Optimization Model of an Efficient Collaborative Power Dispatching System for Carbon Emissions Trading in China

Abstract

In this paper, a collaborative power dispatching system (CPDS) was developed to maximize the profit of a regional biomass power system consisting of an independent power grid. A power generating, dispatching and carbon emissions trading system (CETS) could be engaged in joint strategic planning and operational execution. The principal of CPDS is interactive planning of generating units in power generation and carbon emissions trading. An efficient carbon emissions trading plan for a CPDS would lead to optimized power generation levels under available power production capacities and carbon emissions. In a case study, four generator policies are proposed by considering basic CETSs to comparatively analyze the function of each generator in the CPDS. Results of four scenarios are compared, showing that biomass energy could replace thermal units to a certain extent, the carbon emissions and coal consumption of the CPDS would lie at a lower level, and a pumped storage unit could adjust the load fluctuations. The results of a carbon trading analysis show that the CETS has no significant impact on the CPDS, but along with the increase in trading price or the decrease in the free quota, the economic interests of power plants will be reduced accordingly. This may lead to carrying out low-carbon projects and reducing carbon emissions. Therefore, it is imperative to reduce carbon emissions by replacing power units with high energy consumption, and improve the consumption capacity of renewable energy.

1. Introduction

The main purpose of developing renewable energy in China is to reduce the proportion of coal consumption and carbon emissions. However, coal remains the largest contributor to China’s energy structure. The total coal consumption of China in 2016 reached 3.9 billion tons, of which about 50% was consumed in the power industry. The combustion of fossil fuels produces much carbon emissions, which could lead to a series of environmental problems. Global warming has become the world’s most urgent issue that needs to be solved. The impact on the air has also become a factor in power generation planning [1]. Thus, it is desirable to build an environmental-friendly electricity dispatching system with low carbon emissions, and explore a scientific method which can increase the proportion of renewable energy power generation. In order to utilize energy resources sustainably and reduce carbon emissions, the proportion of coal energy consumption should be decreased, and at the same time, the share of renewable energy needs to be greatly increased [2]. According to the National Energy Development Plan’s design—which was proposed by the Chinese government—by 2020, the proportion of domestic non-fossil energy and natural gas consumption will increase to 15% and 10%, respectively, while the proportion of coal consumption is expected to fall to 58%. Carbon emissions have a negative effect on the environment and human health. An effective CETS could encourage the greatest use of low-carbon energy generation and the resulting decrease in carbon emissions per unit power generation.

The CETS has been carried out at seven pilot cities in China, and the national carbon trading market will be officially launched in 2017. According to the plan, the first batch of industries in the national carbon trading market will include five traditional manufacturing industries (steel, nonferrous metals, cement, chemical and electricity) and the transportation industry [3,4]. This policy has been validated by foreign countries, and the influence of similar policies on reducing carbon emissions has been analyzed [5,6]. On the specific implementation of the trading system, there are many researchers focusing on the different quota allocation mechanisms, discussing the setting of carbon price and the free quota. At the same time, the government will also introduce some detailed measures [7,8,9]. CETS is important for power generation dispatching; it may affect the sequence and efficiency of power generation, and subsequently have an impact on the strategic choice of power plants.

Recently, the majority of studies have focused on renewable power generation dispatching. Li et al. proposed an optimization scheduling model for wind power and thermal power with an energy storage system [10]. Bekele et al. studied the feasibility of a small renewable hybrid power system for off-grid rural electrification in Ethiopia [11]. Wang et al. designed a hydro/photovoltaic hybrid power system in a microgrid [12]. In order to better solve the problem of renewable power consumption, Campana et al. built a dynamic model of a photovoltaic pumping system, and proposed a new approach to optimize this system [13,14]. Ma et al. studied the feasibility of a stand-alone hybrid solar–wind–battery system [15]. Ma et al. also proposed a techno-economic optimization model of a pumped storage and photovoltaic power system [16]. Ou et al. studied the dynamic operation and control of hybrid power systems [17,18,19]. As for the effect of a renewable power generation system, Olsson et al. considered the environment effect of photovoltaic water pumping [20]. Beluco et al. evaluated the effects of a hybrid hydro–PV system, and indicated that most dispatching systems use a renewable resource backed up by a fossil fuel-based generator [21]. Zhang et al. studied the use of thermal power to assist wind power in maximizing the realization of wind power consumption [22]. There are also some studies that address the strategy of a microgrid-scheduling optimization system [23,24,25].

