Optimal Operation of CCHP System Combined Electric Vehicles Considering Seasons

Abstract

Energy shortage has always been a problem that the world needs to face. The combined cooling, heating, and power (CCHP) system, as a multi-level energy utilization system that can provide cooling, heating, and electric energy simultaneously, is considered to have good development prospects in alleviating energy problems. In addition, because of the rapid development of electric vehicles (EVs), using EVs as power supply devices has become a hot topic of research. In this paper, EVs are combined with the CCHP system as new power supply equipment, and the influence of the season on the user’s cooling, heating, and power demand is considered. Aiming at the minimum economic cost, the system is optimized by using the PSO algorithm in two operating modes: following electricity load (FEL) and following thermal load (FTL). The final results show that the participation of EVs can reduce costs in both operating modes, especially in FTL mode, which can reduce costs by 4.58%, 13.61%, 12.74%, and 3.57% in spring, summer, autumn, and winter, respectively. In addition, the FEL mode is more economical in spring and winter, and the FTL mode is more economical in summer and winter. In addition, the $C{O}_2$ emissions in FEL mode are always less than in FTL mode.

1. Introduction

Due to the increase in human activities, global energy consumption is also growing, and the demand for energy in urban areas is much higher than in other areas. According to the International Energy Agency, globally, urban areas consume more than 70% of fossil energy and emit more than 60% of greenhouse gases [1,2]. It is impossible to stop the global urbanization process. Hence, the key to solving the problem is to determine how to increase energy efficiency and reduce greenhouse gas emissions. The combined cooling, heat, and power (CCHP) system is a multi-generational system that can provide cooling, thermal, and electrical energy at once [3,4,5]. With its significant benefits in terms of improved energy efficiency and lower emissions, the CCHP system is thought to be one of the most promising solutions in this area [6,7,8,9,10].

In addition, the enormous popularity of photovoltaic (PV) energy generation has caused a huge imbalance between power generation during the day and at night, which will undoubtedly increase the grid’s operational burden and have a negative impact on system stability and planning [11,12]. One of the best solutions to this issue is the electrical energy storage system ($EESS$) [13]. Even if the difference between on-peak and off-peak electricity price is only considered from an economic point, an $EESS$ is necessary [14], and then, to reduce the conflict between the different energy demands of users and the supply of the CCHP system, thermal energy storage systems ($TESS$) and cold energy storage systems ($CESS$) are also required. Hence, the combination of a CCHP system and energy storage system (ESS) is necessary to improve energy utilization efficiency, reduce pollution emissions, and improve grid stability.

Regarding the CCHP system, researchers have conducted a great deal of related research. Some researchers have combined different ESSs with CCHP and optimized and evaluated them. In [15], the combination of a CCHP system with a $TESS$, $EESS$, or hybrid energy storage system (HESS) was considered, respectively. The system performance of these different configurations was compared, and the suitable user boundaries were calculated. In [16], a HESS consisting of electric and thermal energy storage systems was proposed. Additionally, a rule-based energy management strategy was designed to realize the decoupling regulation of electric energy and thermal energy in a CCHP system. Razmi et al. [17] proposed a CCHP system using compressed air energy storage (CAES), and the organic Rankine cycle was used to improve the efficiency of the CAES system. In addition, the impact of the CASE vessel’s min and max pressure ratios on the CASE volume and round-trip efficiency was taken into account. Shan et al. [18] proposed a CCHP model with a HESS, using the updated Artificial Bee Colony algorithm based on the Beetle Antennae Search method to solve the problem of minimizing the daily generation dispatch cost and daily environmental pollutant control cost of the micro-grid. However, most CCHP systems only consider $EESS$ and $TESS$ and ignore CESS. In [19], based on the supply–demand energy matching principle, an improved design of an active $TESS$ is proposed. The results show that, in most situations, the higher the cooling/heating-to-electricity ratio, the longer the peak electricity load lasts, and the smaller the peak-to-valley ratio of power load, the better thw performance of the system will be. Abdalla et al. [20] focused on the two-stage optimization strategy of the microgrid system, including CCHP and HESS. The capacity and operation plan of each type of equipment are optimized.

