posted 4 hours ago
Eric, I appreciate the foray into underappreciated chemistries!
For readers, please keep two things clear: energy is how much, power is how fast.
I spent about a year in a professional engagement working with LTO, and have (or my then-employer has) the experimental data backing up all the numbers you note, though I will say that -40degC is pretty harsh on the cells, and they do not perform well, but they don't die. You can hit them very hard at -20degC and they perform pretty well, though. (These were engine-start profiles I was testing, very high power, moderate energy.) They act more like capacitors than like batteries, sometimes. This suggests that they may have a role slightly different from an LFP battery, and they do excel at short-burst work. For home use, however, they might be a good choice as an uninterruptible power supply application, where you are handing over to a generator within a relatively short period. You don't need a big bank to get the power you need, as long as they are not really needed for substantial energy storage. I should also mention that their cycle life (for a well-built cell) can be 2-5x that of LFP, which is really cool, and I have indeed cycled them out to many thousands of cycles to verify this.
Couple comments on their weaknesses are appropriate, though:
1) Their supply chain is terrible. There are very few vendors, and the best one (Toshiba) emphasizes industrial uses like forklifts (a great use case!). The Chinese vendors make them in more DIY-friendly form factors than Toshiba, but there still are not many, and US-sourced? Don't ask. As a result, the cell prices are not generally competitive with other chemistries, though the system lifetime cost can make sense due to the long life.
2) LTO is an anode. This is paired against (typically) an NMC (nickel, manganese, cobalt) or NCA (nickel, cobalt, aluminum) cathode, so you don't get away from the rare metals penalty, which is one of the best arguments for LFP (a non-rare-metal cathode, which is paired with graphite for the anode).
3) Related to the above, the electrochemistry of LTO means you are fixed at a nominal cell voltage of 2.3V, compared to 3.2V for LFP and 3.7V for NMC/graphite (most "lithium batteries"). So, if you want a 12V battery, you need 6 cells in series instead of 4.
4) Related to the above, the energy density of LTO in practical batteries is closer to 50% that of LFP, and so not only do you need 1.5x the cells to get your voltage, you need 2x the strings to get your energy. It's more like a factor of 3x "more battery" if you want the same energy storage.
5) Lastly, though LTO performs cold, it does not perform well hot. This is driven by the electrolyte, and is akin to how motor oil can be picked for cold or picked for hot, but it takes a fancy oil to do both. Well, it's challenging and expensive to concoct an electrolyte that performs well over a wide range. I have seen cells begin to develop internal gas pockets at 45degC and up, which (as a former resident of the Southwest) is not that hot. LTO is chemically hard on many electrolyte chemicals, and over time, low quality cells can chemically crack the electrolyte molecules (again, Toshiba is exceptional here, this is part of their secret sauce for these cells). So, cool cells are happier cells, these reactions being worse at elevated temperatures.
So, if you are gifted a large quantity of LTO cells, there are very cool things to do with them. You can use them as a drop-in for lead-acid start batteries, they will perform in that high-power role, and this is something you should NOT use LFP for! I have seen the burned wreck of a golf cart that was "converted" to LFP, and the cells were not capable of handling the charge or discharge loads the cart had - it was meant for lead-acid. For the time being, I think LFP remains a very useful chemistry for home energy storage, even if you have to tuck it up at night.
That said, I am hopeful that sodium-ion will develop further and become more affordable and widely-distributed. For stationary batteries, they have all the benefits of LTO without any of the lithium or rare metals. They are bulky, but their cycle life and low-temp performance are pretty interesting on paper. I have not developed much data on them, just have a few cells to play with. They are also a low-voltage cell, 2.1V or so.
Thanks for bringing up an interesting and under-known chemistry! I would encourage anyone doing a system design to look at all the (affordable) options, and decide what you need most - power or energy. For most home-grid use, I would bet it's energy, but people can come up with all kinds of scenarios, and for some it might be power that matters the most. If sizing for power, it is important to size everything to the max current you could draw, not just the batteries. I have used LTOs as a source for verifying wire-melting currents, and they will do this cheerfully - so make sure the balance of the system is sized correctly! As in, it's no good to have a crazy-big engine in a tiny car where you'll wreck the drivetrain with it.
Happy homesteading,
Mark
.png)
4
