Porsche Taycan EPA range

[S4WRXTTCS]

I guess Porsche has successfully lowered the bar by testing the car in some horribly inefficient manner during their internal EPA testing.

My theory of what's going on it the low EPA numbers are both conservative, but not overly conservative. Where if you're doing city driving with lots of stop and go the efficiency is going to be worse than a lighter Tesla. Where if you're on the freeway doing 75+ then it's closer to it's sweet spot. After all it was meant for the autobahn, and not rush hour LA traffic like the Tesla.

I tend to favor more conservative range estimates due to all the factors that can change it. Instead of causing the customer grief in trying to figure things out it's better to adjust down a bit. Their entire market is customers that expect things to just work. They probably also realized there was no way they were going to compete range wise with a Raven refresh Model S.

I also wonder if they didn't study the range. From the Teslafi Leader board its clear that people aren't really exploiting the range the various Tesla models offer. In fact I can likely get on the top 50 board (for the Model 3) with my P3D if I really wanted to. Even for the Model S the minimum to get on is just over 266 miles, and yet the Raven refresh of the S has an EPA range of 373.

Sure I get that there aren't THAT many people on TeslaFi. But, to be honest one thing that would suck with the Taycan is giving up the functionality the TeslaFI website offers.

 

[AlanSubie4Life]

My theory of what's going on it the low EPA numbers are both conservative, but not overly conservative. Where if you're doing city driving with lots of stop and go the efficiency is going to be worse than a lighter Tesla. Where if you're on the freeway doing 75+ then it's closer to it's sweet spot. After all it was meant for the autobahn, and not rush hour LA traffic like the Tesla.

My theory is they completely disabled regen during the test and used the highest performance gear shift mapping.

While I agree that the “maximum range speed” of the Taycan is very likely considerably higher than that of the Model 3 (19-20mph), I very much doubt it is optimal near freeway speeds. The range is going to suffer if you drive continuously at 75mph. I thought someone posted the model details above somewhere...remember seeing it somewhere. Not sure if that model is accurate though. Really just need someone to take three Wh/mi data points at constant speed, several miles, averaged traveling in both directions, on a controlled stretch of road, from the Taycan (25mph, 45mph, 65mph) and just fit the cubic equation to it.

 

[gowthamn]

My theory is they completely disabled regen during the test and used the highest performance gear shift mapping.

While I agree that the “maximum range speed” of the Taycan is very likely considerably higher than that of the Model 3 (19-20mph), I very much doubt it is optimal near freeway speeds. The range is going to suffer if you drive continuously at 75mph. I thought someone posted the model details above somewhere...remember seeing it somewhere. Not sure if that model is accurate though. Really just need someone to take three Wh/mi data points at constant speed, several miles, averaged traveling in both directions, on a controlled stretch of road, from the Taycan (25mph, 45mph, 65mph) and just fit the cubic equation to it.

"TheTaycanDude" in the taycanevforum said he will soon have access to the car for an entire day. He said he will do such range related tests. He will be driving from Seattle to Portland and back.

 

[diamond.g]

Doesn't the EPA range test basically run the highway cycle until the vehicle stops? Seems way more conservative than most peoples freeway/highway experience (IE higher speed steady state driving).

 

[WarpedOne]

One question:
Why would taycan in "real world driving" had more range than EPA ratIng while EVERY OTHER EV (and non EV) gets even less range i.e. one needs to try really hard to equal it?

Wishful thinking much?

 

[AlanSubie4Life]

Why would taycan in "real world driving" had more range than EPA ratIng while EVERY OTHER EV (and non EV) gets even less range i.e. one needs to try really hard to equal it?

That’s a good question - the same one I asked above. (And you even oddly disagreed with a post of mine related to this mystery.) It’s kind of the central mystery of this thread at this point. It’s why I am curious about the Seattle to Portland (Bellevue to Beaverton) trip results. I think it will make it, based on anecdotal real-world results available thus far. Which is surprising.

Doesn't the EPA range test basically run the highway cycle until the vehicle stops?

Not quite. Not sure how much latitude there is, but in Tesla’s case they run UDDS four times and weight those results, then run HWY twice. That is only 55 miles total or so. Then they run a steady state exhaustion cycle until the vehicle stops moving. They then take the ratio of the total used energy to the energy used in each of the cycles (possibly weighted), and multiply that ratio (much greater than one) by the length of the individual UDDS and HWY cycles. This gives a UDDS number and a HWY number. This process means they are effectively extrapolated numbers; you run a cycle using a small portion of the battery, then divide the fixed cycle distance by the % of battery used. But there is also some weighting involved with the four UDDS cycles.