In addition, a great deal of research work has also been conducted on the operational strategies and combinatorial design of CCHP systems. In [21], a large urban CCHP system including a gas turbine, absorption chiller, electrical chiller, electrochemical energy storage, thermal energy storage ice storage air conditioning systems, etc., was proposed, and an MINLP model was developed to determine the best combination of technologies to meet the energy demand of district buildings subject to practical constraints. Mao et al. [22] proposed a CCHP system hybridized with photovoltaic/thermal panels and TESS, using the particle swarm optimization (PSO) algorithm to calculate the optimal size of key components. Li et al. [23] proposed an improved method following the balanced thermoelectric load for a CCHP system with heat pumps. When compared to the original hybrid electric-heating load operation strategy, the improved operation strategy improves overall system performance by about 18%. In [24], a new electric load following strategy was proposed. Li et al. [25] proposed an interactive operation strategy based on the energy matching characteristics of the following electric load, following thermal load, and following hybrid electric–heating load operation methods. Melo et al. [26] proposed a multi-objective optimization strategy that integrates renewable and non-renewable energy devices, considering the combined environmental and economic benefits. These operation strategies can effectively improve the operating efficiency of the CCHP system and reduce primary energy consumption. In [27], a modified mayfly algorithm is proposed for the optimization of CCHP. The results show that the proposed optimization algorithm performed 8.15% and 10.06% better compared to the traditional mayfly algorithm and genetic algorithm, respectively. In [28], a conventional waste-driven combined heat and power cycle is combined with a large-scale absorption chiller ($AC$). Taking Danish energy market design and thermodynamic assessment as the case study for this work, the results showed that the thermal and electrical efficiencies of the proposed system are better than the conventional system by 12% and 1.3%, respectively. Lombardo et al. [29] proposed a novel solar driven CCHP system and studied this for three Italian locations, spread along the peninsula, with three different climates. The results show that the effectiveness of the system is greatly affected by solar radiation and weather condition. However, there is great potential for solar-powered micro-CCHP systems based on advanced technology for residential applications.

With the rapid increase in the number of EVs, researchers have conducted many studies on the charging and discharging of EVs. Huang et al. [30] proposed a smart microgrid operation plan combining EVs with photovoltaics and ESS in parallel. The calculation results show that the use of optimized solutions can reduce operating costs and increase overall revenue. In [31], an optimized charging and discharging method of EVs within a campus microgrid is proposed, which would significantly improve grid operation stability. It can be seen that researchers pay more attention to the impact of EV charging on grid load, and the economic benefits that EVs can bring to discharging the grid are not given enough attention. In addition, classical EVs are considered to be an electrical load, with charging being a one-way process from the grid to the EVs [32]. Vehicle to Grid (V2G) technology, as a technology that can transfer electricity from EVs to the grid, has received a great deal of attention from researchers [33]. Amamra et al. [34] present an optimited bidirectional V2G operation that can reduce EV charging costs. In [35], the uncertainty of wind power generation and the state of charge (SOC) of EVs are modeled as uncertainty prediction sets, and an effective strategy to improve the safety and economy of microgrid systems is proposed. Tepe et al. [36] proposed an optimization method to form an optimized combination of pools based on the power and energy capability profiles of commercial EVs. The results show that, in comparison to pools that were built randomly, an aggregator can enhance income per vehicle by up to seven times across markets by choosing smart pool composition. In [37], a practical energy management system for EV charging stations powered by renewable energy sources is proposed. The method is adaptable enough to prosumers, which are typically modest PV power plants for self-consumption, or to specifically designed PV-based EV charging stations. Gong et al. [38] proposed a smart charging strategy for EVs to mitigate the negative impact of fluctuations of renewable energy source outputs. Results show that the proposed charging strategy is effective in alleviating the output fluctuations of RESs. The charging cost of EVs can be reduced by 7.6% and 10.3%, respectively, in winter and summer. Because of the growing number of EVs, new solutions must be provided for basic charging stations. In [39], Vandewetering et al. provided a full mechanical and economic analysis of three novel PV canopy systems to provide a low-cost PV parking lot canopy to supply EV charging. The results show that each system has good economic benefits. Khoso et al. [40] analyzed different types of carport canopies and compared PV power generation at different standard tilt angles. The comparison of PV generation on mono-pitch, duo-pitch, and barrel-arch canopies is also presented.

