Design method and device of multi-energy system for data center considering operation reliability

Abstract

The present invention provides a design method and device for a data center multi-energy system that takes into account operational reliability, determines the capacity of energy facilities with a desired reliability level, develops a reliability improvement algorithm for coupling normal operation and island or equipment failure operation, and achieves a trade-off between economy and high reliability in data center planning problems. The present invention takes into account the high level of operational reliability required by data centers, and at the same time determines the operation mode of the data center energy system under normal conditions, island conditions, and equipment failure conditions, including the way to meet electricity demand and cooling demand and the circulation conditions of each water circulation loop.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-energy systems, and in particular relates to a design method and device for a multi-energy system for a data center taking operational reliability into consideration. Background Art

[0002] Data centers are large-scale, all-weather infrastructures that operate with computing as their main task. They are of great significance to economic transformation and the development of the IT industry. With the development of industrialization and informatization, data centers consume more and more energy. A typical data center can consume as much energy as 25,000 households, and its unit space consumption can reach hundreds of times that of a standard office space. The data center is developing rapidly. By 2030, the energy consumption of data centers is expected to reach 3,000 terawatt hours, accounting for nearly 8% of the global estimated electricity demand. Electricity costs have become a major expense for data centers today. Therefore, energy-saving transformation of data centers can bring greater economic and environmental benefits.

[0003] In terms of energy consumption, the energy consumed by data centers is mainly composed of two parts: energy consumed by IT equipment (servers, storage, etc.) and energy consumed by infrastructure (cooling equipment, etc.). Among them, the energy consumption of air conditioning and refrigeration systems accounts for about 37% of the total energy consumption, which is the second largest source of energy consumption after IT equipment energy consumption. At the same time, water cooling systems are increasingly widely used in data centers due to their high energy efficiency, small footprint, and low noise. Therefore, reducing the power consumption and cooling consumption of data centers, and reducing the carbon emissions of their power consumption and cooling consumption, is an important way to promote energy-saving transformation of data centers.

[0004] At the same time, with the advent of the information age, more and more companies rely on the Internet to conduct business, and their data processing and other tasks need to be completed in data centers. On the one hand, real-time access services have a low tolerance for delays and interruptions. On the other hand, data leakage, data loss, business interruption, etc. caused by abnormal equipment operation will cause huge reputation and economic losses to the company. Therefore, when constructing and managing data centers, redundant power supply, cooling, monitoring, and fire-fighting equipment are often configured in case of emergency. However, the current research on data center planning schemes that consider operational reliability at home and abroad is very limited, and most of the calculations are based on redundancy, which may lead to waste of investment.

[0005] In recent years, hydrogen energy has gradually entered the human vision with its advantages of high energy density, easy storage, high efficiency and cleanliness. The high efficiency and cleanliness of hydrogen fuel cell cogeneration technology make it well integrated and adaptable in energy systems with liquids such as water as the medium. At the same time, data centers are usually equipped with large-capacity fire water pools for use in the event of a fire. The small volume is 300 cubic meters, and the large volume can reach 2,000 cubic meters. Large-volume pools will lead to waste of resources when they are idle. At the same time, stagnant water is easy to breed bacteria and cause water pollution. Therefore, a hydrogen-containing multi-energy system that combines hydrogen energy, electrical energy, cold energy, and thermal energy and stores hydrogen, hot water, and cold water can better meet the energy needs of data centers. However, people currently lack in-depth research on the application effect of such energy systems in data centers. At the same time, how to combine such energy systems with existing facilities in data centers to save costs to a greater extent is still unanswered. Summary of the invention

[0006] In order to solve the above problems, the present invention develops a hydrogen-containing multi-energy system that adapts to the load demand and actual conditions of the data center, and provides a data center multi-energy system planning model that takes into account island conditions and equipment failure conditions. The model determines the capacity of energy facilities with the expected reliability level, and uses Benders decomposition to develop a reliability improvement algorithm for coupling normal operation (as the main problem) and island or equipment failure operation (as a sub-problem), thereby achieving a trade-off between economy and high level reliability in the data center planning problem.

[0007] The design method of a data center multi-energy system considering operational reliability includes the following steps:

[0008] S1. Obtain data: obtain the power demand, cooling demand, and total area where solar panels and solar thermal collectors can be installed of the target data center;

[0009] S2. Establish an energy system planning model with an expected reliability level based on electricity demand, cooling demand, and the total area where solar panels and solar thermal collectors can be installed: the energy system planning model uses the parameters of each device and the operation scheduling strategy as the variables to be solved, and determines the constraints on the variables based on the electricity demand of the energy system under normal conditions, island conditions, and equipment failure conditions, the way to meet the cooling demand, and the flow conditions of each water circulation loop, and takes stable operation in all scenarios and the minimum total cost in normal scenarios as the objective function, so that the system configured according to the solved parameters of each device can meet the requirements of stable operation under normal conditions and the minimum total cost;

[0010] S3. Simplify and split the energy system planning model according to the scenario: simplify and split the energy system planning model obtained in S2 to obtain the main problem and sub-problems;

[0011] S4. Generate stability constraints and add them to the main problem and sub-problems for iterative solution to obtain the planning results of each device that meets all scenarios.

[0012] Furthermore, in S2, the constraints include hydrogen balance constraints in all scenarios, electricity balance constraints in all scenarios, heat balance constraints in all scenarios, cold balance constraints in all scenarios, buying and selling electricity constraints in all scenarios, solar energy production capacity constraints in all scenarios, electrolyzer operation constraints in all scenarios, fuel cell operation constraints in all scenarios, absorption chiller power consumption constraints in all scenarios, hot water tank related constraints in all scenarios, cold storage equipment related constraints in all scenarios, hydrogen storage tank related constraints in all scenarios, and equipment parameter constraints in all scenarios.

[0013] The electrical balance constraints in all the scenarios are:

[0014]

[0015] in, is the power purchased from the grid during period t in scenario s, is the power generated by the solar panel during period t in scenario s, is the power generated by the fuel cell during period t in scenario s, is the power fed back to the grid during period t in scenario s, is the power consumed by the electrolyzer during period t in scenario s, is the power consumed by the cold water distribution unit during period t in scenario s, is the power consumed by the absorption chiller during period t in scenario s, is the power consumed by the power distribution unit during period t in scenario s.

[0016] Furthermore, in S3, the objective function of the subproblem is:

[0017]

[0018] in, and is the introduced slack variable, Used to relax electrical balance constraints, Used to relax the cooling balance constraint. When the objective function is 0, it means that the current system can meet the electricity demand and cooling demand in the current scenario. When it is not 0, it cannot be met.

[0019] Furthermore, in S3, the constraints of the subproblems include the subproblem electrical balance constraint, which is expressed as:

[0020] in, is the power purchased from the grid during period t, is the power generated by the solar panel during period t, is the power generated by the fuel cell during period t, is the power fed back to the grid during period t, is the power consumed by the electrolytic cell during period t, is the power consumed by the cold water distribution unit during period t, is the power consumed by the absorption chiller during period t, The power consumed by the power distribution unit during period t.

