CN113378480B - Condition-based maintenance method and system for underwater Christmas tree based on remaining service life prediction - Google Patents

Abstract

The invention belongs to the field of petroleum engineering, and particularly relates to an underwater Christmas tree condition maintenance method and system based on residual service life prediction. The method for maintaining the underwater Christmas tree according to the situation based on the residual service life prediction comprises the following six steps: the method comprises the steps of establishing a degradation impact model of each component of the underwater Christmas tree, establishing an incomplete maintenance model of each component of the underwater Christmas tree, establishing a residual service life prediction model of each component of the underwater Christmas tree, establishing a spare part model of the underwater Christmas tree system, establishing an optional maintenance model of the underwater Christmas tree system and determining a maintenance decision threshold of the underwater Christmas tree system. The underwater Christmas tree visual maintenance system based on the residual service life prediction comprises five parts: the system comprises an underwater Christmas tree production loop data acquisition module, an underwater Christmas tree annular loop data acquisition module, an underwater Christmas tree chemical agent injection loop data acquisition module, an underwater Christmas tree sensor data collection and storage module and an underwater Christmas tree maintenance decision subsystem.

Description

Condition-based maintenance method and system for underwater Christmas tree based on remaining service life prediction

technical field

The invention belongs to the field of petroleum engineering, and in particular relates to a condition-based maintenance method and system for an underwater Christmas tree based on remaining service life prediction.

Background technique

Subsea Christmas tree is the key facility of subsea production system and has been widely used in offshore oil exploitation. Subsea Christmas tree is mainly composed of wellhead connector, tubing hanger, plug, tree cap, tree body, valve and various passages and other system components. , control and regulate oil well production, guarantee operations, testing and wax removal and other daily production management. Subsea Christmas trees are less affected by the sea level environment and can be suitable for deepwater or ultra-deepwater oil and gas development, so they have attracted much attention and developed vigorously.

Because the underwater Christmas tree works on the seabed for a long time, its structure is complex and the operating conditions are complicated, which leads to the emergence of problems such as difficult installation, high maintenance cost, and difficulty in maintenance. Once the underwater Christmas tree fails, it will bring huge economic losses and even cause damage to the marine environment and casualties. Existing maintenance methods are usually regular maintenance, high maintenance costs, and prone to problems of over-maintenance and under-maintenance. Condition-based maintenance is a maintenance method that makes maintenance decisions based on the degradation status of components, which can effectively reduce the problems of over-maintenance and under-maintenance, and reduce maintenance costs while ensuring system safety. Therefore, there is an urgent need for a condition-based maintenance method and system for an underwater Christmas tree based on remaining service life prediction.

SUMMARY OF THE INVENTION

In order to overcome the defects of the prior art, the present invention provides a condition-based maintenance method and system for an underwater Christmas tree based on remaining service life prediction.

In order to achieve the above purpose, the maintenance method of underwater Christmas tree based on remaining service life prediction includes the following 6 steps:

S1: Establish a degradation impact model of each component of the underwater Christmas tree based on historical fault data, which includes the following steps:

S11: Establish the internal degradation model of each component of the underwater Christmas tree. Model the internal degradation process of each component of the subsea Christmas tree as a gamma process;

S12: Establish an external impact model of the marine environment of each component of the underwater Christmas tree. The external shock process of the marine environment of each component of the underwater tree is modeled as a Poisson process;

S2: According to the maintenance data of the underwater tree system and the degradation data after the maintenance, an incomplete maintenance model of each component of the underwater tree is established, which specifically includes the following steps:

S21: Establish a degradation state reduction model after each component of the subsea tree is incompletely repaired;

S22: Combine the times of incomplete maintenance of each component of the underwater Christmas tree, establish a degradation acceleration model after the incomplete maintenance of each component of the underwater Christmas tree;

S3: According to the historical fault data of the underwater Christmas tree system, a neural network algorithm-based remaining service life prediction model of each component of the underwater Christmas tree is established, which specifically includes the following steps:

S31: Establish a prediction model for the remaining service life of each component of the underwater Christmas tree under the normal degradation state based on the neural network algorithm;

S32: Establish a prediction model for the remaining service life of each component of the underwater Christmas tree under incomplete maintenance based on a neural network algorithm;

S4: Establish a spare parts model of the underwater Christmas tree system. The specific implementation of this step is as follows:

The spare parts model of the subsea tree system adopts the spare parts strategy of (s, S), where s is the minimum sum of spare parts of the subsea tree components owned by the subsea tree system, and S is the subsea tree system owned by the subsea tree system. The maximum number of spare parts combined for each component of the tree, with a maximum of one spare for each subsea tree component;

S5: Establish a maintenance model for the underwater tree system depending on the situation, and take the minimum ratio of the maintenance cost of the underwater tree system at the current maintenance time to the predicted value of the remaining service life of the underwater tree system after the maintenance as the optimization goal to determine the optimal maintenance time at the current time. The maintenance method of each component of the subsea Christmas tree, and the ordering status of the spare parts of the subsea Christmas tree system after the current maintenance time is determined according to the consumption of spare parts, which specifically includes the following steps:

S51: every unit cycle time T, the degradation status of each component of the underwater Christmas tree is obtained by diagnosing and analyzing the pressure, flow, temperature and leakage data collected by the sensors of each component of the underwater Christmas tree;

S52: Input the degradation state of each component of the underwater Christmas tree into the remaining service life prediction model of each component of the underwater Christmas tree under the normal degradation state, and obtain the predicted value of the remaining service life of each component of the underwater Christmas tree and the prediction of the remaining service life of the underwater Christmas tree system value;

S53: According to the degradation state of each component of the underwater tree, the predicted value of the remaining service life of each component of the underwater tree, and the predicted value of the remaining service life of the underwater tree system, and combined with the data of the spare parts of the underwater tree system, make maintenance decisions, which include: The following steps:

S531: When the predicted value of the remaining service life of the underwater Christmas tree system is higher than the safe remaining service life threshold ST of the underwater Christmas tree system, go to S51, otherwise, start the maintenance preparation;

S532: After the maintenance preparation work is completed, use the underwater tree system to maintain the genetic algorithm according to the situation to determine the maintenance method of each component of the underwater tree;

S533: After each component of the underwater Christmas tree is repaired according to the maintenance method determined in S532, according to the usage of the spare parts of the underwater Christmas tree system, determine the order quantity and order type of the spare parts of the underwater Christmas tree system;

S6: Taking the minimum maintenance cost per unit time of the subsea tree system as the goal, determine the optimal maintenance decision threshold of the subsea tree system, namely the safe remaining service life threshold ST of the subsea tree system and the spare parts strategy threshold of the subsea tree system ( s, S).

The subsea tree maintenance system based on the prediction of remaining service life, including five parts: subsea tree production circuit data acquisition module, subsea tree annulus loop data acquisition module, subsea tree chemical injection circuit data Acquisition module, subsea tree sensor data collection and storage module and subsea tree maintenance decision-making subsystem.

The subsea tree production loop data acquisition module includes production main valve sensor group, production wing valve sensor group, production isolation valve sensor group, well surface control downhole safety valve sensor group and production choke valve sensor group.

The subsea Christmas tree annular loop data acquisition module includes annulus main valve sensor group, annulus wing valve sensor group, switch valve sensor group and annulus entry valve sensor group.

The underwater Christmas tree chemical injection circuit data acquisition module includes a methanol injection valve sensor group, a chemical injection valve sensor group 1, and a chemical injection valve sensor group 2.

The subsea tree maintenance decision-making subsystem includes the degradation state diagnosis module of each component of the subsea tree, the remaining service life prediction module of each component of the subsea tree, the maintenance module of the subsea tree system according to the situation, and the spare parts library module of the subsea tree system. And subsea tree system maintenance decision result display module.

Compared with the prior art, the effective gain effect of the present invention is: the method and system for the maintenance of the underwater Christmas tree based on the remaining service life prediction, considering the real-time state of the underwater Christmas tree system during the maintenance, and at the same time the underwater oil production is carried out. The maintenance cost of the tree system is combined with the remaining service life of the underwater tree system, and a more reasonable maintenance method can be adopted for each component of the underwater tree. Compared with the regular maintenance strategy of the underwater tree system, the production of the underwater tree can be guaranteed. At the same time of safety, it can effectively reduce the over-maintenance and under-maintenance, thereby reducing maintenance costs and improving the maintenance level of the underwater Christmas tree system.