Hence, in this paper, a large CCHP system including a TESS, EESS, and CESS combined with EVs is established.

2. Model of System

The model of the CCHP system is shown in Figure 1. The gas turbine ($GT$) can provide electricity by burning gas and emitting flue gas as a by-product. The thermal energy in flue gas can be partly recovered by the heat recovery system ($HRS$) to increase energy efficiency and decrease waste, and the AC can convert thermal energy into cooling energy for users. When the cooling power provided by the $AC$ is not enough to meet the demands of the users, the electric chiller ($EC$) can compensate for this. In the same way, the gas boiler ($GB$) can be used to provide a lack of thermal energy if the HRS cannot support enough heat. In addition, users can also buy electricity from the grid instead of just using the electricity provided by the $GT$. The system is equipped with ESS for all three types of energy, thermal, cooling, and power, which can improve the stability and economic efficiency of the system. All three devices operate independently of each other, and the time when they are charged or discharged needs to be determined by the economics of the whole system. Noticeably, in addition to GT and the grid, EVs can also provide some cheap electricity to the system.

2.1. Mathematical Model of Energy Balance

In this system, the $GT$, the $EV$, and the grid supply the electricity that users need. The $EC$ requires some electricity when it is operating. In addition, when the $EESS$ is charging or discharging, it will also get or provide some electricity, respectively. Hence, the electricity balance is shown in Equation (1).

$$E_{users}+E_{EC}+E_{EESS}^{cha}=E_{grid}+E_{GT}+E_{EESS}^{dis}+E_{EV}$$

(1)

where $E_{users}$ is the electricity demands of the users, $E_{EC}$ is the electricity used by $EC$, $E_{EESS}^{cha}$ is the electricity used to charge the $EESS$, $E_{grid}$ is the electricity purchased from the grid, $E_{GT}$ is the electricity generated by the $GT$, $E_{EESS}^{dis}$ is the electricity generated when the $EESS$ is discharged, and $E_{EV}$ is the electricity provided by an $EV$.

The $GB$ and the $HRS$ supply both the thermal energy required by the users and the thermal energy needed to operate the $AC$, and the cooling demands of users are supplied by the $EC$ and the $AC$. Considering the charging and discharging of the $TESS$ and $CESS$, the thermal balance and cold balance in the system are shown in Equations (2) and (3), respectively.

$$H_{users}+H_{AC}+H_{TESS}^{cha}=H_{GB}+H_{HRS}+H_{TESS}^{dis}$$

(2)

where $H_{users}$ is the heat demanded by the users, $H_{AC}$ is the heat required to operate the $AC$, $H_{TESS}^{cha}$ is the heat used for $TESS$ charging, $H_{GB}$ is the heat supplied by $GB$, $H_{HRS}$ is the heat recovered by $HRS$, and $H_{TESS}^{dis}$ is the heat supplied by $TESS$ when it is discharged.

$$C_{users}+C_{CESS}^{cha}=C_{EC}+C_{AC}+C_{CESS}^{dis}$$

(3)

where $C_{users}$ is the cooling energy demanded by users, $C_{CESS}^{cha}$ is the cooling energy used for $CESS$ charging, $C_{EC}$ is the cooling energy supplied by $EC$, and $C_{CESS}^{dis}$ is the cooling energy supplied by $CESS$ when it is discharged.