Biomass resourcing also has the effect of carbon reduction, and biomass power units could operate continuously, unlike PV and wind power. The utilization rate of biomass energy in China is low; most of the rural areas in China usually deal with biomass energy by direct combustion, and it not only causes environmental pollution, but also wastes the economic value of the biomass resources. The development of biomass power generation has some positive significance in increasing farmers’ income and reducing air pollution [26,27,28]. According to the Thirteenth Five Year Plan for Electric Power Development promulgated by Chinese government departments, it is encouraged to carry out combined coal-fired and biomass energy power-generation projects. Most of the existing multi-energy scheduling studies focus on wind and photovoltaic power [10,11,12,13,14,15,16]. This paper incorporates biomass power generation into a CPDS to study the importance of such units, as biomass power generation is more stable than that derived from wind or photovoltaic sources. In the traditional CPDS, the common practice is modeled on minimized cost, which does not take the CETS cost into consideration [29,30]. Therefore, it is desired that the CPDS should be optimized through the analysis of an operation plan of generators, where CETS cost is indispensable.

Generally, the CPDS has attracted the interest of researchers, but these studies tend to focus on small scale, off-grid power systems in remote areas. In these cases, conventional units assist renewable units to connect the grid in a passive state, ignoring the economic efficiency of conventional units. Moreover, biomass power units are seldom involved in CPDS, and the traditional CPDS model rarely considers national carbon trading systems—which will be implemented in China—systematically. Developing a multi-objective-optimized CPDS model, considering economic benefit and environmental cost, is desired.

Hence, the objective of this study is to develop an optimized model of a CPDS which consists of biomass, thermal and pumped storage units under a carbon emissions trading system in China. Taking the output of units and carbon emissions trading as decision variables, the proposed model is applied to an independent power grid to verify its validity. Different generator policies are used to explore the function of various types of units in the CPDS. It is useful to emphasize the importance of a carbon emissions trading system. The results of the CPDS under four carbon trading scenarios are calculated to figure out the policy effect, which is helpful to reasonably develop power generation planning. Finally, we discuss whether the optimization model proposed can achieve a stable operation of the power generation system, and at the same time cut down coal consumption and carbon emissions.

2. Development of the Collaborative Power Dispatching System

2.1. Complexity of the System

A power system involves a large number of processes, which can be categorized into power generation, transmission and distribution. The power generation system is the focus of this study. The principle of a CPDS is the overall planning; its complexity is reflected in the diversification of generators, and each power generation process is closely linked. Furthermore, the proposed CPDS combines CETS to analyze the impact of carbon trading on power generation. The power industry is one of the main sources of carbon emissions, and it is necessary to analyze the impact of the domestic power industry under the implementation of China’s carbon trading system. According to the trading system, each generator has a free quota. If a generator’s quota is not sufficient to meet its need, it can purchase another generator’s residual quota on the carbon emissions trading market.

In traditional power generation, thermal power—with its reliable operation, stable output, low cost and other merits—occupies a large share of power generation. At present, China’s renewable energy power generation installed capacity is increasing steadily, and the vast majority of the regional power grid dispatching system is composed of thermal power and renewable energy power generation. Wind and photovoltaic power are two of the fastest-growing, clean and renewable sources of power in China, especially in the past five years. Photovoltaic power generation converts solar radiation into electricity directly, without fuel consumption, it does not cause environmental pollution; likewise, wind power does not produce carbon emissions. However, both wind and photovoltaic power use natural resources as energy, which will inevitably be subject to natural conditions and result in the randomness and interruption of power output [31]. These situations will doubtlessly bring challenges to a power generation dispatching system. In order to reduce the impact of natural factors on the consumption of renewable energy, biomass power is added to the proposed CPDS. Furthermore, the coexistence of electricity generation and the consumption of the pumped storage units also increases the complexity of the system.

The present work proposes the optimization of a power generation dispatching system (Figure 1). Three kinds of existing electricity facilities (coal-fired thermal power, biomass power and pumped storage power facilities with different capacities) are available to meet the load demand in a competitive way. Thermal power operates stably with low operational costs, but produces a large amount of carbon emissions. Biomass power has a better carbon reduction effect, but its installed capacity is limited by the fuel supply. Pumped storage units are conducive to a smooth load curve, but its operation cost is too high, and the existing price system may not guarantee its economic benefits. In the case of fixed load demand, the purpose of dispatching is to achieve the optimized rational allocation of these three resources.

2.2. Model Development

To improve the stability of power generation and decrease carbon emissions, an optimization model is built in this paper, in which biomass power and pumped storage power are used in combination with thermal power. Because the operating cost of generating renewable energy units is higher than that of thermal power, the dispatching model with the minimum total cost may use less renewable energy than thermal power. Therefore, the model is not conducive to increasing the proportion of renewable energy. Considering that electric tariffs may stimulate the increase in the proportion of renewable energy generation, the objective function should include economic benefits and environmental costs.

