Gen II Prius Individual Battery Module Replacement

I was being facetious. I have extensive experience with NiMH chemistry in the Honda IMA world and have a lot of chargy-dischargy goodies. I have a total of 11 IMA packs across G1 Insight, HCH1 and HCH2 to services my 2X G1 Insights and 2X HCH2s.

162k miles with most of them here in the Phoenix area. It was originally sold in AK. There are red magic-marker dots at the (+) end, so I expect this is its second pack. Sealing goop everywhere and KOH crystals sneaking past the seal. Pack was flat for about 3 months. Prior owner had it for about 100K miles. Once it started showing battery issues, he parked it, looked for options and decided to list it on Craigslist for salvage value. Unfortunately, I bought it from the buyer who bought it for salvage value at a small premium.

Torque Pro w/OBDII bluetooth reader gave me ample warning, but I didn't realized what I was seeing.

AVERAGE module voltage in the pack was 1.44V. The two that recovered were 6.74V and 7.26V. I started with 1C charging and ~3A discharges.

 

Hey Steve,
It sounds like you've been working with the Honda batteries for quiet some time. I admit that I don't know anything about the IMA sticks, the demand for Honda batteries in my area doesn't compare to the Toyota batteries. For that reason I've never had an IMA pack in my shop.

I got a little lost with your Gen1 Prius, I think you said you got it around 100K and it was already showing signs of problems. You put the other 62K on it in AZ. Which work was done to get it going again at the 100K mark? Did the battery go flat and sit at 100K or at 162K, or both? Either way I feel like the result is the same. Once the modules self discharge past a certain point they are damaged and have to be replaced.

I like your cycle details, it's not the way we do it but its something new that I don't see mentioned very often. One thing that we do is slow the charge rates way down and use some very high power blowers that are ducted into the bottom of the battery assembly. This removes the waste heat in the exact same manner that Toyota designed it, the only exception is that we extract from the bottom up which makes more sense. Lowering the temperature also lowers the voltage as the modules charge and allows for a much more successful balancing of the cells without the risk of damage due to heat and over pressurization.

I would be concerned with HOW deep you are going with your discharges. I'v never tried it like that before but it seems like it could have negative side effects. Like I said before, everything I do is designed to be just enough to get the desired results without putting any added stress on the batteries.

I have no doubt that with all the tests that your doing and data your collecting you will be able to positively identify the healthiest 38 modules from your 2 second gen packs and build a dependable long lasting battery for your Prius. If you email me your address I will ship you a set of new nickel plated bus bars that we have manufactured here in Dallas. I learned a long time ago that cleaning those old bars is miserable and no matter what we did they would still corrode again.

Best of luck to you Steve,

Matt

 

Matt,

To clarify, car was bought new in AK (temp extreme). It currently has 162K on it. It has been in AZ for at least the last 100K miles. I have never driven it. It was towed to my driveway with a dead pack. I realize now I included MANY non-pertinent details

I'm a little confused by your description. Are you doing a whole pack charge, or doing individual modules, or "I can neither confirm nor deny that, but I know the answer."?

When I'm doing whole pack charging, it's @ 350mA or C/18. I'm sure I'm not getting as much cooling as you are, but my box-fan is blowing through an elevated pack, so I'm getting airflow above and below.

I understand your reluctance. I have been very reluctant to do deeply discharge packs/modules as well. Honda IMA "D" cells and prismatic modules are both made by Primearth (Toyota owns > 80%), so I have to believe they share some common positives such as reasonable polarity reversal tolerance. So many examples of IMA sticks with 5 very healthy and robust cells, yet it's limited by one turd that vented at some point and became permanently imbalanced from it's brethren. Multiple that by 120-132 cells within the pack, and it's pretty obvious why the IMA packs aren't as reliable as the Toyotas.

A pioneering member of the IMA community reverse-charged a "D" cell at 2A until it hit -1.8V. He then resumed normal cycling @ 15A and observed no deterioration compared to other individual cells that were not reverse-charged. performance after 50 full cycles showed no deterioration. If it weren't for this evidence, I wouldn't be at all comfortable with deep discharging where I know cells will get reversed.

Thank you for the vote of confidence. I'm logging data and plan to post a summary in the Gen1 forum.

2 day soak in sulfuric acid (battery acid i have leftover from dry-shipped motorcycle batteries). Dilute and rinse. Add baking soda and agitate. Rinse. 10-20 strokes on green Scotchbrite pad per side, rinse, dry and done. Takes about 20 minutes of touch time to do a gen1 pack. Still, would prefer to never do it again... PM will be sent shortly. IMA doesn't use copper. Likely nickel or nickel plated.

Thanks,

Steve

 

Steve, in our shop everything is done on module basis but with the core of the pack assembled. By that I mean it's just the modules and the end plates with the bars connecting them (That's the long metal bars, not the bus bars). I'v never seen the benefit in doing any testing with the modules connected in series, plus it exposes the tech to potentially lethal voltages. I have to be able to record the data from each module or I would never be able to identify those that have been damaged and need replacement. This is also my issue with grid chargers for the purpose of repairing batteries. They are very viable as a maintenance tool as long as the overall health of the pack is good.