2.2. Mathematical Model of Components

The $GT$ generates electricity by burning gas, which emits high-temperature flue gas as the by-product. Hence, the mathematical model of $GT$ is shown in Equations (4) and (5).

$$E_{GT}=G_{GT}\times {\eta }_{GT}$$

(4)

where $G_{GT}$ is the energy provided by gas burning in the $GT$, and ${\eta }_{GT}$ is the generating efficiency of the turbine.

$$H_{GT}=G_{GT}\times (1-{\eta }_{GT})$$

(5)

where $H_{GT}$ is the heat of the flue gas emitted by the $GT$.

The $GB$ provides thermal energy only by burning gas. Hence, the mathematical model of $GB$ is shown in Equation (6).

$$H_{GB}=G_{GB}\times {\eta }_{GB}$$

(6)

where $G_{GB}$ is the energy provided by gas burning in the $GB$, and ${\eta }_{GB}$ is the heating efficiency of the $GB$.

The $HRS$ can recover and reuse part of the thermal energy in the flue gas emitted by the turbine. Hence, the mathematical model of the $HRS$ is shown in Equation (7).

$$H_{HRS}=H_{GT}\times {\eta }_{HRS}$$

(7)

where ${\eta }_{HRS}$ is the recovery efficiency of the $HRS$.

The $AC$ can convert thermal energy into cooling energy for users. Hence, the mathematical model of $AC$ is shown in Equation (8).

$$C_{AC}=H_{AC}\times CO{P}_{AC}$$

(8)

where $H_{AC}$ is the thermal energy consumed by the $AC$, and $CO{P}_{AC}$ is the coefficient of performance ($COP$) of the $AC$.

The $EC$ can convert electricity into cooling energy for users. Hence, the mathematical model of the $EC$ is shown in Equation (9).

$$C_{EC}=E_{EC}\times CO{P}_{EC}$$

(9)

where $E_{EC}$ is the electricity consumed by the $EC$, and $CO{P}_{EC}$ is the $COP$ of the $EC$.

The $EESS$, $TESS$, and $CESS$ can store electric energy, thermal energy, and cooling energy, respectively, but they cannot be charged and discharged at the same time. Hence, the mathematical models of the $EESS$, $TESS$, and $CESS$ are shown in (10)–(12).

$$\begin{array}{c}\hfill \left\{\begin{array}{c}{E}_{EESS}^t=E_{EESS}^{t-1}\times {\eta }_{EESS}+E_{EESS}^{cha}-E_{EESS}^{dis}\hfill \\ E_{EESS}^{cha}\times E_{EESS}^{dis}=0\hfill \end{array}\right.\end{array}$$

(10)

where $E_{EESS}^t$ is the SOC of the $EESS$ at the current moment, $E_{EESS}^{t-1}$ is the SOC of the $EESS$ at the previous moment, and ${\eta }_{EESS}$ is the storage efficiency of the $EESS$.

$$\begin{array}{c}\hfill \left\{\begin{array}{c}{H}_{TESS}^t=H_{TESS}^{t-1}\times {\eta }_{TESS}+H_{TESS}^{cha}-H_{TESS}^{dis}\hfill \\ H_{TESS}^{cha}\times H_{TESS}^{dis}=0\hfill \end{array}\right.\end{array}$$

(11)

where $H_{TESS}^t$ is the SOC of the $TESS$ at the current moment, $H_{TESS}^{t-1}$ is the SOC of the $TESS$ at the previous moment, and ${\eta }_{TESS}$ is the storage efficiency of the $TESS$.

$$\begin{array}{c}\hfill \left\{\begin{array}{c}{C}_{CESS}^t=C_{CESS}^{t-1}\times {\eta }_{CESS}+C_{TESS}^{cha}-C_{CESS}^{dis}\hfill \\ C_{CESS}^{cha}\times C_{CESS}^{dis}=0\hfill \end{array}\right.\end{array}$$

(12)

where $C_{CESS}^t$ is the SOC of the $CESS$ at the current moment, $C_{CESS}^{t-1}$ is the SOC of the $CESS$ at the previous moment, and ${\eta }_{CESS}$ is the storage efficiency of the $CESS$.

