Source-Storage-Load Flexible Scheduling Strategy Considering Characteristics Complementary of Hydrogen Storage System and Flexible Carbon Capture System

Abstract

In the current literature, there exists a lack of analysis regarding the coordination of the spinning reserve and time-shift characteristics of hydrogen storage systems (HSS) and flexible carbon capture systems (FCCS) in terms of low-carbon economic operation. They are presently used solely as a tool to capture carbon dioxide, without fully utilizing the advantages of their flexible operation. The coordination and complementarity of the FCCS and HSS can ensure stable power supply and improve renewable energy (RE) consumption. Combined with demand side response (DSR), these factors can maximize the RE consumption capacity, reduce carbon emissions, and improve revenue. In this paper, a source-storage-load flexible scheduling strategy is proposed by considering the complementary nature of FCCS and HSS in terms of rotating standby and time-shift characteristics. First, the operational mechanisms of FCCS, HSS, and demand side response (DSR) are analyzed, and their mathematical models are constructed to improve flexibility in grid operation and regulation. Next, deficiencies in FCCS and HSS operation under rotating reserve requirements are analyzed to design a coordinated operation framework for the FCCS and HSS. This operational framework aims to enable the complementarity of the rotating reserve and time-shift characteristics of FCCS and HSS. Finally, based on the carbon emission trading mechanism, a three-stage ladder carbon emission trading cost model is constructed, and a source-storage-load flexible scheduling strategy is established to achieve an effective balance between low carbon emissions and economic performance. The simulation results demonstrate that the strategy reduces the overall cost by 8.57%, reduces the carbon emissions by 35.33%, and improves the renewable energy consumption by 3.5% compared with the unoptimized scheme.

1. Introduction

1.1. Research Background

With the introduction of the “dual-carbon” target and the construction of new power systems, China has put forward new requirements for carbon emissions and flexibility in the power industry [1]. Currently, China’s power generation infrastructure is still dominated by coal power generation, which accounts for 55.3% of China’s energy consumption [2,3]. There is an urgent need to vigorously develop renewable energy sources in order to facilitate the increasingly important energy transition to renewable power systems. However, the randomness and volatility of renewable energy (RE) output leads to difficulties in RE consumption and high pressure on spinning reserves. Therefore, it is important to fully coordinate the dispatchable resources of the source, load, and storage to achieve balanced RE consumption for a low-carbon and flexible operation [4].

The replacement of traditional high-carbon emission power generation equipment with flexible carbon capture systems (FCCS) is key to realizing a low-carbon transition. FCCS can significantly reduce greenhouse gas emissions by directly capturing and storing carbon dioxide emitted from coal-fired power plants. Additionally, as carbon regulations become more stringent, companies may have to pay high carbon emission fees. Carbon capture technology is used to reduce the cost of carbon emissions. When RE sources are not generating sufficient power, coal-fired power generation and CO₂ capture via FCCS can be used to ensure a stable power supply, thus improving overall system flexibility and RE consumption capacity. However, these technologies are constrained by their rotating standby, and the adjustable range of their output is limited.

Owing to the high energy density of hydrogen, hydrogen storage systems (HSS) can store large amounts of energy over long periods of time, providing significant advantages over short-term storage systems such as batteries. In addition, hydrogen can be used not only for power generation, but also as an industrial raw material and fuel for fuel cells. This versatility increases the economic efficiency and application flexibility of hydrogen. However, hydrogen storage systems involve multiple energy conversions (electrical energy to hydrogen and back to electrical energy), with energy losses occurring at each conversion. With the current technology, this efficiency loss is large (usually around 50%). Hydrogen storage systems have energy time-shift characteristics. They are particularly suitable for use with intermittent renewable energy sources, such as wind and solar power. For example, in times of excess wind or solar power generation, the excess power can be used to produce hydrogen by electrolysis of water. During periods of under-generation or peak demand, the stored hydrogen can be used to generate electricity, balancing supply and demand. By creating an integrated power generation system with renewable energy sources, the issue of power wastage can be effectively mitigated. However, focusing only on time-shift characteristics without examining the equivalent rotating standby capacity of HSS fails to realize the full potential of renewable energy consumption. In view of the shortcomings of the two, studying the coordinated operation characteristics of FCCS and HSS is of great significance. This could not only enhance the system’s economic and low-carbon attributes, but also significantly boost the level of renewable energy consumption.

1.2. Literature Survey

Converting high carbon-emitting units to FCCS under the dual-carbon objective is an effective way to achieve low carbon emissions [5]. Carbon capture units can improve the deep peaking output of the units, increase the peaking margin, and effectively address high carbon emissions.

Carbon capture units can primarily be categorized into basic modes (flue gas splitter type and solution storage type) and a flexible mode [6,7]. Dispatch models with solution storage-type carbon capture plants are developed to reduce high carbon emissions [8]. Models are also constructed by considering carbon capture equipment of the flue gas splitter type. However, basic carbon capture equipment faces the problem of high operational energy consumption [9]. During peak load hours, the unit output is used to balance the system load, making it impossible to meet the energy consumption requirements of the carbon capture equipment.

FCCS combines low-carbon emissions with a wide peak load shifting capability [10], which can promote RE consumption [11]. In addition, FCCS have a faster response time and deeper regulation range than conventional thermal units, making them more suitable for the system spinning reserve capacity. FCCS assume the system spinning reserve demand [12,13], and the proposed models have been verified by arithmetic case studies.

However, the large-scale grid integration of RE sources places high reserve pressure on FCCS. It is difficult to reduce their net output during load valley hours, and thus it is difficult to recognize the advantages of a wide regulation range [14].

HSS have good development prospects as a form of RE storage [15]. The production of hydrogen and methanol from non-circulating plastic waste has been analyzed in detail. This has helped the EU diversify its energy sources and improve the efficiency of its waste management [16]. In one study on the hydrogen production process, a screw kiln continuous reaction system was used to investigate the production of hydrogen from a representative waste plastic [17]. The introduction of HSS can enable the reuse of locally discarded wind and solar resources, thereby increasing local RE consumption. Additionally, storing hydrogen and generating electricity during peak load hours can reduce the system’s peak-to-valley difference and increase HSS revenue [18]. A micro-grid capacity optimization allocation model considering HSS has been developed. This model realized the capacity optimization of hydrogen storage configuration through the energy time-shift characteristics of HSS [19]. A hydrogen energy storage peaking station design has been proposed, and the possibility of peaking based on the energy time-shift characteristics of HSS was analyzed [20]. A day-ahead electricity-hydrogen energy trading model based on the time-shift characteristics of HSS was also proposed to maximize HSS profitability [21]. Most of the above literature studies on HSS have been based on micro-grids or integrated energy systems. The above literature mainly covers studies the energy time-shift characteristics of HSS, but do not consider the conversion efficiency of HSS.

HSS are characterized by fast regulation. In addition to the energy time-shift, they can also reduce thermal power standby capacity by flexibly adjusting the output to undertake the system spinning reserve capacity [22]. The hydrogen storage equivalent reserve characteristic is defined as the ability of HSS to flexibly adjust the power output to provide reserve requirements. To reduce randomness and volatility, HSS have been utilized to provide up and down spinning reserve capacity for wind power plants [23]. In one study, a low-carbon scheduling method for comprehensive utilization of hydrogen energy was proposed, and HSS were utilized to provide rotating standby capacity [24]. The above literature investigates the effectiveness of HSS in undertaking spinning reserve demand. However, the high construction cost of HSS and the limitations relating to their capacity mean that they cannot undertake reserve requirement alone; they need to be coupled with other reserve equipment.