2.2.1. Objective Function

The objective function is maximizing the profit of thermal, biomass and pumped storage power units in the proposed CPDS:

$$\max Z=Z_{th}+Z_{bio}+Z_{hp}$$

(1)

The profit of thermal power:

$$Z_{th}=p_{th}{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{i=1}^N{Q}_{i,t}-C_{fuel}-C_{c{o}_2}}}$$

(2)

The fuel cost also includes the startup and shutdown cost of thermal power units, as follows:

$$C_{th-fuel}={\displaystyle \sum _{t=1}^T{\displaystyle \sum _{i=1}^N[p_{coal}{u}_{i,t}{f}_i(Q_{i,t})+}}{u}_{i,t}(1-u_{i,t-1})S{U}_i+u_{i,t-1}(1-u_{i,t})S{D}_i],\mathrm{and}$$

(3)

$$f_i(Q_{i,t})=a_{0,i}+a_{1,i}{Q}_{i,t}+a_{2,i}{Q}_{i,t}^2,$$

(4)

where in $p_{coal}$ is the price of standard coal; $f_i(Q_{i,t})$ is function of standard coal consumption of thermal power unit with real-time output $Q_{i,t}$; $u_{i,t}$ is a 0–1 variable, which stands for operation or stop status of thermal power unit i at time t, when $u_{i,t}$ = 0, it means this unit stop and doesn’t have coal consumption, when $u_{i,t}$ = 1, it means this unit start-up. And $a_{0,i}$, $a_{1,i}$, $a_{2,i}$ are parameters of the coal consumption function, and all are exogenous variables, depending on the characteristics of the unit. $S{U}_i$ is start-up cost of thermal power unit i at time t, if and only if $u_{i,t}$ = 1 and $u_{i,t-1}$ = 0, this formula has got to work, $u_{i,t}(1-u_{i,t-1})S{U}_i$ doesn’t equal 0; similarly $S{D}_i$ is shutdown cost of thermal power unit i at time t, if and only if $u_{i,t}$ = 0 and $u_{i,t-1}$ = 1, this formula has got to work, $u_{i,t-1}(1-u_{i,t})S{D}_i$ doesn’t equal 0.

The carbon emissions cost is equal to the carbon emissions trading price multiplied by the purchase quota of carbon emissions, and the purchase quota of carbon emissions is equal to the total carbon emissions of thermal units minus the free quota as follows:

$$C_{c{o}_2}=p_{c{o}_2}[{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{i=1}^N{E}_i(Q_{i,t})}}-E_0],\mathrm{and}$$

(5)

$$E_i(Q_{i,t})=c_{0,i}+c_{1,i}{Q}_{i,t}+c_{2,i}{Q}_{i,t}^2,$$

(6)

where $E(Q_{i,t})$ is carbon emissions function of thermal power unit; $E_0$ is free quota of carbon emissions; $c_{0,i}$, $c_{1,i}$, $c_{2,i}$ are parameters of the carbon emissions function, and all are exogenous variables, depending on the characteristics of the unit.

The Profit of Biomass Power:

$$Z_{bio}=p_b{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^M{P}_{k,t}-C_{b-fuel}+W_{c{o}_2}}}.$$

(7)

The biomass power fuel cost $C_{b-fuel}$ is equal to the fuel price multiplied by the fuel consumption, as follows:

$$C_{b-fuel}=p_{bio}{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^M{f}_k(P_{k,t})}},\mathrm{and}$$

(8)

$$f_k(P_{i,k})=b_{1,k}{P}_{k,t},$$

(9)

where $b_{1,k}$ is the given exogenous parameter of the coal consumption function and depends on the characteristics of the unit. Strictly speaking, this function should be a quadratic function (just like function of the standard coal consumption of thermal power unit $f_i(Q_{i,t})$, but the data obtained in the survey is insufficient.

$W_{c{o}_2}$ is an additional benefit for biomass power-generation enterprises due to the carbon emissions trading system. It is equal to the carbon emissions trading price multiplied by the reduction of carbon emissions. We assume that the entire reduction in the carbon emissions of biomass power units could be traded, and this benefit can be calculated by:

$$W_{c{o}_2}=p_{c{o}_2}{E}_{b-c{o}_2},\mathrm{and}$$

(10)

$$E_{b-c{o}_2}=b_{2,k}{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^M{P}_{k,t}}},$$

(11)

where $b_{2,k}$ is the given exogenous parameter of this function. Actually, there are many ways to calculate reduction of carbon emissions [32,33], and the emphasis is on the different benchmarks. In order to facilitate computation, we simplify it into a linear function according to the survey data in this paper.

