period are calculated

by which the values of the same temperatures are calculated using the models of virtual sensors

and the values of the residuals ε _ij , which are used to generate a diagnostic message, and the volume of oil leakage from the coil is calculated in the case of burnout diagnostics based on the measurement data FT 1, FT 2, TI 4, TI 5 and set A, specified in (2) in as constant values,

According to several measured and corresponding predicted values of the wall temperatures of the radiant and / or convective zones of the furnace, the values of the residuals between the calculated and measured values averaged over certain time intervals t are calculated, and when these residuals exceed the boundary values, the fact and place of leakage is diagnosed. The localization of the leakage area is determined by the installation site of the temperature sensors, for which the discrepancy is maximum.

Taking into account the correlation of the sensor temperatures, the scheme for determining the relationship of the virtual sensor models, for example, the temperature of the radiant zone of the furnace for a fairly typical case of three sensors, is shown in Fig. 2. The principle of simulation does not change for more sensors and a zone of the furnace. In general, the structure of models of virtual sensors is as follows:

моделей получают регрессионным анализом статистических данных.Parameters

models are obtained by regression analysis of statistical data.

For example, in FIG. 2 models M ₁₂ , M ₁₃ will look like

By the values ε _{ij of the} residuals of the models

and the value of the thermal imbalance (discrepancy)

a diagnostic conclusion is formed. When forming a diagnostic conclusion, it is assumed that the simplest flow of events (Poisson's law) is at work, i.e. Only one event can occur at a time. The diagnosis is based on the following rules:

1. If ε _ij ≤ε ^dop,. then the technological situation is normal, i.e. no burnout, sensors are in good order;

ИЛИ (ε_ij>ε^доп.и ε_ki>ε^доп.), то вероятна неисправность датчика температуры T_i;2. If

OR (ε _ij > ε ^add. And ε _ki > ε ^add. ), Then the temperature sensor T _{i is} likely to malfunction;

3. If the conditions under Rule 2 are met, i.e. a malfunction of any sensor is diagnosed, then the diagnostics should be stopped until the malfunction is eliminated, go to step 5 of the diagnostic cycle;

4. If

then burnout is likely in the area of installation of the temperature sensor T _i ;

5. The end of the cycle of the diagnostic algorithm for the serviceability of temperature sensors and (or) the fact of a leak.

Application in the unbalance calculation algorithm according to (5) assumes certain assumptions, namely;

- the calorific value of the fuel (q) is constant or varies within narrow limits;

- the heat capacity of the oil product (raw material) (С _с ) is constant or varies within narrow limits;

- mass consumption of fuel and raw materials (F _τ , F _c ) are measured with sufficient accuracy.

. В этом случае, алгоритм формирования диагноза может быть реализован без правила 2. Выражение (5) позволяет определить объем утечки F_y какIf in the equipment with technical means for measuring the temperature in the radiant and (or) convective zones of the furnace, sensors with self-diagnostic means are used, or an autonomously organized diagnostics of sensors is implemented due to duplication (tripleting) of sensors, then the diagnostics of the burnout of the coil can only be carried out by checking the unbalance violation condition

... In this case, the diagnosis formation algorithm can be implemented without rule 2. Expression (5) allows you to determine the volume of leakage F _y as

To generate a diagnostic message, the following actions (algorithm) are performed:

n - количество измерение на периоде τ.1. Calculate the average wall temperatures, for example, the radiant zone of the furnace as

n is the number of measurements in the period τ.

2. Calculate according to (1) the values of temperatures by models of virtual sensors.

3. Calculate according to (4) the residuals of the models of virtual sensors.

4. Based on the residual values of the virtual sensor models, a diagnostic message is generated based on the residual analysis.

5. If the coil burnout is diagnosed, according to (6), the volume of oil product leakage is calculated.

An example of the implementation of the method.

Consumption of oil product (raw materials), 50,000 kg / h.

The amount of heat in the raw material at the input, 150 0 KJ * gr.

The calorific value of fuel - gas (calorific value) is taken as 44257 kJ / kg.

