Abstract
In recent years, the optimal scheduling of multienergy has become the focus of the research. Against this background, this paper builds a model of a multienergy flow system of cooling, heating, and power, and advanced adiabatic compressed air energy storage (AA-CAES) is introduced to smooth wind power generation (WPG) and supply heating/cooling energy. Simulated annealing algorithm (SAA) is employed to energy-saving scheduling of the system with “exergy assessment” method. The energy-saving index and the exergy efficiency are compared in different cases. SAA is compared with the particle swarm optimization (PSO) algorithm in solving the optimal scheduling strategy. The cooling, heating, and power demands of an industrial park and WPG in typical days are employed to the simulation. The scheduling results of exergy input of SAA are far less than those of PSO in typical days of different seasons. The multienergy system without AA-CAES is also modeled, and energy-saving economic scheduling is carried out. The exergy efficiency of the system with AA-CAES is between 38% and 58% while the exergy efficiency of the system without AA-CAES is merely between 27% and 48%.
1. Introduction
Since the 21st century, the global economy, politics, and science and technology have developed continuously. With economic globalization and construction of the community of human destiny, potential crises have been exposed, including the shortage of fossil energy. The energy demand has increased sharply, and the contradiction between energy supply and demand is increasing day by day. Making full use of renewable energy and improving energy efficiency have become an inevitable requirement for the sustainable development of human society. The multienergy flow system and integrated energy system can effectively improve energy utilization efficiency. It can break the existing mode of separate design and independent operation of traditional energy systems, realize the overall planning and coordinated operation of different types of energy, and effectively realize the multienergy complementarity and cascade utilization of all kinds of energy [1].
The energy-saving dispatch objective function of traditional power system is generally the minimum primary energy consumption. In contrast, there are different kinds of energy in the multienergy flow system. Therefore, the energy scheduling of multienergy flow system should consider not only the saving in quantity but also the saving in quality [2].
Multienergy flow system is complex, and its efficiency can be characterized by energy or exergy, which are, respectively, energy analysis method and exergy analysis method. Energy analysis method and exergy analysis method are both thermodynamic analysis methods, but the former only analyzes the quantitative relationship of energy and energy efficiency based on the first law of thermodynamics, which can only reflect the relationship in quantity of system energy but cannot reflect the waste of energy quality caused by the working process of energy system. On the basis of exergy balance, the latter evaluates the thermodynamic perfection of the equipment involved in energy consumption in the system, so as to reveal the amount, location, and influencing factors of energy quality loss.
When energy is converted from one form to another, there are corresponding changes in both quality and quantity. The first law of thermodynamics describes the law of energy conversion and conservation. Therefore, when the “quantity” of energy is determined, in order to achieve the goal of energy saving, it is necessary to start with the “quality.” The second law of thermodynamics reveals that in the process of energy conversion, the “quality” attribute of energy decreases, and its quality must depreciate in the process of energy conversion and transmission. Exergy is the energy that can be continuously and completely converted into any other form of energy under given conditions, so it can also be called available energy or effective energy [3].
On the basis of the first and second laws of thermodynamics, the exergy analysis method organically combines the quantity and quality of energy and completes a profound analysis of the essence of energy and quality degradation in the process of energy conversion and transfer, thus has become one of the main thermal efficiency analysis methods of energy systems [4]. The exergy economic model of IES is established by combining exergy analysis while the solution process of the adaptive genetic algorithm is given, and then, an IES in Bali is employed to be simulated to verify the effectiveness and feasibility of the exergy analysis method [5]. The cost-exergy multiobjective optimization model of the system considering the economic and energy-saving indicators is established, and it is used the genetic algorithm to solve the model [6]. Energy efficiency, exergy efficiency, and carbon emission of the coupled cold thermoelectric are analyzed. Then, complementary performance of the vacuum tube solar collector and cold thermoelectric is also carried out which can be used for the multienergy flow system. Some research work focuses on IES planning problem [7]. A solar Rankine cycle integrated cascade refrigeration system is proposed to meet a hospital’s cooling, heating, and power demands. The system is analyzed from energy, exergy, and exergoeconomic point of view, and exergy destruction rates of the system components are investigated. Energy and exergy efficiency of the system are 89.39% and 8.70%, respectively [8]. The exergy efficiency is selected as the evaluation index, and the exergy loss of each link in the residential building CCHP system is analyzed to investigate the potential for using heat generated during compression stage of a compressed air energy storage system (CAES) [9]. CAES and combined with thermal storage (CAES-TS) are analyzed connected to a district heating network by using exergy and exergoeconomic analysis [10].
