Comparison of Different Coupling Modes between the Power System and the Hydrogen System Based on a Power–Hydrogen Coordinated Planning Optimization Model

Abstract

Hydrogen is receiving unprecedented momentum and is projected to meet a sizable share of the final energy demand in the future. The coupling between the hydrogen and power systems can help integrate volatile renewable energy, reduce curtailment, and realize long-term energy storage. Different coupling modes are being proposed, yet quantitative comparisons are absent. To fill this gap, this paper firstly summarizes the system composition of potential power–hydrogen coupling modes and analyzes their advantages and disadvantages. Then, a model for power–hydrogen coordinated planning optimization is proposed for quantitative analysis. Southern Xinjiang is chosen as a representative of the northwestern area in China, which has plentiful renewable resources and a relatively small local load at present, for a case study. Through result analysis, it is found that the mode of power–hydrogen coupling at the source side, either for in situ utilization or for long-distance transport via pipelines, shows better economic competitiveness. The discussion provides suggestions and a reference to policymakers for formulating infrastructure or industry plans in advance to better accommodate the rapidly developing renewable energy.

1. Introduction

Green hydrogen, referring to the hydrogen produced through water electrolysis powered by renewable energy, has been regarded as the missing piece of the clean energy puzzle [1]. It is estimated that by 2060, hydrogen will account for more than 15 percent of the final energy consumption [2], which also proves its irreplaceable role in the clean energy system. The reasons are manifold. Firstly, and most importantly, green hydrogen can serve as a high-grade heat source, raw materials, and fuels to assist in the deep decarbonization of many hard-to-abate sectors that heavily rely on fossil fuels, including metallurgy, high-quality heating, aviation, etc. [3,4]. In metallurgy, green hydrogen can be used as a reducing agent, which replaces coal and natural gas to reduce the carbon emission during the process of steel production. In the transportation sector, fuel cell vehicles are more suitable for heavy-load and long-distance transport, which is an important supplement to electric vehicles that are more suitable for short distances. However, when considering the carbon emissions of all equipment used in hydrogen production, blue hydrogen might even be cleaner than green hydrogen [5]. Secondly, with water electrolysis being the key link of the power-to-X technologies, green hydrogen can serve as an important intermediate to promote the coupling between different sectors [6]. For instance, electrolytic hydrogen produced from variable renewable energy (VRE) can be used to synthesize ammonia, methanol, and other chemicals, which promote the coupling between the power sector and the industry sector.

The high expectation of green hydrogen enhances the close interrelations between hydrogen and power systems. In China, renewable power will account for more than 90% of the total installed capacity of power sources [7], which could effectively reduce the adverse environmental effects of fossil fuels and address climate change [8,9], but also poses great challenges to the safe and stable operation of the power system. Hydrogen, as an energy carrier that can be easily and efficiently stored on a large scale, can serve as flexible energy storage for the power system. To be more specific, water electrolysis can function as a flexible load, while along with fuel cells and hydrogen-fueled gas turbines, it can provide long-term flexibility [10,11,12]. However, the main obstacle that limits the widespread use is the relatively high cost [13,14].

Various studies have, to some extent, studied power–hydrogen coupled systems. Concerning planning optimization, it is found that considering generation and transmission planning in a power–hydrogen integrated framework results in a reduction of total system costs [15,16]. Reference [17] develops a generalized framework to co-optimize infrastructure investments across the electricity and hydrogen supply chains and discovers that power–hydrogen coupling could effectively lower the cost of energy system decarbonization. Reference [18] studies the cross-regional electricity and hydrogen deployment in China using a coordinated optimization method. Reference [19] designs the low-carbon transition pathway on a regional scale considering electricity-hydrogen synergy. From a microscopic perspective, References [20,21] study the stand-alone microgrid powered by hydrogen, which proves that hydrogen could enrich flexibility resources. However, the studies mentioned above did not fully exploit the advantages of hydrogen energy storage, especially for long term storage. Reference [22] proposes a planning model for an electricity-hydrogen integrated energy system which has taken seasonal hydrogen storage into account.

The operation and scheduling strategy of power–hydrogen coupled systems has also been the focus of research. References [23,24,25] study the operation of power systems and hydrogen production considering technical and economic specifications on different types of electrolysers and operating modes. Reference [26] establishes a power to hydrogen and heat model to realize the optimal operation of a system, combining an active distribution network and a district heat network. Reference [27] proposes an optimal dispatch strategy for an integrated energy system with hydrogen, which takes the price–load uncertainties and risk-averse factors into account. Furthermore, References [28,29] take the electricity and hydrogen markets into consideration.

The above-mentioned works have made substantial progress in power–hydrogen coordinated planning and operational optimization. However, the penetration of renewable energy is not given enough consideration, and the different modes of power–hydrogen coupling are rarely categorized and analyzed. This paper aims at carrying out a comprehensive comparison of the different power–hydrogen coupling modes. Firstly, the different coupling modes are introduced. Their respective advantages and disadvantages are analyzed. Then, the power–hydrogen coordinated planning optimization model embedding with the function of production simulation is proposed, which lays a solid foundation for the case study. Based on the case study of Southern Xinjiang, the operational and economic characteristics of different power–hydrogen coupling modes are compared and discussed, with a hope to provide reference and guidance for related companies and policymakers in formulating strategies for future development.