[0021] Furthermore, in S3, the constraints of the subproblems include the consistency constraints of the subproblem parameters. The expression of the consistency constraints of the subproblem parameters is as follows:

[0022]

[0023]

[0024]

[0025]

[0026] in, and To calculate the generated Lagrange multiplier, The parameters of each device obtained by the previous iteration of the main problem, The energy in and out flags of the energy storage device obtained by the previous iteration of the main problem, are variables representing the parameters of each device in the sub-problem, is the variable representing the energy in and out flag of the energy storage device in the sub-problem; and is the mark of the heat input and output of the hot water tank in the system during period t. When the hot water tank supplies heat, is 1, is 0; when the hot water tank stores heat, is 0, is 1; and It is the sign of the inlet and outlet cooling capacity of the cold storage equipment transformed from the fire water tank in the system during period t. When the cold storage equipment transformed from the fire water tank supplies cooling capacity, is 1, is 0; when the cold storage equipment transformed from the fire water tank stores cold energy, is 0, is 1; and It is the mark of the amount of hydrogen in and out of the hydrogen storage tank in the system during period t. When the hydrogen storage tank supplies hydrogen, is 1, is 0; when the hydrogen storage tank stores hydrogen, is 0, is 1, and B is the set of energy storage device entry and exit states.

[0027] Furthermore, in S3, the constraints of the subproblems include the subproblem island constraints:

[0028]

[0029] in, It is a flag to indicate whether the interaction with the power grid is normal. It is used to indirectly describe the situation of the system. When the interaction between the system and the power grid is normal, 1, otherwise 0, that is, the system is running in an island scenario is 0.

[0030] Furthermore, in S3, the constraints of the sub-problems include the sub-problem equipment failure constraints:

[0031]

[0032] in, It is the working status flag of the absorption chiller, which is used to indirectly describe the situation of the system. When the absorption chiller of the system works normally, 1, otherwise 0, that is, the system is running in the device failure scenario is 0.

[0033] Further, in S4, the iterative solution of the main problem includes the following steps:

[0034] S401. Solve the main problem and get the solution , let i=1;

[0035] S402, construct sub-problems, for the i-th scenario in the scenario set s containing S scenarios, solve Substitute the subproblem of scenario s[i] and solve the subproblem to obtain the solution of the subproblem , Lagrange multipliers and Lagrange multipliers ;

[0036] S403, if , add stability constraints to the main problem, jump to S401 to continue calculating the new solution ;like , jump to S404;

[0037] S404, if , i=i+1, jump to S402; if , the solution is completed.

[0038] Furthermore, in S403, the stability constraint is:

[0039]

[0040] in, is the solution to a sub-problem, and To calculate the generated Lagrange multiplier, The parameters of each device obtained by the previous iteration of the main problem, The energy in and out flags of the energy storage device obtained by the previous iteration of the main problem, are variables representing the parameters of each device in the sub-problem, is the change of the energy flag bit of the energy storage device in the sub-problem, E is the set of devices e, and B is the set of energy storage device entry and exit states.

[0041] A design device for a multi-energy system of a data center considering operational reliability includes a data acquisition module, a model building module, a model segmentation module and a solution module connected in sequence;

[0042] Among them, the data acquisition module is used to collect raw data and pass the collected raw data to the model building module. The raw data includes electricity demand, cooling demand, and the total area where solar panels and solar collectors can be installed; the model building module is used to establish an energy system planning model with an expected reliability level based on the raw data; the model segmentation module is used to simplify and segment the energy system planning model to obtain the main problem and sub-problems; the solution module is used to generate stability constraints and add them to the main problem and sub-problems for iterative solution to obtain the planning results of each equipment that meets all scenarios.

[0043] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0044] In response to the special load requirements of data centers, the present invention combines the transformation of existing fire water tanks to establish a hydrogen-containing multi-energy system that fully utilizes fuel cells, electrolyzers, solar panels, solar collectors, absorption refrigerators, hydrogen storage tanks, hot water tanks and cold storage equipment to absorb, convert and store hydrogen energy, solar energy, electrical energy, thermal energy and cold energy. The connection method between each device is clarified, so that the system can meet the electricity and cooling needs of the data center in a near-zero-carbon manner.

[0045] The present invention takes into account the high level of requirements for operational reliability of data centers, namely: the energy system adapted thereto should be able to ensure the stable operation of the data center when some special events (natural disasters, abnormal weather, etc.) occur, and summarizes the impact of the above special events on the energy system of the data center as power outages (islands) and equipment failures. At the same time, the operation mode of the energy system of the data center under normal conditions, island conditions, and equipment failure conditions is determined, including the way to meet electricity demand and cooling demand and the circulation conditions of each water circulation loop.

[0046] The present invention combines the operation modes of the data center energy system under normal conditions, island conditions, and equipment failure conditions, describes them into a solvable mathematical form, and provides a planning model that takes into account island constraints and equipment failure constraints. By solving the model, the capacity of each device in the energy system with the desired reliability level can be determined.

[0047] The present invention develops a reliability improvement algorithm for coupling normal operation (as the main problem) and island or equipment failure operation (as the sub-problem), achieving a trade-off between economy and high-level reliability in data center planning problems. The algorithm can divide the original model that is difficult to solve due to the large number of variables into main problems and sub-problems that are easier to solve, and perform iterative solutions, so that the number of variables solved at a single time is greatly reduced, thereby speeding up the solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a schematic diagram of a hydrogen-containing multi-energy system for a data center;

[0049] Figure 2 Flowchart of the design method considering operational reliability for hydrogen-containing multi-energy systems in data centers;

[0050] Figure 3 This is the flow chart of the reliability improvement algorithm in the design method;

[0051] Figure 4 It is a schematic diagram of the iterative composition of the main problem constraints in the design method;

[0052] Figure 5 Generate a schematic diagram for the data center's hydrogen-containing multi-energy system fire water tank renovation plan;

[0053] Figure 6 Schematic diagram of a data center hydrogen-containing multi-energy system when the fire pool size is limited;

[0054] Figure 7 Schematic diagram of cold storage equipment for the transformation of two fire water tanks;

[0055] Figure 8 A schematic diagram of the module structure of the design device provided by the present invention. DETAILED DESCRIPTION

[0056] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] Example 1

[0058] Reference Figure 1 An embodiment of the present invention provides a data center hydrogen-containing multi-energy system that takes operational reliability into consideration. The system consists of an absorption chiller, a cold storage device, a fuel cell, a hot water tank, a solar collector, an electrolyzer, a hydrogen storage tank, a photovoltaic panel, a power distribution unit 1, a power distribution unit 2, a cold water distribution unit, a water pump, and a valve.

[0059] Among them, the functions and effects of each device are as follows: the absorption chiller cools by absorbing heat from hot water and consumes a certain amount of electricity. It is the most important refrigeration equipment in the system; the cold storage equipment is mainly modified from the existing fire water pool in the data center. After adding a water distributor or baffle to the fire water pool, the cold water is distributed according to the rule that the temperature increases as it goes up or the temperature increases as it goes to the left. It is mainly used to store cold energy to save refrigeration costs through scheduling and provide cold energy for the system when the absorption chiller fails; the fuel cell uses cogeneration technology, consumes hydrogen energy and generates electricity and heat energy. The hydrogen energy used is mostly produced by renewable energy electrolysis or industrial by-product hydrogen, which can effectively improve the cleanliness of the energy consumption of the data center; the hot water tank stores heat in the form of hot water for cooling by the absorption chiller, and the hot water is also distributed according to The temperature increases as it goes upwards or as it goes to the left; the solar collector absorbs the heat from the sun's radiation and converts it into heat in the form of hot water, providing a clean heat source for the system; the electrolyzer produces hydrogen through electrolysis when the electricity price is low or the photovoltaic panel generates too much power, so as to increase the utilization rate of renewable energy and save operating costs; the hydrogen storage tank is used to store hydrogen, so as to save power generation costs through scheduling and provide electricity to the system through the electrolyzer when operating on an isolated island; the photovoltaic panel converts light energy into electrical energy to power the system, thereby improving the environmental friendliness of the system's energy consumption; the water pump provides power for each water circulation loop in the system; the valve is used to control the flow distribution of each water circulation loop in the system; the power distribution unit 1 distributes power to the power supply end, the power distribution unit 2 distributes power to each computer room, and the cold water distribution unit distributes cold water to each computer room.