In this paper, the model of the multienergy flow system is firstly established and analyzed according to the second law of thermodynamics. AA-CAES is introduced as an energy storage device to absorb redundant wind power and supply heating energy in the multienergy supply system. The main contributions of this paper are as follows:(1)SSA is used to solve the optimal energy-saving scheduling method of the system with “exergy assessment” as the energy-saving index.(2)The traditional simulated annealing method has some limitations and can only solve nonlinear unconstrained problems. So, it needs to be improved. The penalty function is added to the algorithm to transform the nonlinear constrained problem into the minimum problem of augmented objective function.(3)The PSO algorithm is improved by introducing second-order oscillation and probabilistic suboptimal solution to improve the search accuracy and avoid falling into local optimum. The results of the two algorithms are compared and analyzed.
2. Energy-Saving Scheduling Model of Multienergy Flow System
2.1. Model of Multienergy Flow System
As shown in Figure 1, the input energy of the multienergy flow system in this paper is WPG and natural gas, and the demand side is cooling, heating, and power. In order to cope with the uncertainty of WPG, AA-CAES is employed in the system, which can not only be used as an energy storage system to balance the randomness of WPG [11] but also can be used as a heating storage or supply equipment.
2.1.1. Model of AA-CAES
The compression process of air in the CAES is close to adiabatic process, which generates a large amount of compression heating. The stages of compressors or expanders have a great influence on the efficiency of AA-CAES. The whole system consists of air compressor, air storage cavern, expander, and heat storage system (heat exchanger and heat storage system). In the expansion stage, heating is supplied coupled with electric power generation by releasing high-pressure air and compressed thermal energy. Compared with nonadiabatic compressed air, the comprehensive efficiency of AA-CAES is up to 70% [12].
In this paper, the AA-CAES power station model is established based on the following assumptions:(a)The air is ideal gas, and water is the heat carrier. The specific heat ratio of the air and the water is constant.(b)Since the scheduling interval time is 1 hour, there is heat exchange between the air storage cavern and the external environment, so it is assumed that the temperature in the air storage cavern is approximately equal to the ambient temperature.(c)It is assumed that the pressure ratio of compression/expansion is identical. The pressure loss of the air passing through the heat exchanger is overlooked.
(1) Compression. A multistage compression and an interstage cooling arrangement are adopted in the compressor of AA-CAES. The compression process of each stage is an adiabatic process [13]. At the compression stage, the compressed power can be calculated by
(2) Heat Exchange.. The compressed high-temperature and high-pressure air by the compressor exchanges heating with the heat carrying medium in the heat exchanger. The efficiency of the heat exchanger can be expressed as follows [14]:
For the notations in (2), a subscript “1” represents hot fluid, “2” represents cold fluid, “in” represents the fluid enters the exchanger, and “out” represents the fluid outflows from the exchanger.
(3) Heat Storage.. The heat generated during compression will be exchanged through the interstage heat exchanger and then stored into the heat storage device. The total amount of heat generated during the compression process, which is also to be stored in the heat storage device, is accumulated by the heat of each period as follows:
(4) Heat Release.. The temperature of the inlet air of the first-stage expander is also the temperature of the outlet air of the heat exchanger connected to the expander. During the energy release process, the air released from the storage cavern is heated when it passes through the heating exchanger. Therefore, the inlet temperature of the first-stage expander is related to the air temperature in the air storage cavern and the temperature of the heat carrier as shown follows:
(5) Expansion.. Multistage expansion and interstage reheating mode are adopted in our CAES expander model, and the expansion process can be viewed as a reversible adiabatic process. The air in each stage of expander is heated by heat exchanger, and the power generated by the CAES power plant can be described as follows [15]:
2.1.2. Model of CHP
For CHP unit, when generate heating and power by consuming natural gas, the energy conversion relationship between natural gas and power/heating is
For the notations in (6), a superscript “e” represents electric power, and “h” represents heating power.