The profit of pumped storage power:

$$Z_{hp}=p_{cap}{K}_t{V}_{cap}+p_h{\displaystyle \sum _{t=1}^T{H}_{h,t}-p_p{\displaystyle \sum _{t=1}^T{H}_{p,t}}}.$$

(12)

The pumped storage unit adjusts its status according to load change. While pumping, the unit uses electricity, and therefore the carbon reduction effect is not significant. Although there are still some controversies about the carbon reduction of this unit in the academic field [34], we chose to ignore the reduction effect in this paper. The output of the pumped storage unit can be calculated by:

$$H_{h,t}=m_1{q}_{ht}^2+m_2{q}_{ht},\mathrm{and}$$

(13)

$$H_{p,t}=m_3{q}_{pt}^2+m_4{q}_{pt}.$$

(14)

where $q_{ht}$ is water flow of power station generate electricity at time t; $q_{pt}$ is water flow of power station use electricity at time t. $m_1,m_2,m_3,m_4$ are exogenous variables, depending on the characteristics of the unit.

2.2.2. Constraints

The constraints are mainly divided into two aspects. The first aspect is the load demand side, and the second is the characteristics of the power units as they relate to thermal, biomass and pumped storage power.

Load Demand Side Constraints

(1) Equilibrium constraint of power system

$${\displaystyle \sum _{i=1}^N{Q}_{i,t}+{\displaystyle \sum _{k=1}^M{P}_{k,t}}+H_{h,t}-H_{p,t}=G_t}.$$

(15)

This formula implies that power supply and demand at time t must always be in balance, and the transmission line losses are not taken into account.

(2) Spinning reserve constraints

Due to the uncertainty of load demand, power generation needs to be adjusted in real time to meet the power balance. Thermal power units usually work as backup units to adjust the output by adding and reducing fuel with flexibility. The adjustments should meet some constraints, which depend on the characteristics of the thermal power units:

$$\begin{array}{l}{R}_{ut}\le {\displaystyle \sum _{i=1}^N{u}_{i,t}(Q_{i,\max }-Q_{i,t})}\\ R_{dt}\le {\displaystyle \sum _{i=1}^N{u}_{i,t}(Q_{i,t}-Q_{i,\min })}\end{array}$$

(16)

Power Units Constraints

(1) Output constraint of thermal power unit

The installed capacity and minimum power output of the thermal power unit determines the upper and lower limits as follows:

$$Q_i{}_{,\min }\le Q_{i,t}\le Q_{i,\max }.$$

(17)

(2) Output climbing speed constraint of thermal power unit

The output change of thermal power in a unit period has been limited by the unit’s characteristics. This kind of real-time output change should meet:

$$\left|Q_{i,t+1}-Q_{i,t}\right|\le Q_{i,v}.$$

(18)

(3) The thermal unit startup and shutdown time constraints

The status change of units will influence the power generation cost, and it is detrimental to the performance of the units to startup and shutdown frequently. Accordingly, the continuous startup and shutdown time should be constrained as follows:

$$\begin{array}{l}0\le (T_{i,t-1}^{on}-M_i^{on})(u_{i,t-1}-u_{i,t})\\ 0\le (T_{i,t-1}^{off}-M_i^{off})(u_{i,t}-u_{i,t-1})\end{array}.$$

(19)

(4) Output constraint of biomass power unit

The same as thermal power; biomass power units should also have output limited as follows:

$$P_{k,}{}_{\min }\le P_{k,t}\le P_{k,\max }.$$

(20)

(5) Output climbing speed constraint of biomass power units

In general, the biomass power unit’s operation is stable, but also needs to meet the output climbing speed constraint as follows:

$$\left|P_{k,t+1}-P_{k,t}\right|\le P_{k,v}.$$

(21)

(6) Output constraint of pumped storage power unit:

$$\begin{array}{l}0\le H_{h,t}\le H_{h,\max }\\ 0\le H_{p,t}\le H_{p,\max }\end{array}.$$

(22)

(7) Pumping and generating flow constraints

The real-time output of the pumped storage unit depends on the water flow. Therefore, this unit should meet water flow constraints as follows:

$$\begin{array}{l}0\le q_{h,t}\le q_{h,\max }\\ 0\le q_{p,t}\le q_{p,\max }\end{array}.$$

(23)

(8) The dynamic balance constraints of storage capacity

The pumped storage power unit adjusts its output by charging or discharging water, and therefore it needs to ensure that storage capacity is sufficient to keep dynamic balance all the time, as is shown in the following equation:

$$E_{t+1}-E_t+\Delta t(q_{h,t}-q_{p,t})=0.$$

(24)

(9) Storage capacity constraint

Storage capacity is limited by its design capacity as follows:

$$E_{\min }\le E_t\le E_{\max }.$$

(25)

(10) Pumping and generating power-complementary constraints

Pumped storage units cannot pump and generate electricity at the same time. So they should meet:

$$H_{h,t}\times H_{p,t}=0.$$

(26)

3. Model Application

3.1. Case Study

This CPDS comprises of six thermal units, three biomass units and a pumped storage power plant with 300 MW installed capacity (

Table 1. The generator set of dispatching system (MW).

