Part 2: Technological Comparison, Invinity's History, and Financials.

The Competition

The comparison up until this point has been with LIBs, for obvious reasons. But VRFBs are not the only technology aiming for a share of the BESS market, and it’s important to see how they compare with other upcoming battery types, especially in the use cases where they show most promise. This section will inevitably be more chemistry-heavy, but I tried to keep it readable.

Sodium-ion Batteries (SIBs)

By far the most talked about competitor to LIBs. SIBs currently struggle with all the usual challenges one would expect from a bleeding edge battery technology, but there are more fundamental issues.

Sodium and lithium are both alkali metals and so share most of their chemical properties. Consequently, SIBs and LIBs have largely the same engineering schemes. But sodium has a lower redox potential, meaning it can maintain a smaller cell voltage than lithium, which translates to SIBs suffering from a lower energy density than even LFPs. Sodium ions are also larger, which means slower diffusion rates through the electrolyte, hence higher internal resistance and lower charge rates. Their larger size also means it’s more difficult to get them to intercalate in the electrodes, and that they cause greater volume expansion in the electrodes once they do, leading to increased mechanical stress and issues of stability and longevity.^(27)

One claim that I hear way too often is that SIBs are safer than LFPs. This is just plain false. The only SIBs that are anywhere close to commercialization use flammable organic solvents, just like LIBs. Research consistently places them squarely between LFP and NCM in terms of safety: when compared to LFPs, they exhibit lower thermal runaway onset temperature, faster temperature rising rate, higher maximal runaway temperature, and emit more gases.^(28-31) Moreover, though it varies by chemistry, the gases emitted by SIBs tend to have a wider explosive limit range, meaning they are more likely to combust. Particularly nasty is propylene carbonate, the most common solvent choice, as it releases propylene gas (basically propane on crack).^(32)

https://imgur.com/a/5yRNzL9 (Comparison of safety parameters between an NCM LIB, an LFP LIB, and an NTM SIB. Left: thermal runaway onset temperature, safety venting temperature, separator collapse temperature, and maximal runaway temperature. Higher is better for the first three, lower is better for the last. Right: kinetic analysis of thermal runaway in the three batteries. Lower is better. Reproduced from reference [28] with permission.)

Overall, performance-wise, SIBs can be viewed as a worse version of LFPs.^(33,34) Their only major improvement is their superior performance in low temperatures, which could be significant for EVs in colder climates (since they don’t have HVAC systems supporting the battery 24/7). But considering their intrinsically lower RTEs, it would take truly arctic environments for this alone to close the performance gap with LFPs in BESS applications.

The main selling point of SIBs is that their theoretically lower production costs will justify their diminished performance, particularly in BESS applications. This is a viable assessment, since SIBs contain no lithium and at most tiny amounts of copper, while all their contained materials are cheap. To see how big of an advantage that is, the intensity of lithium in LFPs is ~0.53 kg/kWh LCE equivalent, while that of copper is ~0.48 kg/kWh, so their respective raw material cost contributions are ~11.13 $/kWh and ~6.08 $/kWh, combining to a total of ~17.21 $/kWh—about 25% of the current total pack price.^(35) This percentage is expected to increase as both copper and especially lithium prices grow with demand while production costs continue to decrease.

It should be noted, however, that the above issues with sodium call for high-performance electrodes and more sophisticated cell engineering, and it’s currently unclear how large of a gap will remain between the production costs of the two technologies.^(36) Moreover, their lower RTE, stability, safety, and longevity incur a heavy LCOS tax, which makes it even more challenging to determine whether they’ll actually make for a more economical alternative to LFP.

There is one undeniable advantage of SIBs: abundance. Both lithium and vanadium demand is expected to exceed supply soon, whereas sodium is everywhere. When developers literally cannot get their hands on other technologies, SIBs will almost certainly be the default choice. This alone promises to carve a substantial chunk of market for them. The possibility of SIB use will also mitigate the strategic vulnerability of relying on foreign, possibly hostile countries to supply materials for an industry as critical as this one.

