A data-driven dynamic risk assessment method

Abstract

This invention discloses a data-driven dynamic risk assessment method, belonging to the field of equipment condition risk assessment technology. This method allows for parameter adjustment based on real-time dynamic risk assessment models according to the dynamic changes of equipment, achieving a data-driven dynamic risk assessment technology. This technology can more accurately predict the failure risks and probabilities of equipment failures. Based on data, it performs risk estimation for equipment, and compared to traditional RBI technology, it is more in line with my country's equipment risk assessment standards. Furthermore, it allows for the development of reasonable and effective countermeasures through expert groups. Especially for the petrochemical industry, timely detection of failure risks can prevent oil or fuel leaks and reduce the occurrence of accidents.

Description

Technical Field

[0001] This invention belongs to the field of equipment condition risk assessment technology, specifically relating to a data-driven dynamic risk assessment method. Background Technology

[0002] As is well known, the petrochemical industry is a high-risk, high-return sector, and raw material leaks are a common occurrence during production. Risk-based inspection (RBI) is an optimized inspection strategy built upon the principle of achieving a balance between system safety and economy. Its essence is to analyze and prioritize the probability and consequences of hazardous events, identify key problems and weaknesses, ensure inherent safety, and simultaneously reduce operating costs.

[0003] The existing equipment risk assessment methods have the following main problems: (1) How to more accurately assess the risk level of equipment and reduce the probability of accidents is an urgent problem to be solved. (2) The existing RBI risk assessment inspection cycle cannot be determined, and risk assessment cannot be given in real time according to changes in equipment.

[0004] Therefore, based on the conventional RBI (Risk Based Inspection) risk assessment method, this invention proposes a big data-driven dynamic risk assessment, establishes a real-time dynamic risk assessment model, and obtains risk assessments that change with time and operating conditions.

[0005] Petrochemical plants are always in a dynamic process, and the medium environment in which the equipment is located is also dynamically changing, such as changes in temperature, flow rate, and medium composition. These changes can affect the risk of the equipment and cause it to change. This method can adjust parameters based on the dynamic changes of the equipment through a real-time dynamic risk assessment model, thus realizing a data-driven dynamic risk assessment technology.

[0006] This technology allows for more accurate prediction of equipment failure risks and the probability of failure. Based on data, it enables risk estimation of equipment and provides a more suitable evaluation standard for equipment risk in my country compared to traditional RBI technology. Furthermore, it allows for the development of reasonable and effective countermeasures through expert groups. In particular, for the petrochemical industry, timely detection of failure risks can prevent oil or fuel leaks and reduce the occurrence of accidents. Summary of the Invention

[0007] In view of the above-mentioned technical problems in the prior art, the present invention proposes a data-driven dynamic risk assessment method, which is reasonably designed, overcomes the shortcomings of the prior art, and has good results.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A data-driven dynamic risk assessment method includes the following steps:

[0010] Step 1: Determine equipment boundaries;

[0011] Step 2: Failure Mode and Effects Analysis;

[0012] Step 3: Analysis of the dominant failure mechanism under the interaction of multiple failure modes;

[0013] Step 4: Dynamic characteristic analysis of factors influencing the dominant failure mode;

[0014] Step 5: Analysis of the dynamic interaction characteristics of influencing factors;

[0015] Step 6: Failure probability analysis and consequence assessment;

[0016] Step 7: Conduct real-time dynamic risk assessment.

[0017] Preferably, in step 1, the purpose of determining the equipment boundary is to define the analysis scope so that factors affecting the safety and reliability of the equipment can be analyzed within a limited range, with the controlled unit as the object of analysis. After the controlled equipment boundary is selected, risk analysis is carried out, mainly analyzing the factors affecting the reliability of the equipment within the boundary range of the controlled unit. The information interaction between the controlled unit and the outside world can be used as the input or output parameters of the controlled unit. Once the equipment boundary is determined, the impact of the boundary and the input and output parameters on the real-time risk is only considered in the real-time risk analysis.

[0018] Preferably, in step 1, the equipment boundary can be defined by a specific piece of equipment and its connected piping system, or by a single piece of equipment with an independent function and its associated piping, or by multiple pieces of equipment and their associated piping that perform a specific function.

[0019] Preferably, in step 2, the failure mode and influencing factor analysis is to analyze the possible causes, influencing factors and consequences of equipment failure.

[0020] Preferably, in step 3, the various influencing factors that cause equipment failure are analyzed, the dominant failure mechanism is identified, and it is characterized as equipment risk.

[0021] Preferably, in step 4, the key parameters of the dominant failure mode affecting equipment failure in the petrochemical unit are constantly and dynamically adjusted. The dynamic characteristics are either the fluctuation of parameters over time or the change of the combination of multiple key parameters. Conducting dynamic analysis on the influencing factors of the dominant failure mode is a key step in realizing dynamic hazard source risk analysis.

[0022] Preferably, in step 5, dynamic interaction analysis mainly refers to analyzing the impact of the interaction behaviors between different key parameters, between key parameters and manual operation, and between devices on the real-time risk of the equipment; dynamic interaction analysis mainly studies the interrelationships of different influencing factors and their impact on the system.

[0023] Preferably, in step 6, the probability of failure is analyzed based on the dominant failure mode and influencing factors; when conducting risk analysis, the worst-case scenario is assessed based on the principle of conservatism to determine the magnitude of the risk.