Type	Thermal Power		Biomass Power	Pumped Storage
Installed capacity (MW)	600	400	30	300
	300	150	30
	100	60	30
Electric tariffs ($/MW·h)	55.10 ($p_{th}$)		108.75 ($p_b$)	55.10($p_h$)	105,850 $/MW ($p_{cap}$)
Fuel cost ($/t)	123.25 ($C_{fuel}$)		43.50 ($C_{b-fuel}$)	41.33 $/MW·h ($p_p$)

). In order to compare the results of different units in the dispatching system, the main types of thermal power units are included. To ensure biomass power could operate stably, there is a sustained supply chain of biomass fuel. Due to the long response time of pumped storage units, this study simplifies the dispatching process and takes one hour as a dispatch period. It is assumed that the entire CPDS could operate independently. The flow diagram of the optimization model can be seen in Figure 2, and the corresponding power generation dispatching model can be formulated as follows:

$$\begin{array}{l}MaxZ=p_{th}{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{i=1}^N\begin{array}{l}{Q}_{i,t}-{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{i=1}^N[p_{coal}{u}_{i,t}{f}_i(Q_{i,t})+}}{u}_{i,t}(1-u_{i,t-1})S{U}_i+u_{i,t-1}(1-u_{i,t})S{D}_i]\\ -p_{c{o}_2}[{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{i=1}^N{E}_i(Q_{i,t})}}-E_0]\end{array}}}\\ +p_b{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^M{P}_{k,t}-p_{bio}{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^M{f}_k(P_{k,t})}}+p_{c{o}_2}{E}_{b-c{o}_2}}}\\ +p_{cap}{K}_t{V}_{cap}+p_h{\displaystyle \sum _{t=1}^T{H}_{h,t}-p_p{\displaystyle \sum _{t=1}^T{H}_{p,t}}}\end{array}$$

(27)

This is subject to the equilibrium constraint of the power system:

$${\displaystyle \sum _{i=1}^N{Q}_{i,t}+{\displaystyle \sum _{k=1}^M{P}_{k,t}}+H_{h,t}-H_{p,t}=G_t},i=1,2,\cdots ,6;t=1,2,\cdots ,24;k=1,2,3,$$

(28)

and other power unit constraints like Equations (16)–(26).

3.2. Data Collection

Data were collected by field survey and data mining research [35,36]. The relevant price data for units are shown in Table 1. The load data in the power system of a typical day are shown in

Table 2. A typical day’s system load data (MW).

Period	Load	Period	Load	Period	Load
1	775	9	1375	17	1175
2	850	10	1450	18	1300
3	925	11	1525	19	1425
4	1000	12	1475	20	1500
5	1075	13	1400	21	1400
6	1150	14	1350	22	1250
7	1225	15	1275	23	1100
8	1275	16	1225	24	900

, and all detailed power units’ operating parameters are shown in

Table 3. Parameters of thermal power and biomass power units.

Units	${\mathit{a}}_{0,\mathit{i}}$	${\mathit{a}}_{1,\mathit{i}}$	${\mathit{a}}_{2,\mathit{i}}$ × 105	${\mathit{c}}_{0,\mathit{i}}$	${\mathit{c}}_{1,\mathit{i}}$	${\mathit{c}}_{2,\mathit{i}}$ × 105	${\mathit{b}}_{1,\mathit{k}}$	${\mathit{b}}_{2,\mathit{k}}$
600 MW	11.71	0.274	0.644	29.04	0.680	1.60	None
400 MW	8.48	0.297	1.84	21.54	0.754	4.67
300 MW	6.17	0.308	3.66	15.8	0.788	9.37
150 MW	3.54	0.332	4.15	9.35	0.876	10.96
100 MW	1.43	0.337	9.01	3.82	0.900	24.06
60 MW	0.67	0.345	11.02	0.88	0.9084	27.13
1#	None						1.1	0.776
2#							1.2	0.776
3#							1.3	0.776

. It is worth noting that China’s carbon emissions trading market is not well developed, and the carbon trading volumes and prices of pilot cities are low. In order to better achieve the national emissions reduction targets, it is assumed the ratio of free quota is 90% and trading price is about $14.5/t, which is a feasible basic carbon emissions trading system that the early market may accept [37].

