Articles/megaposts by DaveT

[Zzzz...]

Cells in the pack do have some space in between them. Not much currently, but if that space could be increased(by using slightly taller cells for example or more energy dense cells) then as object rip through the battery pack individual cylindrical cells would be forced sideways instead of smashed. Might require stronger cell casing and contact welds that break or bend if force applied to the cell.

 

[Raffy.Roma]

@Zzzz...

The thread that you mentioned is very interesting. I also learnt a feature of the Model S that I didn't know. The 16 modules forming the battery pack can be visualized on the screen! Amazing!
IMO, also if I am not an expert of battery, the danger of fire could be very much mitigated if as soon as a module is damaged it would be immediately switched off via firmware. Ok lithium catches fire when in contact with the air. But if no current passes through the damaged module the risk of having a fire starting from the damaged module would be very much lowered.

No?

 

[VolkerP]

No, not at all. The fire is started by heat that results from an electric short circuit, when the metallic debris punctures the cell and forms an electric path from anode to cathode within the cell. It is irrelevant if the cell is asked to deliver current to the outside after this, the reaction will go on nevertheless.

 

[Raffy.Roma]

Then, assuming that the anode is already grounded, the cathode could be immediately grounded (via firmware) when the corresponding module is damaged. This way the short circuit caused by the road debris couldn't start the fire anymore.

No?

 

[JRP3]

You can't control an individual cell through firmware. Also, there is no metallic lithium in the cell to react with anything.

 

[Lessmog]

Then, assuming that the anode is already grounded, the cathode could be immediately grounded (via firmware) when the corresponding module is damaged. This way the short circuit caused by the road debris couldn't start the fire anymore.

No?

So, to mitigate an internal short-circuit you recommend an external short-circuit? How does that work - fight fire with fire? Sounds sort of like magic to me.

 

[Raffy.Roma]

@JRP3

Ok you are right. But it could be done at level of all the cells of a single module. It's just a matter of grounding all the cathodes of the single module contemporarily by integrating in the module such a control driven by the firmware in emergency situations.

No?

- - - Updated - - -

So, to mitigate an internal short-circuit you recommend an external short-circuit? How does that work - fight fire with fire? Sounds sort of like magic to me.

If you discharge the battery externally and not through the lithium the battery shouldn't catch fire IMO. No?

 

[Lessmog]

@JRP3

Ok you are right. But it could be done at level of all the cells of the modules. It's just a matter of grounding all the cathodes of the single module contemporarily by integrating in the module such a control driven by the firmware in emergency situations.

No?

- - - Updated - - -



If you discharge the battery externally and not through the lithium the battery shouldn't catch fire IMO. No?

Now, I'm not an electrical engineer nor a chemist, but I suspect the danger here lies in the rapid energy release producing too much heat, rather than a traditional chemical reaction ("fire") involving Li. The same amount of energy would have to be dissipated via your external short, so I still don't see how that helps much.

But I confess this is speculation on my part; let's hope the people behind the battery's construction are more sophisticated.

 

[kenliles]

Dave

Thank you for your mega posts, I carefully read all of them.

I do not comment often as I seem to be out of sync with other contributors due to living in a different time zone, and that makes it difficult to engage in live discussions.

This time I just would like to expand on your idea of testing road debris impact on battery. It is possible to do that testing by using battery only, not the whole car. The debris impact on battery can be simulated in a test rig whose design allows for varying the relative speed of battery vs debris and for varying the distance between the battery and various pieces of debris. Car weight can be simulated by placing the equivalent weight load on the battery. The cost of building such test rig would not be excessive and I would be surprised if Tesla does not already have one.

In fact such testing is not at all needed, or practical. They can simply take a battery shell filled with the appropriate weight and tow it over objects and see how far they penetrate. If they go deep enough to project into the cell area you can assume cell shorting and fire. Force sensors would calculate the forces involved and a decision could be made as to how likely such an impact is going to be and what reinforcement would mitigate it, and if it's reasonable or not to do so. Also, they may very well already have this data and it's simply a matter of what is practical to do. They can't build a tank.