2. Overview of Different Power–Hydrogen Coupling Modes

In general, there are two power–hydrogen coupling modes, power–hydrogen coupling at the power source side and power–hydrogen coupling at the power load side.

2.1. Power–Hydrogen Coupling at the Load Side

In the mode of power–hydrogen coupling at the load side, as shown in Figure 1, centralized renewable power generation bases are constructed at places rich in renewable energy resources, such as the northern and western regions in China. The generated power is thus transmitted to the central and eastern regions through ultra-high-voltage (UHV) transmission lines. Then, at the load side, hydrogen is produced for either local utilization or hydrogen-fueled turbines, which can serve as voltage support or capacity reserve for the power grid at the load center.

The main advantages of power–hydrogen coupling at the load side are as follows.

First, hydrogen production factories can flexibly choose the site. For the mode of power–hydrogen coupling at the source side, most hydrogen production factories would choose to install PV or wind power stations by themselves for a relatively cheap power supply. Utility-scale PV and wind power stations are usually built away from the city, which is also far from the demand center. As for the mode of power–hydrogen coupling at the load side, the electrolysers use electricity from the grid. By building factories closer to regions with a large hydrogen demand, the costs of conversion, storage, and long-distance transport of hydrogen can be saved. Furthermore, the potential safety risk caused by large-scale hydrogen transport over long distances can be avoided.

Second, it could make full usage of UHV power transmission lines, as a means of large-scale optimized allocation of renewable energy. From previous research [30,31,32], when the transport distance increases from 400 km to 1000 km, the cost of hydrogen transport via pipelines increases from 0.56 USD/kg to 1.26 USD/kg, while that of electricity transmission via UHV lines increases from 0.62 × 10⁻² USD/kWh to 1.10 × 10⁻² USD/kWh. After conversion to the caloric value, the unit cost of power transmission is around one-fifth to one-fourth of that for hydrogen transport. Therefore, for energy transmission, UHV lines are economically more competitive than hydrogen pipelines.

Third, the peak-valley price differences for electricity are larger near the load center. This could provide economic compensation to some extent for losses in the energy conversion processes of power–hydrogen–power.

However, this mode also has some potential drawbacks, which are listed as follows.

First, a comparatively lower capital investment for UHV lines does not necessarily result in a lower cost for energy transmission in total. Due to the high volatility of renewable energy, it has to be bundled with coal power, hydropower, and power from electro-chemical storage and pumped storage to form a relatively smooth power curve for UHV lines. This extra investment in adjustable power sources and energy storage cells might result in a high system cost.

Second, it cannot make full utilization of the outstanding dynamic response characteristics of electrolysers since the high volatility of renewable energy is already smoothed by adjustable power sources and energy storage cells at the source side. The power transmitted to the load side is relatively stable, which limits the performances of electrolysers to provide grid-balancing services and absorb the fluctuation of VRE at the source side.

Third, power–hydrogen–power conversion could result in low efficiency and large energy losses. The efficiency of alkaline and proton exchange membrane electrolysers is no larger than 70%. The efficiency of hydrogen-fueled gas turbines is around 50%. The overall efficiency of power–hydrogen–power is only 35%, and 65% of the energy is wasted during the energy conversion processes.

2.2. Power–Hydrogen Coupling at the Source Side

To be more specific, the mode of power–hydrogen coupling at the source side can be divided into three sub-modes: source side coupling for power transmission, source side coupling for hydrogen transport, and source side coupling for in situ utilization.

2.2.1. Power–Hydrogen Coupling at the Source Side for Power Transmission

In this mode, electrolysers, hydrogen storage tanks, and hydrogen-fueled gas turbines are deployed at the source side, aiming at reducing VRE curtailment and supporting the stable power transmission over long distances using UHV lines, as shown in Figure 2. The volatile renewable power can be utilized by electrolysers due to its excellent dynamic responses. The produced hydrogen can either be used for local consumers, or for gas turbines to generate power again. It can serve as an important supplement to electro-chemical energy storage, pumped storage, and regulatory power sources, which can provide grid-balancing services and ensure a relatively stable power transmission by UHV lines.

The main advantage of this mode is that a relatively high utilization rate of both the renewable energy and the power transmission lines can be guaranteed. In the future, coal and gas power plants will be gradually phased out. Electrolysers, hydrogen storage, and hydrogen-fueled gas turbines can serve as an important means of peak regulation and providing grid-balancing services. When the output of renewable power is large, electrolysers operate at full power, and hydrogen is stored in tanks. When the output of renewable power is small, gas turbines fueled by hydrogen are turned on. This could ensure a relatively stable power transmission through UHV lines, whose utilization rate is improved as much as possible under the premise of ensuring the high utilization rate of renewable energy.

The potential problem of this mode is the low efficiency and large energy loss caused by the power–hydrogen–power conversion processes, similar to the mode of power–hydrogen coupling at the load side.

2.2.2. Power–Hydrogen Coupling at the Source Side for Hydrogen Transport

The schematic diagram for this mode is shown in Figure 3. In this mode, hydrogen produced from renewable power at the source side is transmitted over long distances by pipelines to the demand center.