So where does this all place SIBs in relation to VRFBs? Nowhere different than LIBs, really. They don’t fare any better in any of the metrics that VRFBs excel at—in fact they fare worse, in exchange for possibly lower cost. The only scenario I can think of where a developer would choose VRFBs over LIBs but not over SIBs is one in which the cost advantage of the latter would be so great as to offset the considerations that gave VRFBs the edge. It’s hard to believe that this would be the case, and in some use-cases (safety in particular) it will be impossible. SIBs therefore don’t threaten to take any larger a market chunk from VRFBs than LIBs.

Zinc-Bromine Batteries (ZBBs)

ZBBs have existed for over a century and are currently seeing a revival due to promising technological advancements. They can come in either static or hybrid flow variants. The hybrid flow types have fallen out of favor, and all their former manufacturers are now defunct (Primus Power are still technically alive but have not been operating for years). I’ll therefore focus on static ZBBs, championed outside of China primarily by New Jersey-based Eos Energy Enterprises.

Starting with the advantages, static ZBBs currently run circles around any other battery technology when it comes to BESS energy density. Their electrochemical density is only a third that of LFP’s, but Eos recently announced their new Indensity architecture, which allows to stack the batteries up to twelve units high, netting them a staggering maximal areal density of 1 GWh/acre. This makes ZBBs a very attractive choice for any project with rigid spatial constraints. They also have an impressive operating temperature window, ranging from -10 to 50 C, meaning they require only minimal cooling (if any) in most climates.

Another significant advantage is material costs, since both zinc and bromine are common and cheap, together requiring about 8 $/kWh.^(37) The main material cost factor is probably the electrolyte itself, which needs to contain complex mixtures of additives and buffering agents to reduce the known problems of the chemistry. Nevertheless, ZBBs can theoretically compete with sodium ion when it comes to cost once their production is streamlined.

When it comes to RTE, static ZBBs lie neatly between VRFBs and LFPs, with cells in lab conditions attaining efficiencies of up to 90%.^(38,39) Examining real world deployments, in their latest earnings presentation Eos claimed an average deployed RTE of 84.6% for their latest Z3 batteries. They don’t say either in the presentation or in the recorded meeting whether that’s DC or AC-AC efficiency, which almost certainly means it’s the former (also the alternative would be ludicrous). Furthermore, these figures were given for 20-80-20% depth of discharge (DOD) windows, which miss the most inefficient parts of the operation. This is confirmed in their product sheet where they say “the maximum DoD can be reduced for applications demanding round trip efficiency in the mid-80s”,^(40) which implies that DC RTE is at most ~80% in deep discharge deployments, of most relevance to LDES (this is why I hate using company data). Taking all this into account, the fully deployed RTE can be expected to be around ~70% for LDES, which is in line with the literature values.

Longevity is tricky. Historically, ZBBs suffered from significant longevity issues, stemming from reactions like zinc dendrite growth on the anode (basically tiny snowflake-shaped stalactites), hydrogen evolution, and corrosion from the free bromine in the battery.^(37) Great strides have been made in mitigating these issues, however, and modern ZBBs can remain stable for over a thousand cycles.^(42) Eos claims a cycle life of 6,000, which would place them competitively against ion batteries. They again don’t specify how number was attained, which leads to suspicion that the conditions were highly favorable, like shallow cycling near 50% SOC and slow C-rates where many of the problematic reactions are negligible. That being said, it’s entirely feasible for ZBBs to reach this figure in realistic deployments given the rapid technological advancements.

https://imgur.com/a/wWMUJM2 (Zinc dendrites in an anode. Reproduced from reference [41] with permission.)

One key challenge of ZBBs is their self-discharge rate, caused by the diffusion of bromine and polybromides from the cathode to the anode.^(43) This is particularly problematic for LDES applications, where the battery is expected to hold its capacity for many hours if not days. An unmitigated ZBB will discharge about 50% of its charge capacity within 2 hours. Luckily, advancements involving the trapping of the problematic bromine within the cathode have worked to ameliorate this effect, with some lab cells boasting a self-discharge of only 3.9% over 24 hours.^(44) It remains to be seen how small this can get for scaled batteries in realistic deployments. Eos say nothing about self-discharge in their published materials.

Lastly, ZBBs face some significant safety issues. On the plus side,