[0024] Preferably, in step 7, the real-time dynamic risk assessment retains the failure probability P in traditional RBI. f With C(t) and failure consequences C(t) remaining constant, the failure probability P is adjusted based on real-time corrosion monitoring parameters. f By considering the influence of (t), we can obtain the changing real-time dynamic failure probability factor k of the equipment, thereby realizing the dynamic change of risk.

[0025] Current dynamic change in failure probability k·P f (t) After corresponding to the original level division range, a new failure probability level is obtained. Combined with the original failure consequence level, a dynamic risk matrix is ​​formed as the real-time dynamic failure probability influencing factor k of the equipment changes, thus realizing dynamic RBI risk assessment.

[0026] Preferably, step 7 specifically includes the following steps:

[0027] Step 7.1: Determine the influencing factor k of the real-time dynamic failure probability of the equipment;

[0028] Corrosion rate, operating hours, and remaining service life directly reflect the corrosion status of equipment, defined as influencing factor f. Changes in each monitoring parameter are directly related to changes in the influencing factor. When calculating the magnitude of influencing factors by acquiring multiple monitoring parameters, α1, α2, α3, ..., α n This represents the real-time monitoring values ​​of each monitoring parameter, and the influencing factors are calculated as f(α1,α2,α3,…,α). n ), α 10 ,α 20 ,α 30 ,…,α n0 The design baseline values ​​for each monitoring parameter are represented by the following: The influencing factor for the real-time dynamic failure probability of the equipment is expressed as follows:

[0029]

[0030] When multiple monitoring parameters are difficult to obtain, or when there are monitoring parameters that have a major impact on corrosion, a single monitoring parameter is selected to fit the relationship between the parameter and the influencing factors. Assuming that monitoring parameters of a pipeline, including operating temperature α1, pH value α2, and operating flow rate α3, all affect the corrosion rate, and that temperature has the most significant impact, then operating temperature is determined as the primary monitoring parameter, and the relationship between it and the influencing factors is fitted as f(α1), α... 10 Representing the design temperature, the influence coefficient k of the failure probability is expressed as:

[0031]

[0032] Step 7.2: Determine the dynamic failure probability level;

[0033] The dynamic RBI failure probability k·P is obtained by correcting the static RBI assessment failure probability using the real-time dynamic failure probability influence factor k. f (t); the current dynamically changing failure probability k·P f (t) is placed into the failure probability level classification criteria determined by the original RBI assessment to obtain a new dynamic RBI failure probability level.

[0034] Step 7.3: Determine the real-time dynamic risk level;

[0035] Using the original static RBI assessment knowledge base to evaluate the failure consequence level, combined with the dynamic failure probability level determined in step 7.2, the risk R(t) = [k·P] is calculated through risk logic. f (t)]·C(t), and then determine the real-time dynamic risk level based on the risk level evaluation matrix determined by the original static RBI assessment;

[0036] Step 7.4: Determine the real-time device health status;

[0037] Equipment health is a normalized metric that measures the health of equipment. It is an assessment of the degree of deviation between the current state and the expected state of the equipment. The normalized metric is generally in the range of (0,1). The expected state refers to the design baseline state of the same equipment or similar equipment under the same operating conditions.

[0038] Preferably, in step 7.1, the influence factor k of the real-time dynamic failure probability of the equipment is defined, specifically including the following steps:

[0039] Step 7.1.1: Analysis of key influencing factors on the likelihood of equipment and facility failure;

[0040] Different corrosion mechanisms correspond to different monitoring parameters. Correlation analysis was conducted on parameters such as operating temperature, operating pressure, operating flow rate, total acid value, crude oil sulfur content, pH value, partial pressure of H2S in gas, and partial pressure of CO2 in gas to screen out key influencing factors that can cause changes in the possibility of equipment failure, which are defined as influencing factors f.

[0041] Step 7.1.2: Determination of factors influencing the probability of real-time dynamic equipment failure:

[0042] With α 10 ,α 20 ,α 30 ,…,α n0 This represents the design baseline values ​​for various monitoring parameters, including the design temperature, design pressure, design flow rate, and design allowable total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas. The influencing factors under design conditions are calculated as f(α). 10 ,α 20 ,α 30 ,…,α n0 );

[0043] Using α1, α2, α3, ..., α n The parameters include real-time monitoring values ​​of various parameters within the equipment and facility operating temperature, operating pressure, operating flow rate, and actual operating conditions such as total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas. The influencing factors under operating conditions are calculated as f(α1, α2, α3, ..., α...). n );

[0044] The impact factor of the real-time dynamic failure probability of the equipment is expressed as:

[0045]

[0046] Preferably, in step 7.2, the dynamic RBI failure probability level classification is as follows:

[0047]

[0048] Preferably, step 7.3 specifically includes the following steps:

[0049] Step 7.3.1: Characterization of the real-time dynamic risk level of the equipment, as shown in formula (2):

[0050] R(t) = [k·P f (t)]×C(t) (2);

[0051] In the formula, R(t) represents the real-time dynamic risk of equipment and facilities, k represents the value of the real-time dynamic failure probability influencing factor of equipment, and P fC(t) is the probability value of equipment failure under operating conditions, and C(t) is the failure consequence.