2.1.3. Other Models
For gas furnaces, electric-driven refrigeration, and absorption chiller, the energy conversion relationship between them and electric power is
2.2. Objective Function
The purpose of energy-saving scheduling for multienergy flow systems is not only to save the quantity of energy but also, more importantly, to quality saving. Under the condition that there is no other heating source except the environment, when the system reversibly changes from any state to a balanced state at a given ambient temperature, the energy that can be converted to useful work to the greatest extent is called exergy. Exergy efficiency refers to the ratio of benefit exergy to cost exergy in the system or equipment. “Quantity” and “quality” in energy are not unified. Energy efficiency reflects the quantitative balance of energy with different qualities, and exergy efficiency reflects the “quantity” and “quality” of energy.
Exergy analysis, i.e., quality assessment, is carried out on various energy sources to achieve the maximum exergy efficiency of the system. Exergy is employed as the energy-saving-economic assessment index to describe the quality of energy supply.
As shown in Figure 1, the cooling, heating, and electric power are supplied by wind power and natural gas. The relationship between the demand side exergy of the system and the input exergy of the system at time t is
In this multienergy system, the electric energy is high-grade energy; that is, electric energy can be completely converted into other kinds of energy, and electric energy is equal to electric exergy. While cooling energy and heating energy are relatively low-grade energy, its exergy value is smaller than its energy value. Therefore, the exergy of the system output at time t is described as follows [16]:
The exergy of the input natural gas of the system can be described as
The wind energy at the input side of the system is a renewable and clean energy, so in this manuscript, the exergy of wind power at the energy input side is not considered, but the exergy of WPG is focuses on. AA-CAES is introduced to compensate the fluctuation of WPG. For WPG-AA-CAES as a whole, the exergy analysis can be expressed as
The exergy efficiency of the multienergy system, which is the ratio of the demand exergy to the input exergy, is
The objective function of the system in this paper is to minimize the input exergy which is described as
2.3. Constraints
2.3.1. Constraints of Energy Balance
The scheduling aim is energy-saving. First, the energy balance constraints of cooling, heating, and power must be satisfied. For absorption refrigeration unit, electric refrigeration unit, and load, the energy balance constraint of cooling energy is represented by
The heating is supplied by AA-CAES, CHP, and gas furnace. The thermal balance constraints between the suppliers and the demand are expressed as
For the whole system, the constraint of electric power balanced is described as
The first constraint is to ensure that the CAES system cannot work at the compression mode and the expansion mode simultaneously. is a binary variable indicating whether the CAES system is working at the expansion mode or not. When , the CAES is working at the expansion mode and is otherwise if . is binary variable stands for compression mode. . is the generation power of CHP at time t, is the WPG at time t, is the power consumption of electrical refrigeration unit at time t, and is power demand at time t.
2.3.2. Operation Constraints of the Equipment
For the multienergy flow system shown in Figure 1, the operation constraints of each unit are as follows.
CHP can provide electric power to the system, and its output limit of electric power and ramp up and down constraint are considered as
For gas furnace, electric refrigeration unit, and absorption chiller unit, when converted to electric power, their energy production limits at all times are
The heat exchanger constraint considering the input-output relationship and the capacity limit is expressed as
The compression/expansion power bounds of the AA-CAES, the capacity constraints of the air storage cavern, and the thermal storage device are listed as
3. Solution Methodology
In order to solve the aforementioned nonlinear mix integer optimal scheduling model to obtain the minimum exergy input of the system, simulated annealing algorithm (SAA) is employed to solve the problem. The simulated annealing algorithm originates from the study of statistical thermodynamic phenomena in the process of cooling objects heated to a certain high temperature. SSA can solve various nonlinear multiobjective programming problems very well. Constraints can’t be solved by the traditional SAA. Penalty function will be introduced to deal this problem.