[0060] Among them, the connection method of each water circulation loop in the hydrogen-containing multi-energy system of the data center is as follows:

[0061] In this system, there are three cold water circulation loops, which can be used at the same time. The cold water circulation mode and flow rate are controlled by controlling the switch status of valve 1, valve 2, valve 3, valve 4 and the working power of the primary water pump and the secondary water pump. The three loops are:

[0062] Cold water circulation loop 1: The water with lower temperature at the bottom of the fire water tank passes through valve 4 and the primary water pump in turn, and is reasonably distributed to the air-conditioning system of each computer room through the cold water distribution unit. After heat exchange and temperature rise in the air-conditioning system of each computer room, it flows back to the top of the fire water tank through valve 3, that is, the cold energy stored in the fire water tank is used to cool the computer room of the data center.

[0063] Cold water circulation loop 2: The water from the top of the fire water tank passes through the secondary water pump and valve 1 in sequence, is cooled by the absorption chiller, and then flows back to the bottom of the fire water tank through valve 2, that is, the cold energy generated by the absorption chiller is stored in the fire water tank.

[0064] Cold water circulation loop 3: The water cooled by the absorption chiller passes through valve 2, valve 4 and the primary water pump in turn, and is reasonably distributed to the air-conditioning system of each computer room through the cold water distribution unit. After heat exchange and temperature rise in the air-conditioning system of each computer room, it flows back to the absorption chiller through valve 3 and valve 1 for the next cooling, that is, the cold energy generated by the absorption chiller is used to cool the computer room of the data center.

[0065] In this system, there are five hot water circulation loops, which can be used at the same time. The hot water circulation mode and flow rate are controlled by controlling the switch status of valves 5, 6, 7, and 8 and the working power of the third-stage water pump, the fourth-stage water pump, and the fifth-stage water pump. The five loops are:

[0066] Hot water circulation loop 1: The low-temperature hot water at the bottom of the hot water tank passes through valve 8 and the five-stage water pump in turn, obtains the heat collected by the solar collector and is heated up, and then flows back to the top of the hot water tank through valve 7, that is, the heat collected by the solar collector is stored in the hot water tank.

[0067] Hot water circulation loop 2: The low-temperature hot water at the bottom of the hot water tank passes through valve 8 and the three-stage water pump in turn, obtains the heat generated by the fuel cell cogeneration, and after heating, flows back to the top of the hot water tank through valve 7, that is, the heat generated by the fuel cell cogeneration is stored in the hot water tank.

[0068] Hot water circulation loop 3: The high-temperature hot water at the top of the hot water tank passes through the four-stage water pump and valve 5 in sequence, is cooled and absorbed by the absorption refrigerator, and then flows back to the bottom of the hot water tank through valve 6, that is, the heat stored in the hot water tank is used as the heat source for cooling by the absorption refrigerator.

[0069] Hot water circulation loop 4: After obtaining the high-temperature hot water collected by the solar collector, it passes through valve 7 and valve 5 in turn. After being cooled and absorbed by the absorption refrigerator, it passes through valve 6, valve 8 and the five-stage water pump in turn and flows back to the solar collector end to continue collecting heat. That is, the heat absorbed by the solar collector is used to power the absorption refrigerator to cool it.

[0070] Hot water circulation loop 5: After obtaining the high-temperature hot water generated by the fuel cell cogeneration, it passes through valve 7 and valve 5 in turn, and after being cooled and absorbed by the absorption refrigerator, it passes through valve 6, valve 8 and the three-stage water pump in turn, and flows back to the fuel cell end to collect heat, that is, the heat generated by the fuel cell cogeneration is used to power the absorption refrigerator to cool it.

[0071] The energy forms involved in the energy system of this data center include: electricity, hydrogen energy, heat and cold. Their direct sources and direct uses in the operation of the energy system are as follows: the direct sources of electricity for the multi-energy system of this data center include electricity purchased from the power grid, power generation by solar panels and electricity generated by fuel cells. The direct uses of the required electricity include electricity consumed by electrolyzers to produce hydrogen, electricity consumed by absorption chillers, power feedback to the power grid, power distribution losses, electricity consumed by cold water distribution units and meeting the electricity needs of the data center; the direct sources of hydrogen energy for the multi-energy system of this data center include hydrogen produced by electrolyzers, hydrogen output from hydrogen storage tanks and hydrogen purchased from the hydrogen market, and the direct uses include hydrogen consumption by fuel cells and storage in hydrogen storage tanks; the direct sources of heat for the multi-energy system of this data center include heat produced by fuel cell cogeneration, heat absorbed by solar collectors, and heat flowing out of hot water tanks, and the direct uses include heat consumption by absorption chillers and storage in hot water tanks; the direct sources of cold energy for the multi-energy system of this data center include cold energy generated by absorption chillers and cold energy flowing out of cold storage equipment, and the direct uses include meeting the cold needs of the data center and storage in cold storage equipment.

[0072] Due to the high level of operational reliability required by data centers, the energy system they adapt to should be able to ensure the stable operation of data centers when some special events (natural disasters, abnormal weather, etc.) occur. The main energy demands of data centers include electricity and cooling, so the impact of the above special events on the normal operation of data centers is manifested in the energy system as power outages (islands) and equipment failures.

[0073] The working conditions are used to describe the internal structure of the system, including whether each device is working properly and whether it can interact with the power grid. The scenarios are used to describe the external environment and working conditions of the system at a certain time, including: electricity demand, cooling demand, hydrogen price, electricity price and working conditions.

[0074] The energy system of the data center meets the power demand and cooling demand in three working conditions (normal condition, island condition, and equipment failure condition) and the circulation conditions of each water circulation loop as follows:

[0075] Normal situation: that is, the situation where the hydrogen-containing multi-energy system of the data center interacts normally with the power grid and all equipment operates normally. At this time, the sum of the power consumption of the data center itself and the power consumption of the multi-energy system of the data center includes: electricity demand, electrolyzer power consumption, cold water distribution unit power consumption, absorption chiller power consumption, and electricity sold to the power grid. The power consumption can be provided by purchasing electricity from the power grid, photovoltaic panels, and fuel cells; the sum of the cold consumption of the data center itself and the cold consumption of the multi-energy system of the data center is the cold demand of the data center, and the cold consumption is provided by the cold storage equipment and absorption chiller transformed from the fire water tank; all loops in the above hot water circulation loop and cold water circulation loop can flow.

[0076] Island situation: the situation where the interaction between the energy system and the power grid is interrupted, while the rest of the equipment operates normally. At this time, the system cannot sell electricity to the power grid. The total power consumption of the data center and the energy system is composed of power demand, electrolyzer power consumption, cold water distribution unit power consumption, and absorption chiller power consumption. The system cannot buy electricity from the power grid. The power consumed by the data center and the energy system is only provided by photovoltaic panels and fuel cells. The total cooling consumption of the data center and the energy system is the cooling demand of the data center. The total cooling consumption is provided by the cold storage equipment converted from the fire water tank and the absorption chiller. All loops in the above hot water circulation loop and cold water circulation loop can flow.

[0077] Equipment failure: that is, the absorption chiller fails, while the rest of the parts operate normally. At this time, the total power consumption of the data center and the energy system includes: electricity demand, electrolytic cell power consumption, cold water distribution unit power consumption, absorption chiller power consumption, and power sold to the power grid. The power consumed can be provided by purchasing electricity from the power grid, photovoltaic panels, and fuel cells; the total cooling consumption of the data center and the energy system is the cooling demand of the data center, and the cooling consumed is only provided by the cold storage equipment transformed from the fire water tank; in the above-mentioned hot water circulation loop, only hot water circulation loop 1 and hot water circulation loop 2 can circulate, that is, the hot water tank does not release heat, but only stores heat to be used after the absorption chiller failure is repaired; in the cold water circulation loop, only cold water circulation loop 1 can circulate, that is, all cooling is provided by the cooling stored in the cooling storage equipment.