These ar similar to comments I would make. As a former test engineer in Aerospace I'll throw my 2c - It's a an excellent direction of thinking (DaveT's suggestion), but can be accomplished with test rigging much more cost effectively- in fact I'd be more than willing to bet, has already been done to a large extent during development. You can follow these test up with some limited real-world testing (similar to JRP3s description) that validate the rig tests and offer some 'around-the-edges' adjustments to the design. and again, I'll bet most if not all of this has already been done- it's a very common practice and requirement in Aerospace (something I believe Elon is familiar with )

also, Auzie - I wouldn't discourage comments posted even out of sequence (with a proper quote). The board is full of people from all over and readers are accustomed to that sort of non-sequential post imo. I would rather see your comments personally

 

[JRP3]

Raffy, if a cell or cells gets penetrated and shorted internally that's where the current will flow and where the heat and sparking will occur. Nothing can be done about that externally, other than trying to cool it down and/or smother it. The cells are welded to conductive plates that connects them together, there is no way to physically disconnect them. Modules are already physically separated by barriers. Maybe Tesla needs more "goo" in the packs that smothers and absorbs heat.
Hopefully at some point non flammable electrolytes can be used, along with a non oxide chemistry, but for now it's not possible if you want high energy density cells.

 

[InTheShadows]

So, to mitigate an internal short-circuit you recommend an external short-circuit? How does that work - fight fire with fire? Sounds sort of like magic to me.

This may be a plausible path. Not to short the batteries, but to have their energy dumped to a high wattage resistor that would discharge the compromised fire walled section in a few minutes. This would quickly remove as much of the energy as possible. Normal C discharge rates could be violated because it's an emergency situation with no hope of preserving the battery life. The batteries would get really hot but it could possibly be done.

I'm sure Tesla engineers are already deep into looking into possibilities of doing some sort of emergency power dump. Jettisoning the energy so to say.

 

[Lessmog]

This may be a plausible path. Not to short the batteries, but to have their energy dumped to a high wattage resistor that would discharge the compromised fire walled section in a few minutes. This would quickly remove as much of the energy as possible. Normal C discharge rates could be violated because it's an emergency situation with no hope of preserving the battery life. The batteries would get really hot but it could possibly be done.

I'm sure Tesla engineers are already deep into looking into possibilities of doing some sort of emergency power dump. Jettisoning the energy so to say.

As I mentioned I myself am not really qualified, but I believe that JRP3 is, to judge what happens in a battery runaway:

Raffy, if a cell or cells gets penetrated and shorted internally that's where the current will flow and where the heat and sparking will occur. Nothing can be done about that externally, other than trying to cool it down and/or smother it. The cells are welded to conductive plates that connects them together, there is no way to physically disconnect them. Modules are already physically separated by barriers. Maybe Tesla needs more "goo" in the packs that smothers and absorbs heat.
Hopefully at some point non flammable electrolytes can be used, along with a non oxide chemistry, but for now it's not possible if you want high energy density cells.

 

[Causalien]

As I mentioned I myself am not really qualified, but I believe that JRP3 is, to judge what happens in a battery runaway:

So, what about injection of concrete to fill up the whole cell? Will that stop the chemical reaction?

 

[Raffy.Roma]

This may be a plausible path. Not to short the batteries, but to have their energy dumped to a high wattage resistor that would discharge the compromised fire walled section in a few minutes. This would quickly remove as much of the energy as possible. Normal C discharge rates could be violated because it's an emergency situation with no hope of preserving the battery life. The batteries would get really hot but it could possibly be done.

I'm sure Tesla engineers are already deep into looking into possibilities of doing some sort of emergency power dump. Jettisoning the energy so to say.

You improved my thought. I think that such a solution could be easily and cheaply implemented. Maybe, depending upon the judgment of the NHTSA because this is a matter that doesn't concern the safety of people inside the car but it concerns the risk that the Model S can be totalled in case of fire after hitting a road debris, that it wouldn't even be necessary a recall because it could be retrofitted during normal service.