3.1. Treatment of Constraints
SAA is a nonlinear unconstrained problem-solving method. The constraints in this paper are not only inequality constraints but also equality constraints. Therefore, it is necessary to transform the nonlinear constrained programming problem into a nonlinear unconstrained programming problem.
According to the penalty function method, the constrained optimization problem is transformed into an augmented objective function minimum problem. The objective function and constraints can be expressed as
(24) and (25) are converted to the minimum problem of augmented objective function which can be expressed aswhere is a monotonically increasing positive sequence, is the objective function without penalty term, is the penalty function, is the penalty factor which is related with the iteration time, and is the penalty term. If X does not meet the constraints, the penalty term , and it increases with the increase of the penalty function. When X satisfies the constraints, , indicating no penalty. Therefore, when is small enough, will be sufficiently close to the boundary of the constrained region, and can be considered as the minimum point that satisfies the constraints.
For the purpose of achieving the form of , equations (15), (16), (18)–(21), the inequality constraints of equations (22) and (23) are converted as equations (27) and (28):
Similarly, for the purpose of achieving the form of , equations (14), (15), and (17) are converted as follows:
3.2. Process of SAA
The key idea of the simulated annealing algorithm is to randomly select a solution as a start first and then generate a random disturbance. If a solution is found that is closer to the optimal solution than the previous solution, then this solution is accepted. If the found solution is deviate from the optimal solution, it does not matter, accept it with a certain probability.
The simulated annealing algorithm is shown in Figure 2.
The overall procedure of the simulated annealing algorithm can be described as follows: Step 1: Initialize the number of iterations. The temperature is T = T0 which should be large enough to make the acceptance probability of any initial solution close to 1. Step 2: Let the current solution X be the initial solution, and calculate the corresponding objective function f. Step 3: If the equilibrium condition is satisfied at this time, skip to step 6. Otherwise, go to steps 4. Step 4: Conduct random perturbation to generate an experimental solution Xnew as the neighbourhood solution of X, and fnew as its corresponding objective function value. Step 5: Execute the acceptance criterion: if fnew ≤ f, accept the experimental solution, make X = Xnew and go back to step 3. If , let X = Xnew, go back to step 3. Otherwise, go back to step 3 directly. Step 6: The iteration is terminated when the termination condition is met. Otherwise, decrease the temperature parameter T, and then go back to step 3. Termination: If and number of iterations is reached, return to the experimental solution. Then, stop the iteration.
3.3. Particle Swarm Optimization Algorithm
PSO is employed to compare the energy-saving optimization results with SAA. PSO is proposed more than 20 years ago [17]. In order to avoid wasting physical strength, birds transfer location information to each other in the process of looking for food to determine the food sources that can meet the needs of the whole birds. Finding the optimal food source is finding the optimal solution of the problem. Many similar intelligent algorithms such as ant colony algorithm, [18] cuckoo search algorithm, [19] fish swarm algorithm [20], and moth fire extinguishing algorithm [21] are inspired by the characteristics of biological populations.
The goal of PSO is to make all particles find the optimal solution in multidimensional hyper-volume. The group of swarms (called population) is randomly generated, and each particle (individual swarm) moves in N-dimensional search space with randomly generated velocity.
For each particle, proposed objective function is calculated by its position, and each particle’s position and velocity can be updated by the following equations:where is the flying speed of particle m in the n-th iteration, is the flight position vector of particle m in the n-th iteration, is the former optimal position of the particle, is the global optimal position, a1 and a2 are the acceleration coefficient used to modify the maximum learning step size, b1 and b2 are the random function (0∼1) used to enhance searching randomness, and h is the non-negative inertia weight used to modify the search range.
The flowchart of PSO is shown in Figure 3.