The advantage of this mode is to alleviate the pressure of the power grid by transporting renewable power over long distances as hydrogen. This could further support the large-scale development and utilization of renewable energy. Unlike electric power, hydrogen has a higher tolerance for fluctuations since its supply and demand does not need to be instantly balanced. Besides, pipelines can also serve as a means of storage.

The potential issue is that hydrogen pipelines are quite expensive and are still in the initial stage of construction in China, which would limit its application in the short term. At present, the most commonly used method for hydrogen transport is with trucks, which is only suitable for small-scale transport over short distances. Long-distance hydrogen pipelines are still in the phase of design. The cost is as high as 1,000,000 USD/km. By blending hydrogen in natural gas pipelines, the cost could be reduced. However, the maximum percentage should not exceed 10%, which limits the scale of hydrogen transport.

2.2.3. Power–Hydrogen Coupling at the Source Side for In Situ Utilization

In this mode, hydrogen produced from renewable power at the source side is used locally, either for chemical plants or for fuel cell cars.

The main advantage of this mode is its higher efficiency and lower cost. Since there is no need to convert hydrogen back to power, the energy losses in energy conversion processes can be saved. However, this mode is extremely dependent on the local demand for hydrogen. If there is not enough demand, hydrogen produced from renewable power will inevitably be transported to other areas. Therefore, in regions with abundant renewable energy, related industries that use hydrogen as either feedstock or fuels should be planned, for instance, chemical, steel, fuel cell cars, etc.

3. A Power–Hydrogen Coordinated Planning Optimization Model

In order to quantitatively compare the above-mentioned modes, a power–hydrogen coordinated planning optimization model is proposed. The basic principle of this model is shown in Figure 4. This model could optimize the configuration of both the power system and the hydrogen system, simultaneously. In the power system, various generation sources (wind power, PV, and hydropower), transmission lines, and energy storage units (electro-chemical energy storage cells and pumped storage) are taken into consideration. Similarly, in the hydrogen system, electrolysers, hydrogen/natural gas pipelines, hydrogen-fueled turbines, and hydrogen storage tanks are considered.

Besides, this model is also embedded with the function of production simulation, which could optimize the output of each equipment on an hourly basis. Based on the optimized construction and operation plan, economic parameters, including total investment, levelized cost of investment, and cost per unit, can be calculated. Furthermore, the model could also assist in the analysis of operational characteristics and the comparison of the utilization rates of key equipment in different power–hydrogen coupling modes.

3.1. Objective Function

The objective function is to minimize the total expenditure, which is shown as follows. Expenditures on the generation, transmission, and storage equipment of both power and hydrogen are considered.

$$\begin{array}{ll}\hfill \mathrm{Min} \mathrm{F}=& ({\displaystyle \sum _{r\in {\mathrm{\Omega }}_r}}{\displaystyle \sum _{g\in {\mathrm{\Omega }}_g}}{Q}_{g,r}\cdot C_{g,r}+{\displaystyle \sum _{r\in {\mathrm{\Omega }}_r}}{\displaystyle \sum _{s\in {\mathrm{\Omega }}_s}}{Q}_{s,r}\cdot T_{s,r}\cdot C_{s,r}+{\displaystyle \sum _{i,j\in {\mathrm{\Omega }}_r}}{Q}_{ij}\cdot D_{ij}\cdot C_{ij}+{\displaystyle \sum _{r\in {\mathrm{\Omega }}_r}}{\displaystyle \sum _{g\in {\mathrm{\Omega }}_g}}{Q}_{g,r}\cdot F_{g,r}\cdot U_{g,r}\cdot C_{\mathrm{c}})\hfill \\ & +\left[{\displaystyle \sum _{i,j\in {\mathrm{\Omega }}_r}}{D}_{ij}\cdot H_{ij}+{\displaystyle \sum _{r\in {\mathrm{\Omega }}_r}}\left(Q_{e,r}\cdot C_{e,r}+Q_{s,r}\cdot C_{s,r}+Q_{t,r}\cdot C_{t,r}\right)\right]\hfill \end{array}$$

(1)

where Ω_r is the set of regions, including the source side and the load side, Ω_g is the set of power sources, including PV, wind power, hydropower, coal power, gas power, etc., Q_g,r and C_g,r are the capacity and cost per unit capacity for the gth type of power source in the rth region, and Ω_s is the set of energy storage units, including electro-chemical energy storage cells and pumped storage. Q_s,r, T_s,r, and C_s,r are the capacity, working time duration at the maximum charging capacity, and cost per unit capacity per hour of the sth type of energy storage in the rth region, respectively, and Q_ij, D_ij, and C_ij are the capacity, distance, and cost per unit capacity per distance of power transmission lines from the ith region to the jth region, respectively. F_g,r and U_g,r are the carbon emission factor and the utilization hour of the gth type of power source in the rth region, C_c is the cost per unit of carbon emission, D_ij and H_ij are the distance and cost per unit distance of hydrogen pipelines from the ith region to the jth region, Q_e,r and C_e,r are the capacity and cost per unit capacity for the electrolysers in the rth region, Q_s,r and C_s,r are the capacity and cost per unit capacity for hydrogen storage tanks, and Q_t,r and C_t,r are the capacity and cost per unit capacity for hydrogen-fueled gas turbines.

