Tesla CTO JB Straubel Responds To (Trolls?) Tesla Horsepower Critics

A hot issue of complaint with a number of Tesla P85D customers over the past several months or so has been what they deem to be an inflated horsepower figure. Part of this is that the horsepower figure of the P85D is much higher than some other Tesla vehicles (like the P85+ that many owners upgraded from) while not seeming to provide an equivalent extra amount of actual power.

Tesla has finally responded, with a blog post from CTO JB Straubel. Depending on how you read it, you could either say JB “burned” or “trolled” the complainers. More neutrally, you could say that he provided a logical, cool-headed response that clears up some technical confusion and provides an official word on how Tesla came up with the numbers and why they vary like they do — I personally think this best describes the story.

No matter how you take the blog post, though, there’s still the separate matter as to whether Tesla provided the horsepower numbers (and 0–60 mph numbers, btw) that it should have (ethically speaking). Did Tesla use the right approach for determining the numbers? Did it try to deceive buyers or did it simply not consider that buyers may be confused and misled? Did it inadvertently use methodology that inadequately compares its various Model S options? Or did it know that it was going to make many buyers think the P85D was much more powerful than it is?

I can’t answer those questions, but I think the very last line in JB’s article captures a couple of important points. JB wrote, “The true measures for any performance EV driver are acceleration times and driving performance of the vehicle.” I think buried in this line is an implication that buyers test drove the P85D before buying it, and that they made a decision to buy based on how the car drove, not based on some numbers on the website. I think that’s logical, yet in some cases incorrect. Many buyers did buy the car without test driving it, or did so considering the experience from a short test drive as well as the horsepower and acceleration figures (and how they compared to other Tesla Model S options). I personally don’t think I’d buy a car without test driving it a couple of times, as well as test driving the other cars I’m considering, but that doesn’t mean everyone is the same. Furthermore, it’s hard to get a good sense for some things in a test drive, and relying on provided specs to help make a decision is very reasonable.

Getting back to the blog post, some people who have had an issue with these figures think the post didn’t fully address the complaints people have, skirted around the issue, and wasn’t specific enough when it comes to the P85D figures. I’ll let you decide (or argue about it, if that’s what you’d like to do).

But I’ve just provided a bunch of synopsis and commentary and haven’t even quoted the bulk of JB’s post yet. You can read the full thing + comments over on the Tesla blog, or you can just read the blog reposted below. It’s quite interesting from a technical point of view, imho, as it shows yet again how different ICE vehicles and EVs are, and further explains how an EV works. But, yes, I’m thinking it doesn’t fully provide the critics with what they wanted.


Tesla All Wheel Drive (Dual Motor) Power and Torque Specifications

By JB Straubel, Chief Technical Officer

Attempting to directly correlate horsepower ratings in petroleum burning vehicles to horsepower in an electric vehicle is a difficult challenge. The physics of an electric vehicle propulsion system are very different from a gasoline one. In an EV, electrochemical reactions in the lithium ion cells create electricity. That electricity flows through power electronics that control the voltage and current, then it flows to electromagnets in the motor that create powerful magnetic fields rotating the shaft to turn the wheels. The power required to rotate this shaft has the most correlation to traditional measures of horsepower. However, the chain actually begins in the electrochemical reactions that happen in the battery pack. Depending on the battery’s temperature, state of charge and age, the amount of electricity extracted can vary widely.

There is some confusion about our methodology for specifying “equivalent” horsepower ratings for our all-wheel drive, dual motor vehicles – the “D” versions of Model S. This document will hopefully answer those questions.

Electrical “Horsepower”

Defining electric power in terms of horsepower is not very intuitive. Kilowatts or Megawatts are a much more useful unit. Electricity alone can’t generate physical motion the way a horse or a fuel-burning engine does. An electric motor converts electricity into motion. Think of electric power as flowing much like fuel flows from a tank to an engine. Various situations (low state of charge, cold temperatures, etc.) can reduce this flow of electrons below the ultimate capability of the electric motor. In other cases, the potential flow of electricity may exceed the capability of the electric motor (warm battery, short duration accelerations, etc.). Since the battery electric horsepower rating varies it is not a precise number to use for specifying the physical capability of an EV. The motor shaft horsepower, when operating alone, is a more consistent rating. In fact, it is only this (single or combined) motor shaft horsepower rating that is legally required to be posted in the European Union.