[0078] Example 2

[0079] Reference Figure 2 The design method of the hydrogen-containing multi-energy system for data centers considering operational reliability includes four steps: obtaining data through investigation, establishing an energy system planning model with the expected reliability level, simplifying the segmentation model according to the scenario, generating stability constraints and solving them iteratively. The specific description is as follows:

[0080] S1. Data acquisition through investigation: Use simulation or field investigation methods to obtain the hourly electricity demand of the data center throughout the year (hereinafter referred to as electricity demand), hourly cooling demand throughout the year (hereinafter referred to as cooling demand), hourly solar radiation intensity throughout the year, the total area that can be installed with solar panels and solar collectors, and the shape and parameters of the fire water pool.

[0081] S2. Establish an energy system planning model with the expected reliability level: This model uses the parameters and operation scheduling strategies of each device as the variables to be solved, and determines the constraints of the variables based on the electricity demand, cold demand satisfaction method and the circulation conditions of each water circulation loop of the energy system under normal conditions, island conditions and equipment failure conditions, so that the system configured according to the solution (i.e. the planned parameters of each device) can run stably and with the minimum total cost under normal conditions, and can still run stably for n hours in the event of sudden specific failures (island conditions and equipment failure conditions), that is, it can run stably in all scenarios and with the minimum total cost under normal scenarios. Among them, the parameters of each device include the area of ​​solar panels, the area of ​​solar collectors, the power of electrolyzers, the power of fuel cells, the power of absorption chillers, the power of cold water distribution units, the capacity of hot water tanks, the capacity of cold storage equipment, the capacity of hydrogen storage tanks, and the total cost includes the sum of the purchase cost of the equipment and the cost of the energy system under normal conditions, both of which are converted to years for calculation. This model is called the original model.

[0082] Each scenario specifies the system's work start time i, work conditions (normal, isolated, equipment failure), and work duration n. According to the work conditions, it can be divided into three categories: normal scenario, isolated scenario, and equipment failure scenario. The normal scenario is a scenario in which the system starts from any period i and lasts for n hours in normal conditions. The isolated scenario means that the system starts from any period i and lasts for n hours in an isolated situation. The equipment failure scenario means that the system starts from any period i and lasts for n hours in an equipment failure situation. The work duration n also represents the expected reliability level. The larger n is, the longer the system can meet the energy consumption requirements of the data center when it is isolated or the equipment fails, and the higher the reliability level. The constraint models of all normal scenarios are the same, but the power demand and cooling demand are different due to different work start times, so different variables need to be used and solved when solving. The same is true for the isolated and equipment failure scenarios. Assume that there are s1 normal scenarios, s2 island scenarios, and s3 equipment failure scenarios. Each normal scenario has x1 variables, each island scenario has x2 variables, and each equipment failure scenario has x3 variables. All scenarios share a set of equipment parameters, for a total of 9 variables. Then the complete original model has variables.

[0083] S3. Simplify and split the model according to the scenario: Since the number of variables in the original model obtained in S2 is too large to be solved, the model obtained in S2 needs to be simplified and split, so as to obtain a main problem and sub-problems that are easier to solve, so that the number of variables to be solved at a single time is greatly reduced. That is, the original model is split according to different scenarios, the planning problem under the typical normal scenario is regarded as the main problem, and the operation problem under each scenario is regarded as a sub-problem, and all sub-problems share a set of equipment parameters, which are obtained by the main problem planning. The stable operation of the system under the typical normal scenario is taken as the main problem constraint, the minimum sum of the investment cost and the operating cost of the system under the typical normal scenario is taken as the main problem objective function, the stable operation of the system under each scenario is taken as the sub-problem constraint, and the minimum sum of the slack variables under each scenario is taken as the sub-problem objective. Among them, the typical normal scenario refers to selecting a representative week in each of the four seasons of spring, summer, autumn and winter, and the system works normally during the week. Among them, the minimum sum of slack variables means the sum of electricity demand and cooling demand that cannot be met by the current energy system is the minimum. Only when the sum of slack variables is 0 does it mean that the system can operate stably under this scenario.

[0084] S4, generate stability constraints and solve iteratively: For the simplified split model in S3 (the sum of the main problem and sub-problems), since some scenarios in the sub-problems often require higher equipment power or capacity, for example, in the scenario of absorption chiller failure, larger capacity cold storage facilities are often required to meet the cooling demand. Therefore, it is difficult to obtain equipment parameters that can operate stably in all scenarios after solving the main problem once, so it is necessary to perform multiple iterative solutions based on the reliability improvement algorithm. In order to iterate faster and more accurately, the solution obtained from the sub-problem is used to provide a basis for modifying the main problem, that is, after solving the main problem and sub-problems obtained in S3, the stability constraints (i.e., infeasible cuts) are generated according to the generalized Benders decomposition principle and added to the main problem for iterative solution, and finally the equipment planning results that meet all scenarios and have a high level of reliability and economy are obtained.

[0085] Among them, for the energy system planning model with the expected reliability level described in S2, the following constraints are included: hydrogen balance constraints in all scenarios, electricity balance constraints in all scenarios, heat balance constraints in all scenarios, cold balance constraints in all scenarios, buying and selling electricity constraints in all scenarios, solar energy production capacity constraints in all scenarios, electrolyzer operation constraints in all scenarios, fuel cell operation constraints in all scenarios, absorption chiller power consumption constraints in all scenarios, hot water tank related constraints in all scenarios, cold storage equipment related constraints in all scenarios, hydrogen storage tank related constraints in all scenarios, and equipment parameter constraints in all scenarios.

[0086] The objective function of the original model is:

[0087]

[0088]

[0089] The objective function means that the total cost of the system under normal conditions is minimized. Among them, e represents a certain device, and E is a set of devices e, including SP, ST, EL, FC, AC, CD, HW, CW, and HS. Among them, SP is a solar panel, ST is a solar collector, EL is an electrolyzer, FC is a fuel cell, AC is an absorption chiller, CD is a cold water distribution unit, HW is a hot water tank, CW is a cold storage device, and HS is a hydrogen storage tank. is the capital recovery factor of equipment e, is the unit cost of equipment e, is the planning parameter of device e, N is the scaling factor from typical day to year, s1 is the index of normal scenario, S1 is the set of normal scenarios, is the probability of the normal scenario s1 occurring, is the price of hydrogen at the hydrogen refueling station, is the amount of hydrogen added from the hydrogen refueling station during period t in the normal scenario s1, is the price of electricity purchased from the power grid during period t, is the amount of electricity purchased from the grid during period t under the normal scenario s1, is the price of electricity sold to the grid during period t, is the amount of electricity sold to the grid during period t under the normal scenario s1, The length of the period.

[0090] The original model is subject to 14 constraints as follows:

[0091] (1) Hydrogen balance constraints under all scenarios:

[0092] in, is the mass of hydrogen purchased from the hydrogen market during period t in scenario s, is the mass of hydrogen produced by electrolysis in the electrolyzer during period t in scenario s, is the mass of hydrogen output from the hydrogen storage tank during period t in scenario s, is the mass of hydrogen consumed by the fuel cell cogeneration in period t in scenario s, t is the period index, s is the scenario index, S1 is the set of normal scenarios, S2 is the set of island scenarios, and S3 is the set of equipment failure scenarios.