[0052] Step 7.3.2: Criteria for determining the probability level of real-time failure;

[0053] The real-time failure probability judgment criterion draws on the static RBI assessment failure probability level judgment criterion, where the original RBI assessment failure probability P f (t) using [k·P f (t)] is used instead, according to [k·P] f The numerical range of (t)] is used to determine the level of real-time failure probability of equipment and facilities;

[0054] Step 7.3.3: Real-time failure consequence level discrimination matrix;

[0055] The criteria for judging the consequences of real-time failures are based on the criteria for judging the severity of failures in static RBI assessments.

[0056] Step 7.3.4: Dynamic RBI Risk Assessment Matrix;

[0057] The dynamic risk assessment matrix for equipment and facilities draws on the static RBI assessment risk assessment matrix.

[0058] Preferably, step 7.4 specifically includes the following steps:

[0059] Step 7.4.1: Health characterization of equipment and facilities;

[0060] By utilizing the k-value, which is an influencing factor for the real-time dynamic failure probability of equipment, the health status of static equipment can be expressed in terms of numerical value.

[0061] If the magnitude of the real-time dynamic monitoring process parameters is directly proportional to the probability of failure, then the health status is expressed as H = 1 / k; if the magnitude of the real-time dynamic monitoring process parameters is inversely proportional to the probability of failure, then the health status is expressed as H = k.

[0062] The health status of static equipment is characterized by formula (3):

[0063]

[0064] In the formula, α represents the magnitude of the real-time dynamic monitoring process technology parameter, f represents the magnitude of the failure probability, k represents the influence factor of the real-time dynamic failure probability of the equipment, k∈(0,1), then H∈(0,1);

[0065] Step 7.4.2: Health assessment criteria for equipment and facilities;

[0066] The fuzzy comprehensive membership method is adopted to define the health evaluation criteria for equipment and facilities. The health of equipment and facilities is scored using four indicators: excellent, good, permissible, and unacceptable. The membership function is defined as ([1,0.75),[0.75,0.5),[0.5,0.25),[0.25,0)).

[0067] Preferably, in step 7.4, it is assumed that the main monitoring parameter of a certain pipeline is the operating temperature α1, and the design temperature is α. 10 Then the operating temperature α1 ≥ α 10 ;

[0068] The factor affecting the likelihood of failure is the corrosion rate; if the corrosion rate increases with increasing operating temperature.

[0069] When the pipeline operating temperature is lower than the design temperature

[0070] When the pipeline operating temperature equals the design temperature As the operating temperature of the pipeline gradually increases, the corrosion rate f(α1) gradually increases, and the health status H=1 / k gradually decreases. That is, the worse the health status of the equipment, the higher the risk level and the cause of the risk can be calculated.

[0071] The beneficial technical effects of this invention are as follows:

[0072] Petrochemical plants are always in a dynamic process, and the medium environment in which the equipment is located is also dynamically changing, such as changes in temperature, flow rate, and medium composition. These changes can affect the risk of the equipment and cause it to change. This method can adjust parameters based on the dynamic changes of the equipment through a real-time dynamic risk assessment model, thus realizing a data-driven dynamic risk assessment technology.

[0073] This technology allows for more accurate prediction of equipment failure risks and the probability of failure. Based on data, it enables risk estimation of equipment and provides a more suitable evaluation standard for equipment risk in my country compared to traditional RBI technology. Furthermore, it allows for the development of reasonable and effective countermeasures through expert groups. In particular, for the petrochemical industry, timely detection of failure risks can prevent oil or fuel leaks and reduce the occurrence of accidents. Attached Figure Description

[0074] Figure 1 This is a flowchart of a dynamic equipment risk assessment method.

[0075] Figure 2 A schematic diagram illustrating the principle of a method for determining real-time dynamic risk levels.

[0076] Figure 3 (a) is a schematic diagram of the real-time dynamic risk level discrimination matrix for equipment and facilities;

[0077] Figure 3 (b) is Figure 3 (a) Legend illustration.

[0078] Figure 4 To analyze the schematic diagram.

[0079] Figure 5 This is a schematic diagram of a metering separator.

[0080] Figure 6 (a) is a schematic diagram of the economic risk matrix;

[0081] Figure 6 (b) is Figure 6 (a) Legend illustration.

[0082] Figure 7 (a) is a schematic diagram of the security risk matrix;

[0083] Figure 7 (b) is Figure 7 (a) Legend illustration. Detailed Implementation

[0084] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0085] A data-driven dynamic risk assessment method, the process of which is as follows: Figure 1 As shown, it includes the following steps:

[0086] Step 1: Determine equipment boundaries;

[0087] The purpose of selecting the equipment boundary is to define the scope of the analysis so that factors affecting equipment safety and reliability can be analyzed within a limited range. Generally, the analysis is performed on the controlled unit (“Equipment Under Control”, EUC). The equipment boundary can be defined by a specific piece of equipment and its connected piping system, by a single piece of equipment with an independent function and its associated piping, or by multiple pieces of equipment performing specific functions and their associated piping.

[0088] Once the equipment boundary is selected, risk analysis primarily focuses on factors affecting equipment reliability within the EUC boundary. Information exchange between the EUC and external systems can serve as input or output parameters. Once the equipment boundary is determined, real-time risk analysis considers only the impact of the boundary and input / output parameters on real-time risk. Because finite objectives are set, real-time risk analysis becomes more targeted.