3.2. Constraints

3.2.1. Balance of Electric Power

The production and consumption of power is described as follows:

$${\displaystyle \sum _{g\in {\mathrm{\Omega }}_g}{p}_{g,r}^t}+{\displaystyle \sum _{s\in {\mathrm{\Omega }}_{sr}}{p}_{s,r}^t}=L_{e,r}^t+p_{e,r}^t+{\displaystyle \sum _{i\in {\mathrm{\Omega }}_r}{p}_{ri}^t}\hspace{1em}\left(t=1,2,3\dots T\right)$$

(2)

where $p_{g,r}^t$ is the output of the gth type of power source at the tth hour in the rth region, including wind power, PV, hydropower, coal power, gas power, and hydrogen-fueled power generation units, $p_{s,r}^t$ is the net power from the sth type of energy storage units, $L_{e,r}^t$ is the local demand for electric power, except for electrolysers, $p_{e,r}^t$ is the power demand of electrolysers, and $p_{ri}^t$ is the power to be transmitted to the ith region by UHV lines.

The spinning reserve is as follows:

$${\displaystyle \sum _{g\in {\mathrm{\Omega }}_g}{Q}_{g,r}}+{\displaystyle \sum _{s\in {\mathrm{\Omega }}_s}{Q}_{s,r}}-\left({\displaystyle \sum _{g\in {\mathrm{\Omega }}_g}{p}_{g,r}^t}+{\displaystyle \sum _{s\in {\mathrm{\Omega }}_s,r\in {\mathrm{\Omega }}_r}{p}_{s,r}^t}\right)={\displaystyle \sum _{g\in {\mathrm{\Omega }}_g}{\pi }_g\cdot Q_{g,r}}\hspace{1em}\left(t=1,2,3\dots T\right)$$

(3)

where ${\pi }_g$ is the reserve coefficient of the gth type of power source.

3.2.2. Balance of Hydrogen

The production and consumption of hydrogen is described as follows:

$$h_{e,r}^t+h_{s,r}^t=L_{h,r}^t+h_{t,r}^t+{\displaystyle \sum _{i\in {\mathrm{\Omega }}_r}{h}_{ri}^t} \left(t=1,2,3\dots T\right)$$

(4)

where $h_{e,r}^t$ is the mass of hydrogen produced from electrolysers at the tth hour, $h_{s,r}^t$ is the mass of hydrogen retrieved from hydrogen storage tanks, $L_{h,r}^t$ is the hourly local demand for hydrogen, $h_{t,r}^t$ is the hydrogen used for hydrogen-fueled gas turbines, and $h_{ri}^t$ refers to the mass of hydrogen pressed into natural gas pipelines or hydrogen pipelines to be transported to the ith region.

3.2.3. Energy Conversion Constraints

In the model, there are two types of equipment involved in the energy conversion processes between power and hydrogen: electrolysers, which convert power to hydrogen, and hydrogen-fueled turbines, which can convert hydrogen back to power.

• Power converted to hydrogen through electrolysers:

  $$p_{e,r}^t\cdot {\eta }_{eh}\cdot {\lambda }_{eh}=h_{e,r}^t \left(t=1,2,3\dots T\right)$$

  (5)

  where ${\eta }_{eh}$ is the efficiency of the electrolysers, and ${\lambda }_{eh}$ is the caloric conversion coefficient with a value of 0.0253 kg/kWh, which is calculated by dividing the caloric value of electricity (3600 kJ/kWh) by that of hydrogen (142,500 kJ/kg).
• Hydrogen converted to power through hydrogen-fueled turbines:

  $$h_{t,r}^t\cdot {\eta }_{he}=p_{t,r}^t\cdot {\lambda }_{eh} \left(t=1,2,3\dots T\right)$$

  (6)

  where ${\eta }_{he}$ is the efficiency of the hydrogen-fueled turbines.

3.2.4. Energy Storage Constraints

In the model, electro-chemical energy storage units, pumped storage units, and hydrogen storage tanks are taken into consideration. Since their operational characteristics are quite similar, only constraints concerning electro-chemical energy storage units are shown, as follows, for illustration.

Firstly, within a certain period of time, the sum of power used for charging multiplied by the efficiency, η_s, is equal to the sum of the discharging power, which could depict the energy loss during the charging and discharging processes:

$${\displaystyle \sum _{t\in {\mathrm{\Omega }}_t}{\eta }_s\cdot p_{s\mathrm{in},r}^t}={\displaystyle \sum _{t\in {\mathrm{\Omega }}_t}{p}_{s\mathrm{out},r}^t}$$

(7)

where ${\eta }_s$ is the efficiency of the sth type of energy storage, $p_{s\mathrm{in},r}^t$ is the charging power at the tth hour, and $p_{s\mathrm{out},r}^t$ is the discharging power at the tth hour.

Secondly, the state of charge between adjacent hours is as follows:

$$R_s^{t+1}=R_s^t+{\eta }_s\cdot p_{s\mathrm{in},r}^t-p_{s\mathrm{out},r}^t \left(t=1,2,3\dots T-1\right)$$

(8)

where $R_s^t$ is the state of charge at the tth hour.

