Hydrogen station location optimization based on multiple data sources

https://doi.org/10.1016/j.ijhydene.2019.10.069Get rights and content

Highlights

  • Multiple data sources are used for hydrogen station location optimization.

  • Data includes stations network, GIS and economy data from Beijing.

  • The results provide hydrogen station planning solution for Beijing.

  • This research has social significance and economic benefits for hydrogen economics.

Abstract

A hydrogen station is one that fills or stores the hydrogen, which is critical to the commercial development of hydrogen energy and fuel cell vehicle industry. Therefore, its location planning becomes an important issue. Similar to the electric vehicle (EV) charging station's planning, several factors are considered including the location, the demand of the fuel, the driving distance, etc. In this paper, multiple data sources are applied to the site selection model, including the existing petrol-refueling station network data, geographic information system (GIS) data, population data and regional economic data. Based on the operation of the genetic algorithm, combined with the idea of the greedy algorithm and the annealing algorithm, we propose a multi-algorithm hybrid solution, which not only can avoid local optimal, but also can converge quickly. On the basis of the site selection scheme of the hydrogen station in California, we have optimized the location scheme in Beijing. Finally, we present the feasibility proposals for hydrogen station location in Beijing, including the appropriate number of hydrogen stations in different regions, the reasonable coverage distance of hydrogen stations, etc. Due to the huge development prospects for hydrogen energy and the urgent need to reduce the construction cost of hydrogen stations in China, this research can quickly optimize the location of the hydrogen station and further explore potential mathematical relationships, which has certain social significance and economic benefits.

Introduction

Since fossil fuels are non-renewable and can cause serious environmental pollution, people are looking for alternative fuels, including hydrogen, ethanol, biodiesel, natural gas, or electricity. As a universal energy source and secondary energy source, hydrogen energy [1] is becoming increasingly significant, and the hydrogen industry is developing rapidly all over the world. The promotion of hydrogen energy includes hydrogen fuel vehicles and hydrogen stations. There are two types of hydrogen energy vehicles, the hydrogen internal combustion engine (HICEV) [2] and the hydrogen fuel cell vehicle (FCEV) [3]. At present, the faster development of the two is the FCEV.

In Germany, the government built a new station every two weeks on average in 2018 and a total of 100 hydrogen filling stations will be built by 2019 [4]. In California, the existing hydrogen stations are concentrated in Los Angeles, Sacramento and San Francisco [5], and hydrogen highway with stations spaced 20 miles apart is going to be built [6]. In Japan, FCEV is launched with high cruising range [7]. Compared to Germany, Japan and other countries, hydrogen energy has greater development potential in China. In the next decade, China plans to deploy more hydrogen stations and public hydrogen service vehicles [8], and in 2020, there will be 100 hydrogen stations and 5000 public hydrogen service vehicles. While in 2030, the number of hydrogen stations and public hydrogen service vehicles will achieve 300 and 50,000. Beijing's 2018 new energy vehicle policy announces that the government will subsidize FCEV for a long time. Before the promotion of FCEVs, a network of hydrogen stations needs to be preliminarily established [9].

During the fast development of hydrogen energy, the location planning of hydrogen station is a critical issue [10]. The high investment cost of the hydrogen station hinders the construction of the hydrogen refueling station [11]. Further promotion of hydrogen energy requires the current hydrogen station network to reach a certain scale. Therefore, the location of the hydrogen station is not only related to its long-term revenue profit, but also plays an important role in the promotion process of “hydrogen economic” [12]. In addition, it is necessary to consider a variety of factors during the location planning.

There are several models for facility location, including GIS models, operations research models [13], flow-capturing models [14], multi-objective optimization models, etc. Factors affecting the location of a hydrogen station include travel distance, fuel demand, traffic flow and so on [15]. The goal of site selection is based on these factors, which include minimizing the number of facilities and minimizing the travel distance or travel time, maximizing the traffic flow captured, maximizing the demand points covered, etc. There is no uniform standard for deciding which model outperforms other models, instead, it depends on our measurement perspective [16].

Section snippets

Related work

At present, various countries around the world are accelerating the construction of hydrogen stations. However, the construction cost and operation cost of the hydrogen station are much higher than those of the traditional gas station, which has become a research problem for many scholars. There have been related researches for the location of hydrogen stations at home and abroad.

The process of promoting hydrogen stations and FCEVs can refer to the promotion process of New Zealand gas-filled

Study area

The government divides Beijing into several regions according to their regional functions: core districts of capital function (Dongcheng District and Xicheng District), urban function development districts (Haidian District, Fengtai District, Chaoyang District and Shijingshan District), new urban development districts (5 districts) and ecological preservation development districts (5 districts). In summary, we select core districts of capital function and urban function development districts

Estimation of demand

The locations of the stations are related to the demand. Some analyses of the demand for hydrogen stations are based on the census tracts. While in California, besides the census tract, Sacramento also uses traffic analysis zones (TAZs) [38] for demand estimation. Traffic modelers in Sacramento have identified areas with roughly homogeneous travel characteristics as TAZs. However, Beijing does not have a similar regional division besides the census tract. So finally we further subdivide each

Model description

In terms of the station's location, there are some differences between the hydrogen station, EV charging station and petrol-refueling station. The locations of EV charging stations have to plan with the structure of the current power grid. And some drivers tend to charge electric cars in their homes rather than at public charging stations due to the long charging time. Existing petrol-refueling station distribution is a competitive distribution, which means there are several petrol-refueling

Model solution

Most of the location problems are classified as NP-Hard problems, which only have exponential algorithm complexity. Large instances of such problems cannot be solved with exact algorithms, and effective approximation algorithms for such problems must be sought.

The hydrogen station location problem can be classified as an integer programming problem. The feasible domain of integer programming problem is discrete. Therefore, it is necessary to combine some heuristic algorithms to optimize the

Result analysis

In this section, the results analysis includes analysis of the proposed solution, analysis of the number of hydrogen stations, and analysis of the coverage distance of hydrogen station, which is compared with the California area.

Conclusion

In account of the high construction cost and lack of network infrastructures, the hydrogen station cannot be deployed in the same way as the existing petrol-refueling station network. It is particularly significant to optimize the location of the hydrogen station. Combined with the researches on the location of the hydrogen station in California, our model is further optimized for Beijing, including the demand estimation and distance measurement. The proposed model is involved with multiple

Acknowledgments

This work was supported by the Beijing Natural Science Foundation (L171010) and the State Grid Corporation of China (520940180016).

References (41)

Cited by (44)

  • Locating hydrogen fuel stations: A comparative study for Istanbul

    2024, International Journal of Hydrogen Energy
View all citing articles on Scopus
View full text