Energy Storage Europe 2019

Part 1

Alfacomp

17 March 2019

Last week I had the opportunity to attend the Energy Storage Europe 2019 conference and exhibition, held in Dusseldorf. This was held in conjunction with the International Renewable Energy Storage conference, hosted by Eurosolar, and is also part of a series of Energy Storage conferences in the US, Japan, China and India.

As a survey of all the different technologies on offer a great place to start is with the very nice analysis of projected future technology costs undertaken by Oliver Schmidt of Imperial College, London. He has analysed the historical pricing of a variety of different energy storage technologies - assessing their levelised cost as a function of the total global amount of that technology installed with the aim of establishing 'experience' rates akin to the way done for photovoltaic panel production.


The first thing Oliver noted is that pumped-hydro energy storage appears to have a slight negative experience benefit. It seems to get more expensive to build as more plants are installed - this seems to go completely against the principle underlying experience rates that the more you do something the easier it should be to do it - so how do we understand this ?

This might be as a result of the limited number of installation sites available - the easy sites get done first and then the harder and more expensive ones are attempted - eventually you run out of sites to use. It is this unavoidable fact that has of course driven the demand for alternative 'locate anywhere' battery solutions described here.

Secondly, whilst there is significant data for Lithium-ion and Nickel-Metal Hydride batteries (i.e. individual low voltage cells) there is much less data for complete systems. In the diagram above these are shown as filled circles on the top left and I detail these (except for the high temperature molten sodium-sulphur) below:-


As you can see there are only 4 points available for utility-scale Vanadium redox-flow and residential Lead-Acid and Lithium-ion, and only 6 points defining the experience curve for utility-scale Lithium-ion. Of these technologies it is Vanadium redox-flow which appears to be the cheapest, and the rates of reduction in price are closely comparable at between 11% and 13% per doubling of installed capacity. This relative paucity of historical information should be borne in mind when considering the subsequent predictions on the penetration of different energy storage technologies.

Oliver's subsequent analysis looked at 12 different energy storage applications, ranging from the very occasional seasonal and black-start applications through tertiary and peaker response and then the very regularly used primary and secondary response functions that every grid requires every day to maintain grid balance (interested readers should consult the related paper for an explanation of these application areas).


As you can see VRFB batteries and hydrogen electrolysis/fuel-cell consumption are considered to be the only technologies that can cover all of these application areas - I am assuming that self-discharge of Lithium-ion, Sodium-Sulphur and Lead-acid precludes them all from very long term energy storage. Such a long timescale application is currently purely theoretical though - and at present only the generation of hydrogen or renewables-derived biofuels appears to offer realistic economics for inter-seasonal storage.

Turning to a specific application that is currently being procured by grid operators, we can look at the secondary response application - this typically kicks in after 30 seconds of a grid frequency fluctuation and may run for periods of up to 1 hour (for an excellent explanation of primary, secondary and tertiary grid balancing, sometimes referred to as frequency regulation, refer to this article).


In this application area (100MW power, 1 hour duration, >10 seconds initial response time) Oliver's analysis (above) shows that by 2020 VRFB's should have something like 30% of the chance of being valued as the Lowest Cost of Storage technology. They do this by stealing market share from pumped-hydro schemes. As 2025 and 2030 roll-on it appears that Lithium-ion technologies reduce in price to join VRFB's and to fully push out pumped-hydro from the picture. A similar qualitative picture (though marginally less generous to VRFB's) emerges from 10 out of 12 of Oliver's application scenarios. When battery manufacturers talk about 'stacking value' this is what they mean - participating in multiple electricity grid markets to offer grid balancing services over multiple timescales and with different availability criteria.

How then do we understand the apparently strong performance of Lithium-ion after 2020 when it appears (from the first graph) that the experience rates for both Lithium-ion based and VRFB based systems are comparable at 12% and 11% respectively ?

One point is that these are system prices - the experience rates for Lithium-ion battery cells appear to be much stronger - up to 30% per doubling in installed capacity. In principle the system prices can be broken into battery cell components and a balance of system components and a derived system experience rate modelled. This would be a rather complicated procedure and might be skewed by the fact that many Lithium-ion battery cell structures may not be appropriate for application at Utility scales due to their safety issues. 

Should such a procedure be applied to Lithium-ion cell pricing to derive system pricing then to avoid favouring one technology over any other a similar procedure would need to be applied to VRFB's. Given the very large Vanadium price component in VRFB's this would need to consider the possibility of reducing Vanadium and VRFB electrolyte production prices, as well as reductions in stack component prices such as membranes, bipolar plates and carbon felt electrodes that may also be used in large volumes for Hydrogen fuel cells or electrolysers.

The era of significant VRFB implementation would necessarily see a very different Vanadium market to the one we currently possess (75% by-product production from Vanadium Slag and only 3 primary Vanadium mines outside of China) - the experience rates that we see in VRFB's between 2008 and 2015 clearly have seen no such evolution in the Vanadium supply market and thus we could expect the VRFB system experience rate of 11% to perhaps by strongly improved on in a post-VRFB-adoption era.

Given the very large Vanadium deposits associated with many Vanadium-Titanate-Magnetite resources, and the fact that Vanadium is an element more common than Nickel, Chromium, Zinc AND Copper there appears little chance that the Vanadium supply would be systemically constrained unlike rarer 'battery-metal' elements such as Cobalt and Lithium.

Overall VRFB's stack up well against Lithium-ion in this analysis despite there being little direct comparison of comparative system pricing since 2015. Much further work needs to be done in this area and VRFB manufacturers need to step forward more actively to show their long-term pricing assumptions and current system prices achieved. In the past this may have been hampered by the separation of VRFB manufacturing and Vanadium production - VRFB manufacturers may have been simply unable to predict what their future unit pricing might be in a market in which they had no idea what the Vanadium price was going to be next year.

The introduction of a vertically integrated Mine-to-Battery value chain obviously completely changes this situation and underwrites long-term electrolyte leasing as a practical route to Capex reduction.

The vertically integrated mine-to-VRFB model represents a tipping point which could completely change the economic outlook for VRFB's and would allow them to really take on Lithium-ion as the technology of Utility-scale electrical energy storage.

This article only conveys the personal opinion of the author. Whilst every effort is made to ensure the content is accurate, we cannot guarantee the accuracy of the data shown. This article does not constitute professional, financial or investment advice and must not be used as a basis for making investment decisions.

Site content is not authorised by the FCA and you are not safeguarded by the Investor Protection measures of the Financial Services and Markets Act 2000. See our full disclaimer