Why Tesla's Model 3 Will Use Liquid Cooling

| About: Tesla Motors (TSLA)


Liquid cooling is objectively superior, while air cooling would cause the Model 3 to fall short of performance specifications.

Implementing an air cooling system could be more expensive.

Such a design is not consistent with common sense, nor with the way Tesla has behaved in the past.

This article is intended to counter speculation that Tesla (NASDAQ:TSLA) would use an air-cooling system, as presented in Mr. Carlson's lengthy and high-quality article "Tesla's GigaFactory: A Christmas in July?" I recommend reading that article before reading this one if you have time.

Air cooling is not an innovation.

In his article, Mr. Carlson calls air cooled batteries an "Innovation". Well, take it from Elon himself: he called the air cooled batteries on the Nissan (OTCPK:NSANY) Leaf primitive, and frankly, he is 100% correct. Air cooling is objectively inferior, and does not fit Tesla's goals for the Model 3, which could be summed up as:

Cost, Performance, Lifespan

In the article, Mr. Carlson believes that "more aggressive" battery designs will lead to efficiency gains, leading to weight and cost reductions. I can certainly see this happening, though not in the manner he described (his cavalier and, in my opinion, inappropriate use of Monte Carlo simulations is a topic for another time).

These hypothetical "more aggressive" battery designs he suggests are only achievable with liquid cooling. The further the limits of the materials are pushed, the more important proper cooling becomes to the integrity of the cells. The laws of thermodynamics do not let us have something for nothing, unfortunately. Mr. Carlson's efficiency argument is thus predicated upon a liquid-cooled system... which is precisely why Tesla used liquid cooling in the Model S.

Air's specific heat capacity is much lower than liquid coolant. All that work making the batteries more efficient is for naught if you sacrifice the quality of the cooling system.

As the Department of Energy's research shows, air-cooled batteries run at low resistance to reduce heat "has the highest costs of the four options shown due to the cost of its high excess power". The Model 3's objective is to be the cheaper mass-market offering, meaning that choosing the highest-cost option is highly unlikely.

The importance of liquid cooling is compounded by how heat affects the lifespan of Li-ion batteries. Because of how important it is to keep the batteries cool, the car's governor will likely be designed to detect hot batteries and will throttle the amperage that is sent to the motor. The only way to compensate for that in terms of power would be to have more batteries, which would run counter to Mr. Carlson's argument that the Model 3 is projected to be too light to have a liquid-cooled system.

Mr. Carlson seems to think a closed-air system with actively cooled air will solve the problem (as opposed to using an open-air cooling system). I am curious to see how a Model 3 that must power an A/C unit to cool the batteries will also do 0-60 in 6 seconds. Something's gotta give, and I'm thinking it's not going to be the 0-60 time proudly listed on Tesla's website.

In addition, air is notorious for expanding and contracting when subject to changing temperatures. This means that the air-cooled system would have to withstand wide pressure swings that liquid cooling systems do not. The materials would need to be much more durable and robust to implement a closed air cooling system, resulting in higher costs.

If the air used for cooling were to have direct contact with the cells as Mr. Carlson proposes, the battery packs themselves would have to be air-tight. This means that the battery packs would have to be pressurized in the Gigafactory, then re-pressurized as they are connected to the car's cooling pipes! You can see how this quickly gets overly complicated, especially compared to the liquid cooling system Tesla already has that has been scientifically proven to work better.

One last piece on performance: The flip side to an air cooling is that the same system is also inferior for heating the batteries. The batteries need to be heated cold weather to perform well (see DoE link). The Model 3 would take longer to reach operating temperatures and would perform worse if an air system were used.

Aside from cooling, there are other metrics that give the air-cooling idea a failing grade:

Production Efficiency

Why would Tesla implement a separate cooling system, and thus a separate battery manufacture, assembly, and installation process for the Model 3, when they are going to be assembling the car in the same plant as the Model S and Model X? Doing so would hurt the economies of scale the Gigafactory is supposed to deliver, as Panasonic (OTCPK:PCRFY) would have to manufacture both the 18650 and the 20700 cells, and then Tesla would have to pressurize the 20700 cells, after spending an exorbitant sum designing pressure-resistant cases.