4. Results Analysis

In this paper, the CPDS is optimized for biomass, thermal and pumped storage power under carbon emissions trading in China. The framework of the results analysis is listed in

Table 4. Factors analyzed in scenario analysis.

Types	Generator Policy	Carbon Trading Conditions
1	Basic power units	Basic, Conservative, Positive, Aggressive
2	Thermal units only	Basic trading system
3	Thermal + biomass	Basic trading system
4	Thermal + pumped storage	Basic trading system

. Seven scenarios, considering different generator policies and carbon trading conditions, are analyzed.

4.1. Generator Policy Analysis under Basic Trading System

To study the function of each unit in the proposed CPDS, scenario analysis is used to discuss the results of different generator policies under a basic carbon trading system.

4.1.1. Dispatching Basic Power Units

The output results of the original power units under a basic carbon emissions trading system are analyzed. Figure 3 shows the summary power generation and carbon emissions flow of the entire system in one dispatching period, and the detailed results are shown in Appendix A. The generating proportion of renewable energy reached 15%, which meets the requirements of China’s power planning. Thermal and biomass power generation provided 2496 MW·h for meeting the pumping needs of the pumped storage unit, which confirms that this unit will contribute to power consumption. Moreover, the presence of biomass units helped to reduce the average carbon emissions of the system. Figure 4 shows the detailed generating proportion of different units. From this figure, three of the largest-capacity installed thermal power units accounted for 77% of total power output, and their installed capacity accounted for only 65% of the CPDS. The remaining three thermal power units accounted for 15.5% of installed capacity, but only accounted for 8% of the total generation, most of which accounted for by the three largest-capacity installed thermal and biomass power units. This shows that the ratio of power generation to installed capacity does not have a simple linear relationship, and with the characteristics of the unit, the proposed optimization model includes the influence of the carbon trading cost on the output sequence of the power units.

The operating status of the power units could be described by output curves. From Figure 5, the 600 MW thermal unit ran stably in full-load status, as did the biomass units; but the units with outputs of 400 MW and 300 MW had large fluctuations. In the case of the 400 MW thermal unit, this may be caused by load change; and the fluctuations in the 300 MW thermal unit may be caused by the spinning reserve constraint. The output of the 400 MW unit changed more frequently, and its fluctuation was mainly due to lack of load demand. The 300 MW unit is usually used as a reserve capacity unit, and in order to meet the spinning reserve constraint, the unit can only work at peak load. Thus, during the rest of the optimization period, it worked in a low-load state. The pumped storage unit also fluctuated. As for the other small capacity thermal units, they could run stably but just in a low-load state so as to meet load demand. Even though they had output, their power generation was not enough to ensure their profits.

4.1.2. Dispatching Thermal Units Only

Figure 6 shows the thermal power units’ dispatching results when only the thermal power units are operated in the power generation system. The inner ring is the thermal power output structure of the thermal units only, and the outer ring is the result of the basic system. Based on this comparison, it is obvious that in the case of thermal power units only, with the exception of the 400 MW and 300 MW units, the generating output of the other units was smaller than that of the basic system, which indicated that the share of the generating units of the two units was taken up by the other units in the basic system.

Compared to thermal-only power system, the basic system was friendlier to the environment. From

Table 5. The thermal power comparative results in both situations.

System	Thermal Power Units Output (MW·h)	Carbon Trading Volume (t)	Coal Consumption (t)
Thermal only	29,400	23,811	9449
Basic	27,084	20,413	8737

, the total power generation was 29,400 MW·h in the case of thermal power units only, which is equal to the total load demand of the day. However, in the basic system, the total power generation of the thermal units decreased by 7.88% and the coal consumption reduced by 7.54%. Furthermore, due to the carbon-reduction effect of biomass power, the reduction rate of the carbon trading volume was 14.27% when compared with the system of thermal power only, which is more beneficial for power producers and the environment.

In consideration of carbon emissions costs, large-capacity units with their own advanced technology have smaller emissions parameters, so they were given priority for power generation. In the case of thermal units only, the sequence of power generation was sorted by their capacity. Due to the limits of the small load demand, some small-capacity thermal power units almost had no output, and unit operation or stop status was completely determined by the load data. In the basic system, the thermal power units could operate stably on account of the role of the pumped storage units.

4.1.3. Dispatching Thermal and Biomass Units

In order to study the role of the pumped storage units in the power generation dispatching system, the following element was taken into account: the thermal units’ and biomass units’ dispatching system.

The comparative results of power output in different situations are listed in

Table 6. The power output comparative results in both situations.

