Hydrogen storage solution : an assessment of its technological innovation system. Insights from patent data.

Keywords:

Abstract:

Research question
Transport faces increasing pressure to reduce its greenhouse gas emissions. For this to happen, low- and zero-carbon energy solutions need to be developed. Hydrogen is one of them. The extent of its implementation for the future of the mobility depends on a set of barriers that must be overcome. Among them, a key issue is related to its storage. Several storage technologies are competing, such as liquid hydrogen, methanol, ammonia, etc. Each technology has its specific advantages in terms of storage density, cost of storage, and safety.
In order to better understand the current state of development of these technologies, their comparative positioning and their chances of success, the Technological Innovation System (TIS) approach can be relevant. This theoretical approach has proven its capability to assess what hampers the development and implementation of a particular technology (Jacobsson and Johnson, 2000; Hekkert et al. 2007 ; Bergek et al., 2008 ; Bergek, 2019). One of the main idea is that around a technology a strong innovation system needs to be developed thanks to several system functions. Historically, the TIS was created and used as a tool for helping policy making, but it can be also helpful for decision makers in R&D firm (Flamand and Frigant, 2020). Until now the methodology of TIS analysis is mainly based on qualitative opinions from experts or/and on historical events analysis (Hekkert et al. 2007). Purely quantitative approaches, like data mining, are less employed. And yet, several data are considered useful for studying innovation, as patent.
Based on it, our research problem is twofold:
- We want to give an overview of the current stage of evolution of hydrogen storage technologies based on an assessment of its technological innovation system. This specific step (the storage) is quite underestimates in the study of hydrogen value chain, while the latter depends on it.
- To do so, we propose to establish a structured methodological framework covering the full potential of patent data to analyze a TIS.
Methodology
Our empirical methodology is based on a patent analysis. From technical codes, we were able to identify about 9100 protected inventions related to hydrogen storage solutions. This corpus is our database to evaluate the level of development of the technologies under study and to realize our methodological demonstration.
For this demonstration, we proceeded to an extensive literature review of, first, the technological innovation system approach and, second, patent analysis. The objective is to combine both. TIS study generally requires to follow 6 steps (named sub-analysis), among which the study of 6 key functions constitutes a key moment. For the whole process, we expose what the patent data can tell us, how (which relevant indicators?) and also the limits of this data.

Findings
Very preliminary results:
With our database, we identify almost 3000 applicants that represent inventive actors (actors being a structural component of an innovation system). Not all of them are equally important for understanding the development of energy storage solutions, and we can assess their role. 20% of the applicants have more than 3 patents filed, accounting together for 75% of the inventions. Among the main applicants are car manufacturers (Toyota, Honda and BMW in the lead) but also some industrials in gas sector as Air Liquid.
Networks and any kind of relationships constitute another of the structural components of TIS approach. Thanks to co-patent data, we can study the technological collaborations. We identify 5 mains networks of partners, each being structured by nationality (Japan, South Korea, USA, France and Germany).
Concerning the functions (key process needed for the development and diffusion of a technology), the first can be enlightened with patent data: « production and diffusion of knowledge ». From patent, we can characterize the technological knowledge base and determine which new knowledge are created (which knowledge?) and determine who contributes to (which actors?). TIS leads us to go further and ensure that technological developments match with current technological lock-in. For this purpose, we have built a map of technological barriers covering the several solutions of storage. 15 technological sub-categories are considered. The next step is to compare this technology map with the patented inventions in order to identify the presence of gaps in the technological development.
We will study diffusion of knowledge with patent citations, and several method and metrics are available. We planned several crossings here: mainly study of citations between actors (according to their nationality or their nature) and between the various technological subsystems.
For the other parts of TIS analysis, significant result can be expected for « Guidance of the search » for example. This function concerns the study of technological variety in time and patent data can permit to see some technological trajectories. Concerning the « market formation » function dedicated to relation between suppliers and users of technologies, we want to test if co-patent can also be relevant.
Practical and theoretical implications
Implications of our research are manifolds. The findings will cover not only a general assessment of technological innovation system for storage hydrogen, an important step of hydrogen value chain. But also we will show that the patent data can be much richer for TIS analysis than literature suggests.