LiFePo4 traction battery replacement ?

They will not go boom. Even cheap Lithium can go 10C for 10 seconds, HV current would never be that big for that long. I have seen it peak around 100amps or so. I think the HVbat fuse is only 125 amps. You could use cheap 10AH cells and it would probably work just fine. EVs are MUCHore demanding from batteries than hybrids and the Lithium I use in them seem to do just fine, even at 1000 amps. Most Lithium has a similar charge/discharge curve. Most decent cells can do 20C 10 second bursts. That's 200 amps for 10 seconds using 10 amp hour cells for a battery replacement options. More than enough for a hybrid application. How do you think I make plug in hybrids? Why do you think all plug in hybrids use Lithium?

XT1585 ?

 

we have a lot of ee's here. some are brighter than others, you're all brighter than me. don't take offense at my skepticism, we get a lot of people who come and ask questions, and then never come back.
as i said, i wish you all the best, and i hope you'll do a write up with pics, diagramming your step by step for others to follow.

 

It's your car. I guarantee I've played with more lithium than you, in millions of vehicles around the planet. I'll assume you've burned to death if there isn't an update in a year or two.

 

If it were going to happen, it would have long before now. No please tell me what your personal experience has been that leaves you so concerned? Tell me exactly what you think causes Lithium to "blow up". My personal experience tells me that most people blow up Lithium because the cells were not packaged properly, or there was no BMS to protect against under voltage/cell reversal for the knee of the discharge curve. Other than that, can you please explain what exactly you mean by " blasts of voltage" from the hybrid system. Maybe I'm not understanding exactly what you mean.

XT1585 ?

 

With the correct batteries and module configuration they can easily accept "blasts" of energy from the Prius and would not go "boom".

For example, the 3.3v 2.3ah A123 ANR26650 cell.
It can handle 70 amp load continuous. 120 amp pulse discharge (10 sec)
Assemble a module/pack in an adequate series/parallel combination and you would can easily handle the input/output needs of the Prius.

But just because it can be done, doesn't mean anyone will really do it.
I have many kwh A123 cells at my shop and could have something put together by the end of the week. But why bother? For me the time, expense, and effort would be better spent in other ways.

 

I would kindly refer you to the Boeing engineers who drew the same conclusion on the Dreamliner 777. I am not saying it won't work. I am saying it is poor engineering because it does not work for all situations the car is reasonably expected to see.

I live on the front range of the Rocky Mountains. My Prius drives the steepest interstate section in the country routinely, sections of road where the grade is so steep most vehicles don't have enough power on a clear and sunny day to maintain speed up the hill. In these cases the Prius discharges the pack completely very quickly. Then on the way back down, it charges it at a high voltage lower energy for prolonged (many minutes) at the same time it spins MG1 to burn off excess energy (B Mode).

I have constructed a few years ago a CAN Reader of my own design that I plug into the OBDII port for curiosity and one screen I always leave it on gives me the instantaneous pack voltage and current readings with sign into/out of the pack. It is not unusual to see 240VDC in at 85A.

Yes a lithium pack can take this no problem... Until it is charged. Then it is way too high a voltage for a nominal ~200VDC pack. Unless you make your pack nominally > 240VDC which would then tell the car that the battery pack is overcharged and it would spin the engine until the pack voltage dropped. With LiFePO4's a feature is that the voltage doesn't degrade much over discharge until it falls off a cliff. Which means it would spin the engine to get rid of power (thinking it was overcharged) until the pack was dead. And then it would be so low it would think it was too severely discharged and start charging it again.

No need for the childish quotes. When you tell someone about how lithium reacts chemically and then the smoke and fire, eyes glass over. When you say "boom" people understand. Lithium will easily let out the magic smoke when overcharged. Depending on the exact chemistry you choose this can be a fizzle and some smoke to a rather extraordinary fireworks show when you have cobalt in there. In any case, usually after the heat build up the plastic casings or insulation is what catches on fire first, and that is just normal toxic plastic fire which makes the whole pack melty and then "boom" (or fizzle).