System	Thermal Power (MW·h)	Biomass Power (MW·h)	Pumped Storage Power (MW·h)	Total Profit ($)
Thermal and biomass	27,240	2160	Not considered	555,975.53
Basic	27,084	2160	2652	659,278.75

. It was apparent that the thermal units’ output in the basic system was less than the other units’ output, which also indicated that they had lower carbon emissions and coal consumption. Due to the added renewable energy sources, the share of thermal power declined from 92.65% to 84.91% by calculation.

The increase in carbon emissions was not only reflected in the increase in thermal power output, but was also caused by frequent changes in the thermal power units’ start–stop statuses and the resulting increased energy consumption. The thermal units’ start–stop statuses in both situations are shown in Figure 7. In the thermal units’ and biomass units’ dispatching system, all thermal power units were in operation except for the 150 MW unit. As for the frequency of start–stop conversion, the 600 and 400 MW units both had zero instances, the 300 MW unit had two, the 100 MW had three and the 60 MW unit had six instances. We can make the conclusion that the smaller the installed capacity, the higher the frequency of start–stop conversion in the dispatching process. For the basic dispatching system, Figure 7b shows that all units had startup and shutdown conversions due to the presence of the pumped storage unit, which fully illustrates that the pumped storage unit played an important role in the stable operation of the power generation system. However, what cannot be ignored is that the small-capacity thermal power units were running in low-load conditions. The reason for this was that the carbon emissions parameters and the energy consumption parameters of these generators were too large, while the marginal efficiency of power generation was low. Therefore, a higher priority power generation sequence can be obtained.

The pumped storage units worked well to adjust the load demand and help the power generation system operate more stably. Figure 8 shows the real-time output of the pumped storage unit in the basic dispatching system. We learned that when the load demand is little, the pumped storage unit will start to pump water, and when the load rises, it will release the water to generate electricity. The conversion of the output of the pumped storage unit is related to the change of the load. During the optimization period, there were four changes in the state of the unit, corresponding to the overall trend of the load curve.

4.1.4. Dispatching Thermal and Pumped Storage Units

In order to study the role of the biomass units in the power generation dispatching system, the dispatching results in the absence of biomass units were analyzed.

The output structure of the thermal power units in different systems is listed in

Table 7. The comparative results of power output structure of thermal power units (MW·h).

Dispatching System	600 MW	400 MW	300 MW	150 MW	100 MW	60 MW	Total Output
The absence of biomass units	14,142	8696	3753	1376	792	480	29,239
Basic system	14,052	7441	3111	1208	792	480	27,084

. It was evident that the most of thermal units’ output in the basic dispatching system was less than the absence of the biomass units dispatching system. This part of the output was the share of biomass power units occupied. Less thermal power means less coal consumption and a lower carbon emissions level. According to calculations, in the thermal and pumped storage unit dispatching system, carbon emissions per power generation were 746 kg/MW·h, but in the original system were 640 kg/MW·h. This strongly indicated that the biomass power generation had a good effect on carbon reduction.

The output of the 400 and 300 MW thermal units had the largest increase due to the absence of biomass units, followed by the 600 and 150 MW units, while the two other small-capacity units remained virtually unchanged. The installed capacity of the biomass units determined their occupied share of the thermal power units (because of the high installed capacity of the biomass units in this paper, which had a great impact on the output of the thermal power units). Moreover, the 100 and 60 MW thermal power units always operated at low load because the overall load demand and power generation capacity did not match, which is not conducive to the healthy operation of the power generation system. Based on the above analysis, in consideration of the high energy consumption, high carbon emissions and low-load operability of small-capacity thermal power units, they should be eliminated or converted into biomass power units when the local area has abundant biomass resources.

4.2. Carbon Trading System Analysis

In this paper, the carbon emissions trading system was a necessary factor. To study the impact of different CETSs on the results of the optimization model, four CETS scenarios were set (

Table 8. Different scenario settings.

Scenarios	Conservative	Neutral	Positive	Aggressive
Free quota (%)	97	90	90	80
Carbon price ($/t)	7.25	14.50	29.00	14.50

).

A business-as-usual (BAU) scenario was set; namely, a basic carbon emissions trading system (neutral scenario). In the Conservative scenario, the free quota was higher than in the BAU scenario, and the carbon emissions trading price was less than BAU. The carbon emissions trading price in the Positive scenario was twice that of BAU, but the free quota did not change. Finally, in the Aggressive scenario, the free quota of carbon emissions was 80%, which may significantly improve the carbon emissions trading cost.

Results under different scenarios were analyzed. The detailed data are listed in Table 8 and Figure 8. According to

Table 9. Dispatching system optimized results under different scenarios.