These devices have power constraints in actual operation, and hence they should satisfy the following constraints shown in (13).

$$\begin{array}{c}\hfill \left\{\begin{array}{c}{E}_{G{T}_{min}}\le E_{GT}\le E_{G{T}_{max}}\hfill \\ H_{G{B}_{min}}\le H_{GB}\le H_{G{B}_{max}}\hfill \\ H_{HR{S}_{min}}\le H_{HRS}\le H_{HR{S}_{max}}\hfill \\ C_{A{C}_{min}}\le C_{AC}\le C_{A{C}_{max}}\hfill \\ C_{E{C}_{min}}\le C_{EC}\le C_{E{C}_{max}}\hfill \\ E_{EES{S}_{min}}\le E_{EESS}\le E_{EES{S}_{max}}\hfill \\ H_{TES{S}_{min}}\le H_{TESS}\le H_{TES{S}_{max}}\hfill \\ C_{CES{S}_{min}}\le C_{CESS}\le C_{CES{S}_{max}}\hfill \end{array}\right.\end{array}$$

(13)

2.3. EV

In order to get more accurate information about the power available to the EVs at each period, the specific behavior of the EVs needs to be simulated. This includes the arrival and departure times of the EVs and the initial SOC of the EVs at the time of arrival, which is usually simulated using the truncated Gaussian distribution [41]. These are shown in Figure 2, Figure 3 and Figure 4, which refer to [42,43]. In order to make the simulation of EV behavior more realistic, some constraints need to be placed on it, which are shown in (14), referring to [42]

$$\begin{array}{c}\hfill \left\{\begin{array}{c}{T}_{arr}=f_{TG}(x;{\mu }_{arr},{\sigma }_{arr}^2,(T_{arr}^{min},T_{arr}^{max}))\hfill \\ T_{dep}=f_{TG}(x;{\mu }_{dep},{\sigma }_{dep}^2,(max(T_{dep}^{min},T_{arr}),T_{dep}^{max}))\hfill \\ SO{C}_{ini}=f_{TG}(x;{\mu }_{ini},{\sigma }_{ini}^2,(SO{C}_{ini}^{min},SO{C}_{ini}^{max}))\hfill \end{array}\right.\end{array}$$

(14)

That is, the departure time must be greater than the arrival time, and at least 20% of the SOC needs to be left when the EV leaves.

In addition, EVs must arrive at the parking lot and connect to the charging pile, and then they can supply energy to the main grid; however, the parking spaces are limited, so the fullness of the parking lot must be considered, which is shown in (15).

$$\begin{array}{c}\hfill \left\{\begin{array}{c}{N}_t=N_{t-1}+N_t^{arr}-N_t^{dep}\hfill \\ N_t\le N_{max}\hfill \end{array}\right.\end{array}$$

(15)

where $N_t$ is the number of EVs parked in the parking lot at time t, $N_{t-1}$ is the number of EVs parked in the parking lot at time $t-1$, $N_t^{arr}$ is the number of EVs arriving at the parking lot at time t, and $N_t^{dep}$ is the number of EVs leaving the parking lot at time t. $N_{max}$ is the maximum capacity of the parking lot; the number of EVs staying in the parking lot cannot be greater than $N_{max}$.

The battery parameters of EVs are also critical factors affecting the available power of EVs.

Table 1. Some parameters of EV battery packs.

