All engineering topics (out of main)

[mongo]

@Fact Checking @KarenRei

3 years and over 460 posts:
Why does Tesla use a Resistance Heater instead of Heat Pump

There is also a Y centric thread, but it is only 5 posts.

 

[Fact Checking]

The problem is worse when you try to use the same compressor to be both AC and heating. dT for cabin cooling is unlikely to be more than -15 - -20°C. But for cabin heating it could easily be more than +40°C. So in heating mode it not only has to do more work, it does so much less efficiently. But the compressor doesn't get more powerful just because you're using it in heating mode.

I agree with your other points, but the heating limit is just a question of compressor sizing. Even if efficiency drops from 400% to 200%, that's still twice as efficient as a resistance heater. A larger compressor is, of course, more expensive, so there are trade-offs - but it's not a primary limit to how much heating power a heat pump can provide.

Btw., I'd expect Tesla to use stronger compressors in the Plaid and maybe the Model 3 P, for better racing performance.

But in any case you are right that they'll have other sources of heat (such as the drive units) in case of extreme heat generation shortfall, which is always a possibility with a heat pump - such as in Camp Mode when there's no good airflow to the heat pump. Most heat pumps have resistance heaters as emergency backup.

 

[bhtooefr]

Incidentally the ID.3 apparently has a CO₂ based heat pump:

VW will roll out CO2 MAC in new electric car series

"German car manufacturer Volkswagen will opt for CO2 mobile air-conditioning (MAC) systems in another series of cars by the end of 2019 – its electric car ‘ID’ series"​

I believe Mercedes also has CO₂ based heat pumps too. Maybe CO₂ heat pumps are out of patent protection already?

So what's going on here with VW and Mercedes using CO2 is essentially...

The EU mandated that HFC-134a be removed from use due to its extremely high global warming potential (it has a GWP-100 of 1430), and replaced with either HFO-1234yf (which performs similarly to HFC-134a as a refrigerant (and can therefore be used with minimal modification) and has a GWP-100 of less than 1) or CO2 (which by definition has a GWP-100 of 1, as it's the reference for GWP).

Mercedes found out that HFO-1234yf is far more flammable than predicted in a crash. (It was already known that the combustion byproducts of HFO-1234yf are incredibly nasty - hydrogen fluoride and carbonyl fluoride are not fun - it was just believed that combustion wouldn't happen). VW and Mercedes subsequently just refused to use HFO-1234yf. And, as CO2 has much higher working pressures and therefore requires a rigidly mounted compressor, they couldn't easily use it on their ICE vehicles (which were designed around an engine-mounted compressor and flexible lines), but they were pushing for it to be the sole choice anyway.

This means that VW and Mercedes simply violated the EU directive, but Germany was still type-approving the cars. France actually tried to block the import of the first Mercedes model that was required to use HFO-1234yf and had HFC-134a, and the courts found that Germany's type approval overrode the directive, and France had to let the cars in.

If I had to guess, VW and Mercedes are simply paying the patent licensing fees on CO2.

 

[StealthP3D]

The problem is that heat pumps not only lose efficiency with increasing dT, but also lose output power. Great for marginal temperature differences, not so much for huge temperature differences (where they matter the most). I've never heard of a car heating system that's able to run without any resistive heater. There has to be a resistive element somewhere.

Absolutely! In very cold temperatures a heat pump without some form of resistive heat will have crippled heat output. This is a very real problem because when it is very cold is exactly when you need the most heat (for example to quickly melt ice off the windshield or prevent the windshield from icing). It's also a problem for vehicles more than for buildings because in buildings the cold weather problem is mitigated by using over-sized heat exchangers and fans (which is obviously a problem in vehicles where weight and size matter).

A guess (if this leak is correct): the inverter in the drive units is the resistive heating, when run in "hold position and just burn power" mode. Heat is pumped from this loop up to cabin-heating temperature.

Would need to be rapidly responsive; you can't make people wait for the full battery loop to heat up before giving them meaningful cabin heat. The battery heating loop would surely need to be shut off / flow limited. Still would likely be laggier than a PTC heater.

