More range?

[epirali]

If you want to accelerate with a certain rate, you will spend the electrons accordingly. The firmware will affect the way the pedal responds to encourage smoother driving, but it will not change the actual power consumption. The regen can be made a bit more efficient, but the gains would be so minuscule, they are not worth compromising the driving experience.

Yes but how quickly the pedal curves and how quickly regen kicks in will moat the efficiency. For example I will bet that regen/recharge is not equally efficient at how much of the power ends up in the battery. So the regen curve can potentially make small but significant differences in the overall range. Also no matter how carefully you control the pedal there is no one-to-one relationship to the current draw from battery, the control software is using the pedal input, after averaging and filtering, to determine the power draw. So again there can be theoretical improvements.

 

[epirali]

Another update for anyone who may be interested: so I made a 26.6 mile trip (1/2 of daily commute) at 50f. Same roads, same speeds, eco pro, mostly auto cruise. Very familiar with power use. What I saw is very interesting. I have never seen a GOM range reported of 80 miles since I got this car, and today with preconditioning (which I always do) I have an 82 mile GOM range in Eco Pro with climate on.

At the end of my trip the SOC meter showed 68% remaining (which again I had now seen in these temperature before, I had programmed the car to have SOC hold so I could monitor a percent from that page before). So the GOM range matches what I managed to get from an actual trip (83 miles or so). The more interesting part is that my efficiency reported for trip is now very different. I used to get very high efficiency at start of trip and it would slowly settle down to 3.7-4 KW/mile before. It no longer does that. It started at low efficiency (2.7 or so) and slowly settled to a final 3.8 at the end of trip. Obviously the car does not have a 21 KWHr or usable capacity, so something is not altogether right here. Unless there is now actually almost 20 usable of the 22 KWHr of capacity (90%), which is doubtful.

So in brief: no increase in efficiency, but a decrease in SOC % used and an increase in total reported and measured range. I did notice the ACC response is different than before (less jumpy).

I am going to try to run the battery all the way down, the only issue is with multiple starts the initial drop for climate and trip efficiency loss will cause some error. But I have a pretty good idea what SOC I should be in on a full round trip of almost 60 miles.

 

[jadnashuanh]

Bottom line, you as the driver, have the ultimate control over how efficient you are in a particular vehicle. Same thing with an ICE...how you drive it affects how efficiently it runs. The EPA test procedure is a fixed plan that enables repeatable results across multiple platforms...that does not mean you will get the same results.

 

[epirali]

Bottom line, you as the driver, have the ultimate control over how efficient you are in a particular vehicle. Same thing with an ICE...how you drive it affects how efficiently it runs. The EPA test procedure is a fixed plan that enables repeatable results across multiple platforms...that does not mean you will get the same results.

I am a trained engineer and I own multiple electric vehicles. I am very familiar with this concept, along with electric motors, battery technology, charging circuits, etc. I was not talking at the global level. Batteries are not as linear in their behaviour as most people think, the drain/charge rate and various characteristics will impact the total available capacity. LiIon is less susceptible to this than other chemistries but it is not immune. Most people have a simplistic view of "total x available how you drive determines rate." But it actually is a bit more complicated than that. This doesn't interest most people, and the simple model is just fine.

The EPA test plan is used to eliminate all variables. But that doesn't mean an engineering approach can not create real world data that measures results. Temperature/air density, route (elevation change), speed, acceleration curve, tire pressure and wear (rolling resistance) can be used to compare data points.

So yes the biggest efficiency factor is the driver pattern + air resistance from speed + temperate (capacity useable + air density). But other control/design factors do impact the overall range. I believe there may be a 10% increase in useful range from the firmware update but need more data.

 

[epirali]

Ok I think I'm figuring it out (maybe others have to). I drained the battery on a round trip and routes I know. On the outbound trip I used 33% of reported SOC for 27 miles, efficiency of 3.7 KW/mile. On return trip (which is slightly more uphill) I used 41% of reported SOC efficiency of 3.5 KW/mile (which is typical for return trip). But the reported SOC loss was much larger than the efficiency difference. And by the time the battery reached Rex startup my range was no different than before the update as others have reported.