[0089] Step 2: Failure Mode and Effects Analysis;

[0090] Failure Mode and Effect Analysis (FMEA) is a commonly used analytical method that primarily analyzes the possible causes, influencing factors, and consequences of equipment failure. In actual installations, equipment is affected by various process media, process controls, and human operations, resulting in diverse causes, modes, and effects of failure. API 581 analyzes and summarizes four failure modes (thinning, environmental cracking, material degradation, and mechanical damage) for refining and chemical plants, which can serve as a basis for conducting equipment failure analysis.

[0091] Step 3: Analysis of the dominant failure mechanism under the interaction of multiple failure modes;

[0092] Petrochemical plants often exhibit multiple failure modes under complex operating conditions, posing challenges to quantitative risk analysis. Therefore, it is necessary to analyze the various influencing factors contributing to equipment failure, identify the dominant failure mechanism, and characterize it as equipment risk.

[0093] Step 4: Dynamic characteristic analysis of factors influencing the dominant failure mode;

[0094] In petrochemical plants, the key parameters affecting the dominant failure modes of equipment are constantly and dynamically adjusted. These dynamic characteristics manifest as fluctuations in parameters over time or changes in the combination of multiple key parameters. Conducting dynamic analysis of the influencing factors of the dominant failure modes is a crucial step in achieving dynamic hazard risk analysis.

[0095] Step 5: Analysis of the dynamic interaction characteristics of influencing factors;

[0096] Interaction analysis primarily refers to analyzing the impact of interactions between different key parameters, between key parameters and human operations, and between different devices on the real-time risks of equipment. Interaction analysis mainly studies the interrelationships of different influencing factors and their impact on the system. Because a petrochemical plant is a complex system, all equipment is interconnected and mutually influential through process media, and the impact of operational processes on the system is significant, conducting interaction analysis is crucial. For example, taking the reactive distillate from a hydrotreating unit as an example, the chlorine content in crude oil affects the crystallization of ammonium salts in the reactive distillate; water injection operations also directly affect ammonium salt crystallization; ammonium salt crystallization, in turn, affects the efficiency of heat exchangers and, consequently, the composition of downstream circulating media. Therefore, conducting interaction analysis on equipment related to major hazard sources is essential.

[0097] Step 6: Failure probability analysis and consequence assessment.

[0098] Based on the dominant failure mode and influencing factors, the probability of failure can be analyzed. In risk analysis, when conducting failure consequence analysis for the dominant failure mode, a conservative approach is generally adopted, requiring an assessment of the worst-case scenario to determine the magnitude of the risk.

[0099] Step 7: Conduct real-time dynamic risk assessment.

[0100] Real-time dynamic risk assessment retains the probability of failure (P) in traditional RBI. f With C(t) and failure consequences C(t) remaining constant, the failure probability P is adjusted based on real-time corrosion monitoring parameters. f By considering the influence of (t), we can obtain the changing real-time dynamic failure probability factor k of the equipment, thereby realizing the dynamic change of risk.

[0101] Current dynamic failure probability k·P f (t) After corresponding to the original level division range, a new failure probability level is obtained. Combined with the original failure consequence level, a dynamic risk matrix is ​​formed as the real-time dynamic failure probability influencing factor k of the equipment changes, thus realizing dynamic RBI risk assessment.

[0102] Real-time dynamic risk assessment process, such as Figure 2 As shown, the specific steps include the following:

[0103] Step 7.1: Determine the influencing factor k of the real-time dynamic failure probability of the equipment;

[0104] Corrosion rate, operating hours, and remaining service life directly reflect the corrosion state of the pipeline, defined as influencing factor f. Changes in each monitoring parameter are directly related to changes in the influencing factor. When calculating the magnitude of influencing factors by acquiring multiple monitoring parameters, α1, α2, α3, ..., α n This represents the real-time monitoring values ​​of each monitoring parameter, and the influencing factors are calculated as f(α1,α2,α3,…,α). n ), α 10 ,α 20 ,α 30 ,…,α n0 The design baseline values ​​for each monitoring parameter are represented by the following: The influencing factor for the real-time dynamic failure probability of the equipment is expressed as follows:

[0105]

[0106] When multiple monitoring parameters are difficult to obtain, or when there are monitoring parameters that have a major impact on corrosion, a single monitoring parameter is selected to fit the relationship between the parameter and the influencing factors. Assuming that monitoring parameters of a pipeline, including operating temperature α1, pH value α2, and operating flow rate α3, all affect the corrosion rate, and that temperature has the most significant impact, then operating temperature is determined as the primary monitoring parameter, and the relationship between it and the influencing factors is fitted as f(α1), α... 10 Representing the design temperature, the influence coefficient k of the failure probability is expressed as:

[0107]

[0108] Step 7.2: Determine the dynamic failure probability level;

[0109] The dynamic RBI failure probability k·P is obtained by correcting the static RBI assessment failure probability using the real-time dynamic failure probability influence factor k. f (t); the current dynamically changing failure probability k·P f (t) is placed into the failure probability level classification criteria determined by the original RBI assessment to obtain a new dynamic RBI failure probability level; as shown in Table 1.