They then take these numbers and multiply them by 0.7032 to give Highway and city range (this is agreed ratio of 5-cycle to 2-cycle results which is allowed to be used for EVs but not for other vehicles anymore). And for the composite number it is a 55/45 weighting of these results.

The highway test is conducted at a very low average speed (~40mph?) and I think does not exceed about 55mph, and the speed varies a lot. So typically at 75 mph aero losses would dominate and you would always do worse than the test results in spite of the various reasons why you might think it is conservative (which it might well be, but remember they also multiply the result by ~0.7).

I don’t think there is any way that aero losses would not also dominate in the Taycan’s case, so the only way that it would greatly exceed the EPA range on the freeway is because Porsche somehow sand-bagged the results by using some crazily non-optimal vehicle configuration. They do have many more modes (they can turn off regen I think) and the transmission introduces a wrinkle too.

 

[ralph142]

That’s a good question - the same one I asked above. (And you even oddly disagreed with a post of mine related to this mystery.) It’s kind of the central mystery of this thread at this point. It’s why I am curious about the Seattle to Portland (Bellevue to Beaverton) trip results. I think it will make it, based on anecdotal real-world results available thus far. Which is surprising.




Not quite. Not sure how much latitude there is, but in Tesla’s case they run UDDS four times and weight those results, then run HWY twice. That is only 55 miles total or so. Then they run a steady state exhaustion cycle until the vehicle stops moving. They then take the ratio of the total used energy to the energy used in each of the cycles (possibly weighted), and multiply that ratio (much greater than one) by the length of the individual UDDS and HWY cycles. This gives a UDDS number and a HWY number. This process means they are effectively extrapolated numbers; you run a cycle using a small portion of the battery, then divide the fixed cycle distance by the % of battery used. But there is also some weighting involved with the four UDDS cycles.

They then take these numbers and multiply them by 0.7032 to give Highway and city range (this is agreed ratio of 5-cycle to 2-cycle results which is allowed to be used for EVs but not for other vehicles anymore). And for the composite number it is a 55/45 weighting of these results.

The highway test is conducted at a very low average speed (~40mph?) and I think does not exceed about 55mph, and the speed varies a lot. So typically at 75 mph aero losses would dominate and you would always do worse than the test results in spite of the various reasons why you might think it is conservative (which it might well be, but remember they also multiply the result by ~0.7).

I don’t think there is any way that aero losses would not also dominate in the Taycan’s case, so the only way that it would greatly exceed the EPA range on the freeway is because Porsche somehow sand-bagged the results by using some crazily non-optimal vehicle configuration. They do have many more modes (they can turn off regen I think) and the transmission introduces a wrinkle too.

iirc the highway portion tops out at 58 mph and averages 42.

 

[diamond.g]

iirc the highway portion tops out at 58 mph and averages 42.

Interestingly US06 hits higher speeds, but somehow averages about the same as the highway test.

 

[AlanSubie4Life]

Interestingly US06 hits higher speeds, but somehow averages about the same as the highway test.

Thanks. I was a little off on my recollection. But anyway, it's way slower than 75 mph and the aero losses are less than half of what they would be at that speed so would normally be optimistic.

But, they multiply the result by 0.7. And in this specific case it seems likely (if the Taycan actually makes the SEA-PDX trip no problem) that somehow Porsche stacked the deck against themselves when they did their range testing.

Lots of disagreements with earlier posts of mine. Not sure why. If the test hasn't been sand-bagged, I agree this Seattle-Portland trip would be a tough one to make for the Taycan. There appear to be a few anecdotes that indicate the test has been sand-bagged. We'll find out - the guy is very active on Twitter!

 

[PhantomX]

One question:
Why would taycan in "real world driving" had more range than EPA ratIng while EVERY OTHER EV (and non EV) gets even less range i.e. one needs to try really hard to equal it?

Wishful thinking much?

I don't think that's always true. When I had the i3, I was constantly beating the 114 miles EPA rated range. And this is with mixed local and freeway driving here in So. Cal. In comparison, I have a much harder time getting my Model 3 MR to match EPA range with the exact same commute, even though Model 3 is technically more efficient.

 

[AlanSubie4Life]

The detailed EPA documents showing what was done in the EPA test of the Taycan are now available. The mystery persists. I suspect that subsequent to these submissions, Porsche has implemented massive drive efficiency improvements (and possibly changed the shifting strategy during the test though I would not expect that to make that large a difference), if discussion about these "250 mile" range display numbers in the vehicle are to be believed (and if that display in the car represents something related to EPA miles) If so, these improvements will show up eventually, probably first in CARB.