According to the characteristics of the position update formula, particle swarm optimization is more suitable for solving continuous optimization problems. If there is a higher requirement for the algorithm convergence speed, the acceleration constant can be increased, but the increase of the convergence speed may cause premature. Inertia weight can be increased to increase the “enthusiasm” of searching new positions to prevent the algorithm from appearing “premature” and falling into local optimum. But it will inevitably lead to a decrease of convergence speed. Random terms are employed to add perturbations to the speed updates to balance the convergence speed and the tendency of falling into local optimum.
3.4. Improved PSO
The traditional PSO algorithm is easy to fall into local optimum which is premature and needs to be improved. Disturbances are often added to the speed update. Second-order oscillation is employed in the speed update in this manuscript, and z1, z2, z3, and z4 are oscillation parameters. The improved speed update equation is
The convergence mode can be controlled by adjusting the four parameters. When the parameters satisfy , , , within a certain number of iterations, the algorithm oscillated converges. The global optimal search ability is increased, and the possibility of falling into local optimum is avoided. After a certain number of iterations, when the parameters satisfy , , , , the algorithm converges gradually. Both the local search ability and the search accuracy are increased.
In the iterative process, the fitness of the particle is calculated by the fitness function and compared with the previous best position. The previous best position of the particle m is updated. An acceptance probability is introduced to the second-order oscillation to avoid falling into local optimum. The probability is used to decide whether to update the best position when the current position of the particle is not better than the previous best position. Finally, the previous best position of particle m is compared with the global best position , and the global best position is updated. The best position of the particle in the past is the minimum exergy input of the system in our work. If the current fitness of the particle is greater than the fitness fq of the previous best position, the previous best position is accepted and updated with a probability of , where M represents the current number of iterations, and N represents the number of populations. The probability gradually decreases as the number of iterations M continues to increase. The algorithm tends to be stable.
3.5. Process of Improved PSO
The scheduling problem is solved by the second-order oscillatory PSO with the introduction of update probability. The entire optimization procedure is as follows which is also seen in Figure 4: Step 1: Initialize the particle swarm Step 2: Update particle velocity and position Step 3: Calculate the fitness of each particle of the particle swarm Step 4: If the current particle position fitness f is smaller than the previous best position fitness fq, update the particles and . If or , update the particles and , too. Otherwise, skip to step 2. Step 5: If the number of iterations is reached, output the result. Otherwise, skip to step 2.
4. Case Studies
4.1. Data of the Case Study
Taking an industrial park as an example of the multienergy flow system to illustrate the exergy-energy-saving technique proposed in this paper, the exergy coefficient of natural gas is 0.95 [5]. The low heat value is 50 MJ/kg [22]. The compression efficiency of AA-CAES is 80%, and the expansion efficiency is 85% [23] with the median exergy efficiency of 0.63 [24]. The temperature of the hot and cold media on the load side is 280.52 K and 312.89 K, respectively. The price of the natural gas is 0.065$/kWh.
The WPG forecast of different seasons is shown in Figures 5–7. The cooling, heating, and power demands of typical day in different seasons are shown in Tables 1–3.
4.2. Simulation Results and Analysis
4.2.1. The Results of SAA
The hourly cooling generation is drawn in Figures 8–10. AC and EC are cooling energy generated by absorption refrigeration and electric refrigeration, respectively. In winter, the cooling demand of the system is 0, and electric refrigeration and absorption refrigeration do not generate cooling energy. It can be seen from Figures 8 and 10 that the output of absorption refrigeration is lower than that of electric refrigeration. Because electric energy is high-grade energy, electric energy is equal to electric exergy.
Figures 11–13 are the hourly thermal energy dispatch for the typical days of the multienergy system in three seasons. GB means the heating energy generated by the gas furnace, CHP means the heating energy generated by the cogeneration device, and CAES means the heating energy output by the heat storage device of AA-CAES. As shown in 3 figures, the heating generation of gas furnaces and CHP is generally higher. AA-CAES could supply more heating energy when there is more compression heating stored previously.