Manufacturer	Chevrolet	EnerDel	Tesla	Nissan	Lishen
Model	Volt Generation 1	Hybrid Bus	Model S85	Leaf Generation 1	EV-LFP
Build year(s)	2015	2017	2013 & 2014	2012 & 2015	2012
Average discharge voltage (V)	360	630	346	368	335
Nominal capacity (Ah)	45.0	31.5	222.0	65.0	115.5
Calculated energy (kWh)	16.2	19.5	76.7	23.9	38.7

shows the battery parameters of the five different manufacturers used in this research [44]. These five batteries were applied to different EVs at a 1:1:1:1:1 ratio in the simulation.

2.4. Objective Function and Cost

This system takes the minimization of the economic cost as the optimization goal, as shown in Equation (16).

$$min{S}_T=S_{inv}+S_{CCHP}+S_{C{O}_2}$$

(16)

where $S_T$ is the total cost, $S_{inv}$ is the cost of investment, $S_{CCHP}$ is the operating cost of the system, and $S_{C{O}_2}$ is the environmental cost of emitting greenhouse gases such as $C{O}_2$.

These three costs are specifically composed of the following parts shown in Equations (17)–(19).

$$S_{inv}=S_{GT}+S_{GB}+S_{HRS}+S_{AC}+S_{EC}+S_{EESS}+S_{TESS}+S_{CESS}$$

(17)

This is the investment cost of the $GT$, $GB$, $HRS$, $AC$, $EC$, $EESS$, $TESS$, and $CESS$. The cost of EVs is not considered because owners purchase them at their own expense.

$$S_{CCHP}=(G_{GT}+G_{GB})\times P_{gas}+E_{grid}\times P_{grid}+E_{EVs}\times P_{EVs}$$

(18)

where $P_{gas}$ is the unit price to purchase gas, $P_{grid}$ is the unit price to buy electricity from the grid, and $P_{EVs}$ is the unit price to purchase electricity from EVs.

$$S_{C{O}_2}=\{(G_{GT}+G_{GB})\times {\mu }_{gas}+(E_{grid}+E_{EVs})\times {\mu }_{grid}\}\times P_{C{O}_2}$$

(19)

where ${\mu }_{gas}$ is the emission factor for burning gas—because the grid generates the electricity purchased from EVs, their $C{O}_2$ emission factors are the same, which is ${\mu }_{grid}$, and $P_{C{O}_2}$ is the price of emitting a unit volume of greenhouse gas.

3. Optimization Method

This research used the particle swarm optimization (PSO) algorithm to solve the optimization problem. The PSO algorithm has the advantages of rapid convergence, remarkable efficiency, and random global search [45]. The mathematical formula for PSO is shown in Equations (20) and (21).

$$V_{id}^{k+1}=\omega V_{id}^k+c_1{r}_1(P_{id}^k-X_{id}^k)+c_2{r}_2(G_{id}^k-X_{id}^k)$$

(20)

$$X_{id}^{k+1}=X_{id}^k+V_{id}^{k+1}$$

(21)

Matlab was used to model and optimize this research, and the entire optimization process is shown in Figure 5. The cooling, heating, electrical loads, and other parameters were input into the system, and then we set the necessary constraints, used the PSO algorithm to optimize the solution, and obtained optimized economic consumption, greenhouse gas emissions, and other results.

4. Case Study

In this research, a commercial area in China was used as a case study to analyze the economic benefits of this system [2]. Since the energy demand of users varies in hours, 8760 h per year would result in calculations that are too complex and difficult to solve. In order to make the calculation simpler, in this research, we planned to replace the energy demands of users with a typical day. In addition, considering that the energy demands of users vary according to the seasons, four typical days in spring, summer, autumn, and winter were designed to optimize the CCHP system, asshown in Figure 6, and the

Table 2. The technical parameters of equipment for the system.

Parameter	Symbol	Value	Unit	Reference
Efficiency of GT	${\eta }_{GT}$	36	%	[46]
Efficiency of GB	${\eta }_{GB}$	85	%	[46]
Efficiency of HRS	${\eta }_{HRS}$	80	%	[46]
COP of AC	$CO{P}_{AC}$	70	%	[46]
COP of EC	$CO{P}_{EC}$	400	%	[46]
Efficiency of EESS	${\eta }_{EESS}$	95.6	%	/
Efficiency of TESS	${\eta }_{TESS}$	95.6	%	[47]
Efficiency of CESS	${\eta }_{CESS}$	95.6	%	/

shows some of the technical parameters used in this case study.