The charging power, $p_{s\mathrm{in},r}^t$, and the discharging power, $p_{s\mathrm{out},r}^t$, should be within the limitation of its maximum capacity, Q_s.

$$0\le p_{s\mathrm{in},r}^t,p_{s\mathrm{out},r}^t\le Q_s \left(t=1,2,3\dots T-1\right)$$

(9)

The maximum-energy state of storage, $R_s^t$, should not exceed its maximum capacity.

$$0\le R_s^t\le Q_s\cdot T_s \left(t=1,2,3\dots T\right)$$

(10)

where Hesc is the working time duration at the maximum charging capacity.

Furthermore, the relationship between $p_{s,r}^t$, $p_{s\mathrm{out},r}^t$, and $p_{s\mathrm{in},r}^t$ is as follows:

$$p_{s,r}^t=p_{s\mathrm{out},r}^t-p_{s\mathrm{in},r}^t \left(t=1,2,3\dots T-1\right)$$

(11)

4. Case Study: The Power–Hydrogen Coordinated Planning of Southern Xinjiang Area

In this section, a case study of power–hydrogen coordinated planning of the Southern Xinjiang area is carried out. Quantitative analysis of the different power–hydrogen coupling modes was conducted using the proposed model. Some background information concerning Southern Xinjiang is firstly introduced. Then, some basic settings, including the design of different schemes, scenarios, and key values of techno-economic parameters, are presented. Later, the results of different modes are systematically compared and analyzed.

4.1. Southern Xinjiang Area

Southern Xinjiang is one of the demonstration areas for power systems with high penetration of renewable energy due to its plentiful resources, especially solar resources. According to the principle of separating intra-provincial usage and inter-provincial delivery, the energy basis in Southwestern Xinjiang is used for meeting the local energy demand, while that in Southeastern Xinjiang is used for long-distance energy transport to the central and southern areas in China.

In order to facilitate the comparison between different modes of power–hydrogen coupling, the energy basis in Southeastern Xinjiang was chosen as the main target of the case study. According to the planning of the energy basis, 237 million kilowatts of PV, 4 million kilowatts of wind power, and 3.6 million kilowatts of pumped storage will be built.

As for power export, it is planned that power will be transmitted to the central and southeastern area in China through UHV lines.

4.2. Basic Settings

4.2.1. Different Schemes Representing Different Power–Hydrogen Coupling Modes

Four schemes were designed for the comparison between different power–hydrogen coupling modes.

• Basic scheme (BS)

In this scheme, there are PV, wind power, hydropower, electro-chemical energy storage, and pumped storage at the source side. Through UHV lines, power is transmitted to the load side. With the help of electro-chemical energy storage cells, pumped storage, and hydropower sources, the power transmission curve is adjusted to a straight line, which can ensure a stable power transmission at any time.

• Scheme of power–hydrogen coupling at the load side (PHLS)

In this scheme, the system composition at the source side is the same as that in the BS. However, the power transmission curve is a “two-segment” curve, with a peak period at midday when the output of PV is relatively large.

At the load side, in addition to electricity users, electrolysers, hydrogen storage, and hydrogen-fueled gas turbines are taken into consideration. At midday, the excessive power is utilized by electrolyers. The produced hydrogen can be used either by hydrogen users, or by hydrogen turbines for power generation.

• Scheme of power–hydrogen coupling at the source side for power transmission (PHSS-PT)

In this scheme, along with electro-chemical energy storage, pumped storage, and hydropower sources, electrolysers, hydrogen storage, and hydrogen-fueled gas turbines are added at the source side to adjust the curve to a straight line for stable power transmission to the load side.

• Scheme of power–hydrogen coupling at the source side for hydrogen transport and in situ utilization (PHSS-HT)

In this scheme, electrolysers and hydrogen storage are added at the source side. Renewable power generated from PV, wind, and water turbines is used by the electrolysers for hydrogen production. Then, hydrogen is blended into west–east gas transmission pipelines at a rate of 5%, while the rest is consumed by local hydrogen users.

In all schemes, the utilization rate of renewable resources should be no less than 90%, and the equivalent utilization hours of UHV lines should be no less than 4000.

4.2.2. Scenarios

The volatility of renewable energy could significantly vary from week to week. In order to analyze the economic and operational characteristics of each scheme with different extents of volatility of the renewable energy, two typical scenarios were set.

The data of the per-unit output of PV and wind power (WP) shown in Figure 5 are real data from Southeastern Xinjiang. As shown in Section 4.1, the installation capacity of PV (240 million kilowatts) was much larger than that of WP (4 million kilowatts), which would cause more significant influences. Therefore, the two sets of per-unit output of PV were chosen first, with one representing weeks with low volatility and the other representing weeks with larger volatility. Then, the corresponding per-unit outputs of wind power on the same days were chosen.

4.2.3. Key Techno-Economic Parameters

The values of the key techno-economic parameters are listed in

Table 1. Values of key techno-economic parameters ($ = USD).