I agree. It sounds like what they have done to mitigate the laggy behavior to some degree is to more tightly integrate whatever they are using for resistive heat (maybe the inverter) with the cabin heating. The Model 3 uses the inverter/motor waste heat to gradually heat the battery in cold temperatures. This has serious limitations in cold rain because the splashing of cold water around the motor really reduces the amount of heat that makes it back to the battery. Also, in extreme cold there is the same problem. The motor is always radiating some heat to the environment which means there is a limited amount that can be collected by the coolant loop.

I find the announcement of these changes very heartening as an investor because it was one of the efficiency refinements that I was disappointed was not included in the initial release of the Model 3. This shows a company that is determined to continually develop and innovate to a high standard. Using first-principles thinking, Tesla has looked at the area most ripe for efficiency improvement and tackled it head-on. This requires substantial engineering and development at a high level and is a sign of a company that is not struggling but flourishing. This is why the competition will not be able to catch them.

As a side note: the sheer number of people leaking things about MY lately, including pictures, sure makes it seem likely that something is afoot for later this quarter. I hope people's expectations on volumes aren't excessive, of course.

I think the speed of the Model Y roll-out will be moderate, not a trickle but not a blast. Because a trickle runs the risk of Osborning Model 3 sales without having the volume to make up for it with Model Y sales. So I think the roll-out will be solid in terms of volume but not earth-shattering.

 

[StealthP3D]

I agree with your other points, but the heating limit is just a question of compressor sizing. Even if efficiency drops from 400% to 200%, that's still twice as efficient as a resistance heater. A larger compressor is, of course, more expensive, so there are trade-offs - but it's not a primary limit to how much heating power a heat pump can provide.

That's not entirely true. The required temperature differential is more problematic in an automotive application. In a building, you can just circulate more air to make up for the lower temperature differential in very cold weather. But in a car you actually need HOT air, and lots of it, to defrost the windshield. This could be somewhat mitigated by the use of a transparent, thin-film resistive heater built into the windshield but then you have the side windows too. And you are back again to the use of resistive heating, at least for the glass. A more powerful compressor, without changing the refrigerant used, will not really solve the low-temperature differential problem. Changing the refrigerant used to achieve a higher differential would require higher pressure lines and fittings which is not practical for multiple reasons.

I'm OK with the energy consumption to defrost my Model 3 in cold temperatures with resistive heat because I'm usually able to do it with shore power and it's only for 2-10 minutes. I would not want to give up the fast defrost feature of my cars - it is sooo much better than waiting for an engine block and gallons of coolant to heat up before I can melt ice. The heat pump would be welcome for long-distance travel in colder temperatures and I hope the Cybertruck has one also (because the cabin area needing heat is so much larger).

 

[Thumper]

Stealth said
"But in a car you actually need HOT air, and lots of it, to defrost the windshield. This could be somewhat mitigated by the use of a transparent, thin-film resistive heater built into the windshield"

Ford offered such a system as an extra cost option on some early Tauruses. It required so much electricity that those models came with an physically and electrically oversized two-stage alternator. When the windshield defrost was engaged, the second half of the alternator supplied the extra power.

 

[jerry33]

That's not entirely true. The required temperature differential is more problematic in an automotive application. In a building, you can just circulate more air to make up for the lower temperature differential in very cold weather. But in a car you actually need HOT air, and lots of it, to defrost the windshield.

But not so much that the windshield cracks.

 

[KarenRei]

I'm OK with the energy consumption to defrost my Model 3 in cold temperatures with resistive heat because I'm usually able to do it with shore power and it's only for 2-10 minutes. I would not want to give up the fast defrost feature of my cars - it is sooo much better than waiting for an engine block and gallons of coolant to heat up before I can melt ice. The heat pump would be welcome for long-distance travel in colder temperatures and I hope the Cybertruck has one also (because the cabin area needing heat is so much larger).

IMHO, it'll be the most benefit while camping. The energy drain in camping mode is predominantly about heating/cooling efficiency. And you don't need high powers. Being able to heat with a COP > 1 would be a big deal. Could literally double the time you can spend camping without plugging in.

On the road in very cold weather, though, not so much.