On the load side, the power consumption behavior of users can be guided by price signals to alleviate the regulatory pressure on power generation and energy storage [25]. Different demand and response resources can be optimally allocated to achieve optimal system economics and to verify the effectiveness of DSR participation [26,27]. Therefore, the DSR can be co-scheduled with other adjusted ability resources to further optimize system performance under the dual-carbon goal.

1.3. Research Focus and Organization

To date, domestic and international scholars have not analyzed the coordination of the spinning reserve and time-shift characteristics of FCCS and HSS in terms of low-carbon economic operation. They are used only as a tool to capture carbon dioxide, without fully utilizing the advantages of their flexible operation. The coordination and complementarity of FCCS and HSS can ensure stable power supply and improve RE consumption. Combined with demand side response (DSR), RE consumption capacity can be maximized, carbon emissions can be reduced, and revenue can be improved. In summary, the research focuses are proposed as follows:

(1)

The research results of some scholars have transformed thermal power units by adding carbon capture equipment, which alleviates the peaking cost during load valley periods. However, carbon capture units cannot achieve optimal carbon capture levels during load peak periods. The first key problem is in identifying how carbon capture units can cooperate with the time-shift characteristics of HSS, and how to jointly tap the low-carbon characteristics to maximize the potential of RE consumption.

(2)

The spinning reserve allocation plan needs to be restudied in depth with FCCS after considering the equivalent spinning reserve characteristics of HSS. Both FCCS and HSS can be used for energy time-shifting and partial spinning reserve capacity. Therefore, there is an urgent need for an in-depth study of the limitations in their operation modes, as well as their advantages and complementary characteristics.

(3)

In low-carbon economic scheduling, the impact of scheduling results have not been fully explored after the coordinated operation of source, storage, and load. The impact of DSR on low-carbon optimal economic dispatch needs to be considered.

Therefore, a source-storage-load flexible scheduling strategy is established by considering the coordination of spinning reserve and time-shift characteristics for the FCCS and HSS under the dual-carbon objective. In Section 2, the operational mechanisms of HSS, FCCS, and DSR are analyzed, and the operational models are constructed separately. The HSS, FCCS, and DSR are fully utilized to improve the dispatch flexibility. In Section 3, a coordinated operation framework is constructed for FCCS and HSS. The coordinated characteristics of the FCCS and HSS are analyzed to assess the difficulty of FCCS down-regulating output during the load valley period. In addition, the performance of the system is further optimized through the load demand side response. In Section 4, the optimal strategies considering the coordination of spinning reserve and time-shift characteristics of FCCS and HSS are established with the aim of minimizing operation costs and carbon emissions. This strategy achieves an effective balance between low carbon emissions and economic performance. The validity of the model proposed is also verified through the case study in Section 5. Finally, conclusions and limitations are presented in Section 6.

2. Operating Mechanisms and Models of HSS, FCCS and DSR Considering Spinning Reserve and Time-Shift Characteristics

2.1. Operating Mechanisms and Models of HSS Considering Spinning Reserve and Time-Shift Characteristics

2.1.1. Operating Mechanisms of HSS Considering Spinning Reserve and Time-Shift Characteristics

The ES system includes electrolytic cells (EC), hydrogen storage tanks (HST), fuel cells (FC), batteries, and so on. The operating mechanisms of the HSS for RE consumption are shown in Figure 1.

According to the ESS operation mechanism, the process of HSS participation in RE consumption is reflected in two aspects:

(1)

Time-shift characterization of HSS

As shown in Figure 1, excess RE is fed into the EC for hydrogen production, and the generated hydrogen is stored in the HST when the RE output exceeds the load demand. The hydrogen from the HST is fed into the FC and converted into power generation to meet the load demand when the RE output is small and the load demand is high.

(2)

Equivalent spinning reserve characterization of HSS

The HSS system for power generation is characterized by fast response times. It can provide upper and lower spinning reserve capacity by flexibly adjusting output, thereby reducing the thermal power reserve capacity. Thermal power units can operate at lower net outputs, allowing RE installation capacity to be added to the grid.

2.1.2. Operating Models of HSS

The output models based on the HSS operation principle are shown as follows. It is assumed that the HSS equipment operates at a fixed efficiency. The efficiency of HSSC operation has not yet been considered.

$$P_{elH,t}={\eta }_{el}{P}_{el,t}$$

(1)

$$E_{hs,t}=E_{hs,t-1}+({\eta }_{el}{P}_{el,t}-{\displaystyle \frac{P_{fc,t}}{{\eta }_{fc}}})$$

(2)

$$P_{fc,t}={\eta }_{fc}{P}_{fcH,t}$$

(3)

where P_el,t and P_elH,t represent the input and output power of the EC; η_el represents the EC efficiency; E_HSS,t is HST energy stored; P_fcH,t and P_fc,t represent the FC input and output power; and η_fc represents the FC efficiency.

The HSS operation constraints are shown based on the above HSS operation mechanism.

(1)

Hydrogen capacity constraints for the HST:

$$E_{hs,0}=E_{hs,T}$$

(4)

$$E_{hs,\min }\le E_{hs,t}\le E_{hs,\max }$$

(5)

where E_HSS,0 and E_HSS,T represent the initial and final HST hydrogen capacity, respectively; and E_HSS,min and E_HSS,max represent the minimum and maximum HST hydrogen capacity limits, respectively.

(2)

Operating power constraints for the EC and FC:

$$\begin{array}{l}0.1P_{el,N}\le P_{el,t}\le P_{el,t\max }\\ 0\le P_{fc,t}\le P_{fc,t\max }\end{array}$$

(6)

$$P_{el,t\max }=\min \left(P_{el,N},{\displaystyle \frac{E_{hs,\max }-E_{hs,t}}{{\eta }_{el}}}\right)$$

(7)

$$P_{fc,t\max }=\min \left(P_{fc,N},{\eta }_{fc}(E_{hs,t}-E_{hs,\min })\right)$$

(8)

where P_el,tmax represents the maximum input power of the EC; P_fc,tmax represents the maximum output power of the FC; and P_el,N and P_fc,N represent the rated powers of the EC and FC, respectively. The minimum technical output of the EC is 10% of the rated power.

2.2. Operating Mechanisms and Models of FCCS Considering Spinning Reserve and Time-Shift Characteristics

2.2.1. Operating Mechanisms of FCCS Considering Spinning Reserve and Time-Shift Characteristics

The conversion of high carbon emission units to the CCU is an effective way to achieve low carbon emissions under the dual-carbon objective. The installation of a diversion device and solution storage device can make it more flexible (FCCS). The operating mechanisms of FCCS are shown in Figure 2.

The operation mode of the FCCS can be described as follows. A portion of the CO₂ produced by conventional thermal power units enters the absorbed tower via the flue gas bypass. The solution storage tank contains an aqueous monoethanolamine solvent (MEA), in which CO₂ is soluble. At this time, CO₂ and MEA combine to form a rich liquid with a high CO₂ concentration, which is stored in the rich liquid tank (Absorber 1). The rich liquid enters the regeneration tower when the CO₂ is needed. The MEA solvent separates CO₂ by high temperature action, and reuse of CO₂ is realized. After separation, the rich liquid forms a poor liquid stream with low CO₂ concentration. It then returns to the poor liquid storage (Absorber 3) and the absorption tower for recycling.