Parameter	Value	Unit
CAPEX for wind power stations	700.0	$/kW
OPEX for wind power stations	34.4	$/(kW-yr)
CAPEX for PV stations	285.0	$/kW
OPEX for PV stations	2.8	$/(kW-yr)
CAPEX for pumped storage	2202.0	$/kW
OPEX for pumped storage	18.1	$/(kW-yr)
CAPEX for electro-chemical energy storage	247.0	$/kWh
OPEX for electro-chemical energy storage	5.9	$/(kWh-yr)
CAPEX for electrolysers	250.0	$/kW
OPEX for hydrogen production through electrolysis	0.2	$/kg
Efficiency for power–hydrogen conversion	50.0	kg/kWh
Dynamic responding speed for electroylsers	100%	S−1
CAPEX for hydrogen-fueled gas turbines	1562.0	$/kW
OPEX for hydrogen-fueled gas turbines	31.2	$/(kW-yr)
Efficiency for hydrogen–power conversion of hydrogen turbines	60%	-
Ramping up/down speed for hydrogen turbines	100%	h−1
CAPEX for hydrogen storage tanks	450.0	$/kg
OPEX for hydrogen storage tanks	18.0	$/(kg-yr)
CAPEX for retrofitted hydrogen pipelines	600,000	$/km
OPEX for retrofitted hydrogen pipelines	140,000	$/yr

, which were determined based on reports published by international research institutes [14,33,34,35,36,37,38,39,40,41,42,43].

4.3. Result Analysis

4.3.1. Total Expenditure

Firstly, the construction plans and the total expenditures are shown in

Table 2. The optimized construction plan for different schemes under different scenarios.

Capacity	Scenario 1				Scenario 2
	BS	PHLS	PHSS-PT	PHSS-HT	BS	PHLS	PHSS-PT	PHSS-HT
Electro-chemical energy storage/(×106 kW)	122.68	85.12	66.54	0	137.33	116.16	55.11	0
Electrolysers/(×106 kW)	0	37.30	65.65	161.24	0	36.98	76.44	158.41
Hydrogen-fueled gas turbines/(×106 kW)	0	2.46	7.14	0	0	2.44	8.42	0
Hydrogen storage/(×106 kg·h−1)	0	0.37	0.68	1.86	0	0.37	0.71	1.70

and Figure 6, respectively.

In BS, only electro-chemical energy storage could serve as a flexibility resource and contribute to grid-balancing. In PHLS and PHHS-PT, electrolysers, hydrogen-fueled gas turbines, and hydrogen storage were added, which assisted in grid-balancing and alleviated the pressure for electro-chemical energy storage. In PHSS-HT, a steady hydrogen delivery was realized with the help of electrolysers and hydrogen storage.

Compared to the BS, the addition of electrolysers, hydrogen-fueled gas turbines, and hydrogen storage in PHLS, PHHS-PT, and PHSS-HT resulted in an obvious decrease of capacity for electro-chemical energy storage cells. Electrolysers, hydrogen-fueled gas turbines, and hydrogen storage could also function as valuable flexibility resources.

From Figure 6, in both scenarios, the capital expenditure and total expenditure for the BS were the largest, followed by PHLS, PHSS-PT, and PHSS-HT. In the first scenario, the capital expenditure was reduced by 18.94%, 29.01%, and 70.18%, respectively, for PHLS, PHSS-PT, and PHSS-HT, compared to that of the BS. As for total expenditure, it was reduced by 15.77%, 27.61%, and 64.58%, respectively, for PHLS, PHSS-PT, and PHSS-HT. Therefore, power–hydrogen coupling at the source side is more economical than that at the load side, in general. Furthermore, the total expenditure for PHSS-HT was the lowest. After converting power to hydrogen, it is cheaper to use it, either locally or to be transported to the load center via pipelines, than to convert it back to power to be transmitted through long distances via UHV lines. This is because UHV lines have high requirements for steady power transmission, and much investment is needed for flexibility resources, such as electro-chemical energy storage cells, hydrogen-fueled gas turbines, etc.

In the second scenario with larger volatility of renewable energy, the capacity of electro-chemical energy storage showed an obvious increase in the BS and PHLS, and accordingly, an increase of 8.64% and 22.08% for total expenditure, respectively. In these two schemes, only electro-chemical energy storage and pumped storage serve as flexibility resources at the source side. A greater volatility of renewable energy results in a requirement for more flexibility resources. In PHHS-PT, the capacity of electro-chemical energy storage showed a slight decrease, while that for electrolysers and hydrogen-fueled gas turbines showed a slight increase. The total expenditure was 38.65% smaller than that of BS. Furthermore, the total expenditure for PHHS-HT was 68.54% smaller than that of BS. In scenarios with larger volatility, power–hydrogen coupling at the source side showed better economic competitiveness.

4.3.2. Cost Per Unit Energy

Besides total expenditure, the four schemes should also be analyzed and compared from a microscopic perspective.

The cost for electricity M_t is as follows:

$$M_t=I_t/\left[{\displaystyle \sum L_e^t}+{\displaystyle \sum L_h^t}/\left({\eta }_{eh}\cdot {\lambda }_{eh}\right)\right]$$

(12)

where I_t is the total expenditure, which includes not only capital expenditure, but also operation, maintenance, and replacement costs, and ${\displaystyle \sum L_e^t}$ and ${\displaystyle \sum L_h^t}$ are the total amounts of electricity and hydrogen consumed by users in the complete lifecycle.