Dual Motor vs. Single Motor (P85 vs. P85D)

The shaft horsepower rating of the rear wheel drive single motor Model S is straightforward and roughly 360-470 hp depending on the variant (60, 85 or P85). Also, it is generally similar, but not the same, as the battery electrical “horsepower” output. The difference is most obvious to drivers when the battery is at a very low SoC. In this state, the chemical reactions generate less voltage and less equivalent horsepower, even though the physical electric motor hasn’t changed. The maximum torque the electric motor(s) are capable of is nearly unchanged as the battery horsepower changes even though the maximum shaft horsepower is reduced as the battery horsepower reduces.

When we launched the all-wheel drive P85D, we took the straightforward and consistent approach of specifying the combined capability of the two electric motors, front + back. The torque from the two motors comes together resulting in a huge boost in acceleration, the “g’s” you feel in a P85D. This is why Insane Mode is so delightful. The vehicle takes off slightly faster than 1g of acceleration delivering the amazing 3.1 second 0-60 mph (96.6 kph) performance. This acceleration was verified by Motor Trend using a base vehicle and medium weight driver. It should be noted that a larger occupant and additional options that increase weight will reduce the acceleration. Also, the Motor Trend standard excludes the first 28 cm of rollout. Including this rollout adds approximately 0.2 seconds to the acceleration.

One additional note is that, while gasoline cars get worse with altitude, electric cars actually get faster. All cars experience reduced air resistance, but gasoline cars become increasingly oxygen-deprived the higher they go. The Motor Trend test was done at approximately sea level, so the Model S will outperform a combustion car of the same nominal acceleration as altitude increases.

With the shaft horsepower coming out of the motors the situation is not always as simple as front + rear. As we have pushed the combined motor horsepower higher and higher, the amount of times where the battery chemical horsepower is lower than the combined motor horsepower has increased.

Also, the all wheel drive system in the dual-motor cars distributes available electrical horsepower to maximize torque (and power) in response to road grip conditions and weight transfer in the vehicle. For instance, during hard acceleration, weight transfers to the rear of the vehicle. The front motor must reduce torque and power in order to prevent the front wheels from spinning. That power is fed to the rear motor where it can be used immediately. The opposite happens when braking, when the front motor can accept more regenerative braking torque and power.

All Wheel Drive 85D and 70D

When developing the 85D and the 70D dual motor vehicles, we maintained a consistent strategy of specifying the motor power using front + back methodology. In both of these vehicle variants the rear motor is of a very similar design to the front motor. The 70D motors deliver roughly 165 horsepower each, the 85D motors roughly 211 horsepower each and the front motor on the P85D delivers 221 horsepower. Although these motors are in the same family, their power electronics and control software are tuned differently.

Where some confusion occurs is that in the 85D and 70D vehicles the combined motor shaft power is very similar to the battery electrical horsepower under many normal conditions. With the P85D the combined motor shaft power can often exceed the battery electrical horsepower available. The dual motors utilize the battery horsepower in the widest variety of real world conditions. The true measures for any performance EV driver are acceleration times and driving performance of the vehicle.

JB Straubel

21 thoughts on “Tesla CTO JB Straubel Responds To (Trolls?) Tesla Horsepower Critics

  1. I heard that at least some P85D test drives were conducted with an 80mph speed limiter in effect. It’d be harder to evaluate high speed performance – where a horsepower deficit would be most noticeable – when subject to a limiter with unknown dynamics.

    Some perspective is called for, of course. Model S PxxD is still the best sedan in the world, and arguably the best car (depending on your criteria). I certainly wouldn’t want to send it back if I owned one.

    Several things do seem off about how its performance was communicated, though – all the more bewildering when there was no need to exaggerate anything.

    One thing that’s not in dispute is that the P85D is the first Model S where the advertised power was not actually delivered anywhere in the powertrain, due to the battery power limit being lower than the combined motor power. Whether this is misleading or not has been argued back and forth on the TMC forums for months now; I’m strongly leaning towards “yes,” especially given how actual output is given for all other models but the PxxD.

    There’s also the peculiar matter of 1-foot rollout in 0-60 times, and why Tesla would adopt it for PxxD models ONLY and not bother footnoting this for ten months.

    Then there’s the non-delivery of the “high speed” performance update promised in a (redacted) footnote. Some argue that the April inverter firmware update (that also gave a huge boost to the 85D) was this update, but it still doesn’t unambiguously achieve the initial claims for P85D. Actual 0-60 is ~3.5s without rollout and peak power is still about 550HP at the battery.

    Having been following this issue rather closely, I really don’t think JB’s post addresses the core of the issues. Not that I expected it to, under the circumstances.