Figures 14–16 are the hourly power dispatch of the typical day. CAES means the power generation of AA-CAES. CHP is the power generation of cogeneration. WIND is the power of wind power generation, and EC is the power consumption of electric cooling unit. In order to meet the power consumption of the load of the system, it can be seen that the wind power output is the major composition of the dispatching electricity. Electric energy is a high-grade energy, and wind power is renewable energy. WPG has the highest exergy efficiency. Compared with CHP power generation which requires purchase of natural gas, WPG does not consume primary energy. Regardless of equipment construction and maintenance costs, it is the most economical and energy-saving option to give priority to WPG. In summer and transition seasons (spring/autumn), CHP power generation is still an important part of the power supply resource assuming a higher proportion of power supply. This is because the power consumption of the system is relatively high in summer and transitional seasons. WPG cannot meet the total power demand, requiring CHP to supply more power. Power supply of AA-CAES is also less due to not so much wind power surplus.
Figure 17 is exergy input of system with AA-CAES of typical day in different seasons. The system maintains a high and stable exergy input from 7:00–21:00 in typical day of three seasons. During this time period, the demand for energy of users is high every day. From 00:00 a.m.–6:00 a.m., the exergy input remains at a low level, because users generally have relatively low energy demand during this period. From 5:00–8:00 is the period when the system exergy input increases the fastest. After 22:00, the system input begins to decrease rapidly. The results verify the rationality of the energy-saving and economic-dispatching strategy of the system. According to the hourly data of day ahead scheduling, the hourly efficiency of the multienergy flow system in this paper can be calculated to be between 38% and 58%.
Figures 18–20 are the daily cost of the multienergy system in different seasons. When wind power generation and AA-CAES are not enough to meet the energy demand, natural gas is purchased to compensate the energy insufficient. The cost of WPG is 0 without considering the fix cost and maintaining cost. Electric energy is a high-grade energy source. The exergy efficiency of using wind power is the highest. When the exergy efficiency of the system is the highest, the exergy input is the smallest. The energy saving and economy of the system in this paper are coincident which means that the most economical is the most energy-saving.
4.2.2. Optimization Result of PSO
The improved second-order oscillatory PSO is applied to the optimal scheduling problem as aforementioned. Day-ahead dispatch is performed on the time scale of 1 hour in the typical to optimize the exergy efficiency and exergy input. The exergy input and exergy efficiency of the two algorithms in typical summer day are compared and analyzed.
Figures 21 and 22 show the comparison of exergy input and exergy efficiency of the two methods. In this paper, the energy saving and economy are consistent, so only the exergy efficiency and exergy input of the two algorithms are compared. The exergy efficiency of improved PSO is higher than that of the SAA for only 3 hours; other times, the SAA exergy is more efficient. From the perspective of the upper and lower limits of exergy efficiency, the maximum exergy efficiency of the two algorithms is almost the same, but the lower limit of the exergy efficiency of the improved PSO is lower. The total exergy input of the SAA is 5.21 million kWh. The total exergy input of the improved PSO is 5.68 million kWh.
Tables 4 and 5 list the exergy efficiency and exergy input of the two algorithms in typical winter and spring/autumn day, where the (1) represents the improved SAA and (2) represents improved PSO. Obviously, it can be seen that the exergy efficiency of the improved PSO is lower than that of the improved SAA at most times. The exergy efficiency of the improved PSO is higher than that of the improved SAA at some times. It is difficult to measure the pros and cons of the two algorithms by comparing the exergy efficiency at each moment. It is necessary to compare the total exergy input of the system. The total exergy input of the system with the improved PSO is 4.27 million kWh in typical winter day. The total exergy input using the improved SAA is 4.14 million kWh. In the typical spring/autumn day, the total system exergy input using the improved PSO is 5.05 million kWh, and the total exergy input of the improved SAA is 4.83 million kWh.
It can be clearly seen that the SAA is not only more stable but also more energy saving and economic. The improvement SAA has a good performance in solving the optimal scheduling strategy of the multienergy flow system in this paper.