By deducting the investment cost of wind power and PV in the denominator, the system cost for electricity M_s is obtained.

The cost per caloric value, M_c, is as follows:

$$M_t=I_t/\left[{\displaystyle \sum L_e^t}\cdot c_e+{\displaystyle \sum L_h^t\cdot c_h}\right]$$

(13)

where c_e and c_h are the caloric values for electricity and hydrogen, respectively.

The M_t, M_s, and M_c for different schemes under both scenarios are shown in Figure 7. In both scenarios, PHSS-PT had the smallest total cost and system cost for electricity, followed by PHLS and the BS. In S1, the system cost for PHSS-PT was 0.06 USD·kWh⁻¹, while that for PHLS and the BS was 0.06 USD·kWh⁻¹ and 0.07 USD·kWh⁻¹, respectively. In S2, the system cost for PHSS-PT was 0.05 USD·kWh⁻¹, while that for PHLS and the BS was around 0.08 USD·kWh⁻¹. The results have again proven the economic advantage of power–hydrogen coupling at the source side.

As for the cost per unit caloric value, PHHS-HT had the lowest M_c, which was less than half of that for the BS, PHLS, and PHSS-PT. This is because by avoiding the process of converting hydrogen back to power, energy loss was also avoided, which could help reduce the cost per caloric value to a large extent.

4.3.3. Operation Characteristics of Equipment

The operation characteristics of the main equipment in the four schemes in S2 are shown in Figure 8 and Figure 9. Figure 8 demonstrates electric power, which includes the output of power resources, the net output of energy storage units, the power consumed by electrolysers, and the power transmitted by UHV lines. Figure 9 shows hydrogen production and consumption.

In the BS, the electro-chemical energy storage and pumped storage units were charged, usually between 11:00 and 20:00, when the output of PV is large, and discharged between 20:00 and 11:00 the next day, in order to guarantee a stable power transmission through UHV lines. In PHLS, the power transmission follows a “two-segment” curve, with a peak period between 12:00 and 18:00, according to the settings. The pressure for peak modulation was alleviated compared to the BS, which reduced the capacity needed for electro-chemical energy storage units. However, on the fifth day in the week, the maximum output of PV was only around one-third of its capacity, power shortage occurred during 12:00 to 16:00, and electro-chemical energy storage units were discharged to fill the gap. Since electrolysers are installed at the load side, their advantages of flexible operation when coping with the volatility of renewable energy at the source side are not fully utilized.

At the load side, the hydrogen generation and consumption are shown in Figure 9a. Electrolysers are turned on between 12:00 and 18:00 to produce hydrogen, which is then stored in hydrogen tanks. During 18:00 to 12:00 the next day, the stored hydrogen is used by hydrogen-fueled gas turbines to generate power again, which equivalently realizes a stable power transmission from the source side to the consumers, similar to that in the BS.

In PHSS-PT, electrolysers are operated in a flexible way. The power of electrolysers is adjusted timely according to the output of PV, even with sharp peaks occurring from time to time, which fully demonstrates the outstanding dynamic responding capabilities of electrolysers, as shown in in Figure 8c. Furthermore, electrolysers, along with hydrogen storage tanks and hydrogen turbines, can function as long-term energy storage, which is an important supplement to the electro-chemical energy storage units. As shown in Figure 9b, on the fifth day, there was no excess renewable power for hydrogen production. By using hydrogen from storage tanks, which is accumulated from the previous days, hydrogen turbines are turned on during the nighttime to avoid power shortage.

In PHSS-HT, there is no regulating pressure to ensure a stable power transmission. Except for a small proportion of curtailment, the rest of the renewable power is used by electrolysers. From Figure 9c, the produced hydrogen were either used locally or pressed into pipelines to be transported to the central areas. In order to ensure a relatively stable hydrogen transmission, the excess hydrogen produced between 11:00 and 21:00 was firstly stored in tanks, and then pressed into the pipelines between 21:00 and 11:00 the next day.

4.3.4. Capacity Factor of Key Equipment

The capacity factors for key equipment in the four schemes are shown in

Table 3. The capacity factor for key equipment in different schemes under different scenarios.

Capacity Factor	Scenario 1				Scenario 2
	BS	PHLS	PHSS-PT	PHSS-HT	BS	PHLS	PHSS-PT	PHSS-HT
Electro-chemical energy storage	47.29%	50.11%	58.82%	-	37.32%	34.72%	60.09%	-
Electrolysers	-	25.00%	19.34%	31.63%	-	25.00%	16.42%	32.22%
Hydrogen storage	-	75.00%	44.37%	55.81%	-	75.00%	37.36%	54.83%
Hydrogen-fueled gas turbines	-	50.00%	47.19%	-	-	50.00%	44.71%	-

.

In S1, for electro-chemical energy storage units, its capacity factor was the largest in PHSS-PT, which was 8.71% larger than PHLS and 11.53% larger than the BS. For electrolysers, hydrogen storage tanks, and hydrogen-fueled gas turbines, the capacity factor was the largest in PHLS, followed by PHSS-HT and PHSS-PT.