[0093] (2) Electricity balance constraints under all scenarios:

[0094]

[0095] This balance takes into account the power consumption of the power distribution unit and the cold water distribution unit, and is more suitable for the high power and cold water consumption characteristics of data centers. is the power purchased from the grid during period t in scenario s, is the power generated by the solar panel during period t in scenario s, is the power generated by the fuel cell during period t in scenario s, is the power fed back to the grid during period t in scenario s, is the power consumed by the electrolyzer during period t in scenario s, is the power consumed by the cold water distribution unit during period t in scenario s, is the power consumed by the absorption chiller during period t in scenario s, is the power consumed by the power distribution unit during period t in scenario s.

[0096] (3) Thermal balance constraints under all scenarios:

[0097] in, is the heat recovery efficiency of the fuel cell heat recovery device, is the heat generated by the fuel cell cogeneration during period t in scenario s, is the heat absorbed by the solar collector during period t in scenario s, is the heat flowing out of the hot water tank during period t in scenario s, is the heat absorbed by the absorption refrigerator during period t in scenario s.

[0098] (4) Cold equilibrium constraints in all scenarios:

[0099] in, is the cooling capacity generated by the absorption chiller during period t in scenario s, is the cooling capacity of the cold storage equipment transformed from the fire water tank in time period t in scenario s, is the cooling capacity required by the data center during period t in scenario s.

[0100] (5) Constraints on buying and selling electricity in all scenarios:

[0101]

[0102]

[0103] Among them, M is the maximum value, and is the flag bit of the interaction between the system and the power grid during period t in scenario s. When the system purchases electricity from the power grid, is 1, is 0; when the system feeds back power to the grid, is 0, is 1;

[0104] (6) Solar energy capacity constraints under all scenarios:

[0105]

[0106]

[0107] in, is the photoelectric conversion efficiency of the solar panel, is the solar panel area, is the solar radiation intensity at time t in scenario s, is the heat recovery efficiency of the solar collector heat recovery device, is the solar collector area, and D is the total area where solar panels and solar collectors can be installed.

[0108] (7) Electrolyzer operating constraints under all scenarios,

[0109] in, is the hydrogen production efficiency of the electrolyzer.

[0110] (8) Fuel cell operation constraints under all scenarios,

[0111] in, is the hydrogen-to-electricity conversion efficiency of the fuel cell, is the electrical-to-thermal conversion efficiency of the fuel cell.

[0112] (9) Power consumption constraints of cooling-related equipment under each scenario:

[0113]

[0114]

[0115] in, , , They are the heat-to-cold conversion efficiency of the absorption chiller, the cold-to-electricity conversion coefficient of the cold water distribution unit, and the cold-to-electricity conversion coefficient of the absorption chiller.

[0116] (10) Hot water tank related constraints under each scenario:

[0117]

[0118]

[0119]

[0120]

[0121] in, and is the mark of the heat in and out of the hot water tank in the system during period t in scenario s. When the hot water tank supplies heat, is 1, is 0; when the hot water tank stores heat, is 0, is 1; and are the maximum and minimum amount of heat supplied by the hot water tank per unit time, and are the maximum and minimum amount of heat stored in the hot water tank per unit time, and are the heat stored in the hot water tank at period t and period t+1 in scenario s, is the heat loss coefficient of the hot water tank.

[0122] (11) Relevant constraints of cold storage equipment (after fire water tank transformation) under each scenario:

[0123]

[0124]

[0125]

[0126]

[0127] in, and is the sign of the inlet and outlet cooling capacity of the cold storage equipment transformed from the fire water tank in the system during period t in scenario s. When the cold storage equipment transformed from the fire water tank supplies cooling capacity, is 1, is 0; when the cold storage equipment transformed from the fire water tank stores cold energy, is 0, is 1; and They are the maximum and minimum amount of cold energy supplied per unit time by the cold storage equipment transformed from the fire water tank. and They are the maximum and minimum amount of cold storage per unit time of the cold storage equipment transformed from the fire water tank, and are the cold storage capacity in the cold storage equipment of the fire water tank transformation in period t and period t+1 in scenario s, The cooling loss coefficient of the cold storage equipment for the fire water tank renovation.

[0128] (12) Hydrogen storage tank related constraints under each scenario:

[0129]

[0130]

[0131]

[0132]

[0133] in, and is the mark of the amount of hydrogen in and out of the hydrogen storage tank in the system during period t in scenario s. When the hydrogen storage tank supplies hydrogen, is 1, is 0; when the hydrogen storage tank stores hydrogen, is 0, is 1; and They are the maximum and minimum amount of hydrogen supplied by the hydrogen storage tank per unit time, and are the maximum and minimum amount of hydrogen that can be stored in a hydrogen storage tank per unit time, respectively. and are the amounts of hydrogen stored in the hydrogen storage tank at period t and period t+1 in scenario s, respectively.

[0134] (13) Parameter constraints of each device in all scenarios:

[0135]

[0136]

[0137]

[0138]

[0139] in, , , , , , , , , They represent the solar panel area, solar collector area, electrolyzer power, fuel cell power, absorption chiller power, cold water distribution unit power, hot water tank capacity, cold storage equipment capacity, and hydrogen storage tank capacity, that is, the planned equipment parameters.

[0140] (14) Calculation formula for investment recovery coefficient of each equipment:

[0141] in, is the investment recovery coefficient of equipment e, I is the interest rate, is the useful life of equipment e.

[0142] Among them, the main problem objective function in S3 is:

[0143]

[0144]

[0145] Among them, s4 is the scenario index, is the set of all typical normal scenarios.

[0146] The main problem constraints include: hydrogen balance constraints under typical normal scenarios, electricity balance constraints under typical normal scenarios, heat balance constraints under typical normal scenarios, cold balance constraints under typical normal scenarios, buying and selling electricity constraints under typical normal scenarios, solar energy production capacity constraints under typical normal scenarios, electrolyzer operation constraints under typical normal scenarios, fuel cell operation constraints under typical normal scenarios, power consumption constraints of cold-related equipment under typical normal scenarios, hot water tank related constraints under typical normal scenarios, cold storage equipment related constraints under typical normal scenarios, hydrogen storage tank related constraints under typical normal scenarios, capacity constraints of each device under each scenario, and calculation formulas for the investment recovery coefficient of each device. The meaning of each parameter is the same as in the original model. The main difference is that the scope of scenarios in which the constraints are satisfied is different. The constraints in the original model need to be satisfied in all scenarios, while the constraints of this main problem only need to be satisfied in typical normal scenarios. The specific description is as follows:

[0147] (1) Hydrogen balance constraints under typical normal scenarios:

[0148] (2) Electricity balance constraints under all scenarios:

[0149]

[0150] (3) Thermal balance constraints under typical normal scenarios:

[0151] (4) Cold balance constraints under typical normal scenarios:

[0152] (5) Constraints on buying and selling electricity under typical normal circumstances:

[0153]

[0154] (6) Solar energy capacity constraints under all scenarios:

[0155]

[0156] (7) Electrolyzer operation constraints under typical normal scenarios,

[0157] (8) Fuel cell operation constraints under typical normal scenarios,

[0158]

[0159] (9) Power consumption constraints of cooling-related equipment under typical normal scenarios:

[0160]

[0161]

[0162] (10) Hot water tank related constraints under typical normal scenarios:

[0163]

[0164]

[0165]

[0166] (12) Relevant constraints of cold storage equipment (after fire water tank transformation) under typical normal scenarios:

[0167]

[0168]

[0169]

[0170]

[0171] (12) Hydrogen storage tank related constraints under typical normal scenarios:

[0172]

[0173]

[0174]

[0175] (13) Parameter constraints of each device under typical normal scenarios:

[0176]

[0177]

[0178]

[0179] (14) Calculation formula for investment recovery coefficient of each equipment:

[0180] Among them, for the subproblems described in S3, the mathematical descriptions of each subproblem are the same, only the demand data and some working state parameters are different. Since the operation problem in each scenario is regarded as a subproblem, the number of subproblems is the same as the total number of all scenarios, and there is no coupling between the subproblems, so they can be solved simultaneously. For any subproblem, its objective function is:

[0181]

[0182] Among them, the objective function transforms some originally unsolvable problems into solvable problems by introducing slack variables, so that the solutions of these problems can convey certain information, which is then fed back to the main problem for the next iteration. At this time, the meaning of the objective function is that the total amount of electricity demand and cooling demand that cannot be met by the system is minimized. and is the introduced slack variable, Used to relax electrical balance constraints, Used to relax the cooling balance constraint. When the objective function is 0, it means that the current system can meet the electricity demand and cooling demand in the current scenario. When it is not 0, it cannot be met.