Description of drawings

Fig. 1 is a block diagram of the maintenance method of the underwater Christmas tree based on the remaining service life prediction;

Figure 2 is a model diagram of the degradation impact model of each component of the underwater Christmas tree;

Fig. 3 is the prediction model diagram of the remaining service life of each component of the underwater Christmas tree based on the neural network algorithm;

Fig. 4 is the maintenance flow chart of the underwater Christmas tree system depending on the situation;

Fig. 5 is the optimization flow chart of the genetic algorithm for maintenance of the underwater Christmas tree system according to the situation;

Fig. 6 is the coding diagram of the genetic algorithm for maintenance of the underwater Christmas tree system according to the situation;

Fig. 7 is the schematic diagram of the crossover operation and mutation operation of the genetic algorithm for maintenance of the underwater Christmas tree system according to the situation;

8 is a schematic diagram of an underwater Christmas tree system;

FIG. 9 is a schematic diagram of a condition-based maintenance system for an underwater Christmas tree based on remaining service life prediction.

In the figure, 101, subsea tree production circuit, 102, subsea tree production main valve, 103, subsea tree well surface control downhole safety valve, 104, subsea tree production wing valve, 105, subsea oil production tree production choke valve, 106, subsea tree production isolation valve, 107, subsea tree annulus circuit, 108, subsea tree annulus main valve, 109, subsea tree annulus wing valve, 110, Subsea Christmas Tree Conversion Valve, 111, Subsea Christmas Tree Annular Inlet Valve, 112, Subsea Christmas Tree Chemical Injection Circuit, 113, Subsea Christmas Tree Methanol Injection Valve, 114, Subsea Christmas Tree Chemical Injection Valve 1 , 115, underwater Christmas tree chemical injection valve II, 201, underwater Christmas tree production loop data acquisition module, 202, production main valve sensor group, 203, production main valve pressure sensor, 204, production main valve temperature sensor, 205 , production of main valve flow sensor, 206, production of main valve acoustic emission sensor, 207, production of wing valve sensor group, 208, production of wing valve pressure sensor, 209, production of wing valve temperature sensor, 210, production of wing valve flow sensor, 211, Production of wing valve acoustic emission sensor, 212, production of isolation valve sensor group, 213, production of isolation valve pressure sensor, 214, production of isolation valve temperature sensor, 215, production of isolation valve flow sensor, 216, production of isolation valve acoustic emission sensor, 217, Well surface control downhole safety valve sensor group, 218, well surface control downhole safety valve pressure sensor, 219, well surface control downhole safety valve temperature sensor, 220, well surface control downhole safety valve flow sensor, 221, well surface control downhole safety valve Acoustic emission sensor, 222, production throttle valve sensor group, 223, production throttle valve pressure sensor, 224, production throttle valve temperature sensor, 225, production throttle valve flow sensor, 226, production throttle valve acoustic emission sensor, 227, Subsea Christmas Tree Annulus Loop Data Acquisition Module, 228, Annular Main Valve Sensor Group, 229, Annular Main Valve Pressure Sensor, 230, Annular Main Valve Temperature Sensor, 231, Annular Main Valve Flow Sensor, 232 , annular main valve acoustic emission sensor, 233, annular wing valve sensor group, 234, annular wing valve pressure sensor, 235, annular wing valve temperature sensor, 236, annular wing valve flow sensor, 237, annular wing Valve acoustic emission sensor, 238, switch valve sensor group, 239, switch valve pressure sensor, 240, switch valve temperature sensor, 241, switch valve flow sensor, 242, switch valve acoustic emission sensor, 243, annulus entry valve sensor group, 244, annular inlet valve pressure sensor, 245, annular inlet valve temperature sensor, 246, annular inlet valve flow sensor, 247, annular inlet valve acoustic emission sensor, 248, underwater Christmas tree chemical injection loop data acquisition module , 249, methanol injection valve sensor group, 250, methanol injection valve pressure sensor, 251, methanol injection valve Temperature sensor, 252, methanol injection valve flow sensor, 253, methanol injection valve acoustic emission sensor, 254, chemical injection valve-sensor group, 255, chemical injection valve-pressure sensor, 256, chemical injection valve-temperature sensor, 257, chemical injection valve, a flow sensor, 258, chemical injection valve, acoustic emission sensor, 259, chemical injection valve, sensor group, 260, chemical injection valve, pressure sensor, 261, chemical injection valve, temperature sensor, 262. Second flow sensor of chemical injection valve, 263, Second acoustic emission sensor of chemical injection valve, 264, Subsea Christmas tree sensor data collection and storage module, 301, Subsea Christmas tree maintenance decision-making subsystem, 302, Subsea oil production Degradation status diagnosis module of each component of the tree, 303. Remaining service life prediction module of each component of the subsea tree, 304, maintenance module of the subsea tree system depending on the situation, 305, spare parts library module of the subsea tree system, 306, subsea oil production The tree system maintenance decision result display module.

Detailed ways

The present invention will now be described in further detail with reference to the accompanying drawings.

As shown in Figure 1, the maintenance method of underwater Christmas tree based on remaining service life prediction includes the following 6 steps:

S1: As shown in Figure 2, the degradation impact model of each component of the underwater Christmas tree is established according to the historical fault data, which includes the following steps:

S11: Establish the internal degradation model of each component of the underwater Christmas tree. The internal degradation process of each component of the subsea tree is modeled as a gamma process, then the degradation amount x _T of each component of the subsea tree per unit cycle time T is independent of each other and obeys the gamma distribution, as shown below:

Among them, f(x _T ,α,β) is the gamma distribution density function, α is the shape parameter of the gamma distribution, β is the inverse scale parameter of the gamma distribution, Γ(x _T ) is the gamma function, and the gamma distribution The shape parameters and inverse scale parameters of are determined from historical data.

The total internal degradation X _t of each component of the subsea tree in t unit cycle time is:

Among them, x _T ^k is the degradation amount of each component of the subsea tree at the kth unit cycle time, and k is the unit cycle time number;

S12: Establish an external impact model of the marine environment of each component of the underwater Christmas tree. The external shock process of the marine environment of each component of the underwater tree is modeled as a Poisson process. For any time period t ₁ , t ₂ ≥ 0, we have

Among them, n is the number of times that each component of the subsea tree is subjected to external shocks from the marine environment, N _c (t ₁ +t ₂ ) is the number of external shocks to the marine environment during the time period t ₁ +t ₂ , and N _c (t ₁ ) is The number of external shocks to the marine environment in the time period t ₁ , P{N _c (t ₁ +t ₂ )-N _c (t ₁ )=n} is the probability of n times of external shocks to the marine environment in any time period t ₂ , λ is a parameter of Poisson distribution, which is determined by historical data.

The intensity x _s of the external impact of the marine environment on each component of the underwater Christmas tree, that is, the amount of degradation caused by the external impact of the marine environment, obeys a normal distribution, as shown below:

Among them, f(x _s ) is the density function of the normal distribution, μ is the mean value of the external impact intensity of the marine environment, and σ is the variance of the external impact intensity of the marine environment, which is determined by historical data.

The total amount of external shocks to the marine environment X _S is:

Among them, x _Sh is the amount of the ^h -th external shock to the marine environment, Ns is the number of external shocks to the marine environment, and h is the number of external shocks to the marine environment.

The degradation state X of each component of the underwater Christmas tree is the sum of the total internal degradation X _t and the total external impact X _S of the marine environment, namely:

X=X _t +X _S

In the absence of maintenance and replacement, the degradation status of each component of the subsea tree will only increase with time.

S2: According to the maintenance data of the underwater tree system and the degradation data after the maintenance, an incomplete maintenance model of each component of the underwater tree is established, which specifically includes the following steps:

S21: Establish a degradation state reduction model after incomplete maintenance of each component of the subsea Christmas tree, which specifically includes the following contents:

According to the maintenance data of the underwater tree and the degradation data after the maintenance, it is obtained that after the incomplete maintenance of the components of the underwater tree, the degradation state X _j is lower than the degradation state X before the maintenance, but higher than the degradation state after the last incomplete maintenance State X _j-1 .