In S2 with larger volatility of renewable energy, the capacity factors for all equipment were smaller than those in S1, in general, except for PHSS-HT, in which the capacity factor of electrolysers was 0.59% larger and that of hydrogen storage was 0.98% smaller. This also proves that PHSS-HT has good adaptability for high volatility of renewable energy, which could avoid compromising the capacity factor of key equipment.

4.4. Sensitivity Analysis

Due to the uncertainty of scalability and marketization for related technologies, a sensitivity analysis should be carried out to study the variation of the impacts of key parameters on the optimization results. Since the capacities of electro-chemical energy storage units, electrolysers, hydrogen-fueled turbines, and hydrogen storage units are variables in the optimization, their related economic parameters (CAPEX, O&M costs of equipment) were set as the objects of the sensitivity study. It is assumed that O&M costs vary accordingly with CAPEX. The optimization results, including capital expenditure, total expenditure, levelized cost of electricity, system cost of electricity, and cost per unit energy, are presented, as shown in Figure 10.

From Figure 10a, the cost of electro-chemical energy storage cells had a large influence on the optimization results. When it decreased by 40%, the capital and total expenditure decreased by 20.19% and 30.84%, respectively. The levelized cost of electricity and the system cost decreased by 21.48% and 27.18%, respectively. The cost per unit energy showed a 21.39% decrease. When the cost of electro-chemical energy storage cells increased by 40%, the increasing range of optimization results was comparatively smaller, mostly between 18% and 24%. The costs of electrolysers, hydrogen-fueled turbines, and hydrogen storage units had a relatively smaller impact on the optimization results, as shown in Figure 10b–d. The variation range of capital, total expenditure, levelized, and system costs of electricity, and the cost per unit energy, were all less than 5%.

5. Discussion

Through the comparison between the different coupling modes using a case study, it was easy to discover the following patterns of power–hydrogen coupling.

Firstly, power–hydrogen coupling can assist in the flexible operation of the power system and help in reducing VRE curtailment. On the one hand, electrolysers can serve as an adjustable load with a fast ramping-up/down speed, which can provide timely grid-balancing services. On the other hand, electrolysers, hydrogen storage, and hydrogen-fueled gas turbines together can function as long-term energy storage, which is an important supplement to electro-chemical energy storage units.

Secondly, power–hydrogen coupling is more economically competitive than no coupling at all. Power–hydrogen coupling at the source side is more economical than that at the load side. In S1, the total investment cost was reduced by 14.44%, 23.82%, and 59.48%, in PHLS, PHSS-PT, and PHSS-HT, respectively, compared to the BS. The cost per caloric value was reduced by 8.63%, 16.69%, and 55.74% in PHLS, PHSS-PT, and PHSS-HT, respectively, compared to that of the BS.

Thirdly, in scenarios with larger volatility of renewable energy, the economic competitiveness of PHSS was more obvious. In S2, the total investment cost was reduced by 35.87% and 63.95% in PHSS-PT and PHSS-HT, respectively, compared to BS. The magnitude of decline was larger than that in S1, which proved that power–hydrogen coupling at the source side had better adaptability for volatile renewable energy.

After producing hydrogen at the source side, the more economical method is to utilize it locally instead of transporting it to other areas via pipelines, which requires the extra investment of infrastructure. However, the local demand for hydrogen might be limited. To solve this issue, more industries with large demands for hydrogen, such as chemical, steel, and long-haul transport, should be planned near areas with plentiful renewable resources.

With more and more renewable energy systems being built in China, the optimal utilization of renewable power is of great importance. Power–hydrogen coupling at the source side provides a feasible solution to cope with the high volatility of renewable energy without exerting much pressure on grid-balancing. As for the specific mode of coupling, a combination of PHSS-PT and PHSS-HT might be more widely adopted in the future. Electrolysers, hydrogen storage, and hydrogen-fueled gas turbines will be installed near the energy systems to assist in the stable power transmission through UHV lines. In the meantime, the excess hydrogen produced will either be used locally or transported to other areas via pipelines.

6. Conclusions

In this paper, different modes for power–hydrogen coupling were firstly introduced, along with a systematic analysis of their respective strengths and weaknesses. Then, the power–hydrogen coordinated planning optimization model was proposed, which served as a tool for the case study on the Southeastern Xinjiang area. Through the comparison on key indicators characterizing different power–hydrogen coupling modes, the following conclusions were drawn:

(1)

Electrolysers, along with hydrogen storage and hydrogen-fueled gas turbines, can function as either flexible loads or long-term energy storage, which supplement electro-chemical energy storage in supporting the safe and stable operation of the power system.

(2)

For the Southeastern Xinjiang area, the most economical mode was power–hydrogen coupling at the source side for in situ utilization, followed by coupling at the source side for hydrogen transport via pipelines, coupling at the source side for power transmission via UHV lines, coupling at the load side, and no coupling at all.

(3)

In scenarios with larger volatility of renewables, the mode of power–hydrogen coupling at the source side showed better economic competitiveness.

The study presented in this paper serves as a first step. In the future, the research group plans to conduct research on a coordinated planning of the power and hydrogen systems on a national scale. Different power–hydrogen coupling modes will be combined in order to achieve an optimal and coordinated planning of both the power system and the hydrogen system.