 

[Curt Renz]

A key takeaway here is that Tesla needs to move public relations up several notches in importance. It's simply inexcusable to take 10 days to get the written company response to a public issue like the second debris fire. It would have taken a concerted SWAT team no more than 24 hours to develop a strong response that address public concerns. If they want to amplify the response, they can always do so.

Indeed that is so, Robert. DaveT and TMC members with public relations backgrounds have been saying the same thing. It won't do to say that our PR methodology is to wait for the public reaction to mishaps. A PR pro is needed who knows what the reaction will be and how to calm it before it gets out of hand. An effective PR chief is not simply a communications major who is able to put a CEO's thoughts into proper written form. He must be someone who has the CEO's ear and is not merely a yes man but is a trusted adviser. He should be paid like the company's managing engineers.

An engineer's mentality to double-check everything first does not work when sanfus are quickly spread by the media and gobbled up by a highly politicized public. The CEO needs to delegate responsibility in a realm that is not his forte. That's especially true if the CEO is CEO of two companies and chairman of another. He may be wise to give up one of his CEO positions, since unexpected events can frequently occur that demand his full attention. I realize that SpaceX was his earlier baby, but it is Tesla Motors that sells to the public and has public shareholders. He has special duties to both of those constituencies. If he does not want to devote all of his time to Tesla, then he really needs to be willing to delegate more fully, especially in the realm of public relations. Proper concern regarding public perception of his company and its products is vital to the success of the company.

 

[hoang51]

DaveT, thanks for your mid-day mega post. What do you see now that TSLA went through $120 and bounced back up a bit? The way I see things, TSLA can rebound after NHTSA clears TSLA of any fault and deems their car is safe despite the "flaw". When would this be, I have no idea. It appears that any quick fix, such as raising the Tesla Model S higher and Elon's posting doesn't stop TSLA from bleeding. It seems to allow shorts to cover and the next day the stock tanks. Elon posting/tweeting used to carry a lot of credibility and allows TSLA to rebound and stay high. Nowadays, it's just selling after selling. I think even with 4Q earnings, the stock would trade sideways. December looks like TSLA could be low $100s or high $90s at the rate it is falling.

 

[JRP3]

You improved my thought. I think that such a solution could be easily and cheaply implemented.

No it can't. I'm sorry but you clearly have no understanding of the power involved or the current flow pathways. Some of this is madness people, please stop. Each cell is connected with a fuse, that fuse will fail at any attempt to dump high current through it, that's it's purpose. The current will flow through the internal short, period, end of story. You can't stuff enough resistors in each module to dump the power in a dead short condition anyway.

 

[kenliles]

No it can't. I'm sorry but you clearly have no understanding of the power involved or the current flow pathways. Some of this is madness people, please stop. Each cell is connected with a fuse, that fuse will fail at any attempt to dump high current through it, that's it's purpose. The current will flow through the internal short, period, end of story. You can't stuff enough resistors in each module to dump the power in a dead short condition anyway.

yeah- the only way I can see something along those lines working at all, would be to short all the cells to ground (spike a rod if the car is stationary); that's probably frought with more danger than just a slow burn - especially given the downward heat direction- apparently the cab is fairly safe if the fire dept doesn't flame the fire. I really think the current (no punn) design is excellent. I just don't see a problem here - safety is clearly not an issue as the results attest, Tesla covers the financial loss now. we're done- just need the world to catch up

 

[DaveT]

TSLA at $121.11

DaveT, thanks for your mid-day mega post. What do you see now that TSLA went through $120 and bounced back up a bit? The way I see things, TSLA can rebound after NHTSA clears TSLA of any fault and deems their car is safe despite the "flaw". When would this be, I have no idea. It appears that any quick fix, such as raising the Tesla Model S higher and Elon's posting doesn't stop TSLA from bleeding. It seems to allow shorts to cover and the next day the stock tanks. Elon posting/tweeting used to carry a lot of credibility and allows TSLA to rebound and stay high. Nowadays, it's just selling after selling. I think even with 4Q earnings, the stock would trade sideways. December looks like TSLA could be low $100s or high $90s at the rate it is falling.

