On the optimal planning of a hydrogen refuelling station participating in the electricity and balancing markets

https://doi.org/10.1016/j.ijhydene.2020.10.130Get rights and content

Highlights

  • Techno-economic model of an HRS able to inject hydrogen into the gas grid.

  • Optimal sizing and operation of an HRS in energy and ancillary services markets.

  • The nonlinear model of the electrolyser and grid costs components.

  • Determining the minimum hydrogen price for profitable business cases.

  • Selling hydrogen to vehicles together with optimal reserve provision is preferred.

Abstract

This paper presents an optimisation model to assess the techno-economic feasibility of a hydrogen refuelling station, which purchases power from the electricity market, supplies the mobility sector with hydrogen, and participates in the ancillary service market. The problem is formed as a mixed-integer nonlinear programming model to investigate the optimal operational plans considering the nonlinear behaviour of an electrolyser and grid costs calculation model. Obtained results from various scenarios in 2020 and 2030 show that participation in the reserve market considering optimal sizing and dispatch of components increase revenues up to 16%, and as a result, decrease the hydrogen break-even price by up to 4.7% and 6.4% in 2020 and 2030, respectively. Exemption from tax and levies for connection to the grid reduces the hydrogen break-even price by up to 13%. Plant operators could benefit from the proposed approach to schedule components reliably while meeting the hydrogen demand and maximising the annual profits.

Introduction

The road transportation was responsible for nearly 21% of the total CO2 emissions of the European Union (EU) in 2016. In 2019, the EU set out a new target of 23% emissions reduction from road transport by 2030 compared to 2005 [1]. To reach this target, the growing use of alternative fuels is indispensable, and hydrogen as one of these alternatives has begun to take its place in the transportation sector. In recent years, hydrogen-powered vehicles and hydrogen refuelling stations (HRSs) have become more available to the public. The major factors to be considered for the implementation of an HRS are financial issues and technical operation of the system, which make the techno-economic analysis of an HRS crucial for investors and decision makers.

Many studies have analysed the techno-economic feasibility of hydrogen systems and HRSs around the world. The technical potential of hydrogen production by placing wind turbines next to the existing fuelling stations has been studied in Ref. [2]. Ref. [3] has examined the techno-economic viability of small-scale HRSs, which produce hydrogen via alkaline electrolysers. In Ref. [4], a strategy has been proposed for the operation of an electrolyser, which consumes energy at times of low electricity prices in the spot market. The proposed method has considered the predictions of electricity prices and wind energy production. However, optimal sizing has not been included, and electricity grid costs and taxes were assumed as constant per unit values. Ref. [5] has used an optimisation strategy to minimise the operation costs of an HRS considering electrolysis process and wind energy production. However, investment costs were not included in the analysis. With a focus on reducing design costs, a Simulink model consisted of an electrolyser, a compressor, storages, and a hydrogen consumer block has been developed in Ref. [6]. While some detailed aspects of an HRS were described in this reference, its simulations lack elaborate operational details such as optimal planning of the electrolyser electricity consumption. Moreover, the authors have not considered sizing of subcomponents, and other sources of revenue, such as participation in the ancillary service markets.

The reasonable sizing and siting of HRSs both improve the hydrogen infrastructure and reduce the production cost of hydrogen. In Ref. [7], a deployment plan has been developed to find the optimal location and the required number of refuelling stations in the Republic of Korea. Ref. [8] has considered the growing hydrogen demand and offered an optimised design for an HRS network. The economic feasibility of an HRS network in Romania considering the annual hydrogen consumption, investments, and net present value at the end of the period has been studied in Ref. [9]. The results showed that the primary barrier for the expansion of hydrogen infrastructures is the unfavourable economic issues rather than technical requirements. While Refs. [8,9] have done thorough analyses, they made assumptions about the capacity of the stations, mainly focused on economic evaluation of HRS rollout plans, and technical aspects and hourly operation of stations were not taken into account. In Ref. [10], an HRS siting optimisation model has been proposed, highlighting that attention should be paid to all aspects of hydrogen cost including the cost of production, transport, storage, and the annual investment and operational costs.