[0183] For any sub-problem, the solar panel area, solar collector area, electrolyzer power, fuel cell power, absorption chiller power, cold water distribution unit power, hot water tank capacity, cold storage equipment capacity, and hydrogen storage tank capacity during operation are known quantities, that is, the parameters of each device obtained by solving the main problem. Therefore, sub-problem parameter consistency constraints are added to unify the main problem and sub-problems, and slack variables are introduced to change the electrical balance and cold balance equations in the original model so that the sub-problem is necessarily solvable. At the same time, sub-problem island scenario constraints and sub-problem equipment failure scenario constraints are added to describe the island scenario and equipment failure scenario respectively. Therefore, the constraints it is subject to include:

[0184] (1) Sub-problem: Hydrogen balance constraint:

[0185] in, is the mass of hydrogen purchased from the hydrogen market during period t, is the mass of hydrogen produced by electrolysis in the electrolyzer during period t, is the mass of hydrogen output from the hydrogen storage tank during period t, is the mass of hydrogen consumed by the fuel cell cogeneration during period t, where t is the period index.

[0186] (2) Sub-problem: Electric balance constraint:

[0187] This balance introduces To relax the power balance, the model can still be solved when the system configured according to the currently planned equipment parameters cannot meet the power demand, especially in the case of an island. is the power purchased from the grid during period t, is the power generated by the solar panel during period t, is the power generated by the fuel cell during period t, is the power fed back to the grid during period t, is the power consumed by the electrolytic cell during period t, is the power consumed by the cold water distribution unit during period t, is the power consumed by the absorption chiller during period t, The power consumed by the power distribution unit during period t.

[0188] (3) Sub-problem: thermal balance constraint:

[0189] in, is the heat generated by the fuel cell cogeneration during period t, is the amount of heat absorbed by the solar collector during period t, is the heat flowing out of the hot water tank during period t, is the heat absorbed by the absorption refrigerator during period t.

[0190] (4) Sub-problem cold balance constraint:

[0191] This balance introduces To relax the cold balance, the model can still be solved when the system configured according to the currently planned equipment parameters cannot meet the cold demand, especially in the case of equipment failure. is the cooling capacity generated by the absorption chiller during period t, is the cooling capacity of the cold storage equipment transformed from the fire water tank during period t, is the cooling capacity required by the data center during period t.

[0192] (5) Consistency constraints on sub-problem parameters:

[0193]

[0194]

[0195]

[0196]

[0197] This part of the constraint is used to ensure the consistency of parameters between the main problem and the sub-problems, that is, the equipment parameters used in each sub-problem are the same and are obtained from the planning of the main problem. and As the input of the reliability improvement algorithm, it participates in the next iteration. and To calculate the generated Lagrange multiplier, The parameters of each device obtained by the previous iteration of the main problem, The energy in and out flags of the energy storage device obtained by the previous iteration of the main problem, are variables representing the parameters of each device in the sub-problem, is the variable representing the energy in and out flag of the energy storage device in the sub-problem. and is the mark of the heat input and output of the hot water tank in the system during period t. When the hot water tank supplies heat, is 1, is 0; when the hot water tank stores heat, is 0, is 1. and It is the sign of the inlet and outlet cooling capacity of the cold storage equipment transformed from the fire water tank in the system during period t. When the cold storage equipment transformed from the fire water tank supplies cooling capacity, is 1, is 0; when the cold storage equipment transformed from the fire water tank stores cold energy, is 0, is 1. and It is the mark of the amount of hydrogen in and out of the hydrogen storage tank in the system during period t. When the hydrogen storage tank supplies hydrogen, is 1, is 0; when the hydrogen storage tank stores hydrogen, is 0, is 1, and B is the set of energy storage device entry and exit states.

[0198] (6) Sub-problem island constraints:

[0199]

[0200] This part restricts the use of To describe the changes of the model in the case of an island, that is, when When is 0, the power purchased from the grid and the power fed back to the grid are both 0. This constraint converts the impact of the island situation on the multi-energy system into a mathematical form that can be added to the existing model for solution. It is a flag to indicate whether the interaction with the power grid is normal. It is used to indirectly describe the situation of the system. When the interaction between the system and the power grid is normal, is 1, otherwise it is 0, that is, the system is running in an island scenario is 0.

[0201] (7) Sub-problem equipment failure constraints:

[0202]

[0203] This part restricts the use of To describe the changes in the model under equipment failure conditions, that is, when When is 0, the cooling capacity of the absorption chiller is 0. This constraint converts the impact of equipment failure on the multi-energy system into a mathematical form that can be added to the existing model for solution. It is the working status flag of the absorption chiller, which is used to indirectly describe the situation of the system. When the absorption chiller of the system works normally, 1, otherwise 0, that is, the system is running in the device failure scenario is 0.

[0204] (8) Sub-problem: Electricity purchase and sale constraints:

[0205]

[0206] in, and It is the flag of the interaction between the system and the power grid during period t. When the system purchases electricity from the power grid, is 1, is 0; when the system feeds back power to the grid, is 0, is 1;

[0207] (9) Sub-problem: Solar energy capacity constraints:

[0208] ;

[0209] (10) Sub-problem: electrolytic cell operation constraints, ;

[0210] (11) Sub-problem: fuel cell operation constraints, ;

[0211] (12) Sub-problem: Power consumption constraints of cold-related equipment: ;

[0212]

[0213]

[0214] (13) Sub-problem: Hot water tank related constraints:

[0215]

[0216]

[0217]

[0218] in, and are the heat stored in the hot water tank during period t and period t+1 respectively.

[0219] (14) Sub-problem: Constraints related to cold storage equipment (after fire water tank transformation):

[0220]

[0221]

[0222]

[0223] in, and They are the cold storage capacity in the cold storage equipment transformed from the fire water tank during period t and period t+1 respectively.

[0224] (15) Sub-problem: Constraints related to hydrogen storage tanks:

[0225]

[0226]

[0227]

[0228] in, and are the amounts of hydrogen stored in the hydrogen storage tank during period t and period t+1 respectively.

[0229] (16) Capacity constraints of each device in the sub-problem:

[0230]

[0231]

[0232]

[0233] Reference Figure 3The iterative solution of the main problem and the sub-problems according to the reliability improvement algorithm described in S4 includes the following steps:

[0234] S401. Solve the main problem and get the solution , let i=1. Among them, The parameters of each device obtained from the previous iteration of the main problem include solar panel area, solar collector area, electrolyzer power, fuel cell power, absorption chiller power, cold water distribution unit power, hot water tank capacity, cold storage equipment capacity, and hydrogen storage tank capacity.