2.2.2. Operating Models of FCCS

The output models of FCCS are as follows:

$$P_{FC,t}=P_{JC,t}+P_{r,t}+P_D$$

(9)

where P_FC,t represents the FCCS total power in the flexible operation mode; P_JC,t represents the FCCS net output power; P_r,t represents the FCCS operating power consumption; and P_D represents fixed consumption of the carbon storage system.

The FCCS operating power consumption P_r,t is as follows:

$$P_{r,t}=\lambda E_{JC,t}=\lambda (\eta {\delta }_t{E}_{C,t}+E_{V,t})$$

(10)

where λ represents the unit set of CO₂ captured power; E_JC,t represents the captured CO₂ capacity; η represents the CO₂ captured efficiency; δ_t represents the flue gas split ratio at time t; E_C,t represents the FCCS captured CO₂ capacity; and E_V,t represents the CO₂ capacity supplied by the storage tank.

In addition, the parameter E_C,t and E_V,t constraints are as follows:

$$E_{C,t}\le \alpha P_{FC,t}$$

(11)

$$E_{V,t}=V_{v,t}\theta {\rho }_V{\delta }_V{\displaystyle \frac{M_{CO2}}{M_V}}$$

(12)

where α represents the emission factor required by the emission policy; V_v,t is the MEA solution volume; M_V and M_CO2 represent the MEA solution and CO₂ molar mass; θ represents the resolution; and ρ_V and δ_V represent the concentration and density of the MEA solution, respectively.

2.3. Operating Mechanisms and Models of DSR

The load-side regulation methods should not be ignored. The introduction of the DSR mechanism can improve the flexible regulation characteristics of load-side resources and give full play to the load-side advantages. Price-based DSR is used to change the load’s electricity consumption behavior by means of tariff guidance. The policy of “peak-flat-valley” time-sharing tariff is applied to guide the load transfer. After adjusting from fixed tariff to time-of-day tariff, the load DSR model is shown in (13):

$$\left[\begin{array}{c}{P}_{DSR,P}\\ P_{DSR,F}\\ P_{DSR,V}\end{array}\right]=\left[\begin{array}{c}{P}_{L,P}\\ P_{L,F}\\ P_{L,V}\end{array}\right]+\left[\begin{array}{ccc}{P}_{L,P}& 0& 0\\ 0& P_{L,F}& 0\\ 0& 0& P_{L,V}\end{array}\right]K\left[\begin{array}{c}\Delta c_P/c_A\\ \Delta c_F/c_A\\ \Delta c_V/c_A\end{array}\right]$$

(13)

where P_DSR,P, P_DSR,F, and P_DSR,V represent the load power in the peak, flat, and valley hours after the implementation of DSR; P_L,P, P_L,F and P_L,V represent the peak, flat, and valley power before the implementation of DSR; K represents the elasticity matrix of tariff demand; Δc_P, Δc_F, and Δc_V represent the tariff changes in the peak, flat, and valley hours before and after the implementation of DSR, respectively; and c_A represents the fixed tariff before DSR.

The elasticity matrix of the tariff demand K is determined by the elasticity coefficient of tariff demand, as shown in (14):

$$K=\left[\begin{array}{ccc}{k}_{P,P}& k_{P,F}& k_{P,V}\\ k_{F,P}& k_{F,F}& k_{F,V}\\ k_{V,P}& k_{V,F}& k_{V,V}\end{array}\right]$$

(14)

where P, F, and V represent the load period in the peak, flat, and valley hours. The main diagonal elements k_P,P, k_F,F, and k_V,V are the coefficients of self-elasticity; and the remaining elements are coefficients of mutual elasticity.

The change scale of load power on electricity price is represented by the price elasticity of the demand coefficient k, as shown in (15):

$$k={\displaystyle \frac{c_{A,t}}{\Delta c_t}}{\displaystyle \frac{\Delta P_{L,t}}{P_{L,t}}}$$

(15)

where P_L,t and c_A,t represent the power and tariff before the implementation of DSR; and ΔP_L,t and Δc_t represent the power and tariff changes after the implementation of DSR.

3. The Characteristics Complementary Framework of FCCS and HSS

3.1. The Characteristics Complementary Framework of FCCS and HSS

The characteristics complementary framework of FCCS and HSS is shown in Figure 3. It contains five energy drives: electricity flow, gas flow, hydrogen flow, heat flow, and carbon flow. Renewable electricity generated from wind and PV during the valley load period is converted to generate electricity during the peak load period through EC, HST, and FC. Its time-shifted characteristics enable additional consumption of RE. The heat flow generated by the EC and FC can also satisfy heating loads. In addition, carbon captured by the FCCS is used to produce methane for profit in the gas market.

The characteristics complementary framework of the FCCS and HSS is analyzed from three perspectives: source, storage, and load.

Source side: Conventional thermal power units, as well as decarbonized and modified carbon capture units, can capture the generated carbon dioxide and reduce carbon emissions from thermal power units on the source side. In addition, carbon captured by FCCS is used to produce methane for profit in the gas market.

Load side: The DSR mechanism guides consumers’ electricity consumption patterns, increasing load demand in the valley period.

RE and energy storage: RE is coupled with HSS to form a co-generation system. This system can realize the reuse of local RE discarded resources and increase the local RE consumption. Spinning reserve demand is transformed from thermal power units to hydrogen storage systems coupled with carbon capture units.

As a result of the above analysis, both FCCS and HSS can be used for energy time-shifting and to perform some of the system’s spinning reserve tasks. Therefore, the limitations of the two systems and their complementary characteristics need to be studied in depth.

3.2. The Coordinated Operation Characteristics of FCCS and HSS

3.2.1. The Spinning Reserve Characteristics of FCCS

FCCS reduce net output power by increasing the amount of CO₂ captured during the valley period. The output power adjustment range is shown as (16) for carbon capture units. However, the FCCS has a wider adjustment range, as shown in (17):

$$P_{HC\min }\le P_{JC,t}\le P_{HC\max }$$

(16)

$$P_{HC\min }-\lambda \eta {\delta }_{\max }\alpha P_{HC\max }-P_D\le P_{JC,t}\le P_{HC\max }-P_D$$

(17)

where P_{HC max} and P_{HC min} represent the maximum and minimum output power for conventional thermal part of carbon capture units; and δ_max represents the maximum flue gas diversion ratio at time t.

In addition, thermal units need to have sufficient reserve capacity to undertake load and RE output uncertainties. The spinning reserve constraints for FCCS are expressed in (18):

$$\begin{array}{l}0\le P_{C,t,up}\le \min \left(U_{C,t}{P}_{HC\max }-P_{JC,t},R_{C,up}\Delta t\right)\\ 0\le P_{C,t,down}\le \min \left(P_{JC,t}-U_{C,t}{P}_{HC\min },R_{C,down}\Delta t\right)\end{array}$$

(18)

where U_C,t represents the FCCS start-stop state; P_C,t,up and P_C,t,down represent the FCCS maximum and minimum reserve capacity; R_C,up and R_C,down represent the FCCS maximum and minimum climbing rates; and Δt is the scheduling time interval.