I agree that at such a low charge rate the cooling probably isn't as much of an issue. My rule of thumb before we designed our cooling system was that if any module swelled enough that I could slide a credit card between it and the next module, then I didn't like it.

Thanks for the info on the deep discharges. I know from my work that these NiMH batteries are pretty tough but I'd v never thought that you could reverse charge one that far and ever get it back. Very cool. Maybe I'll get some weak modules and run some tests like that to see how they respond.

I'll look for your PM and send you that stuff in the morning.

Matt

 

Got it. Thanks. I will resort to module work for final testing. I'm just hoping to eliminate the bulk of the individual module cycles.

Thanks again,

Steve

 

Matt is true to his word. Nickel buss bars and a pack of new nuts to boot. Quivering in my hot little hands...

Now if I can just gnaw my way through these packs....

Thanks Matt!

Steve

 

Matt, you are using the exact same process I used on mine (ie set a minimum health for modules in the pack and anything less gets replaced) and it worked very well but can you share a bit more on what sort of charge/discharge cycles you use in order to bring them back? If you can't do it in here I would love to have a PM on it if you wouldn't mind just since I wanted to try and see if it worked equally well on the Highlander too.

 

I have a 2007 Prius with ~ 118K miles which, after repeatedly draining the starter battery, showed the red triangle of death and various error codes and warning lights indicating the traction battery needed repair. The car had been mostly sitting since January 2015, and sat some more after the codes showed up in August or so. After researching my options, I found your forum and decided to try to recondition my modules using a Hitec x4 battery charger and the profile suggested by ryousideways, namely:

3 cycles of
Discharge --> Charge
Discharge at 0.7A to 6.0V
Charge at 5.0A to 7250mAh
timer limit 720 minutes, cool down 5 minutes

I have been cycling through 4 modules at a time since 10/28; data from the first 4 was wiped out somehow, but the remaining 16 completed so far show the following:
Module Discharge#1 Charge#1 Discharge#2 Charge#2 Discharge#3 Charge#3
7 (no data)
14 (no data)
21 (no data)
28 (no data)

6 3270 7250 5005 7250 5281 7250
13 2693 7250 4833 7250 5253 7250
20 2972 7251 4755 7250 5236 7250
27 3710 7250 5395 7250 5592 7250

5 2599 7218 4527 7250 5110 7251
12 2381 6491 3892 6739 4200 6986******
19 2860 7250 4616 7250 5103 7250
26 3704 7250 5274 7250 5474 7250

4 3767 7250 5340 7250 5506 7250
11 2940 7250 4827 7250 5212 7250
18 3343 6363 4768 7250 5132 7251
25 2980 7250 4793 7250 4870 7250*******

3 3507 7250 5107 7251 5438 7250
10 3343 7250 4911 7251 5311 7250
17 92 2731 1086 3559 1015 3122*******
24 3453 7251 5074 7251 5426 7250

Module #17 is clearly bad (and it was showing low voltage before I started), and #12 and #25 are suspect, but my question is this: are all of these modules suspect? Based on the writeup of ryousideways and others, I had expected most of the modules to come up to 6500mAh or so, and instead they are in the low 5000's. I can do another conditioning discharge/charge cycle, but if these batteries are hopeless, I don't want to waste my time.

I do not think this is the original battery for the car based on its having missing/non-uniform fasteners when I took it out. If I replace the substandard modules, is it likely that I'll see multiple module failures on a regular basis, or will it just turn on the ICE more often due to the diminished capacity of the modules?

Also, how does one measure the internal resistance of the module?

Thanks for your help!

 

Those modules are def usable. I would replace all suspect modules though.

 

The expectation that cycling can restore a NiMH module to 100% rated capacity is very unrealistic. There is a natural 1-2% deterioration per year from aging.

Steve

 

I would be also interested in them being a bit more consistent with claims. When Matt suggested that they change up to 10-12 modules in a failed pack - it would indicate that the rest of the pack would have degraded significantly and have lost more than 50% of their Design Capacity. Fact.
Matt from Texas Hybrid Batteries decided that they have a right to make statements on behalf of all readers of this thread by labelling very negatively my previous posts.
I have seen that every time I post challenging questions and someone representing snake oil business becomes "unhappy" here - there is a spark of the hackers attacks on New Zealand site coming from US. Not cool...

Matt, please share a light on your "process" of resurrection of degraded modules and bringing them up to specs. What is the final test you are conducting to measure claimed “rebuild” capacity?