The degradation state distribution of each component of the subsea tree after incomplete repair is modeled as a uniform distribution from X _j-1 to (XX _j-1 )·0.6+X _j-1 , as follows:

Among them, f _Xj is a uniformly distributed density function, and j is the number of incomplete repairs of components;

S22: Combine the times of incomplete maintenance of each component of the subsea tree, and establish a degradation acceleration model after the incomplete maintenance of each component of the subsea tree, which specifically includes the following contents:

After incomplete maintenance of each component of the subsea tree, the degradation acceleration is expressed in the parameters of the internal degradation model and the external impact model of the marine environment as follows:

α _j =α+κ·j

μ _j =τ ^j ·μ

Among them, α _j is the shape parameter of the internal degradation gamma distribution after the jth incomplete repair, and μ _j is the average value of the external impact strength of the marine environment for each component of the underwater tree after the jth incomplete repair. τ and κ are degradation acceleration coefficients, τ>1, determined by historical data;

S3: As shown in Figure 3, according to the historical fault data of the underwater Christmas tree system, a neural network algorithm-based remaining service life prediction model of each component of the underwater Christmas tree is established, which specifically includes the following steps:

S31: Establish a prediction model for the remaining service life of each component of the underwater Christmas tree under the normal degradation state based on the neural network algorithm. The prediction model of the remaining service life of each component of the _{subsea tree} based on the neural network algorithm in the normal degradation state The number of cycle times t and t-1 and the number of incomplete repairs j in the current degraded state are used as the input of the neural network, input into the input layer of the neural network, pass through the hidden layer of two layers of 3 nodes R, and output underwater oil production in the output layer The remaining service life data Rul of each component of the tree in the normal degradation state;

S32: Establish a prediction model for the remaining service life of each component of the underwater Christmas tree under incomplete maintenance based on a neural network algorithm. The prediction model of the remaining service life of each component of the subsea Christmas tree based on the neural network algorithm is based on the incomplete maintenance of the components of the _subsea Christmas tree. The degradation state X _j of each component of the Christmas tree after incomplete maintenance and the number of incomplete maintenance j are used as the input of the neural network, input to the input layer of the neural network, and pass through the hidden layer of two layers of 3 nodes R, and output the underwater oil production in the output layer The remaining service life data Rul under the incomplete maintenance of each component of the tree;

After several trainings, a prediction model for the remaining service life of each component of the underwater Christmas tree is obtained with the prediction accuracy meeting the requirements of use;

S4: Establish a spare parts model of the underwater Christmas tree system. The specific implementation of this step is as follows:

The spare parts model of the subsea tree system adopts the (s, S) spare parts strategy, where s is the minimum sum of spare parts of the subsea tree components owned by the subsea tree system, and S is the subsea tree system owned by the subsea tree system. The maximum number of spare parts combined for each component of the tree, with a maximum of one spare for each subsea tree component. The initial quantity of spare parts for the subsea tree system is S. When the subsea tree system is repaired, the subsea tree components are replaced, that is, the spare parts are used. The predicted value of the remaining service life of each component of the tree is sorted from low to high as the order of ordering spare parts, and the spare parts of each component of the subsea tree are ordered, so that the amount of spare parts of the subsea tree system can be supplemented to S. Ordering spare parts, incurring ordering fees, spare parts can only be used for replacement after the end of the order period, and spare parts storage costs will be incurred until they are not used;

S5: As shown in Figure 4, establish a maintenance model for the underwater tree system depending on the situation, and take the minimum ratio of the maintenance cost of the underwater tree system at the current maintenance time to the predicted value of the remaining service life of the underwater tree system after maintenance as the optimization goal Determine the optimal maintenance method for each component of the subsea Christmas tree at the current maintenance time, and determine the ordering status of the spare parts of the subsea Christmas tree system after the current maintenance time according to the consumption of spare parts, which specifically includes the following steps:

S51: every unit cycle time T, by diagnosing and analyzing the pressure, flow, temperature and leakage data collected by the sensors of each component of the underwater Christmas tree, to obtain the degradation state of each component of the underwater Christmas tree;

S52: Input the degradation state of each component of the underwater Christmas tree into the remaining service life prediction model of each component of the underwater Christmas tree under the normal degradation state, and obtain the predicted value of the remaining service life of each component of the underwater Christmas tree and the prediction of the remaining service life of the underwater Christmas tree system value, including the following:

Input the current degradation status and working life of each component of the subsea tree, the degradation status and working life of the previous moment, and the number of incomplete repairs at the current moment into the neural network algorithm-based neural network algorithm for the remaining components of the subsea tree under the normal degradation state. The service life prediction model is used to obtain the predicted value of the remaining service life of each component of the underwater Christmas tree, and at the same time, the predicted value of the remaining service life of the underwater Christmas tree system Rul _sys is defined as the minimum value of the remaining service life prediction value of each component of the underwater Christmas tree, As follows:

Rul _sys =min(Rul ₁ ,Rul ₂ ,...,Rul _N )

Among them, N is the number of subsea tree components, Rul ₁ is the predicted value of the remaining service life of the first component of the subsea Christmas tree, Rul ₂ is the predicted value of the remaining service life of the second subsea tree component, and Rul _i is The predicted value of the remaining service life of the ith component of the underwater Christmas tree, and Rul _N is the predicted value of the remaining service life of the Nth component of the underwater Christmas tree;

S53: According to the degradation state of each component of the underwater tree, the predicted value of the remaining service life of each component of the underwater tree, and the predicted value of the remaining service life of the underwater tree system, and combined with the data of the spare parts of the underwater tree system, make maintenance decisions, which include: The following steps:

S531: When the predicted value of the remaining service life of the underwater Christmas tree system is higher than the safe remaining service life threshold ST of the underwater Christmas tree system, go to S51, otherwise, start the maintenance preparation;

Repair preparation includes leasing repair boats, preparing repair tools, and hiring repair personnel. Maintenance preparation costs include rental maintenance boat costs, maintenance tool costs and maintenance personnel costs. During preparations for maintenance, the subsea tree system remains in working condition and continues to degrade, causing downtime losses if the subsea tree system fails. The time consumed by the maintenance preparation work is the maintenance preparation time LT;

S532: After the maintenance preparation work is completed, use the underwater tree system to maintain the genetic algorithm according to the situation to determine the maintenance method of each component of the underwater Christmas tree, which specifically includes the following contents:

According to the spare parts of the underwater tree system, the components of the underwater tree are divided into those with spare parts and those without spare parts. There are three options for subsea Christmas tree components with spare parts: no maintenance, incomplete maintenance and replacement. For subsea Christmas tree components without spare parts, there are only two options: no maintenance and incomplete maintenance.

The maintenance cost C _m at the current maintenance time of the subsea Christmas tree system includes the maintenance preparation cost c _S of the subsea Christmas tree system and the maintenance cost c _g of each component of the subsea Christmas tree. The maintenance cost of each component of the subsea Christmas tree c _g includes the preventive incomplete maintenance cost C ^ipm and the preventive replacement cost C ^pr when each component of the subsea Christmas tree is functioning normally, and the post-event incomplete maintenance when each component of the subsea Christmas tree fails Costs C ^icm and ex post replacement costs C ^cr . Because the degradation degree of each component of the subsea tree is more serious when it fails, and the maintenance is more difficult, the cost of incomplete repair and replacement after the event of each component of the subsea tree is higher than the cost of preventive incomplete maintenance and preventive replacement. , the maintenance cost C _m at the current maintenance time of the subsea tree system is as follows:

where i is the subsea tree assembly number, N is the number of subsea Christmas tree assemblies, c _g ⁱ is the maintenance cost of the i-th subsea tree assembly, and c _i ^ipm is the preventive incomplete maintenance cost of component i , ^{ci icm} is the post-incomplete maintenance cost of component _i , ^{ci pr} is the preventive replacement cost of component _i , _ci ^cr is the post-replacement cost of component i, q _i is the preventive incomplete maintenance coefficient of component i, y _i is the post-incomplete maintenance coefficient of component _i , ri is the preventive replacement coefficient of component _i , pi is the post-event replacement coefficient of component _i , when qi is 1, it means that component i adopts preventive incomplete maintenance, When y _i is 1, it means that component _i adopts post-incomplete maintenance; when ri is 1, it means that component _i adopts preventive replacement; when pi is 1, it means that component _i adopts post-event replacement; when qi and y When _i , ri , and p _i are all 0, it means that component _i is operated without maintenance.

For subsea Christmas tree components that take no maintenance operation, the degradation state and the predicted value of remaining service life remain unchanged; for subsea Christmas tree components that take replacement operation, the degradation state is 0, and the predicted value of remaining service life becomes the initial value; The subsea Christmas tree components in operation, according to the incomplete maintenance model of each component of the subsea Christmas tree, the estimated value X _jm of the degradation state of the subsea Christmas tree components after the incomplete maintenance is defined as the average value of the degradation state after the incomplete maintenance is: ,As follows:

X _jm =0.5·[(XX _j-1 )·0.6+2·X _j-1 ]

Input the degradation state X _t before the maintenance of the subsea tree components under incomplete maintenance, the unit cycle time t of the work before the maintenance, the estimated value X _jm of the degradation state after the incomplete maintenance, and the number of incomplete maintenance j input into the neural network-based algorithm The prediction model of the remaining service life of each component of the subsea Christmas tree under the condition of incomplete maintenance can obtain the predicted value of the remaining service life of each component of the subsea Christmas tree after the maintenance. The minimum value of the remaining service life prediction value of each component of the subsea tree is taken as the remaining service life prediction value Rul _sys-m of the subsea tree system after maintenance.