I've attached a spreadsheet comparing to the Model S Performance and Model 3 Performance; it's not a pretty picture for Porsche, so far:

The explanation is not regen - the brake pedal in the Porsche is all regen up to 0.3g, according to these documents. The vehicle was run in 1st gear only according to one note, though:

"Note: Normally the vehicle stays in 2nd gear. First gear is used only for very high acceleration, or in "Sport" or "Sport Plus" modes. Only 1st gear is utilized on the FTP and Highway cycles."

Here are the Porsche documents, anyway; I think there might have been a couple other interesting pieces of info in there too which I can't recall right now.

https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=48943&flag=1

https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=48941&flag=1

Anyway, still no explanation of the mystery of all the videos from people saying they have driven 250 miles no problem in the Taycan. I have to imagine they are using a different version of the software than was used for this testing - there's just no way they would make it if there is not some sort of large efficiency improvement from these numbers.

 

[PhantomX]

The detailed EPA documents showing what was done in the EPA test of the Taycan are now available. The mystery persists. I suspect that subsequent to these submissions, Porsche has implemented massive drive efficiency improvements (and possibly changed the shifting strategy during the test though I would not expect that to make that large a difference), if discussion about these "250 mile" range display numbers in the vehicle are to be believed (and if that display in the car represents something related to EPA miles) If so, these improvements will show up eventually, probably first in CARB.

I've attached a spreadsheet comparing to the Model S Performance and Model 3 Performance; it's not a pretty picture for Porsche, so far:

View attachment 502955

The explanation is not regen - the brake pedal in the Porsche is all regen up to 0.3g, according to these documents. The vehicle was run in 1st gear only according to one note, though:

"Note: Normally the vehicle stays in 2nd gear. First gear is used only for very high acceleration, or in "Sport" or "Sport Plus" modes. Only 1st gear is utilized on the FTP and Highway cycles."

Here are the Porsche documents, anyway; I think there might have been a couple other interesting pieces of info in there too which I can't recall right now.

https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=48943&flag=1

https://iaspub.epa.gov/otaqpub/display_file.jsp?docid=48941&flag=1

Anyway, still no explanation of the mystery of all the videos from people saying they have driven 250 miles no problem in the Taycan. I have to imagine they are using a different version of the software than was used for this testing - there's just no way they would make it if there is not some sort of large efficiency improvement from these numbers.

I wonder if this is the typical under-selling of spec that German makers like to do that's biting them back. My experience with BMW and Porsche ICE's are that they tend to outperform the spec they published, be it performance or efficiency numbers. For example, I was able to get 30+mpg on my BMW F80 M3 on a highway cruise and 27+mpg on my Porsche 987.1 Cayman S. Both are those are quite a bit above the EPA published numbers.

Similarly, my friends and I have been able to easily beat the EPA range for BMW i3, so I wonder if this is a case of Porsche going too far trying to be conservative with EPA tests and went too far?

 

[AlanSubie4Life]

so I wonder if this is a case of Porsche going too far trying to be conservative with EPA tests and went too far?

I personally don't think so, but maybe. Aside from that comment about the FTP & Highway tests being done entirely in 1st gear, Porsche doesn't have a lot of latitude on what they can report here. They have to clearly state their test conditions, and then they have to run the car, and it has to be reproducible by the EPA if they happen to choose to test the Taycan. I attached the link to the documents hoping maybe someone can point out what the issue might be. The car was not run in Sport+ or Sport mode - that is explicitly stated. They did the 2-cycle test and used the 0.7 scale factor to convert to 5-cycle equivalent, as everyone else does.

There is the 0.7 scaling factor - Tesla uses 0.75 for Model S Performance (it's different for other Model S variants) which makes their results look 7% better (which is why I include the true DC efficiency calculation, which compares the raw results scaled by 0.7 across the board). I don't understand why Tesla can use different values for this "0.7" but apparently there is some justification. That's really the only latitude that they have when they're doing these tests. But the Taycan is still way worse even when 0.7 is used for everything. Specifically, the Taycan uses at least 26% more energy per mile. This is in spite of the road load HP at 50mph being very similar for the two vehicles (Taycan Turbo vs. Model S P 21"). Implies a lot of drivetrain inefficiency, I would think.

As I said, my suspicion is that this is just dated software which was used, which has subsequently been improved. If they are actually closer to a rating of 250 miles now, that would put them nearly on par with the Model S Performance (which has a substantially larger battery, when looking at available capacity). We'll see!