Table 3. The economic parameters of equipment for the system.

Equipment	Value	Unit	Reference
GT	1350	$/kW	[48]
GB	93	$/kW	[46]
HRS	120	$/kW	[46]
AC	180	$/kW	[46]
EC	145	$/kW	[46]
EESS	520	$/kW	[6]
TESS	371	$/kW	[6]
CESS	371	$/kW	[6]

shows the economic equipment parameters for the system.

Table 4. The range of operating variables for the equipment.

Parameter	Symbol	Value	Unit
GT capacity	$G_{GT}$	[0, 3000]	kWh
GB capacity	$G_{GB}$	[0, 2000]	kWh
HRS capacity	$H_{GT}$	[0, 2000]	kWh
AC capacity	$H_{AC}$	[0, 1000]	kWh
EC capacity	$E_{EC}$	[0, 550]	kWh
EESS capacity	$E_{EESS}$	[0, 1000]	kWh
TESS capacity	$H_{TESS}$	[0, 1000]	kWh
CESS capacity	$C_{CESS}$	[0, 1000]	kWh

shows the range of operating variables for the equipment.

Table 5. Some parameters of gas, electricity, and $C{O}_2$.

Parameter	Symbol	Value	Unit	Reference
Unit price of electricity	$P_{grid}$	0.04 (0:00–8:00)	$/kWh	[49]
		0.08 (8:00–24:00)	$/kWh	[49]
Unit price of gas	$P_{gas}$	0.03	$/kWh	[50]
$C{O}_2$ emission factor of gas	${\mu }_{gas}$	0.22	kg/kWh	[50]
$C{O}_2$ emission factor of grid	${\mu }_{grid}$	1.2	kg/kWh	[46]
Unit price of $C{O}_2$ emission	$P_{C{O}_2}$	0.0075	$/kg	[50]

shows some parameters about gas and electricity and $C{O}_2$ emissions.

In this commercial area, there are about 2500 parking spaces for EVs. Figure 7 shows how much electricity EVs can supply to the system per hour.

5. Optimization Results

5.1. Operation Mode

In this case, since the capacity of each piece of equipment has been determined, the optimization focuses on optimizing the operation of the equipment. In this study, two modes of operation, followinEDg electrical load (FEL) and following thermal load (FTL), were used to test the system. In FEL mode, the $GT$ first meets the demand of the electric load until it reaches the maximum operating power, and the missing electricity is supplemented by the grid and EVs; the missing thermal is provided by the $GB$. In FTL mode, $GT$ first meets the demand of the thermal load until it reaches the maximum operating power, and the missing electricity and thermal load are supplemented by the grid, EVs, and $GB$, respectively. In addition, in each operation mode, this study also divides the system into two cases, with EV participation and without EV participation, and optimizes and compares them.

5.2. System in FEL Mode

Figure 8, Figure 9, Figure 10 and Figure 11, respectively, show the equipment operation of the system in FEL mode in different seasons. The part below the zero axis of the ESS represents charging, and the part above represents discharging energy.

Table 6. Cost of operating the system in FEL mode.