The optimization objective is to minimize the ratio of the maintenance cost at the current maintenance time of the underwater tree system to the predicted value of the remaining service life of the underwater tree system after the maintenance, as shown below:

The maintenance method of each component of the subsea Christmas tree is determined by using the genetic algorithm of the maintenance of the subsea tree system according to the situation. The coding method is shown in Figure 5, 0 means no maintenance, 1 means incomplete maintenance, and 2 means replacement. A chromosome includes 12 codes, which respectively represent the maintenance methods of 12 corresponding components. The maintenance method of the subsea tree component is affected by the status of the spare parts of the subsea tree component. When the component has spare parts, the spare part status is 1, and when the component has no spare parts, the spare part status is 0. Components with spare parts have three options of 0, 1, and 2, and components without spare parts have two options of 0, 1.

The optimization process of the genetic algorithm for the maintenance of the underwater tree system according to the situation is shown in Figure 6. The specific optimization process is as follows:

First, the population is initialized, and multiple sets of chromosomes are randomly generated.

In order to avoid the possible maintenance combination that is far from the optimal solution due to the randomness of the genetic algorithm, it is necessary to limit the maintenance method of the components that meet certain requirements, and filter the chromosomes that do not meet the requirements, as follows:

Step 1: When the subsea tree assembly has no spare parts, if the predicted value of the remaining service life of the assembly is lower than the safety threshold, an incomplete repair must be carried out;

Step 2: When the subsea tree component has spare parts, if the predicted value of the remaining service life of the component is lower than the safety threshold, incomplete repair or replacement must be carried out;

Step 3: When the subsea tree component has spare parts, if the component fails and the number of incomplete repairs is greater than 3, the component must be replaced;

Step 4: When the subsea Christmas tree component has spare parts, the component degradation state is lower than 0.7 and the component has not been incompletely repaired, and the component is not replaced;

Step 5: When the degradation state of the subsea tree components is lower than 0.4, the components are not fully repaired and replaced;

Then, the maintenance cost of the underwater tree system at the current maintenance time and the predicted value of the remaining service life of the underwater tree system after the maintenance are calculated. The ratio of the maintenance cost at the current maintenance moment of the underwater Christmas tree system to the predicted value of the remaining service life of the underwater Christmas tree system after the maintenance is taken as the fitness value W of each chromosome, as shown below:

Secondly, perform crossover operation, mutation operation and selection operation to select the optimal chromosome for the next iteration operation. The crossover operation and mutation operation are shown in Figure 7. The crossover operation is to randomly generate a crossover point, and exchange the position of the chromosome segments after the crossover point in adjacent chromosomes. A mutation operation is a random change of the coded value at a position on a chromosome. The new maintenance method obtained by the mutation operation must meet the requirements of the spare parts status of the subsea tree components, that is, the components without spare parts cannot be mutated into the replacement maintenance method. The selection operation uses the roulette method.

Finally, it is judged whether the termination condition is satisfied, that is, the maximum number of iterations. If the termination condition is satisfied, the optimal maintenance method of each component of the subsea Christmas tree is output; otherwise, the iterative process is continued until the termination condition is satisfied;

S533: After each component of the subsea tree is repaired according to the maintenance method determined in S532, according to the usage of the spare parts of the subsea tree system, determine the order quantity and order type of the spare parts of the subsea tree system, including the following contents:

The number of spare parts N _zh of the subsea tree system after maintenance is:

_Nzh = _SNr

where N _r is the number of component replacements.

When the number of spare parts N _zh is not lower than s, there is no need to order spare parts; when it is less than s, it is necessary to order spare parts, and the number of spare parts to be ordered N _or is:

N _or =SN _zh

The order type of the spare parts of the subsea tree system is determined according to the order of the predicted value of the remaining service life of each component of the subsea tree after maintenance from low to high;

S6: Taking the minimum maintenance cost per unit time of the subsea tree system as the goal, determine the optimal maintenance decision threshold of the subsea tree system, namely the safe remaining service life threshold ST of the subsea tree system and the spare parts strategy threshold of the subsea tree system ( s, S), including the following:

Determine the total maintenance cost _Cz of the subsea tree system. The total maintenance cost of the subsea tree system includes the maintenance cost C _m of each maintenance time of the subsea tree system during the total operating time of the subsea tree system t _z , the subsea tree system spare parts order cost c _bj and the storage cost c _cc and downtime losses due to subsea tree system failures as follows:

Among them, N _f is the total number of cycles with downtime loss, c _D is the downtime loss per unit cycle time, M is the total number of maintenance times of the subsea tree system, l is the number of maintenance times of the subsea tree system, and C ^l _m is Maintenance cost of the subsea tree system at the first maintenance time;

(ST,s,S)为目标，用遍历法确定水下采油树系统安全剩余使用寿命阈值ST和水下采油树系统备件策略阈值(s,S)，优化目标如下所示：Minimal maintenance cost per unit time of subsea tree system

(ST, s, S) as the target, the traversal method is used to determine the safe remaining service life threshold ST of the subsea tree system and the strategy threshold (s, S) of the spare parts of the subsea tree system. The optimization objectives are as follows:

Among them, ST _min and ST _max are the upper and lower limits of the safe remaining service life threshold ST of the subsea tree system, s _min is the lower limit of the number of spare parts, and S _max is the upper limit of the number of spare parts.

As shown in FIG. 8 , the underwater Christmas tree system includes an underwater Christmas tree production circuit 101, an underwater Christmas tree annulus loop 107 and an underwater Christmas tree chemical injection circuit 112; wherein the underwater Christmas tree production circuit 101 includes water Lower Christmas tree production main valve 102, underwater Christmas tree well surface control downhole safety valve 103, underwater Christmas tree production wing valve 104, underwater Christmas tree production throttle valve 105 and underwater Christmas tree production isolation valve 106, When the liquid is normally produced, the valves of the underwater Christmas tree production circuit 101 are kept open, and the oil from the subsea oil well flows into the Christmas tree pipeline, and passes through the underwater Christmas tree well surface to control the downhole safety valve 103, and the underwater Christmas tree production master. Valve 102, underwater Christmas tree production wing valve 104, then through the underwater Christmas tree production throttle valve 105 to adjust the output of oil, and finally enter the production manifold through the underwater Christmas tree production isolation valve 106, when the underwater Christmas tree production When the temperature or pressure of the circuit exceeds the maximum value of the set value, the main valve 102 for subsea tree production, the subsea tree production wing valve 104 and the subsea tree production isolation valve 106 are closed in turn according to the situation to isolate the subsea tree. A passageway between the production manifold and the production manifold to prevent the occurrence of dangerous accidents; the subsea Christmas tree annulus circuit 107 includes the subsea Christmas tree annulus main valve 108, the subsea Christmas tree annulus wing valve 109, and the subsea Christmas tree switching valve 110 and subsea tree annulus entry valve 111, when leakage occurs between the tubing and casing, if the temperature or pressure of the subsea tree annulus loop 107 exceeds the maximum value of the set value, the subsea tree annulus is opened The empty main valve 108 and the subsea tree annular wing valve 109 discharge the leaked oil and gas through the annular channel. If the pressure value continues to increase, the subsea tree switching valve 110 and the subsea tree production wing valve 104 are opened. , return the leaked oil and gas to the underwater Christmas tree production circuit 101 through the conversion channel; the underwater Christmas tree chemical agent injection circuit 112 includes the underwater Christmas tree methanol injection valve 113, the underwater Christmas tree chemical agent injection valve-114 and the underwater Christmas tree chemical agent injection valve-114 The Christmas tree chemical injection valve 2 115 controls the injection flow of various chemicals by controlling the opening of the underwater Christmas tree methanol injection valve 113 , the chemical injection valve 1 114 and the chemical injection valve 2 115 .

As shown in Figure 9, the underwater Christmas tree maintenance system based on remaining service life prediction includes five parts: the underwater Christmas tree production loop data acquisition module 201, the underwater Christmas tree annular loop data acquisition module 227, the water The lower Christmas tree chemical agent injection circuit data acquisition module 248 , the underwater Christmas tree sensor data collection and storage module 264 , and the underwater Christmas tree maintenance decision-making subsystem 301 .