[0235] S402, construct sub-problems, for the i-th scenario in the scenario set s containing S scenarios, solve Substitute the subproblem of scenario s[i] and solve the subproblem to obtain the solution of the subproblem , and ;

[0236] S403, if , add stability constraints to the main problem, jump to S401 to continue calculating the new solution ;like , jump to S404;

[0237] S404, if , i=i+1, jump to S402 to continue calculation; if , the solution is completed.

[0238] The stability constraint in S403 is:

[0239]

[0240] in, is the solution to a sub-problem, and To calculate the generated Lagrange multiplier, The parameters of each device obtained by the previous iteration of the main problem, The energy in and out flags of the energy storage device obtained by the previous iteration of the main problem, are variables representing the parameters of each device in the sub-problem, is the variable representing the energy in and out flag of the energy storage device in the sub-problem. E and B have been described in the sub-problem parameter consistency constraint.

[0241] Among them, refer to Figure 4, S4 is an iterative solution to the main problem and the sub-problem, and after the i-th scenario is iterated, the constraints of the main problem include: a set of constraints of the typical normal scenario, a set of reliability constraints added by the first scenario, a set of reliability constraints added by the second scenario, ..., a set of reliability constraints added by the i-th scenario. The number of reliability constraints added to each scenario is not fixed and is determined by the optimization solution of the sub-problem after iteration.

[0242] At the same time, in order to further improve the utilization rate of the fire water pool configured in the data center, a method is provided to select the most appropriate fire water pool renovation plan based on the planning results and fire water pool parameters, thereby reducing the construction cost and floor space of the energy system.

[0243] Reference Figure 5 The steps for generating the transformation plan for the fire water pool of the hydrogen-containing multi-energy system in the data center are as follows:

[0244] SA1. According to the above-mentioned planning method of hydrogen-containing multi-energy system for data centers considering operation reliability, the capacity of cold storage equipment is obtained. and the cooling capacity in and out of the cold storage facility under typical scenarios , and investigate the shape and parameters of existing or to-be-built fire water pools;

[0245] SA2. Determine several candidate renovation plans based on the unified technical measures for fire water pool renovation required by national standards, building specifications and construction units. Each renovation plan includes: the shape, size, insulation and waterproof materials of the fire water pool after renovation, the type and installation location of the water distributor, the height and diameter of each water outlet, the height and diameter of each water inlet, the height and diameter of the water hydrant, the standard water flow rate of each water inlet and outlet, and the standard volume flow of each water inlet and outlet;

[0246] SA3. For each set of renovation solutions determined in SA2, use commonly used CFD (computational fluid dynamics) software, such as ANSYS Fluent, to simulate the fire water pool after renovation. The simulated value is the cooling capacity in and out of the cold storage facility under the typical scenario obtained in SA1. , and finally the experiment resulted in the optimal fire water tank renovation plan.

[0247] Among them, the combinations of thermal insulation and waterproof materials in the above SA2 include: molded thermal insulation material mortar waterproof material, molded thermal insulation material board waterproof material, and on-site foamed thermal insulation material waterproof surface coating.

[0248] Among them, the types of water distributors in the above SA2 include octagonal, radial disc, stripe (horizontally connected stripe), and H-shaped. The first two are suitable for cylindrical water storage tanks, while the latter two are more suitable for square water tanks. Their function is to arrange the inflow and outflow of water in a certain working area according to a certain rule, so as to avoid water flow turbulence caused by the inflow and outflow of water, reduce the extent of the increase in the temperature of the water outlet at the bottom of the fire water tank due to the output of cold water, further improve the utilization rate of cold water and reduce the power consumption of the water pump.

[0249] The simulation of the fire water pool in the above SA3 includes the following steps:

[0250] S301, create a model using SolidWorks according to the parameters determined in SA2, and perform meshing using ICEM;

[0251] S302: The cooling capacity in and out of the cold storage facility under the typical scenario obtained in SA1 The parameters in SA2 are used to preliminarily determine the water inlet and temperature, water outlet and temperature in each time period, and the simulation algorithm and calculation conditions are selected in ANSYS Fluent;

[0252] S303, using the parameters selected in S302 to perform simulation calculations, and obtain the temperature field, velocity field, and thermocline layer characteristics;

[0253] S304. Select the solution with the best temperature stratification effect, the smallest cold loss, and the highest cold water utilization efficiency among the renovation solutions to renovate the fire water tank, and obtain the optimal inlet and outlet temperature and flow rate under the solution.

[0254] Among them, when the height of the existing or to-be-built fire water pool is too low, resulting in the water distributor and the natural temperature stratification modification plan not meeting the requirements, a diaphragm or partition is used to vertically stratify the fire water pool to reduce water mixing. At this time, the connection method of the hydrogen-containing multi-energy system is as follows: Figure 6 As shown, at this time, the working mode of the system in each working condition and the connection mode of each hot water circulation circuit are the same as Figure 1 The systems shown are identical.

[0255] At this time, there are still three cold water loops in the system, and the three loops can be used at the same time. The cold water circulation mode and flow rate are controlled by controlling the switch status of valve 1, valve 2, valve 3, and valve 4, as well as the working power of the primary water pump and the secondary water pump. The three loops are:

[0256] Loop 1: The water with lower temperature on the right side of the fire water tank passes through valve 4 and the first-level water pump in turn, and is reasonably distributed to the air-conditioning system of each computer room through the cold water distribution unit. After heat exchange and temperature increase in the air-conditioning system of each computer room, it returns to the left side of the fire water tank through valve three, that is, the cold energy stored in the fire water tank is used to cool the computer room of the data center.

[0257] Circuit 2: The water on the left side of the fire water tank passes through the secondary water pump and valve 1 in turn, is cooled by the absorption chiller, and then flows back to the right side of the fire water tank through valve 2, that is, the cold energy generated by the absorption chiller is stored in the fire water tank.

[0258] Loop 3: The water cooled by the absorption chiller passes through valve 2, valve 4 and the primary water pump in turn, and is reasonably distributed to the air-conditioning system of each computer room through the cold water distribution unit. After heat exchange and temperature rise in the air-conditioning system of each computer room, it flows back to the absorption chiller through valve 3 and valve 1 for the next cooling, that is, the cold energy generated by the absorption chiller is used to cool the computer room of the data center.

[0259] Reference Figure 7 The two types of fire water tanks have at least four water inlets and outlets. Under normal circumstances, the fire water tanks after the transformation are naturally layered horizontally according to the relationship between temperature and density. No partition is added. The closer to the bottom of the fire water tank, the lower the water temperature. The connection method in the hydrogen-containing multi-energy system is as follows Figure 1 As shown; when the height of the fire water tank is too low so that the horizontal stratification effect is poor, a baffle is used to control the flow of water for vertical stratification, so that the water temperature is lower closer to the right side of the fire water tank and higher closer to the left side of the fire water tank.

[0260] Example 3

[0261] Reference Figure 8 , a data center multi-energy system design device considering operational reliability includes a data acquisition module, a model building module, a model segmentation module and a solution module connected in sequence.

[0262] Among them, the data acquisition module is used to collect the hourly electricity demand, hourly cooling demand, hourly solar radiation intensity, total area for installing solar panels and solar collectors, and shape and parameters of the data center throughout the year, and transmit the collected hourly electricity demand, hourly cooling demand, hourly solar radiation intensity, total area for installing solar panels and solar collectors, and shape and parameters of the fire pool to the model establishment module; the model establishment module is used to establish an energy system planning model with an expected reliability level according to the data transmitted by the data acquisition module; the model segmentation module is used to simplify and segment the energy system planning model to obtain the main problem and sub-problems; the solution module is used to generate stability constraints and add them to the main problem and sub-problems for iterative solution to obtain the planning results of each equipment that meets all scenarios.

