openinverter forum

I’ve been wondering it I could fill my battery boxes with a non conductive fluid for a total immersion thermal management solution for my bmw phev modules, perhaps even pump to a radiator, just an idea at this point has anyone else tried something similar?

Something like this

https://www.futurebridge.com/industry/p ... y-cooling/

At hundreds of volts I would think you would struggle to find a liquid that is a good enough insulator.

Damian did a video on his E39 where he just used a big sheet of aluminium as a heatsink if I remember correctly he pulled 100kw and it didn't get even warm.

I also went a bit overboard overthinking it basically all we really care about is ensuring they do not get too hot or charge when it's too cold, the solution is really simple heater pads for when it's cold on charge and big sheet of aluminium as a heat sink and shut down if it gets too hot.

LION Smart flooded the whole battery with 3M Novec.

https://www.lionsmart.com/

So it looks like silicone based transformer oils are up to the job of being fire resistant and non conductive but it’s no cheap and comes by the drum, so may well be a non starter

Think I’ll be testing the setup dry and monitoring temperatures before I decide if further cooling is required

The thing to remember is based on the acceleration over the past 5 years of ev take up and how the prices have come down (I paid £85 &95 per kwhr), think what will be available in 5 more years, you have to think of cells as disposable (or at least useable for solar once their capacity degraded).

I'll admit that it is a very hard pill to swallow, particularly as I am a person whom has serviced all of my cars at less than their service intervals (by myself as it is really hard to find a good mechanic), do the thought of not actively managing the temperature goes against the grain, but all we need to do is actively monitor to make sure we are safe.

What pack did you get rstevens82? Where did you get it also? Ebay prices seem much higher than that.

Got amazingly lucky on eBay with (a cheeky) best offer with a VW Passat gte (9.9kwhr) just after the 1st lockdown, then got the 2nd pack in November/December.

Ironically I have cooling plates built in but currently no balancing, so my plan is to manually balance the cells and see how long it iltakes for them to drift, if it takes 6 months or more I can live with that, worst case I replace the BMS.

SciroccoEV paid £100 kwhr for the i3 pack which is probably better battery:
viewtopic.php?f=11&t=958

Dave black brought a crashed leaf which is actually the most sensible thing to do if you have the space (not sure what he paid but probably most cost effective overall)
viewtopic.php?t=724

Forgot the obvious 120euro/kwhr that's actual capacity, if you go by original (sticker) capacity it's 2000/23=87euro/kwhr
viewtopic.php?f=12&p=23587#p23587

Just keep your eyes peeled and keep your options open.

ps I know I was very lucky as eBay prices fluctuate massively and I expect the b word has messed things up a bit.

rstevens81 wrote: ↑Thu Feb 04, 2021 10:33 am Got amazingly lucky on eBay with (a cheeky) best offer with a VW Passat gte (9.9kwhr) just after the 1st lockdown, then got the 2nd pack in November/December.

Ironically I have cooling plates built in but currently no balancing, so my plan is to manually balance the cells and see how long it iltakes for them to drift, if it takes 6 months or more I can live with that, worst case I replace the BMS.

Good job!

Regarding battery drift i think if you use them from 4.0V to 3.0V per cell you will not have much drift. I have noticed a certain drift between cells gets established probably because of diferent conditions inside battery boxes. Some of them may be in the cented where they are warmer and some are at the ends... Which affects the voltage state, but not SOC! I repeatedly saw cells diverge some and i left the cells for some time without heavy maintenance. And it seems they went into their own SOV (state of voltage ) without affecting SOC. And since i didnt discharge them too much everything stayed the same...

rstevens81 wrote: ↑Thu Feb 04, 2021 10:33 am Got amazingly lucky on eBay with (a cheeky) best offer with a VW Passat gte (9.9kwhr) just after the 1st lockdown, then got the 2nd pack in November/December.

Getting a Golf GTE pack delivered this afternoon for £500.

Have you reused the precharge and contactors in the original pack?

[/quote]

Getting a Golf GTE pack delivered this afternoon for £500.

Have you reused the precharge and contactors in the original pack?
[/quote]
Well done good price.