My professional experience tells me that most people, including engineers at multi-billion dollar auto corporations, blow up Lithium for all sorts of reasons. Over voltage, under voltage, over current, dropping a tool on-top a pack, using the wrong type of bus bar, etc. Lithium chemistry batteries are really easy to use poorly (look at any cheap RC car) but really difficult to use properly.

Explained a little bit above. You said that you are or are going to run a LiFePO4 equivalent pack as a direct replacement for the NiMH pack currently in the Prius. If you use a pack with a nominal voltage similar to the Prius pack around 200V, then you have 40V that can be across the pack for long periods of time including after the pack is fully charged (i.e. going down a mountain). Lithium does NOT like this. Once it is full, it is full and you can really only trickle very little above its fully pack voltage. Not to mention that at 100% you are killing the pack really fast. A low current high voltage event as would be seen by a fully charged LiFePO4 pack during a high speed regen down a mountain for prolonged periods will overheat the cells. This same event for a depleted pack would just charge the pack relying on its internal chemistry which we also know is a bad idea for battery longevity, but no there won't be fire then.

The Prius runs its battery between 180V and 240V which is a very large range that is representative of 40% of its total pack capacity. The Toyota ECU's are all programmed to make sense of voltage readings from the pack as a NiMH cell, and it varies greatly and often. When you throw lithium in there, you are going to be operating at a fairly fixed voltage most of the time, then a drop-off to cutout voltage. If you pick a voltage too low, the engine will always run to charge the pack. If you pick a voltage too high, the engine will always run to burn energy from the pack. To do this correctly, all you need to do is have a small intermediary that converts the actual SOC of your pack into the expected SOC voltages for the Prius and the Prius will then know your SOC through this intermediary conversion.

The Prius expects 40% (1-bar empty) to be 215.4v and 81% (8 green bars full) at 238.5v all the while it allows discharge and charge curves to swing between 180v and 240v. Perfectly acceptable for NiMH and almost for a lead acid conversion. But not for Lithium.

As a real world example a Tesla Model S (85kw) goes from about 370V at about 20% SOC to 400V at 98%-ish SOC.

Some background:
Worked as a consultant and contractor on 2 EV Startup companies, one for an electric sports car and the other for electric delivery trucks
Built a PHEV conversion for my GenII pack using LiFePO4 cells and my own circuit designs.
Work at a semiconductor company in power engineering where we are constantly getting packs from different auto manufacturers to install prototype non-released parts. Basically I have the packs of a MY 2022 and MY 2024 "Big-Name-German" and 3 "Big-Name-Japanese" sitting in our building hooked up to our equipment doing things that as of today in the public domain are marked as "impossible" or "impractical for production" and will be standard when these things release in 8 years. Car design takes a very long time...

 

Yes. I know it does. Doing a modern EV conversion properly takes a long time. I can't imagine designing a whole car (although conversions come with their own issues). I use a BMS with my conversions, there is a CAN module that allows the car to operate with Lithium. I generally use less expensive Lithium, so the cells are fairly large, 40/50amphour. I have always included a charger, I have never used small cells. I have seen a few packs smolder, I watched (as I'm sure you have) a few expensive lithium packs smolder, melt, never flames though. Lithium iron batteries don't have run away thermal events quite like other chemistries, but also tend to suck in cold weather. The biggest problem I have is cold weather charging, in the cold I usually have to keep a trickle of current from charger to keep a little warmth in the pack. I have just built gauges for torque for engine, mg2, mg1, and mgr (for my 400h) in watts and have been more interested in power flow. I have a techstream scanner, but it takes too long to set up, and it is too big to drive around with a military spec laptop. I have seen current over 100 amps to and from the battery, but never for long. The BMS system I use for the plug in hybrids generally allows for operation in full electric mode, minus mg1 burning off energy spinning the engine in fuel cut up to about 70 mph. If smaller cells are used, I parallel the pack with the NiMh pack, if they are larger or have better cold weather charging, I remove the NiMh pack. Usually I use 84, 40-60ah cells. It is not that it can't be done, it's not that it needs something in-between, it is about making sure the HVB and HV ECU's get tricked into operating in a desirable (but obviously not factory) fashion, with the right equipment. I installed some of those crappy ENGINER kits (thanks for the amazing product support Jack). They work OK as long as you can keep the parts working, and deal with not accepting regen. Their DC-DC converters go bad and their chargers suck, ok,,and so do their cells and BMS....right...it all sucks. His idea was good. It was a cheap way to add some extra economy, but the reliability was horrendous. I worked with a company in CT that was developing a hybrid "bolt on" conversion kit. After much protest from me about the way they designed their Lithium packs, they melted 3 packs. Their hub motors worked pretty well though. Anyhow....gen 2 Prius is the easiest car to do a conversion on, the gen 3 cars are harder to regen into, the Lithium has to cut out during regen.