4.2.3. Result without AA-CAES
Day-ahead dispatch is performed on the time scale of 1 hour in the typical to optimize the exergy efficiency and exergy input. Because there are many demands for cooling, heating, and electric power in summer, only summer load data are selected for simulation comparison. The energy input and energy load of the new multienergy flow system without AA-CAES are the same as those of the multienergy flow system including AA-CAES. Energy saving and economic scheduling are carried out for the multienergy flow system without AA-CAES, and SSA in section III is used to solve the problem. The exergy input and exergy efficiency with and without CAES in typical summer day are compared and analyzed.
Table 6 lists the exergy input and cost of the two conditions in summer, where the (1) represents the system with AA-CAES and (2) represents the system without AA-CAES.
The multienergy flow system without AA-CAES cannot absorb wind power, which will lead to waste of wind power. At all the time, if the wind power cannot meet the energy consumption of the system in the next hour, it will need to purchase natural gas for load demand to be supplied, which will increase the exergy input and cost of the system.
It can be seen that the exergy input of the two systems is roughly the same at 6:00, because the wind power is sufficient and there is a certain amount of surplus, and the energy in the heat storage of AA-CAES is basically exhausted at 7:00, both of which can be considered that there is no energy storage equipment to supplement heat and power supply, and other energy needs to be input with little difference. However, the wind power output cannot meet the energy load on the demand side of the system at 9:00. The multienergy flow system without AA-CAES needs to purchase a certain amount of natural gas to meet the energy demand. The system with AA-CAES absorbs the surplus wind power during last hour and then supply for the demand at 9:00. Therefore, the exergy input of latter is naturally smaller than the multienergy flow system without AA-CAES. It can be obtained in the same way that the cost of the system with AA-CAES is far less than the other one.
Figure 23 shows the comparison of exergy efficiency under two conditions. It can be seen obviously that the total exergy input with AA-CAES is far lower than the total exergy input without AA-CAES from Table 6. Therefore, the exergy efficiency of the former is higher than the latter on the whole.
It can be clearly seen that, with AA-CAES, it has a good performance in solving the optimal scheduling strategy of the multienergy flow system in this paper.
5. Conclusion
In this paper, the concept of the second law of thermodynamics is introduced into the systems. The optimal scheduling of multienergy flow systems is carried out with the goal of the most energy saving based on exergy evaluation, and the objective function is established with the minimum exergy input of the system as the objective function. The AA-CAES is used as the energy storage system in the multienergy flow system. The improved PSO and SSA are used to solve the day-ahead scheduling of the system, and the exergy input and exergy efficiency of the results are compared. According to the comparison, SAA is better than PSO when employed to solve the optimal scheduling problem. The main conclusions are as follows:(1)Exergy is introduced into this paper and developed to optimize the energy-efficient scheduling method for multienergy flow systems. Electric energy is a high-grade energy resource; therefore, using wind power in the maximum extent can reach the highest exergy efficiency.(2)Through the energy-saving scheduling method proposed in this paper, the exergy saving characteristics is extremely obvious both in summer and winter. It can be concluded that the energy-saving scheduling method above of the multienergy flow system is feasible and correct.(3)When AA-CAES is used in the multienergy flow system, the energy-saving effect is better than not using it. The efficiency of multienergy flow system in this paper is between 38% and 58% at all times.(4)From the results above, the scheduling results of exergy input of SAA are, respectively, 5.21, 4.14, and 4.83 million kWh. However, the scheduling results of exergy input of PSO are, respectively, 5.68, 4.27, and 5.05 million kWh. Taking the former as the scheduling method reduces the exergy input of 0.13 to 0.47 million kWh. Compared with PSO, it can be clearly seen that the SAA is not only more stable but also more energy saving and economic. The improved SAA has a good performance in solving the optimal scheduling strategy of the multienergy flow system in this paper.