In addition to siting, the HRS subcomponents should be sized correctly to meet the hydrogen demand at any given time. If the size of the subcomponents is not optimised, not all produced hydrogen will be used for the mobility sector and the excess amount will be supplied to the gas grid or will be vented to atmosphere. Besides, capacity optimisation of each unit at the planning stage is of vital importance to decrease huge investment expenses. Recently, some articles have focused on the techno-economic analysis of hydrogen systems together with size optimisation. Ref. [11] has designed a photovoltaic-powered HRS to supply a taxi fleet in a Brazilian city. Results pointed out that the levelised cost of hydrogen is inversely proportional to the hydrogen production volume. However, the authors have only considered a topology for supplying cars, but not heavy-duty vehicles. Besides, a constant value for the efficiency of the electrolyser was assumed, and injection to the gas grid and grid costs were not included in the analysis. In Ref. [12], a method has been proposed to find the optimal size of the electrolysers in a system consisting of wind turbines, electrolysers, and hydrogen fuel cells. The authors studied the trade-offs between selling the hydrogen to consumers or store it at times when the electricity price is high. In Ref. [13], the authors have determined the optimal location, size, and the number of wind turbines, electrolysers, compressors, storages, and fuel cells in a test system. This source has mainly dealt with the excess wind power through the electrolysis process, while the interaction between the power grid and hydrogen production and consumption has not been taken into account. Moreover, hydrogen demand has not been considered for the mobility sector, or authors assumed a constant value for the hydrogen daily demand calculated based on the nominal capacity of the electrolyser in Refs. [12,13].

The authors have analysed different HRS architectures from the economic point of view for 2015 and 2050 in Ref. [14]. They have dimensioned compressor and high-pressure hydrogen storage to minimise capital and electricity costs. However, costs and energy consumption scheduling associated with the electrolyser were not considered, and therefore, grid connection costs excluded the electrolyser. Ref. [15] has introduced a siting and sizing method to achieve the optimal cost for hydrogen consumers. While the proposed model covers various techniques of hydrogen production, transportation, and storage, the authors have mainly focused on the economic features of a group of stations, and the technical aspects and operational scheduling of electrolysers, storages and compressors in an HRS were not taken into account. Ref. [16] has presented a model to evaluate the HRS profitability and to find the optimal size of the electrolyser and storage unit for different car-sharing fleet sizes. However, operational aspects, such as electrolyser scheduling and as a result, the exact model of grid costs have not been included in the analysis.

In addition to sizing, it is also important to explore possible markets that are properly suited for HRS. The first clear business model relies on trading between the markets for selling hydrogen and the electricity market during times of low electricity price. For example, Ref. [17] has designed an HRS in which hydrogen is produced by a proton exchange membrane (PEM) electrolyser using electricity from a photovoltaic plant and the power grid. The authors have evaluated the feasibility of case studies with several objective functions. However, they have studied a topology for supplying cars, but not heavy-duty vehicles, and assumed an amount for the capacity of subcomponents such as hydrogen storage.

Other business models that provide added incomes aim at the provision of ancillary services to guarantee the reliable performance of power systems [18,19]. Contribution of water electrolysers in providing ancillary services has already been investigated in some papers. Ref. [20] has studied the economic profitability of an electrolyser participating in the French primary reserve market. The authors concluded that the provision of the primary reserve with the current requirements is an unattractive option as it increases hydrogen production cost. A model was developed in Ref. [21] to study the performance of an HRS. While the authors have considered participation in the reserve market, their research lacks the optimal sizing of the primary reserve capacities and the nonlinear models of the electrolyser and electricity grid costs. Participation in the secondary reserve market has also been taken into account in detail in Refs. [22,23]. An overview of a 6 MW power to gas plant in which a part of the produced hydrogen is delivered to an HRS has been given in Ref. [22]. Providing secondary reserve was among the different operating modes. They found 1 ₠/kg as the production costs because they did not include either taxes and grid costs on purchased electricity or investment costs of compressors and storages. Furthermore, hydrogen demand has not been included as a limitation in the study. Ref. [23] has proposed a model to evaluate the economic feasibility of hydrogen plants that provide secondary frequency service in addition to supplying hydrogen for different markets. Instead of sizing HRS components, the authors have determined the minimum demand expected from the hydrogen-powered vehicles so that electrolysis facility with 5 MW size and assumed hydrogen selling price returns enough profits. They have considered a constant efficiency for the electrolyser, and hourly operation of the electrolyser and other subcomponents has not been presented.