 

[PhantomX]

I personally don't think so, but maybe. Aside from that comment about the FTP & Highway tests being done entirely in 1st gear, Porsche doesn't have a lot of latitude on what they can report here. They have to clearly state their test conditions, and then they have to run the car, and it has to be reproducible by the EPA if they happen to choose to test the Taycan. I attached the link to the documents hoping maybe someone can point out what the issue might be. The car was not run in Sport+ or Sport mode - that is explicitly stated. They did the 2-cycle test and used the 0.7 scale factor to convert to 5-cycle equivalent, as everyone else does.

There is the 0.7 scaling factor - Tesla uses 0.75 for Model S Performance (it's different for other Model S variants) which makes their results look 7% better (which is why I include the true DC efficiency calculation, which compares the raw results scaled by 0.7 across the board). I don't understand why Tesla can use different values for this "0.7" but apparently there is some justification. That's really the only latitude that they have when they're doing these tests. But the Taycan is still way worse even when 0.7 is used for everything. Specifically, the Taycan uses at least 26% more energy per mile. This is in spite of the road load HP at 50mph being very similar for the two vehicles.

As I said, my suspicion is that this is just dated software which was used, which has subsequently been improved.

I should have worded better. What I meant to say is that Porsche may have designed the car to be more efficient in typical real world driving conditions, rather than EPA conditions. One of the ways this could happen is they have a short 1st gear in their 2-speed transmission. And if the EPA tests were all ran in 1st gear, then it would mean the car is running with a less ideal setup in term of efficiency. Just a speculation though, but I got to think there has to be a reason why some cars can beat EPA numbers and some cannot.

 

[AlanSubie4Life]

And if the EPA tests were all ran in 1st gear, then it would mean the car is running with a less ideal setup in term of efficiency. Just a speculation though, but I got to think there has to be a reason why some cars can beat EPA numbers and some cannot.

Sure, and maybe that was a deliberate choice, or more likely it was an issue with the software with the vehicle in dyno mode only supporting first gear use at the time of the test, or something like that.

I don't think there are ever really substantial differences between real world results and the EPA tests (unless the test was performed fraudulently), assuming you are comparing under conditions similar to the EPA test - however, that mysterious 0.7 scale factor (which can be as high as 0.75!) is one degree of freedom which I don't understand which would explain part of this discrepancy, for EVs (doesn't apply if the vehicle is tested with the 5-cycle test). But the test conditions are pretty carefully defined.

I just don't think it makes sense to be so conservative on such a key metric for EV sales. It's true they are just selling 2000 of these vehicles or so, but people still care about range. My guess is it was just a placeholder test and the software was lagging behind at the time, and their hand was forced. We'll have to get some more info on real-world results - hopefully that Seattle to Portland run will happen soon.

 

[PhantomX]

Sure, and maybe that was a deliberate choice, or more likely it was an issue with the software with the vehicle in dyno mode only supporting first gear use at the time of the test, or something like that.

I don't think there are ever really substantial differences between real world results and the EPA tests (unless the test was performed fraudulently), assuming you are comparing under conditions similar to the EPA test - however, that mysterious 0.7 scale factor (which can be as high as 0.75!) is one degree of freedom which I don't understand which would explain part of this discrepancy, for EVs (doesn't apply if the vehicle is tested with the 5-cycle test). But the test conditions are pretty carefully defined.

I just don't think it makes sense to be so conservative on such a key metric for EV sales. It's true they are just selling 2000 of these vehicles or so, but people still care about range. My guess is it was just a placeholder test and the software was lagging behind at the time, and their hand was forced. We'll have to get some more info on real-world results - hopefully that Seattle to Portland run will happen soon.

Good point on potential software issue causing it to run in 1st gear only. It would definitely be interesting to see what the real world number is once more Taycans are delivered to customers.

 

[diamond.g]

the EPA tests are over a set amount of time, would running the test in second gear really improve things that much?

 

[WarpedOne]

no.

 

[AlanSubie4Life]

the EPA tests are over a set amount of time, would running the test in second gear really improve things that much?

It doesn’t seem like it would make a huge difference at the typical test speeds, no. But the fact that that may have been the case indicates their software was pretty beta at test time, so would not be surprised by massive subsequent drive efficiency improvements. (Remember Tesla has improved efficiency by ~8% over time and they started with something they did not have to rush out.)

We’ll see. Remember the allusions to the new drive software in the CARB documents.

 

[FlatSix911]

Just putting this out there... Official Big Game Commercial 2020 - Extended Cut

Can you find the Hybrid 918 is this video?