Subsea Christmas tree production loop data acquisition module 201 includes production main valve sensor group 202 , production wing valve sensor group 207 , production isolation valve sensor group 212 , well surface control downhole safety valve sensor group 217 and production choke valve sensor group 222 . The production main valve sensor group 202 includes a production main valve pressure sensor 203, a production main valve temperature sensor 204, a production main valve flow sensor 205 and a production main valve acoustic emission sensor 206, which are mounted on the underwater Christmas tree production main valve 102, respectively. It is used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the main valve 102 of underwater Christmas tree production bears. The production wing valve sensor group 207 includes a production wing valve pressure sensor 208, a production wing valve temperature sensor 209, a production wing valve flow sensor 210 and a production wing valve acoustic emission sensor 211, which are mounted on the underwater Christmas tree production wing valve 104, respectively. It is used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the subsea Christmas tree production wing valve 104 bears. The production isolation valve sensor group 212 includes a production isolation valve pressure sensor 213, a production isolation valve temperature sensor 214, a production isolation valve flow sensor 215, and a production isolation valve acoustic emission sensor 216, which are mounted on the underwater Christmas tree production isolation valve 106, respectively. It is used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the subsea Christmas tree production isolation valve 106 bears. The well surface control downhole safety valve sensor group 217 includes a well surface control downhole safety valve pressure sensor 218 , a well surface control downhole safety valve temperature sensor 219 , a well surface control downhole safety valve flow sensor 220 and a well surface control downhole safety valve acoustic emission sensor 221 , mounted on the underwater Christmas tree well surface control downhole safety valve 103, respectively used to monitor the oil pressure, temperature, flow data and valve body leakage that the underwater Christmas tree well surface control downhole safety valve 103 bears. The production throttle valve sensor group 222 includes a production throttle valve pressure sensor 223, a production throttle valve temperature sensor 224, a production throttle valve flow sensor 225, and a production throttle valve acoustic emission sensor 226, and is mounted on the underwater Christmas tree production node. The flow valve 105 is used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the underwater Christmas tree production throttle valve 105 bears.

The subsea Christmas tree annular loop data acquisition module 227 includes an annulus main valve sensor group 228 , an annulus wing valve sensor group 233 , a switch valve sensor group 238 and an annulus entry valve sensor group 243 . The annulus main valve sensor group 228 includes an annulus main valve pressure sensor 229, an annulus main valve temperature sensor 230, an annulus main valve flow sensor 231 and an annulus main valve acoustic emission sensor 232, and is attached to the subsea Christmas tree annulus The main valve 108 is respectively used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the underwater Christmas tree annulus main valve 108 bears. The annular wing valve sensor group 233 includes an annular wing valve pressure sensor 234, an annular wing valve temperature sensor 235, an annular wing valve flow sensor 236 and an annular wing valve acoustic emission sensor 237, and is mounted on the annulus of the underwater Christmas tree The wing valve 109 is used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the underwater Christmas tree annular wing valve 109 bears. The switch valve sensor group 238 includes a switch valve pressure sensor 239, a switch valve temperature sensor 240, a switch valve flow sensor 241 and a switch valve acoustic emission sensor 242, which are mounted on the switch valve 110 of the underwater Christmas tree and are respectively used for monitoring underwater oil production. The pressure, temperature, flow data of the oil that the tree switching valve 110 is subjected to, and the leakage of the valve body. The annular entry valve sensor group 243 includes an annular entry valve pressure sensor 244, an annular entry valve temperature sensor 245, an annular entry valve flow sensor 246 and an annular entry valve acoustic emission sensor 247, which are mounted on the subsea Christmas tree annulus The inlet valve 111 is used to monitor the pressure, temperature, flow data of the oil and the leakage of the valve body that the subsea Christmas tree annular inlet valve 111 bears.

The underwater Christmas tree chemical injection circuit data acquisition module 248 includes a methanol injection valve sensor group 249 , a chemical injection valve sensor group 254 and a chemical injection valve sensor group 259 . The methanol injection valve sensor group 249 includes a methanol injection valve pressure sensor 250, a methanol injection valve temperature sensor 251, a methanol injection valve flow sensor 252, and a methanol injection valve acoustic emission sensor 253, which are mounted on the underwater Christmas tree methanol injection valve 113, respectively. It is used to monitor the pressure, temperature, flow data of methanol liquid and the leakage of the valve body that the underwater Christmas tree methanol injection valve 113 is subjected to. The chemical injection valve-sensor group 254 includes the chemical injection valve-pressure sensor 255, the chemical injection valve-temperature sensor 256, the chemical injection valve-flow sensor 257, and the chemical injection valve acoustic emission sensor 258, which are mounted underwater. The chemical agent injection valve 1 114 of the Christmas tree is used to monitor the pressure, temperature, flow data of the chemical agent liquid and the leakage of the valve body that the underwater Christmas tree chemical agent injection valve 1 114 bears. The second sensor group 259 of the chemical injection valve includes the second pressure sensor 260 of the chemical injection valve, the second temperature sensor 261 of the chemical injection valve, the second flow sensor 262 of the chemical injection valve, and the second acoustic emission sensor 263 of the chemical injection valve. The second lower Christmas tree chemical injection valve 115 is used to monitor the pressure, temperature, flow data of the chemical liquid and the leakage of the valve body that the underwater Christmas tree chemical injection valve 2 115 bears.

The subsea tree sensor data collection and storage module 264 communicates with the production main valve sensor group 202, the production wing valve sensor group 207, the production isolation valve sensor group 212, the well surface control downhole safety valve sensor group 217, and the production choke valve through the signal cable Sensor group 222, annular main valve sensor group 228, annular wing valve sensor group 233, switching valve sensor group 238, annular inlet valve sensor group 243, methanol injection valve sensor group 249, chemical injection valve-sensor group 254 and The two sensor groups 259 of the chemical injection valve are connected to collect and store the signals collected by the sensors.

The subsea Christmas tree maintenance decision-making subsystem 301 includes a degradation state diagnosis module 302 of each component of the subsea Christmas tree, a remaining service life prediction module 303 of each component of the subsea Christmas tree, a subsea Christmas tree system maintenance module 304, and a subsea Christmas tree System spare parts library module 305 and subsea tree system maintenance decision result display module 306 . The degradation state diagnosis module 302 of each component of the underwater Christmas tree receives the sensor data of the sensor data collection and storage module 264 of the underwater Christmas tree through the signal cable, and is used to diagnose the degradation state of each component of the underwater Christmas tree; the remaining components of the underwater Christmas tree are used The life prediction module 303 receives the degradation state of each component of the subsea Christmas tree diagnosed by the diagnosis module 302 of the degradation state of each component of the subsea Christmas tree, and obtains the remaining service life of each component of the subsea Christmas tree through the prediction model of the remaining service life of each component of the subsea Christmas tree Predicted value; the subsea Christmas tree system maintenance module 304 receives the remaining service life prediction value of each component of the subsea Christmas tree predicted by the remaining service life prediction module 303 of each component of the subsea Christmas tree and the subsea Christmas tree system spare parts library module 305 Data, through the maintenance model of the underwater tree system depending on the situation, the optimal maintenance method of each component of the underwater tree and the order quantity and order type of the spare parts of the underwater tree system are obtained. Finally, the maintenance decision result display module of the underwater tree system is used. 306 is displayed to the maintenance operator.

Claims (6)

1. The method for maintaining the underwater Christmas tree according to the situation based on the residual service life prediction is characterized by comprising the following steps of: the method comprises the following 6 steps:

s1: the method comprises the following steps of establishing a degradation impact model of each component of the underwater Christmas tree according to historical fault data:

s11: establishing an internal degradation model of each component of the underwater Christmas tree, modeling the internal degradation process of each component of the underwater Christmas tree into a gamma process, and determining the degradation x of each component of the underwater Christmas tree in unit cycle time T_TIndependent of each other and subject to a gamma distribution, as follows:

wherein, f (x)_Tα, β) is a gamma distribution density function, α is a shape parameter of the gamma distribution, β is an inverse scale parameter of the gamma distribution, and Γ (x)_T) For a gamma function, shape parameters and inverse scale parameters of the gamma distribution are determined from historical data;

total amount of internal degradation X of each component of subsea tree in t unit cycle times_tComprises the following steps:

wherein x is_T ^kThe degradation amount of each component of the underwater Christmas tree in the kth unit cycle time is shown, wherein k is a unit cycle time number;

s12: establishing an external impact model of the marine environment of each component of the underwater Christmas tree, and enabling each component of the underwater Christmas tree to be out of the marine environmentThe partial impact process is modeled as a Poisson process, and for any time period t₁,t₂Is greater than or equal to 0, has