    This is speculation, but the best explanation I can come up with is that Tesla were aiming to deliver the initial performance claims with an OTA update, similar to what they are doing with Autopilot. Somewhere around April/May, they realised it wasn’t possible without hardware changes. (That’s when the HP claim and upgrade promise disappeared from the website.)

    Since then, I guess the good faith efforts stopped and obfuscation began, including re-spinning the upgrade as “Ludicrous Mode.” (Note that Ludicrous does just meet the 3.2s 0-60 claim without rollout, and almost meets the power claim when measured at the battery.)

    1. I took JB’s letter to imply that with today’s battery technology, the D models could well improve upon their current performance simply with larger battery packs in the relatively near future. *shrugs*

      1. That is very likely true. It doesn’t address the alleged issue of mis-selling P85D with the original battery, though.

          1. I also don’t doubt it was a mistake made in good faith. What’s getting people riled up is what happened next.

            The firmware update now offered as Ludicrous Mode would’ve come very close to unambiguously meeting the disputed claims, had the hardware been capable of supporting it. I strongly suspect it was their intent to deliver this for free up until around April, when that footnote disappeared.

            When task impossibility became undeniable, their next move was obfuscation. JB’s press release is a continuation of this; information that both complainants and apologists already know and have been discussing for months. Nothing there settles the dispute.

            The thing with horsepower claims is that, along with the car’s mass, they imply how much acceleration to expect when travelling faster than the traction-limited speed range (beyond ~30mph in this case.) Even accounting for the relative merits of ICE versus direct drive electric powertrains, P85D never met those expectations, and never will without the $5k HV current path upgrade.

            We should note that every Model S variant besides the PxxD has its actual peak power listed. You have to wonder why the exclusion, if not because they’d have to explain why it isn’t the 691hp figure everyone remembers.

            As for 1-foot rollout, again I suspect the whole thing happened by accident. Between product launch in October and first delivery in December, they must’ve realised the car would only meet the announced metric using rollout – a convention they’ve never used before, and still don’t use for other models. I believe it was a convenient dodge, adopted retroactively.

            Unfortunately, it’s locked them into a convention that exaggerates how much extra performance you get when you pay for that “P” badge, and they only recently added a footnote explaining this. If that’s not grounds for a mis-selling complaint, nothing is.

            Granted these were probably their best self-defence moves under the circumstances, but I really don’t like it.

            Hopefully there will be a Gen2 Model S soon, and they can use the opportunity to drop all this baggage.

          2. I think this post by smac on the TMC forums nails it:

            “TBH the performance isn’t why I feel slighted (I don’t have a P85D)

            It’s about expectation management. I feel Tesla should be whiter than white, they have an astounding product and excessive hyperbole hurts them long term.

            If they’d said it was 550bhp from day one, could hit 0-60 in 3.5, etc. etc. we ALL would have thought this a fantastic car. They would have sold just as many, and none of this would have happened.

            As it stands I can’t help get the feeling they have upset loyal customers, then had their hand forced to release the L upgrade (further upsetting those wanting “top of the line”). This in turn ties their hands on product release timing, as I can’t see them squeezing any more out of the platform. They could have kept their powder dry for use next year to bolster Model S sales in light of the X.

            All for what? A number which on it’s own is worthless. (The weight, the gearing, the Cd, the ….. all combine to the overall “system” perfomance)

            We’ve been used to poor expectation management with respect to timing, just never product descriptions before. This is why as someone who wishes Tesla to succeed, I find the whole thing rather frustrating, and I hope at least the letter shows that to senior management.”

          3. I definitely hope they steer away from hyperbole and statements/figures that are more optimistic than realistic. Agreed that there’s no real need, and just sets people up for disappointment and Tesla up for criticism and skepticism.

          4. We do not need to wait for the Gen2 … if the battery capacity is the problem then it seems to me that the P85D version, now we have a 90kw/h battery option, perhaps no longer makes sense.

          5. Agreed about the redundancy of P85D. Market segmentation at the top is a mess now.

            Since the April update, it was already being cannibalised from below by the 85D. P90D with Ludicrous does the same from above.

          6. No mistake. Tesla purposely misled. And JB’s blog is a continuation of this policy.

            The basic problem people are starting to realize is the claimed HP with the given cars weight does not accelerate as would an ICE vehicle. This is not noticeable at low speeds where the traction control and instant acceleration blows away ICE cars. Only at higher speeds does it become noticeable. And since it’s illegal to drive at those speeds it is not an issue for most people. However taken to a race track people wonder why it’s only getting about 60% of the HP it should be getting. 60% is not a hard number, just a ballpark number, that would be my guess.