The output power adjustment range for spinning reserve demand can be derived from (17) and (18) as follows.

$$P_{HC\min }-\lambda \eta {\delta }_{\max }\alpha P_{HC\max }-P_D+P_{C,t,down}\le P_{JC,t}\le P_{HC\max }-P_D-P_{C,t,up}$$

(19)

From (16)–(19), FCCS have a limited net output adjustment range under the spinning reserve demand, which cannot fully utilize the carbon capture advantage. In addition, the development of HSS is still in its infancy in China. Due to capacity limitations, HSS are unable to undertake the regulation load demand alone, whether to provide the spinning reserve or energy time-shift demand.

Therefore, there is an urgent need to explore the coordination and complementarity between FCCS and HSS in terms of energy time-shift and spinning reserve, and to analyze the coupling relationship. This is the key to solving the problem of rapid development of HSS.

3.2.2. The Equivalent Spinning Reserve Characteristics of FCCS Coupled with HSS

First, various constraints on the HSS spinning reserve need to be considered. The EC reserve capacity is limited by its own and the HST constraints. The equivalent spinning reserve constraints of the EC are shown as (20):

$$\begin{array}{l}0\le P_{el,t,up}\le (P_{el,t}-0.1P_{el,N}){\phi }_t\\ 0\le P_{el,t,down}\le \min \left({\phi }_t(P_{el,N}-P_{el,t}),{\displaystyle \frac{E_{hs,\max }-E_{hs,t}}{{\eta }_{el}}}\right)\end{array}$$

(20)

where P_eL,t,up and P_eL,t,down represent the EC maximum and minimum reserve capacity, respectively; and φ_C,t represents the EC start-stop state.

In addition, the FC reserve capacity is limited by its own and the HST constraints. The start-stop characteristics will also affect its standby capacity. The equivalent spinning reserve constraints of the FC in the startup state are shown as (21):

$$\begin{array}{l}\begin{array}{cc}0\le P_{fc,t,up}\le \min \left({\psi }_t(P_{fc,N}-P_{fc,t}),(E_{hs,t}-E_{hs,\min }){\eta }_{fc}\right)& {\psi }_t=1\end{array}\\ \begin{array}{cc}0\le P_{fc,t,down}\le {\psi }_t{P}_{fc,t}& {\psi }_t=1\end{array}\end{array}$$

(21)

where P_fc,t,up and P_fc,t,down represent the FC maximum and minimum reserve capacity; and ψ_C,t represents the FC start-stop state.

The 50% capacity is still available in the shutdown state with fast response capability. The equivalent reserve capacity constraint of the FC in the shutdown condition is shown as (22):

$$\begin{array}{l}\begin{array}{cc}0\le P_{fc,t,up}\le \min \left(0.5P_{fc,N},(E_{hs,t}-E_{hs,\min }){\eta }_{fc}\right)& {\psi }_t=0\end{array}\\ \begin{array}{cc}0\le P_{fc,t,down}\le {\psi }_t{P}_{fc,t}=0& {\psi }_t=0\end{array}\end{array}$$

(22)

It is assumed that the FCCS energy consumption varies linearly with the increase in the FCCS output. Its consumption is highest during the trough period and lowest during the peak period. The changes in the regulation ranges of conventional generation, carbon capture units, and carbon capture units after the reserve of hydrogen storage under spinning reserve demand are shown in (23) and Figure 4, respectively.

$$\begin{array}{l}\Delta P_{J1}=(P_{HC\max }-P_{C,t,up})-(P_{HC\min }-P_{C,t,down})\\ \Delta P_{J2}=(P_{HC\max }-P_D-P_{C,t,up})-(P_{HC\max }-\lambda \eta {\delta }_{\max }\alpha P_{HC\max }-P_D+P_{C,t,down})\\ \Delta P_{J3}=(P_{HC\max }-P_D-P_{C,t,up*})-(P_{HC\max }-\lambda \eta {\delta }_{\max }\alpha P_{HC\max }-P_D+P_{C,t,down*})\end{array}$$

(23)

where P_C,t,up* and P_C,t,down* represent the maximum and minimum reserve capacity of carbon capture units after the reserve of hydrogen storage, respectively (P_C,t,up* < P_C,t,up, P_C,t,down* < P_C,t,down); and ΔP_J1, ΔP_J2 and ΔP_J3 are the net output adjustment ranges of conventional units, FCCS, and FCCS after the reserve of HSS under the spinning reserve demand.

3.2.3. The Time-Shift Characteristics of FCCS Coupled with HSS

Due to the anti-peaking characteristics of RE, RE generation is abandoned more in load valley periods, and there is a lack of RE consumption capacity when considering carbon capture units alone. In addition, carbon capture unit output is large during load peak periods. However, the load demand keeps the power generation cost at a high level. When considering only the energy time-shift characteristics of HSS, more power generation is lost in the energy conversion process. The effect of RE consumption is not ideal for RE abandonment.

FCCS and HSS can be considered to complement each other’s advantages by matching the energy time-shift characteristics after the above analysis.

(1)

FCCS improve RE consumption in load valley periods through the liquid storage tanks to ease the peak pressure of HSS.

(2)

When FCCS reach the lower output limit during the load valley period, the phenomenon of power abandonment still occurs. HSS can consume the remaining electricity and replace the high-cost carbon capture unit output with energy storage output during the load peak period, so as to improve the operation economy.

4. Source-Storage-Load Flexible Scheduling Strategies Considering the Coordination of Spinning Reserve and Time-Shift Characteristics for HSS and FCCS

4.1. Objective Function

The source-storage-load flexible scheduling strategies are optimized to minimize carbon emissions and achieve the lowest possible economic cost under the system operation and output constraints. This strategy achieves an effective balance between low carbon and economic performance. The total system operating cost sub-objective function f includes the thermal power unit generation cost F₁, the thermal power unit start-stop cost F₂, the total spinning reserve cost F₃, the thermal power unit carbon trading cost F₄, the carbon capture unit solution loss cost F₅, the hydrogen storage maintenance cost F₆, and the total power loss cost F₇.

The total electricity loss cost F₇ is derived from the following two parts: (1) RE abandoned electricity loss, and (2) electricity loss caused by the low conversion efficiency of HSS in the process of consuming abandoned RE.

$$\begin{array}{l}\min f_1=F_1+F_2+F_3+F_4+F_5+F_6+F_7\\ F_1=c_{coal}{\displaystyle \sum _{t=1}^T{\displaystyle \sum _{n=1}^N(a_n{P}_{thn,t}^2+b_n{P}_{thn,t}+c_n)}}\\ F_2={\displaystyle \sum _{t=1}^T({\displaystyle \sum _{n=1}^N{U}_{n,t}(1-U_{n,t-1})}}{c}_{sn})\\ F_3={\displaystyle \sum _{t=1}^T{\displaystyle \sum _{k=1}^M(c_k^{up}{P}_{k,t,up}+c_k^{down}{P}_{k,t,down})}}\\ F_4={\displaystyle \sum _{t=1}^T{c}_{ct}(E_{total}-\alpha {\displaystyle \sum _{n=1}^N{P}_{thn,t})}}\\ F_5={\displaystyle \sum _{t=1}^T{c}_{sc}{\eta }_{sc}{E}_{JC,t}}\\ F_6={\displaystyle \sum _{t=1}^T(c_{el}{P}_{el,t}+c_{fc}{P}_{fc,t})}\\ F_7={\displaystyle \sum _{t=1}^T{c}_r(P_{rq,t}+(1-{\eta }_{el})P_{el,t}+\frac{(1-{\eta }_{fc})P_{fc,t}}{{\eta }_{fc}})}\end{array}$$