Season	With or Without EVs	Cost of Gas ($)	Cost of Electricity ($)	Cost of Emitting ${\mathit{CO}}_2$ ($)	Total Cost ($)
Spring	Without EVs	2122	210	140	2472
	With EVs	2150	240	161	2551
Summer	Without EVs	1982	1135	243	3360
	With EVs	2073	813	265	3151
Autumn	Without EVs	1968	922	217	3107
	With EVs	2006	673	234	2913
Winter	Without EVs	2097	341	154	2592
	With EVs	2195	234	162	2591

shows the cost of operating the system in FEL mode. Through these figures and the data in tables, it can be seen that in spring, the total cost of operating with EVs is a little higher than operating without EVs. This is because when operating with EVs, the EESS, TESS, and CESS store more energy and do not discharge. However, in summer and autumn, the economics of operating with EVs are greater than the economics of operating without EVs, and the total cost is reduced by 6.22% and 6.24%, respectively. Although the cost of gas and $C{O}_2$ emissions has increased slightly, the advantage in the cost of electricity has fully made up for these. In addition, in winter, the cost of operating with or without EVs is not much different. The reduction in the cost of electricity is basically the same as the increase in the cost of gas and $C{O}_2$ emissions. However, regardless of the season, the participation of EVs leads to a small increase in $C{O}_2$ emissions, at 15%, 9.1%, 7.8%, and 5.2% in spring, summer, autumn, and winter, respectively.

5.3. System in FTL Mode

Figure 12, Figure 13, Figure 14 and Figure 15 show the equipment operation of the system in FTL mode in different seasons. The part below the zero axis of the ESS represents charging, and the part above represents discharging energy.

Table 7. Cost of operating the system in FEL mode.

Season	With or Without EVs	Cost of Gas ($)	Cost of Electricity ($)	Cost of Emitting ${\mathit{CO}}_2$ ($)	Total Cost ($)
Spring	Without EVs	1865	703	182	2750
	With EVs	2087	360	177	2624
Summer	Without EVs	1310	1633	289	3232
	With EVs	1198	1296	298	2792
Autumn	Without EVs	1224	1485	266	2975
	With EVs	1052	1254	290	2596
Winter	Without EVs	1822	793	189	2804
	With EVs	2034	487	183	2704

shows the cost of operating the system in FTL mode. These figures and the data in the table show that the total cost of operating with EVs is always less than that of operating without EVs, regardless of the season; it is reduced by 4.58%, 13.61%, 12.74%, and 3.57% in spring, summer, autumn, and winter, respectively. However, in spring and winter, the total cost of the system operating in FTL mode is more expensive than that of the system operating in FEL mode. In spring, it is increased by 11.25% when no EVs participated in the operation and by 2.86% when the EVs participated. In winter, it increased by 8.18% and 4.36% when operating without and with EVs, respectively. However, in summer and autumn, the economic benefit of the FTL mode is much higher than that of the FEL mode. In summer, when the system operates without EVs, the total cost is reduced by 3.81% and by 11.39% when the system operates with EVs. In autumn, when the system operates without EVs, the total cost is reduced by 4.25% and by 10.88% when the system operates with EVs. This is because when the heat load is relatively low, the FTL mode can significantly reduce the operating power of the $GT$ to reduce gas consumption, and the participation of EVs in operation can reduce the cost of electricity purchase. However, in terms of $C{O}_2$ emissions, the FTL mode produces more than the FEL mode in any case.

6. Conclusions

This study established a CCHP system including a $GT$, $GB$, $HRS$, $EC$, $AC$, $EESS$, $TESS$, $CESS$, and EVs. Aiming at the minimum economic cost, the PSO algorithm is used for optimization, and two operating modes, the FEL mode and FTL mode, are used to test the economy of the system. The final results show that irrespective of the mode of operation, the economic benefits of the process with the participation of EVs are higher than those without the involvement of EVs, especially in the FTL mode. However, on the premise that EVs participate in the operation, it should be noted that the economic benefits of the operating mode are greatly affected by the season. The FEL operation mode should be adopted in spring and winter because the FTL mode costs 2.86% and 4.36% more than the FEL mode in these two seasons, respectively, and in summer and autumn, the FTL mode is definitely better than the FEL mode. The economic cost being reduced by 11.39% and 10.88% cannot be ignored. Choosing the appropriate operating mode according to the season can obviously lead to better economic benefits. In addition, no matter the situation, the $C{O}_2$ emission cost of the FEL mode is always less than that of the FTL mode.