 

I had exactly this situation and was too lazy to replace #1 & #28 (and was a bit suspicious). I replaced #14, and rebalanced (just wiring in ||, not charge discharge cycling). Six months later, everything is fine and in testing with torque under load, end pairs 1 and 14 are my strongest (max deltaV is 0.3V). I think the end ones can get selectively drained under certain conditions. Not sure how, since current must pass through all in series.

Damien

 

Steve has some good points in here but a few items concern me as well. Everyone doing this as a DIY project (ie no professionals) will likely be using the similar "weak chargers" like you get from hobby shops for use with RC batteries. The chargers have real limitations in that you can usually charge up to 5A and discharge up to 1A, certainly less than ideal when working with Prius batteries. So that being said they are perfectly adequate for providing relative (not absolute) capacity measurements. This works quite fine when creating a baseline for capacity within the battery pack. We also know heat/swelling of the packs can cause damage to otherwise good modules so going beyond the module's ability to receive charge with higher current is not good. Earlier in this thread you can see where someone had barely 2000 mAh across their pack yet managed to replace the damaged module(s) and put the pack back in service.

Remember the point of this exercise is to find relative capacity numbers without damaging any modules and running charge cycles up to 8000 mAh with higher currents is asking for trouble in a DIY situation. This is why I stick to 7000 mAh when doing mine. The batteries can accept very high currents in use (100A+) but remember they also only go up to 80% SOC too in order to do this. Using very low currents (ie less than 350mA to keep heat/pressure down) then you could go up to 8000-9000mAh just fine but that takes time and many people aren't interested in experimenting as much as they are in getting the car back on the road ASAP. Finally the only way to get a good check on internal resistance in a DIY situation is to use your mini VCI to check the modules while still in the vehicle but an equally reliable method is to just load test them for 2 minutes each with a 200w bulb (measuring voltage at the start and end of the test, then calculate voltage drop), that should be more than enough to sort out any bad ones and allow you to match new modules for use in your pack.

 

Lil' bit, but he's been gone over a year...

Insight Central: Honda Insight Forum - Search Results

I had a situation where modules that seemed good by the most typically used measures (voltage and low current discharge capacities of 5000-6000mAh). When I tested them at 20A discharge, they essentially had zero capacity. Of course, this was an IR issue and can be detected with IR measurements. Which proves to me that one or the other MUST be done (high discharge or IR measurements)

Thanks,

Steve

 

Like you mentioned above, here you can see where my modules all tested fine for capacity using a pair of Hitec X4 AC chargers. (Quad chargers are a huge help when doing a rebuild quickly since cycle times are very long at 1A discharge)
Gen II Prius Individual Battery Module Replacement | Page 12 | PriusChat

Yet when I went and load tested it with a headlight (your 200W load) you can see how the bad one showed up quite readily.
Gen II Prius Individual Battery Module Replacement | Page 13 | PriusChat

Once I found the bad one I was able to test the new module and verify it would fit well with those in the current pack before install. Someone more knowledgeable (JeffD, Eric) please correct me if I'm wrong but, you mentioned you didn't understand how the load relates to IR but I believe the internal resistance of the battery is directly proportional to voltage drop under load hence why the load test actually works. Point being that you don't need to calculate the actual internal resistance, you only need to see its effect in order to find the bad one and verify others are in relatively good shape compared to one another (as you alluded to in your final comment above).

 

AH... I get it. You're using the load test to compare voltage drops, NOT compute IR. That makes a lot more sense. Healthy ones will have drop "X" and unhealthy ones will have drop > "X". The load test effectively compares IR without measuring it, much like my 20A discharge test.

Steve

 

I'll be curious to see how your method works as you document and take photos. That charger didn't exist back when I did my rebuild (this is why I went the quad charger route to save alot of time (I could do 8 modules at once). I also went with a plug in AC version so I didn't need to deal with battery banks or additional power supplies which add cost. I would surmise that as there are alot more cheap high power lipos in the market now (like Turnigy, Multistar, etc) for drones and such and I imagine they will also be coming out with chargers that can be more aggressive on the discharge side as well in the future (should be a great help in the rebuild process). I did cycle each module 3 times and more if needed (some continued to improve) so not sure if that is your plan too or not but I guess we'll wait and see what works best with your equipment.

 

I test all of my modules at a 20 amp discharge rate to 5.4 volts. With five dischargers going I can run through a whole pack in about 1 and 1/2 hours. And during that time I can get most of an already tested pack reassembled.

One caution is that for any comparative test to have meaningful results consistency is required. The modules must start at the same full SOC arrived at in the same manner. I grid charge to over full and let them sit overnight before discharging. Voltage alone is not a good indicator of SOC. Paralleling modules will not bring them to the same SOC, it only matches surface charge.

Brad

 

Brad, this sounds like a professional shop.......what equipment are you using for grid charging and discharging?

 

The major downside to this is that a severely imbalanced module will have a cell reverse during the taper at 6V and will maintain a fairly high current as the other 5 cells drop in voltage. This is not healthy, so a certain baseline performance should be established before the tapering current method is applied.

Steve