Either way, it'll be interesting to see what they come up with. I'm sticking with my bet that they're using the drive units as a resistive heat source from which the heat pump pumps heat up to temperatures sufficient to heat the cabin, and that they'll try to minimize the dead mass that has to be heated in order to keep the system responsive (because nobody wants an ICE-like laggy heater)

 

[MitchJi]

Either way, it'll be interesting to see what they come up with. I'm sticking with my bet that they're using the drive units as a resistive heat source from which the heat pump pumps heat up to temperatures sufficient to heat the cabin, and that they'll try to minimize the dead mass that has to be heated in order to keep the system responsive (because nobody wants an ICE-like laggy heater)

I hopefully that they add phase change themal mass, probably in an insulated box that your can preheat when you are connected to power

I read a post by someone with an oil heater who heated it before leaving for work and it kept his car warm for his whole drive to work. It much more efficient to store heat instead of electricity and it’s a very comfortable heat

 

[mongo]

OT Y temp control and architecture

IMHO, it'll be the most benefit while camping. The energy drain in camping mode is predominantly about heating/cooling efficiency. And you don't need high powers. Being able to heat with a COP > 1 would be a big deal. Could literally double the time you can spend camping without plugging in.

On the road in very cold weather, though, not so much.

Either way, it'll be interesting to see what they come up with. I'm sticking with my bet that they're using the drive units as a resistive heat source from which the heat pump pumps heat up to temperatures sufficient to heat the cabin, and that they'll try to minimize the dead mass that has to be heated in order to keep the system responsive (because nobody wants an ICE-like laggy heater)

Regarding camping: a COP of 2 would double camping time, but only if you were using all your extra power up in the first place. If you used 5% SOC for conditioning, and had 30% left when you got to civilization, it wouldn't matter.

If the heat source is the drive unit, then there isn't any COP boost from the heat pump, rather the only gain is in the outlet temperature, so does it help enough to use? Data logs have shown 55C (131 F) at each drive unit stator (7kW total or 20k BTU) which would be sufficient to defrost the windshield/ heat the cabin. 9kW "stator heater" ON when supercharging. Hypocritical waste?

My guess is that, if they did indeed remove the cabin PTC, they routed the coolant output from the drive unit(s) to a heater core in the cabin. Although, from an efficiency POV, I'm not thrilled with the heat loss from the drive unit housing, it would get a boost from the drive unit losses.

The other benefit of drive unit based heating is eliminating the HV wiring and fuse to the cabin PTC (along with the controller). Additionally, the theoretical topology change could put the AC compressor in the rear to reduce wire length (controller integrated in penthouse?) and also improve cabin noise. Also reduces damage in front impacts.

MOD In-Post EDIT:

NONE of this discussion belongs here. Take it to the "All Engineering..." thread or, better yet, out of the Investor Forum altogether. No further posts on this will survive.

 

[KarenRei]

Regarding camping: a COP of 2 would double camping time, but only if you were using all your extra power up in the first place

If you're not using up your power, then we're not talking about your maximum camping time, and it's sort of a pointless conversation. The whole point of bringing it up was to discuss the scenario in which it makes the most difference, and contrast that to scenarios where it makes the least difference. Not to overcomplicate things by blending the two together.

If the heat source is the drive unit, then there isn't any COP boost from the heat pump

You're mixing up two separate things: maximum heat output (kW), of primary interest when initially heating a car (and potentially in very cold weather), vs. maintenance heating in random camping weather. The former requires either a much bulkier, heavier, more expensive heat pump than is needed for air conditioning/pack cooling, or more realistically, a resistive heat source (the latter is the setup of almost all EVs that utilize heat pumps). Normally the resistive heat source is a PTC heater, but in this case, the drive unit seems the likely answer, since a PTC is said to be missing, and Tesla already switched pack heating to be drive unit-based.

But that's only needed when you need high power output. If you only need minimal maintenance heating, there's no reason to generate resistive heating in the drive unit; you'll just use a heat exchanger with the outside air as your cold well. So bringing up the drive unit heating in this context is irrelevant.

Data logs have shown 55C (131 F) at each drive unit stator (7kW total or 20k BTU) which would be sufficient to defrost the windshield/ heat the cabin.