So I believe whether because of reset of data or because of new firmware the reported SOC range is NOT linear, and is tilted towards the higher SOC. This explains both the much more optimistic GOM numbers and why on first leg (100-66%) the numbers lines up to GOM which makes it seem like there was a range increase from firmware. I guess I'll wait to see if the SOC scaling dynamically adjusts over time to become more linear, which may explain the gradual return to the old GOM range numbers.

 

[jadnashuanh]

Someone told me, and I do not know if they had any valid basis for that statement so beware, that it takes the car 500-miles or so of driving to 'learn' both you and the car's actual operational characteristics...up until that point, it's operating on an unoptimized set of parameters and thus not as likely to be accurate.

 

[bwilson4web]

Is there an easy way to check our version?

The reason I ask is I picked up the car May 27, over two weeks after this thread started. Once I got home, we had some outstanding warranty work done including the 'bolt'. If we can check our 'version' easily, members can report what they have and we can propose ways to test the various hypothesis.

Thanks,
Bob Wilson

 

[alohart]

Is there an easy way to check our version?

The software version is included in the user profile data that can be downloaded to a USB flash drive using an iDrive menu command. Insert this flash drive into a computer's USB port and open the only file on the drive in your favorite text editor. The software version is the value of the "i-step" tag.

 

[jadnashuanh]

Different models and countries can have a different series of s/w, so it can be a little misleading. Some of them seem to track across the board, but not all.

 

[epirali]

Someone told me, and I do not know if they had any valid basis for that statement so beware, that it takes the car 500-miles or so of driving to 'learn' both you and the car's actual operational characteristics...up until that point, it's operating on an unoptimized set of parameters and thus not as likely to be accurate.

Right. But what's interesting is at least in my case the "error" is mainly in the non-linearity of the SOC %.

 

[bwilson4web]

Is there an easy way to check our version?

The software version is included in the user profile data that can be downloaded to a USB flash drive using an iDrive menu command. Insert this flash drive into a computer's USB port and open the only file on the drive in your favorite text editor. The software version is the value of the "i-step" tag.

Ok,

This is what I've got:

<i-step>I001-16-07-506</i-step>

Bob Wilson

 

[alohart]

Our 2014 i3 BEV is running software version I001-16-07-506. I haven't noticed any increase in range since this software was installed on 1 November, but the Batt. Kapa. max value has consistently been higher than before the update:

before 19 Oct: 16.5 - 19 kWh
19 Oct.: 16.5 kWh
4 Nov: 20.9 kWh
~13 Nov.: 19.9 kWh
22 Nov.: 20.0 kWh

I don't understand how Batt. Kapa. max is calculated or what I would need to do to get consistent values.

Explanations for the increased values could be that the algorithm used to calculate Batt. Kapa. max was modified, or that the unusable battery pack capacity buffers could have been reduced in size to hide possible battery capacity loss. If the usable battery pack capacity is actually ~20 kWh, that implies that the unusable battery pack capacity buffers are 1 - (20 / 21.6) = 7% split between the low and high charge level buffers. That's so small that it would be easier to damage one's battery pack by either discharging the battery pack as much as the battery management system allows or by leaving it fully for a significant period of time. I hope that the protective buffers weren't reduced in size, but we might never know.

 

[alohart]

Yesterday, I noticed the lowest predicted range I have ever seen in our 2014 BEV: 79 mi. This despite having driven the past 6 miles with an efficiency of 5.5 mi/kWh and an ambient temperature in the mid 70's. I had not driven more aggressively at any time recently, so the low range prediction was odd. It could be explained by a low battery capacity, but Batt. Kapa. max was 20.1 kWh yesterday, consistent with the higher values since the July 2016 software upgrade was installed on 1 November (did BMW actually reduce the sizes of the already small unusable capacity buffers?). 20.1 kWh x 5.5 mi/kWh = 110 mi which is far greater than the 79 mi estimate. The 79 mi estimate is for Comfort mode whereas my 5.5 mi/kWh lifetime average has occurred under Eco Pro mode, so the estimate difference is somewhat better that it appears. The Comfort mode range estimates prior to the July 2016 software upgrade were very accurate for my Eco Pro mode driving which makes those estimates high for actual Comfort mode driving. Maybe the correction that was implemented intended to make the Comfort mode estimate more accurate for Comfort mode driving.