[0263] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0264] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

Claims

1. A design method for a data center multi-energy system considering operational reliability, characterized in that: The following steps are involved: S1. Obtain data: obtain the power demand, cooling demand, and total area where solar panels and solar thermal collectors can be installed of the target data center; S2. Establish an energy system planning model with an expected reliability level based on electricity demand, cooling demand, and the total area where solar panels and solar thermal collectors can be installed: the energy system planning model uses the parameters of each device and the operation scheduling strategy as the variables to be solved, and determines the constraints on the variables based on the electricity demand of the energy system under normal conditions, island conditions, and equipment failure conditions, the way to meet the cooling demand, and the flow conditions of each water circulation loop, and takes stable operation in all scenarios and the minimum total cost in normal scenarios as the objective function, so that the system configured according to the solved parameters of each device can meet the requirements of stable operation under normal conditions and the minimum total cost; S3. Simplify and split the energy system planning model according to the scenario: simplify and split the energy system planning model obtained in S2 to obtain the main problem and sub-problems; The planning problem under a typical normal scenario is regarded as the main problem, and the operation problem under each scenario is regarded as a sub-problem. All sub-problems share a set of equipment parameters, which are obtained by planning the main problem. The typical normal scenario refers to selecting a representative week in each of the four seasons of spring, summer, autumn and winter, and the system works normally in that week; S4. Generate stability constraints and add them to the main problem and sub-problems for iterative solution to obtain the planning results of each device that meets all scenarios.

2. The design method of a data center multi-energy system considering operational reliability according to claim 1 is characterized in that: In S2, the constraints include hydrogen balance constraints in all scenarios, electricity balance constraints in all scenarios, heat balance constraints in all scenarios, cold balance constraints in all scenarios, buying and selling electricity constraints in all scenarios, solar energy production capacity constraints in all scenarios, electrolyzer operation constraints in all scenarios, fuel cell operation constraints in all scenarios, absorption chiller power consumption constraints in all scenarios, hot water tank related constraints in all scenarios, cold storage equipment related constraints in all scenarios, hydrogen storage tank related constraints in all scenarios, and equipment parameter constraints in all scenarios. The electrical balance constraints in all the scenarios are: in, is the power purchased from the grid during period t in scenario s, is the power generated by the solar panel during period t in scenario s, is the power generated by the fuel cell during period t in scenario s, is the power fed back to the grid during period t in scenario s, is the power consumed by the electrolyzer during period t in scenario s, is the power consumed by the cold water distribution unit during period t in scenario s, is the power consumed by the absorption chiller during period t in scenario s, is the power consumed by the power distribution unit during period t in scenario s.

3. The design method of a data center multi-energy system considering operational reliability according to claim 1 is characterized in that: In S3, the objective function of the subproblem is: in, and is the introduced slack variable, Used to relax electrical balance constraints, Used to relax the cooling balance constraint. When the objective function is 0, it means that the current system can meet the electricity demand and cooling demand in the current scenario. When it is not 0, it cannot be met.

4. The design method of a data center multi-energy system considering operational reliability according to claim 1 is characterized in that: In S3, the constraints of the sub-problems include the sub-problem electric balance constraint, which is expressed as: in, is the power purchased from the grid during period t, is the power generated by the solar panel during period t, is the power generated by the fuel cell during period t, is the power fed back to the grid during period t, is the power consumed by the electrolytic cell during period t, is the power consumed by the cold water distribution unit during period t, is the power consumed by the absorption chiller during period t, The power consumed by the power distribution unit during period t.

5. The design method of a data center multi-energy system considering operational reliability according to claim 1 is characterized in that: In S3, the constraints of the subproblems include the consistency constraints of the subproblem parameters. The expression of the consistency constraints of the subproblem parameters is as follows: in, and To calculate the generated Lagrange multiplier, The parameters of each device obtained by the previous iteration of the main problem, The energy in and out flags of the energy storage device obtained by the previous iteration of the main problem, are variables representing the parameters of each device in the sub-problem, is the variable representing the energy in and out flag of the energy storage device in the sub-problem; and is the mark of the heat input and output of the hot water tank in the system during period t. When the hot water tank supplies heat, is 1, is 0; when the hot water tank stores heat, is 0, is 1; and It is the sign of the inlet and outlet cooling capacity of the cold storage equipment transformed from the fire water tank in the system during period t. When the cold storage equipment transformed from the fire water tank supplies cooling capacity, is 1, is 0; when the cold storage equipment transformed from the fire water tank stores cold energy, is 0, is 1; and It is the mark of the amount of hydrogen in and out of the hydrogen storage tank in the system during period t. When the hydrogen storage tank supplies hydrogen, is 1, is 0; when the hydrogen storage tank stores hydrogen, is 0, is 1, and B is the set of energy storage device entry and exit states.

6. The design method of a data center multi-energy system considering operational reliability according to claim 1, characterized in that: In S3, the constraints of the subproblems include the subproblem island constraints: in, It is a flag to indicate whether the interaction with the power grid is normal. It is used to indirectly describe the situation of the system. When the interaction between the system and the power grid is normal, is 1, otherwise it is 0, that is, the system is running in an island scenario is 0.

7. The design method of a data center multi-energy system considering operational reliability according to claim 1, characterized in that: In S3, the constraints of the sub-problems include the sub-problem equipment failure constraints: in, It is the working status flag of the absorption chiller, which is used to indirectly describe the situation of the system. When the absorption chiller of the system works normally, 1, otherwise 0, that is, the system is running in the device failure scenario is 0.

8. The design method of a data center multi-energy system considering operational reliability according to claim 1, characterized in that: In S4, the iterative solution of the main problem includes the following steps: S401. Solve the main problem and get the solution , let i=1; S402, construct sub-problems, for the i-th scenario in the scenario set s containing S scenarios, solve Substitute the subproblem of scenario s[i] and solve the subproblem to obtain the solution of the subproblem , Lagrange multipliers and Lagrange multipliers ; S403, if , add stability constraints to the main problem, jump to S401 to continue calculating the new solution ;like , jump to S404; S404, if , i=i+1, jump to S402; if , the solution is completed.

9. The design method of a data center multi-energy system considering operational reliability according to claim 8, characterized in that: In S403, the stability constraint is: in, is the solution to a sub-problem, and To calculate the generated Lagrange multiplier, The parameters of each device obtained by the previous iteration of the main problem, The energy in and out flags of the energy storage device obtained by the previous iteration of the main problem, are variables representing the parameters of each device in the sub-problem, is the change of the energy flag bit of the energy storage device in the sub-problem, E is the set of devices e, and B is the set of energy storage device entry and exit states.

10. A design device for a multi-energy system of a data center considering operational reliability, characterized in that: It includes a data acquisition module, a model building module, a model segmentation module and a solution module which are connected in sequence; Among them, the data acquisition module is used to collect raw data and pass the collected raw data to the model building module. The raw data includes electricity demand, cooling demand, and the total area where solar panels and solar collectors can be installed; the model building module is used to establish an energy system planning model with an expected reliability level based on the raw data; the model segmentation module is used to simplify and segment the energy system planning model to obtain the main problem and sub-problems; the planning problem under the typical normal scenario is regarded as the main problem, and the operation problem under each scenario is regarded as a sub-problem, and all sub-problems share a set of equipment parameters, which are obtained by planning the main problem; the typical normal scenario refers to selecting a representative week in each of the four seasons of spring, summer, autumn and winter, and the system works normally during this week; the solution module is used to generate stability constraints and add them to the main problem and sub-problems for iterative solution to obtain the planning results of each equipment that meets all scenarios.

Images