Wherein N is the number of times that each component of the underwater Christmas tree is impacted by the external of the marine environment, and N is_c(t₁+t₂) Is t₁+t₂Number of external impacts on the marine environment, N, on a time period_c(t₁) Is t₁Number of times of external impacts on the marine environment, P { N), on a time period_c(t₁+t₂)-N_c(t₁) N is at any value of t₂The probability of the occurrence of n times of external impacts of the marine environment in a time period, wherein lambda is a parameter of Poisson distribution and is determined by historical data;

intensity x of external impact on marine environment on components of subsea tree_sThat is, the amount of degeneration caused by external impact in marine environment follows normal distribution as follows:

wherein, f (x)_s) The density function is normally distributed, mu is the mean value of the external impact strength of the marine environment, and sigma is the variance of the external impact strength of the marine environment, and is determined by historical data;

total amount of external impact X in marine environment_SComprises the following steps:

wherein x is_S ^hThe number is the external impact quantity of the marine environment at the h time, Ns is the external impact times of the marine environment, and h is the external impact times number of the marine environment;

the degradation state X of each component of the underwater Christmas tree is the total internal degradation amount X_tAnd the oceanTotal amount of environmental external impact X_SAnd (c) the sum, i.e.:

X＝X_t+X_S

under the condition of not adopting maintenance and replacement, along with the increase of time, the degradation state of each component of the underwater Christmas tree is only increased but not reduced;

s2: according to the maintenance data of the underwater Christmas tree system and the degradation data after maintenance, an incomplete maintenance model of each component of the underwater Christmas tree is established, and the method specifically comprises the following steps:

s21: the method comprises the following steps of establishing a degradation state reduction model after incomplete maintenance of all components of the underwater Christmas tree, and specifically comprising the following steps:

obtaining the degradation state X of each component of the underwater Christmas tree after incomplete maintenance according to the maintenance data of the underwater Christmas tree and the degradation data after maintenance_jLower than the degraded state X before service, but higher than the degraded state X after the last incomplete service_j-1；

Modeling degradation state distribution of each component of the subsea tree after incomplete maintenance as X_j-1To (X-X)_j-1)·0.6+X_j-1As follows:

wherein f is_XjIs a uniformly distributed density function, and j is the number of times of incomplete maintenance of the component;

s22: the method comprises the following steps of establishing a degradation acceleration model after incomplete maintenance of each component of the underwater Christmas tree by combining the incomplete maintenance times of each component of the underwater Christmas tree, and specifically comprises the following contents:

after the components of the underwater Christmas tree are not completely maintained, the degradation acceleration is shown on the parameters of an internal degradation model and an external impact model of a marine environment as follows:

α_j＝α+κ·j

μ_j＝τ^j·μ

wherein alpha is_jShape of internal degenerate gamma distribution after jth incomplete repairForm parameter, μ_jAfter the jth incomplete maintenance, the average value of the external impact strength of the marine environment borne by each component of the underwater Christmas tree; tau and kappa are degradation acceleration coefficients, tau is more than 1 and is determined by historical data;

s3: the method comprises the following steps of establishing a residual service life prediction model of each component of the underwater Christmas tree based on a neural network algorithm according to historical fault data of the underwater Christmas tree system, and specifically comprising the following steps:

s31: establishing a residual service life prediction model of each component of the underwater Christmas tree in a normal degradation state based on a neural network algorithm, and predicting the degradation state X of each component of the underwater Christmas tree at the adjacent monitoring time by using the residual service life prediction model of each component of the underwater Christmas tree in the normal degradation state based on the neural network algorithm_t-1And X_tThe unit cycle time t and t-1 of work in the corresponding degradation state and the incomplete maintenance times j in the current degradation state are used as input quantity of the neural network, the input quantity is input into an input layer of the neural network, and residual service life data Rul of all components of the underwater Christmas tree in the normal degradation state are output at an output layer through two hidden layers of 3 nodes R;

s32: establishing a residual service life prediction model of each component of the underwater Christmas tree based on a neural network algorithm under incomplete maintenance, and establishing a degradation state X of each component of the underwater Christmas tree before maintenance at the maintenance time by using the residual service life prediction model of each component of the underwater Christmas tree based on the neural network algorithm under incomplete maintenance_tThe unit cycle time t of work at the maintenance time, and the degradation state X of each component of the underwater Christmas tree after incomplete maintenance_jAnd the incomplete maintenance times j are used as the input quantity of the neural network, input into an input layer of the neural network, pass through two hidden layers of 3 nodes R, and output the remaining service life data Rul of the incomplete maintenance of each component of the underwater Christmas tree on an output layer;

obtaining a residual service life prediction model of each component of the underwater Christmas tree with prediction precision meeting the use requirement through multiple times of training;

s4: establishing a spare part model of an underwater Christmas tree system, wherein the specific implementation of the step is as follows:

the model of the subsea tree system spare parts adopts a spare part strategy of (S, S), wherein S is the minimum quantity of the sum of the components and spare parts of the subsea tree owned by the subsea tree system, S is the maximum quantity of the sum of the components and spare parts of the subsea tree owned by the subsea tree system, each subsea tree component has only one spare part at most, the initial quantity of the subsea tree system spare parts is S, when the subsea tree system is maintained and the subsea tree components are replaced, namely the spare parts are used, if the quantity of the subsea tree system spare parts is lower than S, the spare parts of the subsea tree components are ordered according to the sequence of ordering the spare parts from low to high of the predicted value of the remaining service life of the subsea tree components, so that the quantity of the subsea tree system spare parts is supplemented to S, the spare parts are ordered, the ordering cost is generated, and the spare parts can be replaced and used after the ordering period is finished, and spare part storage costs are incurred until unused;

s5: establishing an optional maintenance model of the underwater Christmas tree system, determining the maintenance mode of each component of the underwater Christmas tree with the current maintenance time as an optimization target by using the minimum ratio of the maintenance cost of the current maintenance time of the underwater Christmas tree system to the predicted value of the residual service life of the underwater Christmas tree system after maintenance, and determining the ordering condition of the spare parts of the underwater Christmas tree system after the current maintenance time according to the consumption condition of the spare parts, wherein the method specifically comprises the following steps:

s51: diagnosing and analyzing pressure, flow, temperature and leakage data acquired by sensors of all components of the underwater Christmas tree every unit period time T to obtain degradation states of all components of the underwater Christmas tree;

s52: inputting the degradation state of each component of the underwater Christmas tree into a residual service life prediction model of each component of the underwater Christmas tree in a normal degradation state to obtain a residual service life prediction value of each component of the underwater Christmas tree and a residual service life prediction value of an underwater Christmas tree system, and the method specifically comprises the following steps:

inputting the degradation state and the working life of each component of the underwater Christmas tree at the current moment, the degradation state and the working life of each component of the underwater Christmas tree at the previous moment and the incomplete maintenance frequency at the current moment into a residual service life prediction model of each component of the underwater Christmas tree in the normal degradation state based on a neural network algorithm to obtain the residual service life of each component of the underwater Christmas treePredicting the residual service life of each component of the Christmas tree, and simultaneously predicting Rul the residual service life of the underwater Christmas tree system_sysThe predicted value of the residual service life of each component of the underwater Christmas tree is defined as the following value:

Rul_sys＝min(Rul₁,Rul₂,...,Rul_N)

wherein N is the number of subsea tree components, Rul₁Prediction of remaining useful life for the 1 st component of subsea tree, Rul₂Prediction of remaining useful life for subsea tree 2 nd component, Rul_iPrediction of remaining useful life for the ith sub-assembly of a subsea tree, Rul_NPredicting a residual service life prediction value of the Nth component of the underwater Christmas tree;

s53: according to the degradation state of each component of the underwater Christmas tree, the predicted value of the residual service life of each component of the underwater Christmas tree and the predicted value of the residual service life of the underwater Christmas tree system, and meanwhile, maintenance decision is carried out by combining spare part data of the underwater Christmas tree system, and the method specifically comprises the following steps:

s531: when the predicted value of the residual service life of the underwater Christmas tree system is higher than the safe residual service life threshold ST of the underwater Christmas tree system, turning to S51, otherwise, starting maintenance preparation work;

service preparation including renting a service vessel, preparing a service tool, and hiring a service person; repair preparation costs include rental repair vessel costs, repair tools costs, and repair personnel costs; during the maintenance preparation work period, the underwater Christmas tree system keeps a working state and continues to degrade, and if the underwater Christmas tree system breaks down, the shutdown loss is caused; the time consumed for the maintenance preparation work is the maintenance preparation time LT;

s532: after the maintenance preparation work is finished, determining the maintenance mode of each component of the underwater Christmas tree by using the underwater Christmas tree system according to the situation maintenance genetic algorithm, wherein the method specifically comprises the following steps:

according to the spare part condition of the underwater Christmas tree system, dividing each component of the underwater Christmas tree into a component with a spare part and a component without the spare part; the underwater Christmas tree component with the spare part has three options of maintenance-free, incomplete maintenance and replacement, and the underwater Christmas tree component without the spare part only has two options of maintenance-free and incomplete maintenance;