The calorific value of raw materials (calorific value) is taken as 49908 kJ / kg.

Raw material temperature at the inlet, 30 ° С.

Heat capacity of raw materials, 2140J / kg * K.

Diagnostics is carried out by three temperatures of the wall of the radiant part: T2, T4, T6.

According to statistical data, models of the temperature relationship were obtained:

T ^m 2 = 0.791 * T6 + 156.6; T ^m 4 = 1.008 * T2 + 37;

T ^m 6 = 1.546 * T4-2856235 (or T4 = 0.6468 * T6 + 184.49).

In the absence of burnout, the errors in the reproduction of the measured temperatures T2, T4, T6 in the averaging interval of 10-20 minutes averaged

соответственно.

respectively.

Fluctuations in the relative heat balance in the absence of leakage according to the averaged data are estimated at ± 1.8%.

Calculations have established that for each percentage of raw material leakage, the heat balance changes by 0.15, i.e. by 15%. Therefore, it can be considered that a noticeable (significant) change in the heat balance becomes at leaks above 0.15%.

For example, with a leakage of 0.2%, the errors in calculating temperatures by models relative to the measured values will average

ε _т26 = 4-5%,

ε _т42 = 5-6%,

The thermal imbalance will be more than 2.5%,

получим выполнение условия наличия утечки.

we obtain the fulfillment of the condition for the presence of a leak.

The analysis of temperature errors indicates that the place of burnout is most likely located near the location of the temperature sensor T2, because the maximum deviations are recorded for models T ^m 2 (T6) and T ^m 4 (T2)

According to the heat balance, the estimated leak rate is approximately 100 kg / h.

Claims (19)

A method for diagnosing oil product leakage from a coil during fiery heating in a tubular furnace, which consists in the fact that according to several measured and corresponding predicted values of the wall temperatures of the radiant and / or convective zones of the furnace, averaged over some time intervals τ, the values of the residuals between the calculated and the measured values and when these residuals exceed the boundary values, the fact and place of leakage is diagnosed, while the localization of the leakage area is determined at the place of installation of temperature sensors for which the residual is maximum, for which the wall temperatures of the radiant and / or convective zones of the furnace are measured with at least three temperature sensors in each zone, and the readings are transmitted to the computing device, in which, on the basis of the measured temperature values, the average temperatures over a given period are calculated, according to which the values of the same temperatures are calculated using the models of virtual sensors

according to the expression:

where

- the average for the period τ calculated by the model M _ij value of temperature in the radiant and / or convective zones of the sensor installation T _i ; parameters

models are obtained by regression analysis of statistical data;

- the average over the period τ measured temperature value in the radiant and / or convective zones by the sensor T _j , which is also installed in the radiant and / or convective zones, respectively, with i, j = 1, 2, 3, i ≠ j, and the average values wall temperatures of the radiant and / or convective zones of the furnace are calculated as

where n is the number of measurements over the period τ, and the residuals of the models are also calculated by comparing the calculated values with the measured ones:

then the discrepancy of the heat balance of the furnace ΔВ is determined as

where F _t , F _c - mass flow rates of fuel and oil product, respectively; q is the calorific value of the fuel; η - furnace efficiency; С _с - heat capacity of the oil product, t _out , t _in are the temperatures of the oil product of the raw material at the outlet and inlet to the furnace, respectively, and based on the values of the residuals of the thermal imbalance, the calculated and measured values, a diagnostic conclusion is made about the state of the measuring instruments and the furnace coil, while the burnout of the coil in the zone of the temperature sensor installation Ti diagnosed if the condition

, where ε ^add. - permissible error of the model

virtual sensor; i = 1, 2, 3… N, j = i-1, k = i + 1, while for j <1 take j = N, for k> N take k = 1; N is the number of temperature sensors participating in the diagnostic algorithm;

- permissible thermal imbalance, discrepancy, after which the amount of oil product leakage is calculated as F _y = ΔВ / (q _y * k), where q _y is the calorific value of the petroleum product; k - coefficient of effective use of the heat of combustion of a petroleum product, is assigned empirically.

Images