Tesla's PTC heaters go up to 80°C, and regardless, you're guaranteed to have a temperature drop between the drive unit and the air entering the cabin. You also have to do a liquid-air heat transfer regardless. You can't just take the max stator temperature and treat that as the temperature air would enter the cabin if blown across them; there's several points at which the temperature will drop along the way.

 

[Dreadnought]

I hopefully that they add phase change themal mass, probably in an insulated box that your can preheat when you are connected to power

I read a post by someone with an oil heater who heated it before leaving for work and it kept his car warm for his whole drive to work. It much more efficient to store heat instead of electricity and it’s a very comfortable heat

OT
Even if there are better suited materials than sodium acetate (which is not great in terms of volumetric energy density), they can't do anything else than storing heat. You're always better off spending the weight and volume on battery as that can drive and heat the car.

 

[mongo]

If you're not using up your power, then we're not talking about your maximum camping time, and it's sort of a pointless conversation. The whole point of bringing it up was to discuss the scenario in which it makes the most difference, and contrast that to scenarios where it makes the least difference. Not to overcomplicate things by blending the two together.

Right, your own battery power. If you weren't hitting the limits of pack SOC to get back to somewhere to charge when using resistive heating, the COP boost doesn't impact the user's camping.

You're mixing up two separate things: maximum heat output (kW), of primary interest when initially heating a car (and potentially in very cold weather), vs. maintenance heating in random camping weather. The former requires either a much bulkier, heavier, more expensive heat pump than is needed for air conditioning/pack cooling, or more realistically, a resistive heat source (the latter is the setup of almost all EVs that utilize heat pumps). Normally the resistive heat source is a PTC heater, but in this case, the drive unit seems the likely answer, since a PTC is said to be missing, and Tesla already switched pack heating to be drive unit-based.

But that's only needed when you need high power output. If you only need minimal maintenance heating, there's no reason to generate resistive heating in the drive unit; you'll just use a heat exchanger with the outside air as your cold well. So bringing up the drive unit heating in this context is irrelevant.

I should have better delineated my two thoughts. First paragraph was regarding camping, rest was heat pump in general.
On the low demand camping front, if you don't need a lot of heat, the COP gain is less of an impact. There may be a lower end to useful operating range also.

Tesla's PTC heaters go up to 80°C, and regardless, you're guaranteed to have a temperature drop between the drive unit and the air entering the cabin. You also have to do a liquid-air heat transfer regardless. You can't just take the max stator temperature and treat that as the temperature air would enter the cabin if blown across them; there's several points at which the temperature will drop along the way.

Like you point out above, temperature and power are two separate things. Per the link, the S PTC is 6kW, so on an energy scale, the system could lose 1kW and be equal there. It is also likely the motor could heat the coolant above 55C since the current use case is only battery pack heating. Automotive heater cores normally output air temps above 115 F, preferably above 135 Turn Up Your Heater when the engine is hot, less so during warm up. Design wise, the closer the fluid temp to the desired air temp, the less air it moves, or the larger the heater core. If they drop the PTC, they have more space for the core.

 

[Snapdragon III]

That's not entirely true. The required temperature differential is more problematic in an automotive application. In a building, you can just circulate more air to make up for the lower temperature differential in very cold weather. But in a car you actually need HOT air, and lots of it, to defrost the windshield. This could be somewhat mitigated by the use of a transparent, thin-film resistive heater built into the windshield but then you have the side windows too. And you are back again to the use of resistive heating, at least for the glass. A more powerful compressor, without changing the refrigerant used, will not really solve the low-temperature differential problem. Changing the refrigerant used to achieve a higher differential would require higher pressure lines and fittings which is not practical for multiple reasons.

I'm OK with the energy consumption to defrost my Model 3 in cold temperatures with resistive heat because I'm usually able to do it with shore power and it's only for 2-10 minutes. I would not want to give up the fast defrost feature of my cars - it is sooo much better than waiting for an engine block and gallons of coolant to heat up before I can melt ice. The heat pump would be welcome for long-distance travel in colder temperatures and I hope the Cybertruck has one also (because the cabin area needing heat is so much larger).