[0110] Table 1 Criteria for Determining the Failure Probability Level of Pressure Equipment and Facilities

[0111] Failure probability level Failure probability 1 <![CDATA[0.00000<k·P f (t)≤0.00001]]> 2 <![CDATA[0.00001<k·P f (t)≤0.00010]]> 3 <![CDATA[0.00010<k·P f (t)≤0.00100]]> 4 <![CDATA[0.00100<k·P f (t)≤0.01000]]> 5 <![CDATA[0.01000<k·P f (t)≤0.1000 ]]>

[0112] Step 7.3: Determine the real-time dynamic risk level;

[0113] Using the original static RBI assessment knowledge base to evaluate the failure consequence level, combined with the dynamic failure probability level determined in step 7.2, the risk R(t) = [k·P] is calculated through risk logic. f (t)]·C(t), and then determine the real-time dynamic risk level based on the risk level evaluation matrix determined by the original static RBI assessment;

[0114] Step 7.4: Determine the real-time static equipment health status;

[0115] Static equipment health is a normalized metric that measures the health of equipment. It is an assessment of the degree of deviation between the current state and the expected state of the equipment. The normalized metric is generally in the range of (0,1). The expected state refers to the design baseline state of the same equipment or similar equipment under the same working conditions.

[0116] In step 7.1, the influence factor k of the real-time dynamic failure probability of the equipment is defined, which specifically includes the following steps:

[0117] Step 7.1.1: Analysis of key influencing factors on the likelihood of equipment and facility failure;

[0118] Different corrosion mechanisms correspond to different monitoring parameters. Correlation analysis was conducted on parameters such as operating temperature, operating pressure, operating flow rate, total acid value, crude oil sulfur content, pH value, partial pressure of H2S in gas, and partial pressure of CO2 in gas to screen out key influencing factors that can cause changes in the possibility of equipment failure, which are defined as influencing factors f.

[0119] Step 7.1.2: Determination of factors influencing the probability of real-time dynamic equipment failure:

[0120] With α 10 ,α 20 ,α 30 ,…,α n0 This represents the design baseline values ​​for various monitoring parameters, including the design temperature, design pressure, design flow rate, and design allowable total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas. The influencing factors under design conditions are calculated as: = f(α) 10 ,α 20 ,α 30 ,…,α n0 ).

[0121] Using α1, α2, α3, ..., α n The parameters include real-time monitoring values ​​of various parameters within the equipment and facility operating temperature, operating pressure, operating flow rate, and actual operating conditions such as total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas. The influencing factors under operating conditions are calculated as f(α1, α2, α3, ..., α...). n ).

[0122] The impact factor of the real-time dynamic failure probability of the equipment is expressed as:

[0123]

[0124] Step 7.3 specifically includes the following steps:

[0125] Step 7.3.1: Characterization of the real-time dynamic risk level of the equipment, as shown in formula (2):

[0126] R(t) = [k·P f (t)]×C(t) (2);

[0127] In the formula, R(t) represents the real-time dynamic risk of equipment and facilities, k represents the value of the real-time dynamic failure probability influencing factor of equipment, and P f C(t) is the probability value of equipment failure under operating conditions, and C(t) is the failure consequence.

[0128] Step 7.3.2: Criteria for determining the probability level of real-time failure;

[0129] The real-time failure probability judgment criterion draws on the static RBI assessment failure probability level judgment criterion, where the original RBI assessment failure probability P f (t) using [k·P f (t)] is used instead, according to [k·P] f The numerical range of (t)] is used to determine the level of real-time failure probability of equipment and facilities;

[0130] Step 7.3.3: Real-time failure consequence level discrimination matrix;

[0131] The criteria for judging the consequences of real-time failures are based on the criteria for judging the severity of failures in static RBI assessments.

[0132] Step 7.3.4: Dynamic RBI Risk Assessment Matrix;

[0133] The dynamic risk assessment matrix for equipment and facilities draws on the static RBI assessment risk assessment matrix.

[0134] Figure 2 A schematic diagram illustrating the principle of a method for determining real-time dynamic risk levels.

[0135] Figure 3 (a) and (b) are schematic diagrams of the real-time dynamic risk level discrimination matrix for equipment and facilities.

[0136] like Figure 3 As shown in (a), the horizontal axis represents the failure consequence level, with A to E indicating increasingly higher failure consequence levels; the vertical axis represents the failure probability level, with 1 to 5 indicating increasingly higher failure probability levels; as... Figure 3 As shown in (b), the risk levels are divided into four levels: low risk (L), medium risk (M), medium-high risk (MH), and high risk (H).

[0137] Step 7.4 specifically includes the following steps:

[0138] Step 7.4.1: Health characterization of equipment and facilities;

[0139] By utilizing the k-value, which is an influencing factor for the real-time dynamic failure probability of equipment, the health status of static equipment can be expressed in terms of numerical value.

[0140] If the magnitude of the real-time dynamic monitoring process parameters is directly proportional to the probability of failure, then the health status is expressed as H = 1 / k; if the magnitude of the real-time dynamic monitoring process parameters is inversely proportional to the probability of failure, then the health status is expressed as H = k.

[0141] The health status of static equipment is characterized by formula (3):

[0142]

[0143] In the formula, α represents the magnitude of the real-time dynamic monitoring process technology parameter, f represents the magnitude of the failure probability, k represents the influence factor of the real-time dynamic failure probability of the equipment, k∈(0,1), then H∈(0,1);

[0144] Step 7.4.2: Health assessment criteria for equipment and facilities;

[0145] The fuzzy comprehensive membership method is used to define the health evaluation criteria for equipment and facilities (as shown in Table 1). The health of equipment and facilities is scored using four indicators: excellent, good, permissible, and unacceptable. The membership function is defined as ([1,0.75),[0.75,0.5),[0.5,0.25),[0.25,0)). Table 2 shows the criteria for judging the failure probability level of pressure-bearing equipment and facilities.