It’s really hard to forecast short-term price action, and TSLA doesn’t seem to be in a place where it’s a definite buy or sell to short-term traders. Lots of traders will be looking to short TSLA when/if it breaks $118.6 (or so) on high volume. At that point the selling pressure could likely bring the stock lower to eventually test the next resistance level being around $108-109 (goldman dip level and 200dma). Today TSLA tested it’s previous low of $119.22 by hitting a lower low of $119.06 but it wasn’t able to break through that level lower. So there’s a definite possibility of testing that resistance level again and lots of people are going to be watching it. It’s also possible that $119.06 is the bottom, but it’s too hard to tell right now.

It’s tough to call the bottom with any accuracy because there are just too many factors involved and we don’t know what catalysts lie ahead. Before going long most traders will want to see TSLA set up for a buying point. Traders have different approaches for this but I’m personally looking for a modified version of a false breakout (downward) with divergence (something William Elder preaches). It’s tough to explain without charts but the concept is I’m looking for a point where it looks like TSLA is going to breakdown but some indicators (ie., MACD histogram) show a more bullish picture (ie., momentum shifting positive). It’s tough to say at what price that will happen because we’ll have to wait and see how the price action pans out.

That said, I personally think we might be at the bottom or getting really close to the bottom. My hypothesis lies in my belief that there is a lot of smart money that will be picking up TSLA at what they believe to be bargain prices. Even when TSLA was in the 190s, I felt like there was a legitimate risk of it dropping to the 130s and even hitting the 100s. But, I’ve held a belief since late 2012 that buying TSLA at under $100 is a screaming bargain. So, I am projecting that others too (it doesn’t have to be a lot of people but it needs to be a lot of cash) will be tempted at the bargain prices of TSLA before it hits $100.

 

[NStar]

Dave, thanks. Do you have any thoughts on the possibility of the fires actually affecting the demand? In one of the German interviews Elon said the demand dropped a little bit right after the first fire but quickly rebounded and became even higher. I think that was October. Since then there were 2 more fires and the sentiment/perception turned more negative now I believe. Craig’s VIN chart shows the new VIN assignment rate (last 2 months average) dropped very significantly from ~700 in Mid-Oct to ~500 now. I know there could be other explanations but I start to feel I cannot simply ignore the possibility of the fires putting a meaningful dent on the demand now. I don’t change my view on Tesla’s long term fundamentals but If this is true it will be at least short term negative impact on Q4 ER. I’m wondering if we have other data points about this matter, like people taking tours recently or talking to sale people.

It’s really hard to forecast short-term price action, and TSLA doesn’t seem to be in a place where it’s a definite buy or sell to short-term traders. Lots of traders will be looking to short TSLA when/if it breaks $118.6 (or so) on high volume. At that point the selling pressure could likely bring the stock lower to eventually test the next resistance level being around $108-109 (goldman dip level and 200dma). Today TSLA tested it’s previous low of $119.22 by hitting a lower low of $119.06 but it wasn’t able to break through that level lower. So there’s a definite possibility of testing that resistance level again and lots of people are going to be watching it. It’s also possible that $119.06 is the bottom, but it’s too hard to tell right now.

It’s tough to call the bottom with any accuracy because there are just too many factors involved and we don’t know what catalysts lie ahead. Before going long most traders will want to see TSLA set up for a buying point. Traders have different approaches for this but I’m personally looking for a modified version of a false breakout (downward) with divergence (something William Elder preaches). It’s tough to explain without charts but the concept is I’m looking for a point where it looks like TSLA is going to breakdown but some indicators (ie., MACD histogram) show a more bullish picture (ie., momentum shifting positive). It’s tough to say at what price that will happen because we’ll have to wait and see how the price action pans out.

That said, I personally think we might be at the bottom or getting really close to the bottom. My hypothesis lies in my belief that there is a lot of smart money that will be picking up TSLA at what they believe to be bargain prices. Even when TSLA was in the 190s, I felt like there was a legitimate risk of it dropping to the 130s and even hitting the 100s. But, I’ve held a belief since late 2012 that buying TSLA at under $100 is a screaming bargain. So, I am projecting that others too (it doesn’t have to be a lot of people but it needs to be a lot of cash) will be tempted at the bargain prices of TSLA before it hits $100.