XT1585 ?

 

as long as the o/p keeps all of this in mind, i look forward to the end result.

 

I have been using this to watch powerflow. It has been extremely helpful. It is much easier than using techstream, I wish I knew more about Android development, I'd like to build a nomograph with the data for these gauges. Does anyone have the expertise to help out? For some reason, I cannot seem to get decent answers on the torque forums, it almost seems the developer has lost interest in the app.
Thanks!


XT1585 ?

 

uski, you're going to have to do allot of engineering if you want to change to a completely different battery chemistry. You need to develop the safety and management circuits for cell balancing and charg management. Then you have to spoof the behavior of the original battery to the CAN bus. NiMH batteries are somewhat dumb and don't need all this.

Now that the patents are off NiMH I'm surprised no one has knocked off the Prius battery for aftermarket.

 

You can build a Nickel battery with cells from Walmart if you really wanted. Check out Orion BMS. They have a good kit for plug-in hybrid conversions. It includes battery management, (not just monitoring like the factory set up) a CAN bus controller and a contactor controller to switch the Lithium out during regen if you are using smaller Lithium batteries. It is available at Ewert Energy Systems. It is not for the faint of heart though. In my opinion, they have the best system for the aftermarket.

XT1585 ?

 

I use alot of 12 volt Lithium Ion cells at work to power test equipment whose batteries cost alot of money or are nla. I power them using 12 volt Lithium's.
One Engineer who does RC planes warned me they'll EXPLODE!! Lots of lithium Ion explosions on youtube. Mostly from being dead shorted.
I was unaware how dangerous Ion's are. So I am very careful with them.......but there kinda scary.

 

Lithium chemistry differs vastly from one type to another. Lithium Iron LiFe cells won't pop unless you dead short them. I have a 100 amp hour battery I made running my tractor, one as the 12v battery in my Prius, I have a Cabrio convertible with 26 KWh worth of Lithium in it, a 1969/Subaru, 4 plug in hybrid conversions, an electric VW bus with an on board 14kwh pack and it has a range extending trailer I am in process of building with a 35kwh pack. LiFe will smolder if you don't treat them right, but they won't pop unless you dead short very large cells. I work with cells from 20 amp-hours, to about 210ah (about 5" x2" x18" per 3.3 volt cell) and they are in series or series parallel configuration for 90 up to about 260volts. Nickel is crap, it is heavy and has no energy density. Depending on the year, nickel Prius batteries are between 6-6.5 amp hour at 273.6 volts for gen 1 and 201.6 for gen 2 and 3. The RC batteries are LiPo4 mostly like laptop or cell batteries and yes, they will explode. There have been many college EVs that have experienced severe "thermal events" hahaha... Using LiPo4. Lithium iron is a bit heavier, but isn't prone to run away self discharge, unless you let the cells swell, increasing the internal resistance, making them their own loads, self feeding smoking, smoldering, melting, and popping. A properly tuned, networked BMS with ground fault detection/protection, and properly adjusted charge current limits and discharge current limits based on temperature and SOC is a must. Discharge "knee" shut down or warning will keep a pack from ending up a pile of self melting goo. I watched the x prize guys I was working with melt down 4 packs of LiFe while designing a bolt on hybridizing kit for commuters. They wouldn't listen to me, tried packaging them in heavy plastic, and....GOO!

XT1585 ?

 

fascinating! have you started your lift back replacement pack yet?