[0146] Table 2 Health Evaluation Criteria for Pressure Equipment and Facilities

[0147]

[0148]

[0149] Step 7.4 specifically includes the following steps:

[0150] Assume the main monitoring parameter of a pipeline is the operating temperature α1, and the design temperature is α. 10 Then the operating temperature α1 ≥ α 10 ;

[0151] The factor affecting the likelihood of failure is the corrosion rate; if the corrosion rate increases with increasing operating temperature.

[0152] When the pipeline operating temperature is lower than the design temperature

[0153] When the pipeline operating temperature equals the design temperature

[0154] As the operating temperature of the pipeline gradually increases, the corrosion rate f(α1) gradually increases, and the health status H=1 / k gradually decreases. That is, the worse the health status of the pipeline, the higher the risk level of the equipment and the causes of the risk can be calculated.

[0155] The invention will now be verified using specific examples:

[0156] Taking a risk assessment of a piece of equipment on a CNOOC platform as an example, the corrosion rate is the main object of observation, and the impact of its change with temperature is monitored.

[0157] Conventional RBI risk assessment principles, such as Figure 4 As shown;

[0158] Data collection

[0159] This mainly includes the container's design and as-built documentation, production process data, and maintenance and inspection data. It includes the following documents:

[0160] (1) Design and as-built documents: including container as-built drawings, calculation sheets, quality certificates, inspection data, on-site nameplates and photos of equipment and central control flow diagrams, etc.

[0161] (2) Process data: mainly includes the operating pressure of the vessel, operating temperature, process flow diagram, pipeline and instrumentation flow diagram, media and flow rate at the outlet and inlet, chemical analysis report of sampling points, etc.

[0162] (3) Operation data: including operating procedures and operation records;

[0163] (4) Corrosion investigation: including critical equipment, potential failure modes and past accidents and their causes;

[0164] (5) Consequence-related information: including equipment asset details, platform structure design drawings, number of platform personnel, maintenance costs and production stoppage losses, etc.

[0165] Evaluation Unit Division

[0166] In RBI software, containers are managed according to a hierarchical relationship of platform-equipment-equipment component. Each component refers to a part of the equipment based on its structure; for example, a vertical tank is typically divided into upper head, lower head, and shell as equipment components. Separators, heat exchangers, and filters are managed and analyzed according to multiple equipment components. This division is because data and media may differ between different equipment components, leading to variations in corrosion conditions and requiring separate risk level studies.

[0167] For equipment that has been divided into assessment units, in addition to the equipment data, data for the components resulting from the equipment division must also be entered; equipment that has not been divided into parts is managed as a whole. The equipment segmentation method is as follows:

[0168] Equipment Introduction

[0169] The metering separator V-1301 is a horizontal single-cavity cylindrical container. Natural gas is contained in the upper part of the left end cap, the upper part of the cylinder, and the upper part of the right end cap; an oil-water mixture is contained in the lower part of the left end cap and the lower left part of the cylinder; and oil is contained in the lower right part of the cylinder and the lower right end cap. Based on the end cap-cylinder-end cap structure and the distribution of the media, the tank is divided into 7 evaluation units, as shown in Table 2.

[0170] Table 2. Division of Evaluation Units

[0171] Evaluation Unit Part medium 1 upper part of the left cap natural gas 2 Lower part of left end cap oil, water 3 Upper part of the cylinder natural gas 4 Lower left part of the cylinder oil, water 5 Lower right part of the cylinder Oil 6 upper part of the right cap natural gas 7 Lower part of right header Oil

[0172] Table 5 shows the basic data for the evaluation units:

[0173] Table 5 Basic Data of Evaluation Units

[0174]

[0175] Corrosion circuit division

[0176] The principle for dividing corrosion loops is to group interconnected equipment with the same corrosion mechanism (generally, the same material, medium, temperature within the same corrosion mechanism range, and insulation conditions) into one corrosion loop. Based on the corrosion mechanism of offshore platform containers and corrosion mechanism identification methods, combined with process flow, logistics data, the most recent medium sampling and testing data, corrosion analysis data, equipment material and medium, operating temperature and operating pressure, the corrosion type of the container is determined and corrosion loops are divided.

[0177] Risk calculation principle

[0178] The risk of a container comes from two aspects: the probability of failure of the evaluation unit and the consequences of its failure. The formula for calculating the risk R(t) is shown below.

[0179] R(t)=P f (t)·C(t)

[0180] (1)P f (t) represents the failure probability, reflecting the expected failure probability of the container, and is calculated using the following formula:

[0181] P f (t)=gff·D f (t)

[0182] In the formula, gff is the general failure frequency of the evaluation unit, which is based on the typical representative value of failure data in the petroleum industry as the general failure frequency of different types of equipment.

[0183] (2)D f (t)——Damage factor; The damage factor is determined based on the analyzed damage mechanism (uniform or local thinning, cracking, creep, etc.), which is related to the material, operating conditions, service status, and inspection techniques for quantifying damage.

[0184] (1) C(t) represents the failure consequence, and quantitative evaluation is carried out according to the following process:

[0185] a) Select representative leaked media and their physical properties;

[0186] b) Select the size of the leakage hole;

[0187] c) Calculate the theoretical leakage rate;

[0188] d) Estimate the total leakage amount;

[0189] e) Determine the type of leak;

[0190] f) Estimate the impact of the detection and isolation system on the leakage amount;

[0191] g) Determine the final leakage rate and leakage amount;

[0192] h) Calculate the consequences of the area calculation.