            So what I’m saying is if you put a car with only 60% of the claimed HP of a Tesla, put weights in it until they weighed the same, and then started a race at a rolling speed of about 100 mph they would be about equal. If you performed the same test with a car of equal “HP” weighted so they are the same then the ICE vehicle would blow the doors off the Tesla.

            However at normal speeds the instant acceleration with good tires fool people into thinking it has great HP. Tesla purposely deceives with their HP claims because it is easier than trying to explain things.

          7. It’s amazing that sooo many Tesla owners and fans are going to ANY length to excuse and defend a company that has misled and lied to them. It’s like they don’t want anything tarnish their fantasy picture of Tesla and it’s owner. Just amazing.

          8. It’s like the turn of the last century when automakers were being forced to describe their cars in terms defined by a horse driven world. They did the same thing but in different ways. So the automakers back then had great advantages but buyers were focused in on old-school legacy metrics. As a result Ford and the like were forced to exaggerate, well lie, to get noticed. Tesla is doing the same on their HP claims now. A little HP in an electric motor gives a feeling that requires a LOT of HP in a conventional engine. So they up their HP ratings to get this feeling across. When Tesla gets real EV competition in say 10 years then things will start to become more honest. Until then they’ll continue to try to describe their round pegged superior solution that has an idiot audience which is expecting a square pegged answer as best they can.

    2. No arguments here. But this is what I wrote on TMC last night, and think it’s relevant here:

      I think I understand the frustration of some people here, but my more forgiving take on it is:

      1) Tesla knows that EVs accelerate faster than ICE vehicles but don’t often get credit for it when you simply look at specs (even looking at the 0-60 time doesn’t capture how quick a Tesla is off the line). To try to counter that a little for the people who would simply be comparing the numbers while sitting on their couches at home, and because it was technically accurate and there’s no real standard for EVs, it used the combined figure.

      2) They figured people would buy the cars based on how they drove, not based on some numbers on the website.

      3) They used the same system as they had used before, but didn’t think (or want) to explain to everyone that the P85D would not be able to use all of its rated horsepower.

      4) After people got pissed here, they decided to clarify the matter. However, it would be irresponsible of a company to put its decisions in bad light, so they clarified in a neutral way rather than saying, “we made a big mistake by indicating horsepower for the P85D that it can’t use in the real world.” If they were only talking to Tesla owners, they’d probably be fine saying this, but they know how the media takes things and they don’t want to turn this into a press nightmare… which they’ve had enough of for illogical reasons.

      5) The wanted to use the nice 0-60 time Motor Trend got for the P85D but didn’t think to use the same system for the other models. (Of course, if they thought about the fact that the systems are different and overhype the P85D compared to other Model S options, that would be very crappy. I’ll give them the benefit of the doubt here, but this is the issue that is most sketchy and most likely to be somewhat deliberate deception in my opinion.)

      I imagine these responses won’t be popular here, especially with customers who are pissed off their cars aren’t more powerful, but I think this is a realistic human take on these decisions.

      1. Could you clarify this please? I just checked Tesla’s site and they have mentioned for P85D (motor power) Which i assume is the motor shaft horsepower JB mentioned. ?? So its 259 hp(front) and 503 hp(rear). This is what the motor is capable of? Right?

        Now the issue is that the battery electrical horsepower is not enough for the motor to reach the above mentioned figure??(Am i correct so far?) So my question is “Whats the max motor shaft horsepower available(combined) based on the max battery electrical horsepower available for P85D and P90D respectively”. ??

        1. Tesla won’t discuss this, and PxxD is the only Model S variant where this number is missing from the specifications. It’s not in JB’s press release either, which is ironic given its title.

          P85D owners have done their own investigations. The conclusion is somewhere around 550hp at the battery, based on API dumps. Expect shaft output to be a few percent lower.

  2. It’s not clear if this affects P90D ludicrous mode. If 0-60 mph time is affected (ie, 0.2 sec slower than standing start), that will put P90D at 3 seconds while Corvette Z06 is 2.9 seconds. There’s only two EV that’s quicker than comparably priced gas cars (SparkEV and P90D), but if P90D is slower than Corvette, SparkEV is the only one that’s quicker than comparably priced cars. It sucks to be the only one.


Leave a Reply

Your email address will not be published. Required fields are marked *