Why not just stick to liquid cooled 18650? Adding complexity by creating a whole new battery and cooling system design further has the potential for production mistakes and/or shutdowns. I do not see Tesla incurring production efficiency losses and risking delays to implement an air-cooled system that its leadership thinks is terrible.

Sales, Warranties, and Cost (again)

Tesla likely needs to offer a battery warranty on the Model 3 similar to the one they offered for Model S, if they are to convince on-the-fence reservation holders and other potential buyers. This makes battery life important to Tesla financially - if an excessive number of batteries fail during the 4 year, 50,000 mile warranty, Tesla will lose money. Compared to that outcome, the cost of a liquid cooling system is worthwhile.

It's not as if liquid cooling is expensive. I have experience with these systems - even my computer (relatively cheap compared to a Tesla) makes use of liquid cooling. The systems that I installed cost me just shy of $100. These are retail cooling systems, so we can expect that Tesla is able to acquire these systems at a discount per unit of area cooled compared to what I got my two cooling systems for.

These retail products have to do the same things Tesla's liquid cooling system does - not leak all over the electronics when subjected to temperatures of 70+ degrees C. Heck, Tesla's liquid cooling system doesn't even need to be as robust as my cooling system (though it is likely more robust than my system). Tesla's cooling system cannot allow the batteries to reach much more than 55 degrees C, lest the batteries degrade significantly (per the previously referenced DoE PDF). Let's compare those operating temperatures to my graphics card running a stress test intended for low-end PCs:

The average temperature quickly ramped from 37 C to 47 C in no time, and if I set it to max temp achieved...

My card hit 81 C during its short 2-minute stress test designed for low-mid range PCs. My card often runs much harder than this for hours, pushing close to 90 C on summer nights. I have yet to develop a leak on my coolant.

"What if debris hits a cooling line as the Tesla drives around?" The coolant is completely shielded by the car's body, so I don't see how that's going to be an issue. Tesla's cooling system is likely designed to travel over 100,000 miles without issues.

Leaks are simply a non-issue for liquid cooling systems, as most systems fail before they do. They are a not expensive, especially when compared to the task of designing a whole new airflow scheme to be used in a mass-manufacturing process that risks excessive warranty claims.

Both Tesla's and Panasonic's processes and goals are already geared towards liquid-cooled batteries.

"Tesla-Panasonic are also free to design aggressive, high energy and power density cells using high-energy cell chemistry because the resulting battery design is fault tolerant and robust against cell-to-cell capacity variations. This will further lower the cost of batteries and the cost and weight of Tesla cars while improving performance and reducing SuperCharger charging times."

Before I begin, I can't let the use of "Tesla-Panasonic" slide. As Montana Skeptic observes, Tesla and Panasonic must treat each other as independent contractors. Using the phrase "Tesla-Panasonic" is misleading, considering that names like Fiat-Chrysler (NYSE:FCAU) and Activision-Blizzard (NASDAQ:ATVI) refer to single companies.

I digress. Make no mistake, Panasonic is the one designing the batteries. While I am sure they are open to input from Tesla, there's not much that can be changed at this stage of the game. The Model 3 design is either finished or nearly finished, and that design will require a specific type of battery.

Panasonic undoubtedly has little appetite for changing their battery manufacture plan now, as the manufacturing equipment is likely already en route to meet their ahead-of-schedule November estimate. The last thing Panasonic would want is to randomly decide to change the design of, or push the limits on, their Li-ion batteries only to have them start failing resulting in an expensive recall.

From Tesla' perspective, I don't think they are keen on pushing the limits on their first mass-market car, either. Even if each individual battery pack is relatively robust, reliability matters - having to replace faulty cells even 10% more often is not worth a 10% energy efficiency gain.