(24)

where T represents the scheduling cycle; N represents the number of conventional units; P_thn,t represents the electricity generated by the n-th conventional unit at time t; c_coal is the price of coal per unit weight; a_nP²_thn,t + b_nP_thn,t + c_n represents the coal consumption function; a_n, b_n, and c_n represent coal consumption coefficients; U_n,t represents the n-th unit start-stop status; c_sn represents the start-stop cost; M represents the equipment collection providing spinning reserve capacity in conventional units, FCCS, and HSS; c^up_k and c^down_k represent the maximum and minimum reserve cost factors for the k-th equipment; P_k,t,up and P_k,t,down represent the maximum and minimum reserve provided by the k-th equipment at time t; c_ct represents carbon trading prices; E_total represents total carbon emissions; α represents the carbon trading quota; c_sc represents the MEA solvent cost factor; η_sc represents the MEA solvent operating loss factor; E_JC,t represents CO₂ captured capacity; c_el and c_fc represent the coefficients of operation and maintenance costs for EC and FC units, respectively; P_rq,t represents the RE abandoned power; and c_r represents the RE abandoned cost coefficient.

To effectively curb the disorderly carbon emissions of some enterprises, this paper mimics the ladder tariff model and constructs a three-stage stepped carbon emission cost model as follows:

$$\min f_2=\left\{\begin{array}{c}\lambda (E_p-E_q)\\ (1+\tau )\lambda (E_p-E_q)-\tau \lambda d\\ (1+2\tau )\lambda (E_p-E_q)-3\tau \lambda d\end{array}\right.\begin{array}{c}{E}_p-E_q\le d\\ d\le E_p-E_q\le 2d\\ E_p-E_q>2d\end{array}$$

(25)

where f₂ represents the daily carbon emission cost; λ represents the trading price; τ represents the price growth; and d represents the interval of the step carbon emission cost. The three-stage stepped carbon emission cost model is shown in Figure 5.

If E_p < E_q, the remaining part of the quota can be traded in the carbon trading market to obtain carbon revenue. On the contrary, additional carbon emissions must be purchased. The higher the excess, the higher the transaction unit price in a stepwise distribution.

The traditional use of linear weighting methods, when transformed into single objective optimization, will inevitably use subjective assignments. This results in a lack of fairness in the results. The multi-objective problem is transformed by minimizing the Euclidean distance as follows:

$$\begin{array}{cc}\min F=\sqrt{{[{\displaystyle \frac{f_1(x)-f_1^{*}}{f_1^{*}}}]}^2+{[{\displaystyle \frac{f_2(x)-f_2^{*}}{f_2^{*}}}]}^2}& x\in \mathrm{M}\end{array}$$

(26)

where x represents the scheduling plan; M is the solution set formed by all possible scheduling plans; and $f_1^{*}$ and $f_2^{*}$ are the optimal solutions of a single sub-objective function in the M solution set, respectively.

4.2. Constraints

The source-storage-load flexible scheduling strategy constraints include power grid balance constraint, thermal power unit operation constraint, RE output constraint, system reserve constraint, and other constraints.

(1)

Conventional unit operation constraint:

$$U_{n,t}{P}_{thn,\min }\le P_{thn,t}\le U_{n,t}{P}_{thn,\max }$$

(27)

where P_thn,max and P_thn,min represent the maximum and minimum output limits, respectively.

(2)

RE output constraint:

$$\begin{array}{l}{P}_{re,tpre}=P_{re,t}+P_{el,t}+P_{rq,t}\\ 0\le P_{re,t}\le P_{re,tpre}\end{array}$$

(28)

where P_re,tpre represents the forecasted RE generation; and P_re,t represents the actual grid-connected power from RE sources.

(3)

Power grid balance constraint:

$$P_{re,t}+{\displaystyle \sum _{n=1}^N{P}_{Jn,t}}+P_{fc,t}=P_{\mathrm{DSR},t}$$

(29)

where P_Jn,t represents the net output power; and P_DSR,t represents load power after demand side response.

(4)

System reserve constraint:

$$\begin{array}{l}p\left\{{\displaystyle \sum _{n=1}^N{P}_{Jn,t}+P_{re,t}+P_{fc,t}+{\displaystyle \sum _{k=1}^M{P}_{k,t,up}\ge P_{\mathrm{DSR},t}}}\right\}\ge \sigma \\ p\left\{{\displaystyle \sum _{n=1}^N{P}_{Jn,t}+P_{re,t}+P_{fc,t}-{\displaystyle \sum _{k=1}^M{P}_{k,t,down}\le P_{\mathrm{DSR},t}}}\right\}\ge \sigma \end{array}$$

(30)

where p{} is the confidence expression; and σ is the confidence coefficient required by the system reserve constraint.

(5)

Other constraints

The optimal economy dispatch model also contains thermal power unit start-stop time constraints and climbing constraints, EC constraints, demand response satisfaction constraints, etc., all of which have been studied; the specific formulas are referenced in references [28,29,30]. In addition, the spinning reserve constraint for FCCS is shown in (16)–(19). The spinning reserve constraint for HSS units is shown in (20)–(22).

The solution process for the proposed low-carbon economic operation strategies is illustrated in Figure 6.

5. Case Study

5.1. The Parameter

The improved IEEE-30 node standard system is shown in Figure 7 [31]. The system contains four conventional units, among which the unit G1 can be decarbonized, and the G2, G3, and G4 are conventional units. The installed capacities of the wind and photovoltaic power are 260 MW and 80 MW. The parameters of DSR, HSS, conventional units, unit G1 (FCCS), and other parameters are shown in

Table 1. The parameters of DSR.

Time	Load Status	Tariff
11:00–14:00, 17:00–19:00	Peak	0.186 ($/kWh)
08:00–11:00, 14:00–17:00, 19:00–22:00	Flat	0.124 ($/kWh)
0:00–08:00, 22:00–24:00	Valley	0.072 ($/kWh)

,

Table 2. HSS parameters.

Parameters	Values
EC efficiency ηel/%	65
FC efficiency ηfc/%	60
Minimum capacity limit of HSS in HST EHSS,min/MWh	30
Maximum capacity limit of HSS in HST EHSS,max/MWh	170
Maintenance cost for EC cel/($/MWh)	5
Maintenance cost for FC cfc/($/MWh)	25
HST capacity/MWh	250
EC capacity/MWh	60
FC output power/MW	30

,

Table 3. Thermal power unit parameters before decarbonization.

Units		G1	G2	G3	G4
Maximum limit output/MW		180	80	50	30
Minimum limit output/MW		45	20	15	10
Coefficients of coal consumption function	an	0.04	0.18	0.63	0.9
	bn	20	18.2	10.7	32.8
	cn	386	366	348	331
Climbing speed/(MW/min)		3.7	1.8	1	0.7
Carbon emission intensity/(t/MWh)		0.99	0.98	0.96	1.13

and

Table 4. Decarbonized carbon capture unit G1 parameters, and other parameters.