maintenance cost C of subsea tree system at current maintenance time_mIncluding subsea tree system maintenance preparation cost c_SAnd cost of maintenance of various components of subsea tree c_g(ii) a Maintenance costs of various components of subsea tree c_gCost C for preventive incomplete maintenance including the normal functioning of the components of the subsea tree^ipmAnd preventive replacement cost C^prAnd cost C of incomplete repair after failure of each component of subsea tree^icmAnd cost of replacement after the fact C^cr(ii) a Because the degradation degree of each component of the underwater Christmas tree is more serious when the component of the underwater Christmas tree breaks down and the maintenance is more difficult, the cost of the afterwards incomplete maintenance and the cost of the afterwards replacement of each component of the underwater Christmas tree are higher than the cost of the preventability incomplete maintenance and the cost of the preventability replacement, and the maintenance cost C of the underwater Christmas tree system at the current maintenance time is higher than the cost of the preventability incomplete maintenance and the cost of the preventability replacement_mAs follows:

wherein i is the number of the subsea tree components, N is the number of the subsea tree components, c_g ⁱCost of maintenance for ith subsea tree component, c_i ^ipmFor preventive incomplete maintenance costs of component i, c_i ^icmFor the cost of ex-situ incomplete repair of component i, c_i ^prFor preventive replacement costs of component i, c_i ^crFor the cost of the subsequent replacement of component i, q_iFor a preventive incomplete repair factor, y, of component i_iIs the after incomplete repair factor, r, of component i_iIs a preventive replacement factor, p, of component i_iFor the post-replacement factor of component i, when q_iWhen it is 1, the group is describedPart i adopts preventive incomplete maintenance, when y_iWhen 1, the component i is explained to adopt the incomplete repair afterwards, when r_iWhen 1, the component i is described as being replaced preventively, when p_iWhen 1, the component i is described as being replaced afterwards, and when q is_i、y_i、r_i、p_iWhen the values are all 0, the component i is described to be in maintenance-free operation;

adopting an underwater Christmas tree component without maintenance operation, wherein the predicted values of the degradation state and the residual service life are unchanged; the degradation state of the underwater Christmas tree component adopting the replacement operation is 0, and the predicted value of the residual service life is changed into an initial value; adopting the incomplete maintenance operation to the underwater Christmas tree component, and according to the incomplete maintenance model of each component of the underwater Christmas tree, estimating the degradation state X of the underwater Christmas tree component after incomplete maintenance_j-mThe average of the degradation states after defined as incomplete repair is as follows:

X_j-m＝0.5·[(X-X_j-1)·0.6+2·X_j-1]

degraded State X before maintenance of subsea Tree Assembly that would take incomplete maintenance_tTime t per unit cycle of work before repair, estimated value X of degradation state after incomplete repair_j-mInputting the incomplete maintenance times j into a residual service life prediction model of each component of the underwater Christmas tree based on a neural network algorithm under the incomplete maintenance to obtain a residual service life prediction value of each component of the underwater Christmas tree after maintenance; taking the minimum value in the predicted values of the residual service lives of all the components of the subsea tree as the predicted value Rul of the residual service life of the subsea tree system after maintenance_sys-m；

The minimum ratio of the maintenance cost of the underwater Christmas tree system at the current maintenance time to the predicted value of the residual service life of the underwater Christmas tree system after maintenance is taken as an optimization target, and the following steps are shown:

determining the maintenance mode of each component of the underwater Christmas tree by using an underwater Christmas tree system according to a situation maintenance genetic algorithm; the coding mode is that 0 represents maintenance, 1 represents incomplete maintenance and 2 represents replacement; one chromosome includes 12 codes, which respectively represent the maintenance modes of 12 corresponding components; the maintenance mode of the underwater Christmas tree component is influenced by the state of spare parts of the underwater Christmas tree component, when the component has the spare parts, the state of the spare parts is 1, and when the component has no spare parts, the state of the spare parts is 0; the components with spare parts have three options of 0, 1 and 2, and the components without spare parts have two options of 0 and 1;

the optimization process of the genetic algorithm for the situation-based maintenance of the underwater Christmas tree system is shown as follows:

firstly, initializing a population and randomly generating a plurality of groups of chromosomes;

in order to avoid the possible maintenance combinations that are significantly different from the optimal solution due to the randomness of the genetic algorithm, it is necessary to define the maintenance modes of the components that meet certain requirements and filter the undesirable chromosomes as follows:

step 1, when the underwater Christmas tree component has no spare parts, if the predicted value of the residual service life of the component is lower than a safety threshold value, incomplete maintenance is required;

step 2, when spare parts exist in the underwater Christmas tree component, if the predicted value of the residual service life of the component is lower than a safety threshold value, incomplete maintenance or replacement is required;

step 3, when spare parts exist in the underwater Christmas tree component, if the component fails and the number of incomplete maintenance times is more than 3, the component must be replaced;

step 4, when spare parts exist in the underwater Christmas tree component, the degradation state of the component is lower than 0.7, the component is not subjected to incomplete maintenance, and the component is not replaced;

step 5, when the degradation state of the underwater Christmas tree component is lower than 0.4, the component is not maintained and replaced incompletely;

then, calculating the maintenance cost of the current maintenance time of the underwater Christmas tree system and the predicted value of the residual service life of the underwater Christmas tree system after maintenance, which are generated by maintaining each component according to the maintenance mode represented by each chromosome; the ratio of the maintenance cost of the current maintenance time of the underwater Christmas tree system to the predicted value of the residual service life of the underwater Christmas tree system after maintenance is taken as the adaptive value W of each chromosome, and the adaptive value is as follows:

secondly, performing cross operation, mutation operation and selection operation for selecting a better chromosome and performing next iteration operation; the crossing operation is to randomly generate a crossing point and to exchange the positions of chromosome segments after the crossing point in the adjacent chromosomes; the mutation operation is to randomly change the coding value of a certain position of the chromosome; the new maintenance mode obtained by the variation operation is required to meet the requirement of the spare part state of the underwater Christmas tree assembly, namely, the maintenance mode of replacement cannot be changed for the assembly without the spare part; the selection operation adopts a roulette method;

finally, judging whether a termination condition, namely the maximum iteration number, is met, if the termination condition is met, outputting an optimal maintenance mode of each component of the underwater Christmas tree, and if not, continuing to execute the iteration process until the termination condition is met;

s533: after the components of the underwater Christmas tree are maintained according to the maintenance mode determined in S532, the ordering quantity and the ordering type of the spare parts of the underwater Christmas tree system are determined according to the service condition of the spare parts of the underwater Christmas tree system, and the method specifically comprises the following steps:

number of spare parts N of subsea tree system after maintenance_zhComprises the following steps:

N_zh＝S-N_r

wherein N is_rThe number of components is changed;

when the number of spare parts is N_zhWhen the number is not less than s, spare parts do not need to be ordered; when the number is less than s, the spare parts need to be ordered, and the number N of the spare parts needs to be ordered_orComprises the following steps:

N_or＝S-N_zh

the ordering type of the spare parts of the underwater Christmas tree system is determined according to the sequence of predicted values of the residual service lives of all components of the underwater Christmas tree after maintenance from low to high;

s6: the method comprises the following steps of determining an optimal maintenance decision threshold value of the underwater Christmas tree system, namely a safe remaining service life threshold value ST of the underwater Christmas tree system and a strategy threshold value (S, S) of spare parts of the underwater Christmas tree system, by taking the minimum unit time maintenance cost of the underwater Christmas tree system as a target, and specifically comprising the following contents:

determining total cost of maintenance C for subsea tree system_zThe total maintenance cost of the subsea tree system includes the total working time t of the subsea tree system_zMaintenance cost C of period subsea tree system at each maintenance time_mAnd the cost of ordering spare parts of subsea tree system c_bjAnd storage cost c_ccAnd the shutdown loss caused by the fault of the subsea tree system is as follows:

wherein N is_fTotal number of cycles to produce loss of shutdown, c_DThe shutdown loss in unit cycle time, M is the total maintenance times of the subsea tree system, l is the number of the maintenance times of the subsea tree system, C^l _mThe maintenance cost of the subsea tree system at the first maintenance time is the maintenance cost;

maintenance cost per unit time is minimized with subsea tree system

For the purpose, a safe remaining service life threshold ST of the underwater Christmas tree system and a strategy threshold (S, S) of spare parts of the underwater Christmas tree system are determined by a traversal method, and the optimization goal is as follows:

wherein, ST_minAnd ST_maxUpper and lower limits, s, of a safe remaining service life threshold ST for an subsea tree system_minAs a lower limit of the number of spare parts, S_maxIs the upper limit of the number of spare parts;

the method for maintaining the underwater Christmas tree according to the situation based on the residual service life prediction is applied to an underwater Christmas tree according to the situation maintenance system based on the residual service life prediction, and the system comprises 5 parts: the system comprises an underwater Christmas tree production loop data acquisition module, an underwater Christmas tree annular loop data acquisition module, an underwater Christmas tree chemical agent injection loop data acquisition module, an underwater Christmas tree sensor data acquisition and storage module and an underwater Christmas tree maintenance decision subsystem;

the underwater production tree production loop data acquisition module comprises a production main valve sensor group, a production wing valve sensor group, a production isolation valve sensor group, a well surface control underground safety valve sensor group and a production throttle valve sensor group;

the underwater Christmas tree annulus loop data acquisition module comprises an annulus main valve sensor group, an annulus wing valve sensor group, a change-over valve sensor group and an annulus inlet valve sensor group;

the underwater Christmas tree chemical agent injection loop data acquisition module comprises a methanol injection valve sensor group, a first chemical agent injection valve sensor group and a second chemical agent injection valve sensor group;

the maintenance decision subsystem of the underwater Christmas tree comprises a degradation state diagnosis module of each component of the underwater Christmas tree, a residual service life prediction module of each component of the underwater Christmas tree, an on-demand maintenance module of the underwater Christmas tree system, a standby component library module of the underwater Christmas tree system and a maintenance decision result display module of the underwater Christmas tree system.