Don't forget, they are going to have lasers to defrost the windshield while also burning off the bugs! /S

 

[KarenRei]

Right, your own battery power. If you weren't hitting the limits of pack SOC to get back to somewhere to charge when using resistive heating, the COP boost doesn't impact the user's camping.

This makes no sense. There are two extremes in terms of climate control discussion:

1) 100% of the pack used for climate control while stationary
2) 100% of the pack used for driving with the climate control on.

Blending the two together in this conversation is pointless. Of course in the real world you'll have some combination of the two. But that's not how you make comparisons.

I should have better delineated my two thoughts. First paragraph was regarding camping, rest was heat pump in general.
On the low demand camping front, if you don't need a lot of heat, the COP gain is less of an impact

Of course it does. It determines how many days you can camp for.

Perhaps I'm seeing the problem: are you picturing that "camping" automatically implies a single overnight stay of fixed length? I'm comparing how long you can camp on a single charge.

Like you point out above, temperature and power are two separate things. Per the link, the S PTC is 6kW, so on an energy scale, the system could lose 1kW and be equal there.

Now we're back to mixing up topics again - just different ones. When it comes to using drive unit heat to heat the cabin, total heating power isn't the problem. The drive unit has tons of heating power. The problem is the maximum temperature you can realistically heat air with it, due to the maximum temperature the drive unit is allowed to reach. This maximum air temperature will always be significantly below the maximum temperature that the drive unit can reach; there will always be a temperature drop with each stage of heat transfer. Which is why PTCs go to such high temperatures; that doesn't mean that the air going into the car is 80°C, it's just that the unit going to 80°C in order to heat the air to the (lower) inlet temperatures.

It is also likely the motor could heat the coolant above 55C since the current use case is only battery pack heating

One presumes that the limit is to protect drive unit longevity.

 

[canoemore]

Reading this discussion about heat pumps I can't help but think of Musk's hint about a future Tesla HVAC product during that Joe Rogan interview: Elon Musk Hints That a Smart Air Conditioner May Be On the Way | Digital Trends

 

[bhtooefr]

Also worth noting that there's some different configurations of heat pump that can improve effectiveness down to lower temperatures.

Inner secrets of the Toyota-Denso heat pump revealed

You would still want a resistive heat source of some kind, for even more extreme cold, though - -10 °C ain't enough.

 

[Artful Dodger]

As noted though, if Tesla is combining it with resistive heating from the drive units, they can bypass these power output limitations.

Yeah, they're going to pump heat from the bty pack* to the cabin. Otherwise, that's wasted energy. But it'll work well in Arctic climates. There's about 10kw of heat shed by the powertrain, moving it to the cabin instead of dumping it overboard is an obvious win:

"Tesla’s recent patent for an Aggregated Battery System with multiple small cells arranged in a common cell enclosure providing both structure and a heat transfer fluid conduit"

AGGREGATED BATTERY SYSTEM - Tesla, Inc. | freepatentsonline.com
​

Cheers!

*full power train cooling loop(s), most likely

 

[Artful Dodger]

There's about 10kw of heat shed by the powertrain

Correction: about 1.5kw of heat shed in a Model 3 during highway cruise, not 10kw (Doh!)

Cheers!

 

[AudubonB]

Without squirming under the front of our Models S or X, can someone remind me of the extant radiator coils each model currently has? The follow-up question is: during an extreme atmospheric event like a silicic volcanic eruption - try to imagine the worst conceivable dust storm and you're getting reasonably close - could those coils become compromised enough to reduce their heat-shedding ability so that the cars cannot properly function?

Rim of Fire has enough Usual Volcanic Suspects and Tesla owners that a Vann diagram would demonstrate significant overlap. US residents can harken to the 1980 Mt. St. Helens eruption, whose ashfall absolutely clogged automobiles' air intake filters. Teslas do not have, obviously, that weak link but my question about coolant coils stands.

Placing that question's premise on its head, the extent to which radiator coolers can be affected by ashfall will indeed affect ICE vehicles before EVs....regardless, I think it pertinent to learn the answer and if it is potentially a problem, consider prophylactic responses.