Various parameters such as refuelling pressure make the design of an HRS complex. Heavy- and light-duty vehicles refuel at 350 bar and cars at 700 bar, respectively. The estimated daily and hourly refuelling demand should be considered in the HRS design as well. Considering the above-mentioned factors and the notable impact of investment and operation costs of equipment on the hydrogen production costs, the authors of this paper were inspired to develop a model to identify the optimum size and hourly working points of subcomponents. The introduced method optimises the operation of an HRS to be compatible with the power and balancing markets and provides the lowest hydrogen break-even price while meeting the hydrogen demand for the mobility sector.

While the aforementioned research set a precious foundation for this paper, they have not analysed the size optimisation of an HRS together with optimal hourly frequency containment reserve (FCR) provision to get the minimum hydrogen break-even price. They also have not employed a detailed model for calculating grid costs; instead, they have assumed a constant price for each MWh energy consumption. This work considers a more detailed model of an HRS that supplies hydrogen vehicles at both 350 and 700 bar combined with the possibility of injection into the gas grid. The proposed approach considers FCR provision and employs exact non-linear models of the electrolyser and grid costs components including taxes, levies, public service obligations, etc.

This article is structured as follows: Section Test system and methodology defines the test system and the proposed method based on specifications of the HRS. Section Simulation inputs describes the input data and the frameworks in Belgium that are applied to the model. Then, the results of the sizing and scheduling method for different scenarios are presented in section Results. Finally, section Conclusion draws the conclusions and possible future steps.

Section snippets

Test system and methodology

In this study, an onsite grid-connected HRS, which buys electricity sourced from sun and wind on the electricity market was analysed. Fig. 1 shows the HRS system configuration, which serves both light and heavy-duty vehicles with tanks at 350 bar and cars with tanks at 700 bar. The HRS design considers a refuelling plan in which two compressors feed the medium and high-pressure storages. The first compressor supplies the medium pressure (450 bar) storage with hydrogen coming from the

Simulation inputs

The current study was conducted based on the regulatory situation in Belgium, and therefore all electricity-related input data, including energy prices, grid tariffs and taxes, and FCR remuneration scheme were chosen for Belgium. Table 1 provides up-to-date techno-economic data gathered in projects from suppliers or obtained from companies around Europe. Specifically, given data for the electrolyser and cell stacks, which are provided by Hydrogenics are close to the information in the mentioned

Results

Simulations are run for two situations according to the hydrogen targets [33] and according to the current conditions in Belgium [34]. Sections Results considering the hydrogen targets, Comparing the current situation with targets and Comparing the current situation with targets discuss the simulation outputs in more detail.

Conclusion

This paper has presented an optimisation method to evaluate the techno-economic feasibility of an HRS. The model optimised both the size and hourly operation of the subcomponents to perform an energy arbitrage in electricity, natural gas, hydrogen, and FCR markets.

The study revealed that the operational costs depend extremely on electricity price and grid costs, which are directly associated with the size of the electrolyser. Hence optimal sizing of components and exemptions from tax and levies

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors acknowledge the GREENPORTS-project (Gas from REnewable Energy in PORTS) funded by VLAIO (Flemish Agency for Innovation and Entrepreneurship), and the NorthWest Europe INTERREG project ITEG (Integrating Tidal Energy into the European Grid) led by the European Marine Energy Centre (EMEC) [NWE 613]. VLAIO, Belgium; European Marine Energy Centre, United Kingdom.

References (35)

Cited by (0)

View full text