Parameters	Values	Parameters	Values
Power consumption per unit of CO2 captured λ/(MWh/t)	0.287	Density of MEA solution δV/(g/mol)	1.01
Carbon capture efficiency η/%	88	Concentration of MEA solution ρV/%	30
Carbon emission factor per unit of generation power α	30	Regeneration tower resolution θ/(mol)	0.24
Price of coal per unit weight ccoal/($/t)	127	Molar mass of MEA solution MV	61.08
Carbon trading quota α	0.7	Molar mass of CO2 MCO2	44
Carbon trading price cct ($/t)	15.4	Output power of FC/MW	30
RE abandoned cost coefficient ($/kWh)	0.06	Upper reserve cost of EC ($/MW)	0
Maximum reserve cost ($/MW)	16.8	Lower reserve cost of EC ($/MW)	4
Minimum reserve cost ($/MW)	17	Upper reserve cost of FC ($/MW)	14.7
		Lower reserve cost of FC ($/MW)	0

.

The arithmetic in this paper is improved from the IEEE-30 standard node system. The positions of traditional units G2, G3, and G4 are all based on the standard node system. There are three improvements:

(1)

This study investigates the synergistic operation effect of carbon capture and hydrogen storage systems, so the conventional unit is decarbonized at node 11, and the HSS is set up for synergistic operation with it.

(2)

Because the installed capacity of the PV and WT is relatively large, they are set up at the remote nodes 5 and 14, and the parameters are from the actual measurement at the energy base of the Hami region in Xinjiang, China.

(3)

The parameters of G2, G3, and G4 are derived from the measured energy bases in Hami area, Xinjiang, China.

To verify the effectiveness, the following five scenarios are set up:

(1)

S1: FCCS, HSS, and DSR are not considered.

(2)

S2: The flexible carbon capture system is considered on the basis of S1.

(3)

S3: The hydrogen storage system is considered on the basis of S2, but only its energy time-shift characteristics are considered.

(4)

S4: The equivalent spinning reserve characteristics of HSS are considered on the basis of S3.

(5)

S5: The load DSR is considered on the basis of S4.

5.2. Comparison Results of Optimal Economic Scheduling Model under Different Scenarios

5.2.1. Comparative Analysis of Conventional Units Output under Different Scenarios

The output of conventional units under different scenarios is shown in Figure 8.

The unit G1 in S1 is a conventional unit without a carbon capture device. It is difficult to reduce its net output during the G1 running period. In this case, thermal power units G2 and G3 are used to satisfy the load demand in the valley period and take care of the system spinning reserve plan due to their lower generation costs. Therefore, the reserve costs, carbon emissions and wind abandonment remained high in S1.

In S2, the introduction of decarbonized carbon capture units has significant advantages in terms of carbon emissions, RE consumption, and total system cost. Carbon capture unit G1 can capture and store CO₂ generated by thermal power units, reducing carbon emission costs. Therefore, in S2, the carbon capture unit G1 attempts to generate electricity at all times to reduce the total system cost. The thermal power unit G1 with a carbon capture device can operate at a lower net output during load valley period.

In S3, a hydrogen storage system is introduced, but only the time-shifted energy characteristics of the HSS are considered. Comparing S3 with S2, it can be seen that the peak power output of the carbon capture unit G1 is slightly reduced due to the application of the time-shifted characteristics of the hydrogen storage system.

In S4, the unit G1 net output is further reduced and RE generation is further increased during the 2:00–5:00 period after considering the equivalent spinning reserve characteristics of ESS. Because the hydrogen storage system has a lower reserve cost than the thermal unit, the output of the other thermal units (G2, G3) is not significantly affected due to higher generation costs.

Based on Scenario 4, load-side demand response is introduced to improve the dispatchability of the system in S5. The conventional thermal unit G2 is shut down when demand side response is introduced. After the introduction of DSR in S5, the load is reduced in the load peak period, and the power generation can be supplied by the conventional thermal power unit G3 alone. The generation deficit is covered by the carbon capture units. The carbon emissions are reduced by 62 t compared to S4. In addition, the carbon emission intensity of G4 is significantly higher than that of other units, and the scheduling plan for G4 is not arranged in different scenarios.

The peaks in Figure 8 all represent the blue curve (G1) output power. Since unit G1 in S1 does not have a carbon capture device installed and has lower generation costs, it is used to obtain the peak load. G1 in S2 has a carbon capture device installed, but its power required for carbon capture is higher, and it needs to bear the system spinning reserve. The introduction of HSS in S3 can significantly reduce the output of G1 during the peak hour (20:00) through the interplay of time-shift characteristics. In S4, considering the equivalent rotating reserve characteristic of ESS, the net output of the G1 unit can be further reduced. With the introduction of DSR in S5, the load in the peak load period is reduced, and the generation that would have been required from G2 can be supplied separately by the conventional thermal unit G3, which has a lower generation cost.

In summary, the carbon capture unit G1 is prioritized for power supply in day-ahead dispatch plans. When the higher output of carbon capture unit G1 results in higher generation costs, the lower cost and lower carbon emission thermal units G3 and G2 are prioritized for power supply. The thermal power unit G4 is not scheduled for generation due to its higher carbon emission intensity.

5.2.2. Comparative Analysis of Power Generation from HSS under Different Scenarios

The comparison of power generation from hydrogen storage systems in different scenarios is shown in

Table 5. Stored electricity comparison of HSS in different scenarios.

Different Scenarios	S3	S4	S5
Hydrogen production electricity/(MW·h)	110.3	96.2	94
Hydrogen power generation/(MW·h)	43.1	37.6	36.7
Power loss in process/(MW·h)	67.2	58.6	57.3

and Figure 9. The approximate confidence interval for the data is set at 0.95.

In S3, the hydrogen storage system converts RE abandoned generation into hydrogen for storage during the load valley period. The hydrogen storage system generates electricity from FC during load valley period. Although there are power generation losses during the conversion process, the thermal generation cost is reduced by $3848.20.

In S4, the hydrogen storage system has a faster regulation speed and lower standby costs, and is more suitable for system rotation reserve as it does not compress the new energy feed-in space when considering the demand for system spinning reserve. However, as a result of the HSS taking on the spinning reserve demand, hydrogen-based production of electricity is reduced from 110.3 MWh to 96.2 MWh compared to S3. In addition, due to the increase in RE consumption capacity, the hydrogen power generation space is crowded out, and the hydrogen power generation is also reduced by 5.5 MWh compared with S3. Finally, due to the reduction in hydrogen-based production of electricity, the consumption of the hydrogen storage system in the energy transfer process is reduced by 8.6 MWh.

In S5, the peak hour load is reduced, and the valley hour load is increased due to the load-side demand response. The total amount of generation assumed by the hydrogen storage system decreases due to its energy time-shift property. Finally, hydrogen-based production of electricity and hydrogen power generation are reduced by 2.2 MWh and 0.9 MWh, respectively, compared with S4.

5.2.3. Comparative Analysis of Scheduling Results under Different Scenarios

The results of power system scheduling under different scenarios are shown in Figure 10, and the cost comparisons are presented in

Table 6. Comparison of system cost and carbon emission in different scenarios.

Cost Index	Different Scenarios
	S1	S2	S3	S4	S5
Thermal power unit generation cost F1 ($)	108,237.6	114,234.9	110,386.7	109,524.5	110,826.7
Thermal power unit start-stop cost F2 ($)	3311.6	3487.2	3457.9	3341.7	2507.4
Spinning reserve cost F3 ($)	28,075.7	29,384.1	29,001.9	24,136.9	23,098.7
Thermal power unit carbon trading cost F4 ($)	10,247	−13,254.7	−11,368	−10,157.6	−13,985.6
Carbon capture unit solution loss cost F5 ($)	0	2863.1	2786.4	2725.9	2887.5
Hydrogen storage operation and maintenance cost F6 ($)	0	0	1509	1134	1257
Total electricity loss cost F7 ($)	11,258.6	5984.7	4236.8	3988.1	3876.5
Carbon emission (t)	2556	1131	1108	1121	1059
RE abandoned electricity (MWh)	186.7	96.3	0	0	0
Total cost ($)	161,130.5	142,699.3	140,010.7	134,693.5	130,468.2

. The reserve cost, carbon trading cost, and RE abandoned generation remain at a high level in S1. The system operation is not low carbon and economical because FCCS and HSS are not considered.