The customer is not terribly likely to notice that their batteries are 10% more efficient than promised. What they certainly will notice is if cells start failing because the cooling system is too weak.

The Model 3's designers are already pressed for time if the late 2017 production start is to be adhered to - they're going to avoid re-inventing Tesla's currently functional designs wherever possible. Designing a whole new cooling system runs counter to the logic of this time-crunch situation. The same logic applies to the single-cell replacement idea that Mr. Carlson floated:

On the single-cell replacement idea

We know that air-cooled batteries will have a worse life, but let's say that air cooling will be used anyway, to make the Model 3 cheaper overall. Not because it's cheaper to implement the "innovative" air-cooled system, but because air-cooling allows Tesla to replace individual cells at service centers.

The extra space needed for the air-cooled design enables the replacement of individual cells, which means that more individual cell failures due to lack of even cooling should be offset by the fact that Tesla won't have to replace entire packs, only individual cells.

Well, there's a few problems with that line of reasoning. Even if we simply write off customer dissatisfaction and increased cell failure risk as having no monetary value, diagnosing which cell(s) amongst the dozens in the pack has/have failed would take quite a bit of time in service centers. If you've ever worked with electrical circuits trying to find a failed component with a voltmeter, you will know exactly how time-intensive that can be.

I do understand there are fuses to prevent failures from cascading to other cells. However, even in the ideal situation where obvious electrical burn marks are visible around a specific cell or fuse, a Tesla employee will likely need to test every single cell in the affected pack to ensure that the entire pack is functional regardless. The last thing either party would want is for the customer to have to come back a second time. This process would take time, and thus would cost Tesla money.

Compare that to this more efficient, cheap, and customer-friendly process: replace the pack entirely, and then ship the faulty pack to a central location be repaired. This process favors the design used for liquid-cooled batteries. In-house repairs far less efficient than having a central battery refurbishment shop.

I'll end this section with what convinced me that single-cell replacement is not on the cards, courtesy of Mr. Carlson:

Notice how in this flowchart, the batteries are held as modules, not cells. I find it highly unlikely that Tesla would ship individual cells to their service centers when Mr. Carlson asserts that Tesla would never possess individual cells to send.

A brief interlude

While we're on this flowchart, the context Mr. Carlson used it in was to assert that the safety stock is not a financial issue for Tesla because the Panasonic cells Tesla is buying are held in Tesla modules rather than in cell form.

Rational people think at the margin. It doesn't matter whether the contract measures the number of batteries that need to be purchased in terms of Tesla modules or Panasonic cells. It's a meaningless distinction - the module, as far as the contractual obligation is concerned, is essentially the part where Tesla make the set of Panasonic cells theirs, completing the transaction.

The bottom line is that Tesla is locked into buying a certain number of batteries - whether you count them in cells or modules is up to you - which has the potential to be a problem if demand were to fall short of projections.

The pieces just don't fit.

To conclude, I'd like to leave with this thought: an air-cooled battery system doesn't fit Tesla's "psychological profile", if you will. Tesla's brand value has come from their willingness (or perceived willingness, depending on whom you ask) to do things the right way, regardless of cost.

If there's a hint of a seatbelt defect, they'll recall all affected cars. If there's one relatively small fire due to circumstances outside of Tesla's control, they decide to take control back by adding additional underbody protection.

There's simply no good reason for Tesla sacrifice production efficiency, the performance of the Model 3, and possibly brand quality, all while increasing costs, just to implement a cooling system that takes up more space and is considered to be inferior by both Tesla's insiders and most observers.

Liquid cooling is inexpensive and higher-quality. Thus, I fully expect Tesla to instead stick with the tried-and-true liquid cooling system.

Disclosure: I am/we are short TSLA VIA LONG-DATED PUTS.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I may initiate a SCTY short position in the next 72 hours to hedge against a possible failure of the proposed acquisition by TSLA. I do not plan to take a position in Panasonic at this time.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500. Become a contributor »
Tagged: , Auto Manufacturers - Major
Problem with this article? Please tell us. Disagree with this article? .