DIY liquid cooling/heating for battery modules?

[BigChief]

I have 14 of the Tesla Model S 24v packs that I'm retrofitting to a DIY electric vehicle.

They all have the water-cooling built in. I have 3 of the OEM Tesla coolant pumps, and a 250w resistive inline coolant heater/pump from a Smart ED.

My main question is, can I just make a single loop by running vinyl tubing from pack to pack, or would that require too much pressure? Or am I supposed to take the Tesla pump, run 14 inlet and 14 outlet taps off it? Or use several of the pumps together?

I guess what I'm trying to figure out is "series" or "parallel", but for water instead of electricity

 

[Electric700]

Welcome to the forum! I believe that li-ion batteries for the Model X / S use a type of glycol coolant, not water. I think a parallel configuration would be better if you can manage it. You may want to do testing with a load on the batteries to see what the benefit is.

Interesting project!

 

[miimura]

I disagree. You would be better off running all the modules in series. That way the flow through each module is equal. If you want to know how the Tesla pack is connected, look through the pictures that wk057 posted when he broke down the pack. Also, you should check the chemical compatibility of G48 coolant with Vinyl tubing. I would be more comfortable using Polyethylene or even better, Polyurethane tubing.

 

[Electric700]

So you have no concerns about hot battery coolant inputs with a series configuration (hot coolant going to subsequent batteries from initial ones)? That's why I said in my post "You may want to do testing with a load on the batteries to see what the benefit is". There are 14 battery packs which is a lot, so perhaps a mixed series / parallel configuration might be an option too.

 

[tga]

But can you guarantee equal flow through all strings in a parallel config?

 

[Don85D]

I would build a manifold that allows parallel coolant flow to all of the input ports. A series connected system will have temperature variations that may be beyond acceptable range without carefully measuring this in advance. A parallel system is simplest, in my opinion.

 

[Rocky_H]

Huh. I can see the pros and cons of each of these. I might be a little concerned about going totally parallel because there is so much opportunity for most of the flow to continually go toward other modules if one or two have restrictions/kinks in their flow somehow and just never get any movement of their coolant. But yeah, series does have that problem that you know there is going to be unequal heat levels among the batteries. I'm sort of inclined toward a serial / parallel configuration that @Electric700 mentioned, to avoid the really big problem with each of these. With less in parallel, it would have a bit more pressure on each, so more likely to keep coolant flowing in all branches. And it would make for fewer in series than all of them, so it would lessen the temperature disparity.

Heat buildup in the batteries would be based on the charge and discharge power levels. If your home-built EV has a somewhat lower power motor than Teslas normally do, it would draw power at a lower rate and not heat up the batteries as much. And you won't be using Supercharging, so if you can keep your charging rate to 10kW or less, it won't build up much heat. So I guess I'm leaning toward series not having that much of a temperature difference issue.

 

[Timberframer]

This is a piping method that will equalize flow in every pack. There are no special valves or devices involved. I have used this method in my Geothermal work for many years with complete success. I have used it with as few as 2 loops and as many as 12 loops. You could series two packs and parallel seven sets of two for a total of 14. If there are an odd number of loops they would all have to be parallel.

It is referred to as reverse return. This will give you even flow in each loop.

Rules: Keep each loop the same length.

Use the same number and types of fittings in each loop.

See image

 

[oudevolvo]

I did a flow test with three Tesla modules and compared parallel and series.

What I found was that even when only connecting three modules in series there was quite a significant pressure drop and thus the flow was low. Additionally the last modules in the row will get less heating/cooling than the first ones.
I'll be implementing them in parallel.

 

[Electric700]

I did a flow test with three Tesla modules and compared parallel and series.
...
What I found was that even when only connecting three modules in series there was quite a significant pressure drop and thus the flow was low. Additionally the last modules in the row will get less heating/cooling than the first ones.
I'll be implementing them in parallel.

Really interesting news, thanks! I hope your project goes well.

 

[GyroGypsy]

I did a flow test with three Tesla modules and compared parallel and series.

View attachment 296777

What I found was that even when only connecting three modules in series there was quite a significant pressure drop and thus the flow was low. Additionally the last modules in the row will get less heating/cooling than the first ones.
I'll be implementing them in parallel.

What is the fluid pressure upper limit and nominal flow for model S batteries?