Nomenclature
| : | The specific heat capacity of hot air |
| /: | The coefficient of performance of absorption chiller/electric refrigeration |
| : | The exergy of various energy demands on the demand side at time t |
| : | The exergy of various energy delivered to the system (kWh) |
| /: | The exergy of natural WPG/gas consumed at time t |
| //: | The exergy of cooling/heating/power in the system load at time t |
| /: | The exergy of wind/natural gas on the demand side at time t |
| : | The outlet exergy of AA-CAES |
| : | The exergy delivered to the electric load by wind power |
| : | The exergy of wind power transferred to AA-CAES |
| : | The total exergy input of the system during the day |
| : | The total cost of the system for the whole day |
| : | The air consumption (kg/h) |
| : | The air consumption of natural gas furnace at time t |
| : | The quantity of purchasing natural gas |
| : | The low heat of natural gas |
| //: | The cooling/heating/electric power of system load at time t |
| m: | The quality of the fluid |
| m a: | The mass flow of air pressurized by the i-th compressor c |
| m h/mc: | The quality of the hot/cold fluid |
| : | The expansion power of the i-th stage at time t |
| /: | The upper/lower limit of generator set(kW) |
| /: | The power/heating generated by CHP unit at time t |
| : | The output of the gas furnace at time t |
| /: | The output of electric/absorption chiller unit (kWh) |
| /: | The power/heating consumption of electric/absorption refrigeration unit |
| : | The cooling consumption of cooling load |
| : | The power consumption of load side |
| /: | The heat power of CHP/CAES at time t |
| /: | The heat power consumption of absorption chiller/heat exchanger at time t |
| : | The total heating storage in AA-CAES |
| : | The generation power of AA-CAES |
| : | The generation power of CHP |
| : | WPG |
| : | The price of natural gas and its price units are converted changed from $/m3 to $/kWh |
| /: | The upper and /lower limits of the output of the gas furnace |
| /: | The upper/lower limits of the output of the electric refrigerator |
| /: | The upper/lower limits of the output of the absorption chiller |
| /: | The upper/lower limits of the output of the heat exchanger |
| /: | The upper/lower limits of the compression power |
| /: | The upper/lower limits of the power generation |
| /: | The upper/lower limits of air pressure in the air cavern |
| Q: | The total cost of the system for one hour |
| Q TES: | The heat energy absorbed by the N-th heat exchanger to store in the heat storage |
| Q c,i: | The heat absorbed by the i-th heat exchange |
| Q H,max: | The upper limit of heat storage capacity of heat accumulator |
| T: | The temperature of fluid |
| T c,in/Tc,out: | The inlet/outlet air temperature of compressor c |
| T c,i,in/Tc,i,out: | The inlet/outlet air temperature of i-th compressor c |
| T e,in,h/Te,out,h: | The inlet/outlet hot fluid temperature of the heat exchanger |
| T e,in,c/Te,out,c: | The inlet/outlet cold fluid temperature of the heat exchanger |
| T as/Twc: | The air temperature in the storage cavern and the maximum temperature of the heat carrier |
| T a: | The environment temperature (K) |
| /: | The temperature of cold/hot medium |
| ///: | The start-stop status of the unit/absorption chiller/gas furnace/electric refrigerator |
| /: | The binary variables for compression/ power generation conditions |
| /: | The upper/lower limits of the units’ climbing |
| : | The effectiveness of the heat exchanger |
| : | The specific air heat ratio |
| //: | The efficiency of the compressor/expansion process/gas furnace |
| /: | The power generation/heat production efficiency |
| /: | The rated compression ratio of the compressor c/expander |
| : | The specific exergy of natural gas |
| : | The exergy coefficient of AA-CAES |
| : | The exergy coefficient of natural gas |
| : | The exergy coefficient of system |
| : | The heat transfer efficiency of heat exchanger |
Abbreviations
| CAES: | Compressed air energy storage |
| AA:CAES: | Advanced adiabatic compressed air energy storage |
| CHP: | Combined heating and power |
| CCHP: | Combined cooling heating and power |
| DP: | Dynamic programming |
| IES: | Integrated energy system. |
Data Availability
The data supporting the current study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This work was supported by the National Natural Science Foundation of China under Grant no. U2166211 and Zhejiang Provincial Natural Science Foundation of China under Grant nos. LGG21E060001 and LY20E070002.