Yes am planning to re use as much as I can! I am seriously impressed with the thought out design.
I would like to labotomise the control module if I can to re use the shunt, but is probably beyond my skills set, if I fail I will use an ivts500.

Here's my project, zero ev with the twizzy had a video
viewtopic.php?f=11&t=1111

The Toyota fluid might be up the the job. All the 'Yota hybrid motors are immersed in the fluid and I have not seen reports of a conductive path from it.

Would require validation obviously.

We have done total immersion testing using Mineral Oil. The picture shows a pair of Chevrolet Bolt Cells being tested. We did 100s of hours of cooling tests using different configurations of cooling plates. Immersion cooling was by far the best. In the end, we decided not to proceed due to the logistics of building a fluid tight battery box, overall weight etc., but we will probably use this for a future pack. Total immersion cooling is used extensively in the Bitcoin mining world. They immerse entire servers without problems.

dougyip wrote: ↑Wed Feb 10, 2021 5:19 am We have done total immersion testing using Mineral Oil. The picture shows a pair of Chevrolet Bolt Cells being tested. We did 100s of hours of cooling tests using different configurations of cooling plates. Immersion cooling was by far the best. In the end, we decided not to proceed due to the logistics of building a fluid tight battery box, overall weight etc., but we will probably use this for a future pack. Total immersion cooling is used extensively in the Bitcoin mining world. They immerse entire servers without problems.

Do your battery temps get very high while on track? While far less abusive load than a race course, when doing everything in my power to be in constant violation of posted speed regulations, I rarely see battery temps get 15F above ambient. I did see a fair temp rise in changing from a 3.3kw charger to the 10kw Tesla charger so I would certainly believe that any sort of fast charging is going to benefit from more advanced cooling.

dougyip wrote: ↑Wed Feb 10, 2021 5:19 am We have done total immersion testing using Mineral Oil. The picture shows a pair of Chevrolet Bolt Cells being tested. We did 100s of hours of cooling tests using different configurations of cooling plates. Immersion cooling was by far the best. In the end, we decided not to proceed due to the logistics of building a fluid tight battery box, overall weight etc., but we will probably use this for a future pack. Total immersion cooling is used extensively in the Bitcoin mining world. They immerse entire servers without problems.

Liquids, such as cooling fluids, are often considered "non-conductive" at "safe" voltages, as in below 60v. Once you get into the hundreds of volts, conductivity does not scale well. Tests at 12v will not indicate results at 300v, much less 600v, it's called "breakdown voltage" and it even applies to wire insulation. I will try and find the dielectric constant for mineral oil, but even subtle differences in chemistry can have huge impacts on conductivity.

<edit>: Well air has a dielectric strength of about 3MV/m, so anything greater than that should be just fine, as long as it is not contaminated

Battery temps can get very high road racing. Our Kia air cooled pack would go from ambient to 65 degC in 20 min. Our Bolt pack with OEM style indirect cooling plates would hit 65 degC within a couple of laps. It wasn't until we went to full liquid cooling plates sandwiching each cell pair that we could run an entire 20 min. Now with our Model 3 pack, temperatures barely hit 50 degC.

Mineral oil is used in HV transformers in the KV range, so no problems at our voltages. We were actually going to use AmpCool: https://www.engineeredfluids.com/ampcool mainly due to it's low fluid density. Phase change fluids like 3M's Novec are the the best, but are extremely expensive.

If you want to play safe, you could separate the pack into lower voltage segments with separate cooling oils. Then you can tolerate cheaper oil and more contamination. It would be useful to know how the oil reacts when contaminated by vented gas from the battery, as that's definitely something that can't happen in a transformer.

dougyip wrote: ↑Wed Feb 10, 2021 4:22 pm Battery temps can get very high road racing. Our Kia air cooled pack would go from ambient to 65 degC in 20 min. Our Bolt pack with OEM style indirect cooling plates would hit 65 degC within a couple of laps. It wasn't until we went to full liquid cooling plates sandwiching each cell pair that we could run an entire 20 min. Now with our Model 3 pack, temperatures barely hit 50 degC.

Good info. Numbers sound familiar so you’ve probably already answered that at some point for me . The immersion stuff sounds pretty interesting.