[0193] It can be seen that risk is a function of time accumulation; as thinning, cracking, or other damage mechanisms accumulate over operating time, the damage factor gradually increases. When multiple damage mechanisms exist in a device unit, each damage mechanism is calculated separately, resulting in a time-dependent risk. By comparing the risks corresponding to each damage mechanism, the highest risk is selected as the risk of the device unit.

[0194] Failure probability calculation

[0195] RBI analysis calculates the probability of failure by assuming that if equipment damage occurs faster than expected, an unforeseen catastrophic accident may occur. The confidence level is determined based on the effectiveness of past inspection methods in detecting various forms and rates of damage. The probability of failure is generally obtained using limit state analysis and the reliability index method.

[0196] Failure Consequence Calculation

[0197] Failure consequences include personnel safety consequences, environmental consequences, and economic consequences, such as Figure 5-7 As shown. Safety consequences are expressed in terms of potential loss of life; economic consequences (including costs of production stoppage and equipment repair) are expressed in terms of an appropriate monetary amount; and environmental consequences are expressed in terms of the mass or volume of pollutants released into the environment, or the costs of environmental cleanup due to pollutant spills.

[0198] Risk Matrix

[0199] To facilitate the risk ranking of containers, a 5×5 matrix diagram method was adopted. In the matrix diagram, the vertical failure probability is divided into five levels: 1, 2, 3, 4, and 5 according to the failure probability coefficient, and the horizontal failure consequences are divided into five levels: A, B, C, D, and E according to the economic loss or safety impact after failure.

[0200] The operating temperature of the metering separator is 60°C, and the design temperature is 90°C. Therefore, the operating temperature α1 < α 10 The corrosion rate is a key factor influencing the likelihood of failure. The corrosion rate increases with increasing operating temperature. Currently, the equipment's operating temperature is lower than the design temperature. As the operating temperature of the current equipment gradually increases, the corrosion rate f(α1) gradually increases, and the health status H = 1 / k gradually decreases, meaning the current health status of the equipment deteriorates. The risk level of the current equipment and the causes of this risk are calculated.

[0201] Table 6 presents the current risk level and underlying causes of the containers involved in the assessment.

[0202] Table 6. Current Risk Level and Cause Analysis of Containers Participating in the Assessment

[0203]

[0204] Petrochemical plants are always in a dynamic process, and the medium environment in which the equipment is located is also dynamically changing, such as changes in temperature, flow rate, and medium composition. These changes can affect the risk of the equipment and cause it to change. This method can adjust parameters based on the dynamic changes of the equipment through a real-time dynamic risk assessment model, thus realizing a data-driven dynamic risk assessment technology.

[0205] This technology allows for more accurate prediction of equipment failure risks and the probability of failure. Based on data, it enables risk estimation of equipment and provides a more suitable evaluation standard for equipment risk in my country compared to traditional RBI technology. Furthermore, it allows for the development of reasonable and effective countermeasures through expert groups. In particular, for the petrochemical industry, timely detection of failure risks can prevent oil or fuel leaks and reduce the occurrence of accidents.

[0206] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A data-driven dynamic risk assessment method, characterized in that: Includes the following steps: Step 1: Determine equipment boundaries; Step 2: Failure Mode and Effects Analysis; Step 3: Analysis of the dominant failure mechanism under the interaction of multiple failure modes; Step 4: Dynamic characteristic analysis of factors influencing the dominant failure mode; Step 5: Analysis of the dynamic interaction characteristics of influencing factors; Step 6: Failure probability analysis and consequence assessment; Step 7: Conduct real-time dynamic risk assessment; In step 7, real-time dynamic risk assessment retains the failure probability in traditional RBI. and consequences of failure Unchanged, based on real-time corrosion monitoring parameters to assess the likelihood of failure. The influence of the changes in equipment's real-time dynamic failure probability factor k is obtained, thereby realizing the dynamic change of risk; Current dynamic change in failure probability After corresponding to the original level classification range, a new failure probability level is obtained. Combined with the original failure consequence level, a dynamic risk matrix is ​​formed as the real-time dynamic failure probability influencing factor k of the equipment changes, thus realizing dynamic RBI risk assessment. Step 7 specifically includes the following steps: Step 7.1: Determine the influencing factor k of the real-time dynamic failure probability of the equipment; Corrosion rate, operating hours, and remaining service life directly reflect the corrosion status of equipment and are defined as influencing factors. The changes in each monitoring parameter are directly related to the changes in influencing factors. When obtaining multiple monitoring parameters to calculate the magnitude of influencing factors, This represents the real-time monitoring values ​​of each monitoring parameter, and the influencing factors are calculated as follows: , The design baseline values ​​for each monitoring parameter are represented by the following: The influencing factor for the real-time dynamic failure probability of the equipment is expressed as follows: ； When multiple monitoring parameters are difficult to obtain, or when there are monitoring parameters that have a major impact on corrosion, a single monitoring parameter is selected to fit the relationship between the parameter and the influencing factors. For example, assuming a pipeline's operating temperature is among the parameters... pH value Workflow All monitoring parameters, including temperature, affect the corrosion rate. Temperature has the most significant impact on the corrosion rate, therefore, the operating temperature is determined as the primary monitoring parameter, and its relationship with the influencing factors is fitted as follows: , Representing the design temperature, the influence coefficient k of the failure probability is expressed as: ； Step 7.2: Determine the dynamic failure probability level; The dynamic RBI failure probability is obtained by correcting the static RBI assessment failure probability using the real-time dynamic failure probability influence factor k. The current dynamic change in the probability of failure By placing it into the original failure probability level classification criteria determined by the RBI assessment, a new dynamic RBI failure probability level is obtained. Step 7.3: Determine the real-time dynamic risk level; Using the original static RBI assessment knowledge base to evaluate the failure consequence level, combined with the dynamic failure probability level determined in step 7.2, the risk is calculated through risk logic. Then, based on the risk level assessment matrix determined by the original static RBI assessment, the real-time dynamic risk level is determined. Step 7.4: Determine the real-time device health status; Equipment health is a normalized metric that measures the health of equipment. It is an assessment of the degree of deviation between the current state and the expected state of the equipment. The normalized metric is in the range of (0,1). The desired state refers to the design reference state of similar equipment or equipment under the same operating conditions.

2. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 1, the purpose of determining the equipment boundary is to define the scope of analysis so that factors affecting equipment safety and reliability can be analyzed within a limited range, with the controlled unit as the object of analysis. After the controlled equipment boundary is selected, risk analysis is carried out, mainly analyzing factors affecting equipment reliability within the boundary range of the controlled unit. Information interaction between the controlled unit and the outside world can be used as input or output parameters of the controlled unit. Once the equipment boundary is determined, the real-time risk analysis only considers the impact of the boundary and input / output parameters on real-time risk.

3. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 1, the equipment boundary can be defined by a specific piece of equipment and its connected piping system, by a single piece of equipment with an independent function and its associated piping, or by multiple pieces of equipment and their associated piping that perform a specific function.

4. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 2, failure mode and factor analysis is to analyze the possible causes, influencing factors and consequences of equipment failure.

5. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 3, the various influencing factors that cause equipment failure are analyzed, the dominant failure mechanism is identified, and it is characterized as equipment risk.

6. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 4, the key parameters of the dominant failure mode affecting equipment failure in the petrochemical plant are constantly and dynamically adjusted. Their dynamic characteristics may be manifested as fluctuations in parameters over time or changes in the combination of multiple key parameters. Conducting dynamic analysis on the influencing factors of the dominant failure mode is a key step in realizing dynamic hazard source risk analysis.

7. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 5, dynamic interaction analysis mainly refers to analyzing the impact of interactions between different key parameters, between key parameters and manual operations, and between devices on the real-time risk of the equipment; dynamic interaction analysis mainly studies the interrelationships of different influencing factors and their impact on the system.

8. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 6, the probability of failure is analyzed based on the dominant failure mode and influencing factors. When conducting risk analysis, the worst-case scenario is assessed based on the principle of conservatism to determine the magnitude of the risk.

9. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 7.1, the influence factor k of the real-time dynamic failure probability of the equipment is defined, which specifically includes the following steps: Step 7.1.1: Analysis of key influencing factors on the likelihood of equipment and facility failure; Different corrosion mechanisms correspond to different monitoring parameters. Correlation analysis was conducted on parameters such as operating temperature, operating pressure, operating flow rate, total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas to screen out key influencing factors that can cause changes in the probability of equipment failure, which were defined as influencing factors. ; Step 7.1.2: Determination of factors influencing the probability of real-time dynamic equipment failure: use This represents the design baseline values ​​for various monitoring parameters, including the design temperature, design pressure, design flow rate, and design allowable total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas. The influencing factors under design conditions are calculated as follows: ; use This represents the real-time monitoring values ​​of various parameters within the equipment and facility operating temperature, operating pressure, operating flow rate, and actual operating conditions, including total acid value, crude oil sulfur content, pH value, partial pressure of H2S gas, and partial pressure of CO2 gas. Influencing factors under operating conditions are calculated as follows: ; The impact factor of the real-time dynamic failure probability of the equipment is expressed as: （1）。 10. The data-driven dynamic risk assessment method according to claim 1, characterized in that: In step 7.2, the dynamic RBI failure probability level classification is as follows: Failure probability level Failure probability 1 0.00000＜ ≤0.00001 2 0.00001＜ ≤0.00010 3 0.00010＜ ≤0.00100 4 0.00100＜ ≤0.01000 5 0.01000＜ ≤0.1000。 11. The data-driven dynamic risk assessment method according to claim 1, characterized in that: Step 7.3 specifically includes the following steps: Step 7.3.1: The real-time dynamic risk level of the equipment is characterized as shown in formula (2): （2）； In the formula, Real-time dynamic risk assessment of equipment and facilities. This represents the numerical value of the real-time dynamic failure probability factor for the equipment. It is the probability value of equipment and facilities failing under operating conditions. The consequences of failure; Step 7.3.2: Criteria for determining the probability level of real-time failure; The real-time failure probability assessment criterion draws on the static RBI assessment failure probability level assessment criterion, where the original RBI assessment failure probability... use To replace, according to The numerical range is used to determine the level of real-time failure probability of equipment and facilities; Step 7.3.3: Real-time failure consequence level discrimination matrix; The criteria for judging the consequences of real-time failures are based on the criteria for judging the severity of failures in static RBI assessments. Step 7.3.4: Dynamic RBI Risk Assessment Matrix; The dynamic risk assessment matrix for equipment and facilities draws on the static RBI assessment risk assessment matrix.

Images