 

I even made a 20 amp hour Lithium pack that was my starting battery for my pickup truck. It was small, and only lasted about 6 months, but it weighed about 8 lbs instead on a 45lb FLA battery, plus there was less electrical noise, brighter headlights and higher ignition voltages. I made an aux. 12v battery for my wife's gen 1 Prius, and have one in my gen 2. My 400h will be getting one as well. XT1585 ? I have two full-blown gen 2 conversions out there using orion BMS. There is a CAN module that takes advantage of different operational modes. One with the nickel pack, and one without. 72 mph EV mode with the engine in fuel cut to keep it from over revving mg1. I have also installed 6 "Enginer" PHEV kits on gen 2s. They were 8kwh packs that use a DC converter to keep the nickel battery at a higher SOC. They give about 22 miles of very careful EV driving, then turn off and run the stock Prius NiMh. The parts are crap though, and the company has not supported installers, or customers with warranty parts. Their BMS would die and ruin the pouch cells and chargers. I have a pile of dead Enginer components that Jack Chen wouldn't warranty or even admit the parts were failing. My Subaru uses a split pack of Lithium, two 85ish volt, 60 amp hour packs (30 amp hour cells in parallel) with a very simple shunt management (that tends to blow shunt transistors and ruin cells occasionally) Two chargers. 4 hours to fully charge. When contacted in, the two packs are in series for about 165 volts. This is about as high as the Curtis controller will accept, and I had to modify the DC brush motor to keep it from the "ring of fire" shorts and arcing to the case, since the motor was only designed for 144v. I don't use DC motors for conversions any more though. I replaced the Lead Acid pack I had originally installed in the subi with Lithium and tripled my range, acceleration, top speed and power, not to mention cutting my charge time by about 6 hours. You have to match the batteries to the application. Very little voltage sag, effortless hill climbing with the Lithium pack. The subaru sat for about 5 years in storage at my ex wife's house. jumped the 12v battery, turned it on and drove it 6 miles to the school I teach at for an AFV day we run after it hadn't seen a charger in 5 years. Lithium has very little self discharge. And oh...2 to 3 times the cycle life of NiMh. I charge it about 6 times a year and use it to make the short trip to my garage from my house. (It isn't registered anymore). Lithium really should have a GOOD BMS system, not these crappy little cell modules offered by some companies that tend to burn out, leaving the pack unprotected. I have not started my non BMS conversion yet, I have to finish the VW van first, and I may decide to do the 400h conversion before my own gen 2. I have a gen 1 Prius also that I may remove the engine and do a pure EV conversion on it. I would probably have to remove mg1 from the transaxle if I want highway speeds from it though and just use the empty case with MG2 and the differential. The car is ugly though, lots of New England road salt rot on it. So. Lithium is now inexpensive, much cheaper than NiMh even. I pay about $1 per amp hour per cell, and I myself have ruined a couple of cells hitting the knee in the discharge curve without a warning from a BMS to shut down or limit the discharge. The other cells will ram power through weak cells and cause cell polarity reversion under load. ONE time is all it takes to completely ruin Lithium cells.
Sorry about the long story, but I just wanted to let everyone know that Lithium is used all the time for EVs and phevs. Proper charging/discharging, temperature and cycle life all affect them, so I have to trickle charge the cheap Chinese Lithium I use in the cold weather to keep the batteries above freezing, or they will get ruined when charging in the cold weather.

XT1585 ?

 

Ewert Energy, Orion BMS and PHEV conversion kits. Best available. Not for the novice though.

 

For anyone seeing this now most of what is discussed in this thread is FUD which is discredited by the simple math of cell specs ala ericbecky. The posted videos show how easy it can be. There are issues in those videos that imo show the author is just smart enough to be dangerous but not fully aware of what he's doing, but tbf that's how most people learn.

If you want the best answer, look at Planetaire for inspiration. 70S LiFe matches the voltage curve and those A123 cells are unrivalled in their performance. They also accept high charge current, which is an important concern when choosing cell configuration since most cells do not handle fast charging well (LTO being a notable exception).

 

doc prius looks like a decent turnkey solution