Compared to S1, the introduction of the decarbonized carbon capture unit in S2 provides significant advantages in terms of carbon emissions, RE consumption, and total cost. The carbon capture unit captures and sequesters CO₂ produced by conventional units, reducing carbon emissions by 1425 t. The thermal power unit G1 with a carbon capture device can operate at a lower net output during the load valley period, which reduces renewable energy power abandonment by 90.4 MWh. The carbon capture unit in G1 can capture and store CO₂ generated by thermal power units, reducing carbon emission costs. Therefore, in S2, the carbon capture unit G1 attempts to generate electricity at all times to reduce the total system cost by $18,431.20.

In S3, the hydrogen storage system is introduced, but only the energy time-shift characteristics of the hydrogen storage system are considered. Compared to S2, the electricity loss cost ($5984.70) and the overall cost ($140,010.7) are reduced, owing to the fact that the RE abandoned generation is significantly reduced by considering the hydrogen storage system time-shift characteristics. In particular, there is a significant reduction of 96.3 MWh in RE abandoned generation and a 29.22% reduction in the electricity loss cost. During the load valley hours, the FCCS G1 is at low output and has sufficient capacity to provide upper and lower spinning reserve. The carbon capture units can carry the system’s spinning reserve requirements on their own.

In S4, the hydrogen storage system has a lower reserve cost compared to thermal power units. The spinning reserve cost is substantially reduced compared to the previous three scenarios. Compared to S3, the reserve cost is reduced by 16.78% and the overall cost is reduced by $5317.30. As a result, the carbon capture unit G1 can operate at a lower net output than S3, and more grid-connected electricity from the RE source is allowed during the load valley period. Due to the limitations in regulation capacity, the EC and FC are not able to fully take over the rotating standby of the system, but have to be coupled with a carbon capture unit, which provides the remaining rotating standby demand.

Based on S4, load-side demand response is introduced to improve the dispatchability of the system in S5. The conventional thermal unit G2 is shut down when demand side response is introduced. After the introduction of the DSR in S5, the load is reduced in the load peak period, and the power generation can be supplied by the conventional thermal power unit G3 alone. The generation deficit is covered by the carbon capture units. The carbon emissions are reduced by 62 t compared to S4. Additionally, the total cost of the system is reduced by $4224.70.

5.3. Coordinated Operation Characteristics Analysis of FCCS and HSS

5.3.1. The Equivalent Spinning Reserve Characteristics Analysis of FCCS and HSS under Different Scenarios

The equivalent spinning reserve allocation plans of S3 and S4 are shown in Figure 11. The approximate confidence interval for the data is set at 0.95.

In S3, during the load valley hours, the FCCS G1 is at low output and has sufficient capacity to provide upper and lower spinning reserve. However, taking on all of the down-rotation standby capacity leaves limited space for RE to be connected to the grid, which in turn makes it difficult for wind power to be consumed. During peak load periods, the carbon capture energy used by CCS units to capture carbon dioxide is small due to the limitation of carbon capture energy consumption. Carbon capture units can carry the system’s spinning reserve requirements on their own.

In S4, the hydrogen storage system has a faster regulation speed and lower standby costs, and is more suitable for system rotation reserve as it does not compress the new energy feed-in space when considering the demand for system spinning back-up. The spinning reserve cost is substantially reduced compared to the previous three scenarios. Compared to S3, the reserve cost is reduced by 16.78%, and the overall cost is reduced by $5317.30. The FC in S4 accounted for 22.67% of the upper spinning reserve capacity, and the FC and EC accounted for 85.72% of the lower spinning reserve capacity. As a result, the carbon capture unit G1 can operate at a lower net output than S3, and more grid-connected electricity from the RE source is allowed during the load valley period. Due to the limitations in regulation capacity, the FC and EC are not able to fully take over the rotating standby of the system, but have to be coupled with a carbon capture unit, which provides the remaining rotating standby demand.

5.3.2. Time-Shift Characteristics Analysis of FCCS and HSS under Different Scenarios

The conventional unit outputs under different scenarios are shown in Figure 12.

FCCS and HSS are not considered in S1. The regulatory capacity of the conventional units is limited. The absorption capacity of RE is weak, and RE curtailment is generated. The energy time-shift characteristic of the flexible carbon capture system and hydrogen storage system can improve the RE grid-connected generation and the operational economy. The outputs of the flexible carbon capture system and the hydrogen storage system in S2 and S3 are shown in Figure 13.

As shown in Figure 13, the carbon capture power consumption is high in S2. However, there is still RE abandonment throughout the 3:00–9:00 period, as the hydrogen storage system has not been introduced to store the rich electricity from renewable energy sources. This results in a waste of renewable energy generation. In S3, the electricity stored in the HST is shifted to the peak period due to the time-shifting characteristic in the hydrogen storage system during the 3:00–9:00 period. The hydrogen is used to generate electricity to prevent load-shedding conditions during the 12:00–13:00 and 16:00–20:00 periods. The carbon capture unit G1 has a lower net output power, which allows for more grid-connected electricity from the RE source. In addition, the introduction of the hydrogen storage system allows RE abandonment to be fully utilized, thereby increasing the RE consumption capacity.

5.4. Comparison Results for a High Percentage of RE Penetration Rate

To verify the advantages of dispatch strategies in terms of RE consumption, the economic dispatch results are compared and analyzed under different RE penetration rates. Comparison results of total cost and power loss under different RE penetration rates are listed in

Table 7. Comparison results of total cost and power loss under different RE penetration rates.

Different RE Penetration Rates		Different Scenarios
		S1	S2	S3	S4	S5
50%	Total cost ($)	161,130.5	142,699.3	140,010.7	134,693.5	130,468.2
	RE consumption (MWh)	2432.8	2926.7	2975.8	2998.7	3021.1
55%	Total cost ($)	163,458.7	141,358.9	138,857.6	132,367.9	128,123.7
	RE consumption (MWh)	2432.8	3061.1	3126.9	3157.2	3198.6
60%	Total cost ($)	165,741.5	140,069.8	137,736.9	130,158.9	126,108.9
	RE consumption (MWh)	2432.8	3199.8	3287.4	3308.7	3357.4
65%	Total cost ($)	169,057.9	140,786.8	138,258.8	128,842.7	125,075.9
	RE consumption (MWh)	2432.8	3241.7	3324.1	3427.4	3458.9
70%	Total cost ($)	171,032	141,864.8	139,348.5	127,815.8	124,358.8
	RE consumption (MWh)	2432.8	3241.7	3324.1	3598.4	3621.1

.

In S1, the RE consumption capacity has reached its ceiling. An increase in RE penetration rate leads to an increase in total system cost and power loss. In S2 and S3, RE consumption is increased due to the time-shift characteristics of the flexible carbon capture system and the hydrogen storage system. However, there is still a lack of RE consumption after the RE penetration rate is increased to 60%. In S4 and S5, the equivalent spinning reserve characteristics of hydrogen storage and the load-side demand response mechanism cause the system to retain a large RE consumption capacity. The total system cost shows a decreasing trend with an increase in RE penetration.