2. The method for the on-the-fly maintenance of an underwater Christmas tree based on the prediction of remaining useful life of claim 1, wherein: the production main valve sensor group of the data acquisition module of the production loop of the underwater Christmas tree comprises a production main valve pressure sensor, a production main valve temperature sensor, a production main valve flow sensor and a production main valve acoustic emission sensor, is attached to the production main valve of the underwater Christmas tree and is respectively used for monitoring the pressure, the temperature and the flow data of oil borne by the production main valve of the underwater Christmas tree and the leakage condition of a valve body; the production wing valve sensor group comprises a production wing valve pressure sensor, a production wing valve temperature sensor, a production wing valve flow sensor and a production wing valve acoustic emission sensor, is pasted on the production wing valve of the underwater Christmas tree, and is respectively used for monitoring the pressure, the temperature and the flow data of oil borne by the production wing valve of the underwater Christmas tree and the leakage condition of the valve body; the production isolation valve sensor group comprises a production isolation valve pressure sensor, a production isolation valve temperature sensor, a production isolation valve flow sensor and a production isolation valve acoustic emission sensor, is attached to the production isolation valve of the underwater Christmas tree, and is respectively used for monitoring pressure, temperature and flow data of oil borne by the production isolation valve of the underwater Christmas tree and the leakage condition of the valve body; the surface control underground safety valve sensor group comprises a surface control underground safety valve pressure sensor, a surface control underground safety valve temperature sensor, a surface control underground safety valve flow sensor and a surface control underground safety valve acoustic emission sensor, is pasted on the surface control underground safety valve of the underwater Christmas tree, and is respectively used for monitoring the pressure, the temperature and the flow data of oil borne by the surface control underground safety valve of the underwater Christmas tree and the leakage condition of a valve body; the production throttle valve sensor group comprises a production throttle valve pressure sensor, a production throttle valve temperature sensor, a production throttle valve flow sensor and a production throttle valve acoustic emission sensor, is attached to the production throttle valve of the underwater Christmas tree, and is respectively used for monitoring pressure, temperature and flow data of oil borne by the production throttle valve of the underwater Christmas tree and the leakage condition of the valve body.

3. The method for the on-the-fly maintenance of an underwater Christmas tree based on the prediction of remaining useful life of claim 1, wherein: the annular main valve sensor group of the underwater Christmas tree annular loop data acquisition module comprises an annular main valve pressure sensor, an annular main valve temperature sensor, an annular main valve flow sensor and an annular main valve acoustic emission sensor, is attached to the underwater Christmas tree annular main valve and is respectively used for monitoring the pressure, the temperature and the flow data of oil borne by the underwater Christmas tree annular main valve and the leakage condition of a valve body; the annular wing valve sensor group comprises an annular wing valve pressure sensor, an annular wing valve temperature sensor, an annular wing valve flow sensor and an annular wing valve acoustic emission sensor, is pasted on the annular wing valve of the underwater Christmas tree, and is respectively used for monitoring the pressure, the temperature and the flow data of oil borne by the annular wing valve of the underwater Christmas tree and the leakage condition of the valve body; the conversion valve sensor group comprises a conversion valve pressure sensor, a conversion valve temperature sensor, a conversion valve flow sensor and a conversion valve acoustic emission sensor, is attached to the conversion valve of the underwater Christmas tree, and is respectively used for monitoring the pressure, the temperature and the flow data of oil borne by the conversion valve of the underwater Christmas tree and the leakage condition of the valve body; the annular space entering valve sensor group comprises an annular space entering valve pressure sensor, an annular space entering valve temperature sensor, an annular space entering valve flow sensor and an annular space entering valve acoustic emission sensor, and is pasted on the underwater Christmas tree annular space entering valve and used for monitoring the leakage conditions of pressure, temperature, flow data and a valve body of oil born by the underwater Christmas tree annular space entering valve respectively.

4. The method for the on-the-fly maintenance of an underwater Christmas tree based on the prediction of remaining useful life of claim 1, wherein: the methanol injection valve sensor group of the chemical agent injection loop data acquisition module of the underwater Christmas tree comprises a methanol injection valve pressure sensor, a methanol injection valve temperature sensor, a methanol injection valve flow sensor and a methanol injection valve acoustic emission sensor, is attached to the methanol injection valve of the underwater Christmas tree, and is respectively used for monitoring the pressure, temperature and flow data of methanol liquid borne by the methanol injection valve of the underwater Christmas tree and the leakage condition of a valve body; the first sensor group of the chemical agent injection valve comprises a first pressure sensor of the chemical agent injection valve, a first temperature sensor of the chemical agent injection valve, a first flow sensor of the chemical agent injection valve and a first acoustic emission sensor of the chemical agent injection valve, is attached to the first chemical agent injection valve of the underwater Christmas tree and is respectively used for monitoring the pressure, the temperature and the flow data of the chemical agent liquid borne by the first chemical agent injection valve of the underwater Christmas tree and the leakage condition of the valve body; the second sensor group of the chemical agent injection valve comprises a second pressure sensor of the chemical agent injection valve, a second temperature sensor of the chemical agent injection valve, a second flow sensor of the chemical agent injection valve and a second acoustic emission sensor of the chemical agent injection valve, is pasted on the second chemical agent injection valve of the underwater Christmas tree and is respectively used for monitoring the pressure, the temperature and the flow data of the chemical agent liquid born by the second chemical agent injection valve of the underwater Christmas tree and the leakage condition of the valve body.

5. The method for the on-the-fly maintenance of an underwater Christmas tree based on the prediction of remaining useful life of claim 1, wherein: the sensor data collecting and storing module of the underwater Christmas tree is connected with a production main valve sensor group through a signal cable, a production wing valve sensor group, a production isolation valve sensor group, a well surface control underground safety valve sensor group, a production throttle valve sensor group, an annulus main valve sensor group, an annulus wing valve sensor group, a change-over valve sensor group, an annulus inlet valve sensor group, a methanol injection valve sensor group, a chemical agent injection valve sensor group and a chemical agent injection valve sensor group, and is used for collecting and storing signals collected by the sensors.

6. The method for the on-the-fly maintenance of an underwater Christmas tree based on the prediction of remaining useful life of claim 1, wherein: the degradation state diagnosis module of each component of the underwater Christmas tree maintenance decision subsystem receives the sensor data of the sensor data collection and storage module of the underwater Christmas tree through a signal cable and is used for diagnosing the degradation state of each component of the underwater Christmas tree; the residual service life prediction module of each component of the underwater Christmas tree receives the degradation state of each component of the underwater Christmas tree obtained by the degradation state diagnosis module of each component of the underwater Christmas tree, and a residual service life prediction value of each component of the underwater Christmas tree is obtained through the residual service life prediction model of each component of the underwater Christmas tree; the underwater Christmas tree system condition maintenance module receives the residual service life prediction value of each component of the underwater Christmas tree and the module data of the underwater Christmas tree system spare part database, which are obtained by predicting the residual service life of each component of the underwater Christmas tree by the underwater Christmas tree system condition prediction module, obtains the optimal maintenance mode of each component of the underwater Christmas tree and the ordering quantity and the ordering type of the underwater Christmas tree system spare parts through the underwater Christmas tree system condition maintenance module, and finally displays the optimal maintenance mode, the ordering quantity and the ordering type of each component of the underwater Christmas tree system spare part to maintenance operators through the underwater Christmas tree system maintenance decision result display module.

Images