Scenarios	Carbon Emissions (kg/MW·h)	Coal Consumption (kg/MW·h)	Output (MW·h)	Profit ($)
BAU	640	274	31,896	659,278.75
Conservative	644	276	31,645	672,477.23
Positive	640	274	31,896	651,589.835
Aggressive	640	274	31,896	627,284.645

, prices rising, or the free quota of carbon emissions decreasing, corresponded to a decrease in the total profits of all power units. In the Conservative scenario, carbon emissions per power generation were 644 kg/MW·h, and coal consumption per power generation was 276 kg/MW·h. These two values are both higher than in the BAU scenario, and prove that a more stringent carbon emissions trading system could effectively decrease carbon emissions. However, in the Positive and Aggressive scenarios, it was found that the results were exactly the same as that of BAU scenario.

The detailed power output of the specific units could explain these results. From Figure 9, compared with the BAU scenario, the total output in the Conservative scenario was smaller, as was the output of the thermal units. However, it was obvious that the power generation structure of the units did not show an obvious change in the Positive and Aggressive scenarios. This shows that these two scenarios have only a slight effect on the thermal power output. In fact, this bears a relation to the model constraints. In order to meet the equilibrium constraints, the spinning reserve constraints and the power unit constraints of the power system—with regard to total profit—the differences were not significant between the Positive and Aggressive scenarios, even when the carbon emissions trading system was more stringent, i.e., when the carbon price was higher or the carbon quota was lower.

Limited by the model constraints, the thermal power output structure did not change considerably, making the carbon reduction effect not obvious. However, to achieve the purpose of maximizing benefits, the decision maker will tend to reduce the proportion of thermal power and arrange more renewable energy units, thereby reducing coal consumption and cutting down carbon emissions.

4.3. Discussion of Results

In terms of the results of different generator policies under a basic CETS, the function of three types of generators in the proposed CPDS was studied by scenario analysis. The results showed that the CPDS was efficient, and in this system, each generator cooperated with each other to ensure the stability of renewable energy.

In the case study, the sequence of power generation had thermal units after biomass units. This may have caused the proportion power generation attributed to thermal units to be further reduced, while increasing the installed capacity of the biomass generator. However, the total installed capacity of the biomass generator was limited by its fuel supply chain. The pumped storage unit played an active role in adjusting the load demand, generated electricity at the peak of the load and used the residual electricity of other units at the time of low load to pump the water, which smoothed the load fluctuations and increased the proportion of energy generated by renewable means.

With respect to the impact of the CETS on the CPDS, this analysis showed that the impact was not significant. This is probably due to the complexity of the dispatching system and the operational constraints of the generators. The four CETSs set up in the scenario analysis comprised the assumptions made in this study, and their impact can be further studied according to the specific implementation plan of China’s carbon trading market.

5. Conclusions

This study aimed to explore an optimization model of a CPDS for biomass, thermal and pumped storage power, considering a carbon emissions trading system in China. Biomass and thermal power units were coupled to achieve a stable power output, combined with pumped storage power to adjust the load demand. Taking the carbon emissions trading systems as the decision variables, maximizing power generation profit was set as the objective, after deducting fuel cost and carbon emissions cost. As a result, the validity of the proposed model was demonstrated through a simplified power system. The results proved that the proposed CPDS can work healthily and operate stably. As opposed to the randomness and intermittency of wind and photovoltaic power outputs, biomass power units could work at full capacity. The results of the generator policy analysis showed that the thermal power generation share was reduced, which led to the reduction of coal consumption and a lowering of carbon emissions in the optimized dispatching system. This indicated that the proposed combination of units was effective in lowering carbon emissions. Pumped storage units could play a good role in adjusting load demand, relieving the pressure of thermal power units on output adjustment and alleviating the damage to the power generation system caused by load fluctuation. It was proved that the proposed CPDS is reasonable and effective through comparative analysis, and the collaborative dispatching of these three kinds of generators has been realized. A carbon trading system analysis was used to study power output under different CETSs, and the results showed that the CETS had a great influence on power generation profit, but had no significant effect on power generation structure. The more stringent CETSs may not directly affect the level of carbon emissions and coal consumption in power generation, but from the perspective of economic benefits, power plants change their business strategies due to falling profits. Therefore, CETSs indirectly affect the behavior of power plants, and tend to reduce power generation, coal consumption and carbon emissions. After the implementation of the national carbon emissions trading market, those small-capacity thermal power units with high energy consumption and high carbon emissions should be eliminated from the market, or transformed into biomass units. The proposed model can be extended to other power generation systems. Future work will focus more on long-term collaborative optimization, considering the electricity, thermal and carbon markets collaboratively, development projects and the optimization of the situation of cogeneration.