5.5. Impact Analysis of Carbon Emissions Trading Price on Carbon Emissions and Scheduling Results

The following two methods are set up:

Model 1: Carbon emission trading (CET) costs are not considered.

Model 2: CET costs are fully considered and incorporated into the total system cost.

The two models have the same residual parameters. The scheduling results are listed in

Table 8. Comparison results of different models.

Models	Comparison Results
	CET Cost/¥	Carbon Emissions/t	System Operation Cost/¥	Total Cost/¥
Model 1	-	2556	837,564.2	837,564.2
Model 2	116,490	1059	796,787.4	913,277.4

.

From Table 8, the overall economic cost increases under the influence of the CET policy because Model 2 considers the CET cost. However, the carbon emissions from coal-fired units are significantly reduced. The introduction of CET policy can effectively reduce environmental pollution. In addition, the sensitivity of the CET price λ to carbon emissions and CET costs is analyzed, as shown in Figure 14.

As shown by the blue curve in Figure 14, as the CET price λ rises, the carbon emissions of conventional energy units become progressively smaller and finally remain at the minimum limit. This is because a rise in the CET price λ will induce power generators to reduce carbon emissions and curb the rise in the cost of carbon emissions. When the CET price λ exceeds 120 ¥/t, the carbon emissions become insensitive to changes in the trading price because carbon emissions from power generation have reached the minimum value. The final CET cost is shown by the red curve; as the CET price λ rises, the CET cost decreases to the minimum value and then increases. It reaches the minimum value of CET cost at a CET price of 110 ¥/t, that is, the best economic result. To achieve the best low carbon performance, the CET price can be increased to 120 ¥/t. At this point, the carbon emissions reach the minimum limit. Therefore, the CET price λ can be used as a weight to balance system economy and low-carbon environmental protection in order to maintain sustainable operation of the system under a low-carbon economy.

5.6. Other Index Sensitivity Analysis of Source-Storage-Load Flexible Scheduling Strategy under Different Scenarios

This section explores the impact of different strategies on system reliability, harmonics, voltage, and load balance under different scenarios. The system reliability assessment is expressed in terms of EENS (Expected Energy Not Supplied). The system harmonic assessment is expressed in terms of THD (Total Harmonic Distortion). The aggregated voltage deviation and load balancing index are used to assess the system voltage condition and load balance. The results of sensitivity analysis under different scenarios are listed in

Table 9. The results of sensitivity analysis under different scenarios.

Index	Different Scenarios
	S1	S2	S3	S4	S5
EENS (MWh)	7.44	7.76	8.01	7.56	7.24
THD (%)	0.58	0.61	0.64	0.64	0.66
Aggregated voltage deviation rate (%)	1.13	1.19	1.24	1.23	1.23
Load balance rate (%)	98.24	99.61	99.68	99.62	98.76

.

In S1–S3, the EENS indicator value increases as its RE consumption gradually rises, and the volatility and randomness of RE causes load-shedding conditions in some periods. In S4, the equivalent rotational characteristics of the hydrogen storage system are considered. It has adequate system back-up capacity in conjunction with the flexible carbon capture system. In S5, during peak hours, the load is reduced as the load-side demand response is considered. The peak load demand is reduced to some extent, which ultimately leads to a reduction in the cut load power of 0.32 MWh.

THD (Total Harmonic Distortion) is directly related to the share of RE generation in the power system. Because wind and photovoltaic power generation is related to wind speed and sunlight intensity, the associated volatility and randomness lead to a large proportion of grid harmonics. Therefore, the higher the share of RE generation, the higher the value of its THD index. In S1–S5, the proportion of RE consumption is increasing, and its THD index is also gradually increasing.

Voltage deviation is related to factors such as the diameter of the conductors used for power transmission, the distance to the supply, the percentage of electricity generated from RE sources, the method of voltage regulation, and the capacity of reactive power compensation. An increase in the proportion of RE consumption leads to increasing aggregated voltage rate deviation. In S4, the equivalent rotational characteristics of HSS are considered. It has enough spare capacity to balance the system voltage deviation. Therefore, the voltage deviation does not increase in S4 and S5.

Load balancing is the rational distribution of electricity between supply and demand. The unit G1 is in the load center area, its output is directly related to the load balance. From the output of G1 in Figure 8, it can be found that when the output of G1 is low, some lines will be overloaded. Therefore, the load balance of S1, 98.24%, is lower than the average of S2, S3, and S4. S5 has a relatively unbalanced load distribution because of the shutdown of unit G2. The load balance rate in S5 is 98.76%.

6. Conclusions and Future Work

6.1. Conclusions

A source-storage-load flexible scheduling strategy is proposed considering the coordination of the spinning reserve and time-shift characteristics of FCCS and HSS. This provides a new basis for RE consumption and low-carbon operations. The following conclusions are obtained through simulations under different scenarios.

• The introduction of the source-side carbon capture unit (S2) reduces the carbon dioxide emissions in the traditional scenario system by 1425 t. Meanwhile, due to the storage tank, carbon capture and carbon emissions can be decoupled and operated, and the new energy grid-connected quantity increases by 90.4 MWh, while the total cost is reduced by 11.44%.
• The coordinated operation of the carbon capture unit with a hydrogen storage system increases the range of regulation of the carbon capture unit for spinning reserve requirements. It also reduces the power loss in the hydrogen storage energy time-shift by 29.22%. The efficiency of energy utilization is improved, and the amount of renewable energy forgone is significantly reduced by 96.3 MWh. In addition, the reserve cost of the hydrogen energy storage system is lower than that of the thermal power unit. Compared with S3, the reserve cost is reduced by 16.78%, and the total cost is reduced by $5317.30.
• The coordinated scheduling of source, load, and storage after considering the demand response mechanism can further improve the low-carbon and economic operation of the system. Simultaneously, with the increasing scale of RE installations, source-load-storage cooperative scheduling has more advantages in terms of economy and RE consumption.
• In addition, when the CST price is maintained at 110 ¥/t, the total carbon emission cost of the system is minimized. The CST price can be used as a weight to balance the economic and low-carbon features of the system, in order to maintain a low-carbon economy and sustainable operation of the system.

6.2. Limitations and Future Work

In this study, the coordinated operation characteristics of FCCS and HSS are analyzed, and an optimal economic scheduling strategy is proposed. However, there are some limitations in the examined scenarios, and future research can be improved as follows:

• This study does not consider the impact of hydrogen storage efficiency improvement on the optimal scheduling results after upgrading the hydrogen storage technology. Hydrogen production technology and FC power generation technology may make significant breakthroughs in HSS. Future work can focus on a simulation analysis of the optimal scheduling results after hydrogen storage system efficiency is improved.
• This study has not explored the optimal capacity allocation for HSS. The next step is to study the optimal capacity calculation method of HSS under the optimal economic scheduling model, and analyze the impact of different installed capacities of hydrogen storage on the scheduling results [32].
• This study takes the lowest operating cost as the objective function, and does not consider the influence of investment cost and expected return in the range of the entire life cycle cost. The next step is to analyze the optimal economic scheduling results with maximizing the revenue from “wind-solar-thermal-storage” bundled operations [33].