Wednesday, October 29, 2014

SF Bay to Tahoe in an i3 REx: What was learned?

Donner Summit is the highest point along Interstate 80 in California at 7,228 feet elevation.
Last week we heard from i3 REx owner John Higham in a post he wrote which detailed his thoughts on the i3's range extender restrictions for the North American market. John certainly didn't mince words and offered his reasoning why he believes the range extender on the i3 should have its artificial restrictions (which are in place to satisfy CARB), relaxed a bit. John also promised to do a road trip which would take him from the San Francisco Bay area up to the 7,228ft Donner Summit in Lake Tahoe and report on the range extender's performance under these strenuous conditions. Below are his findings.

SF Bay to Tahoe by the Numbers, Part 2

Nailed it.  Well, nearly.

In Part 1 of this post about all things REx, I declared that a US spec BMW i3 REx could not make the popular weekend getaway of Lake Tahoe from the San Francisco Bay area without being speed limited within 15 miles of Sacramento. How speed limited depended on the slope of the road as you climbed east into the mountains, but top speed would range from 40 to 55 MPH. The alternative was to fully charge in Sacramento before any significant climbing begins and then again in Colfax about halfway up the hill. This makes such a drive impractical.

I also declared that a European spec’d i3 would make it no problem, so long as one kept the diminutive 1.9 gallon tank filled and the feature known as “Hold Mode” engaged. In fact, European cars have made similar drives into the Alps.

The basis of these declarations is simple physics. In Part 2 we test the physics from Part 1.Twice. First with a U.S. spec BMW i3 REx and then with a European spec i3 REx.

OK, I lied. I don't have a European spec i3. But I do have a US spec i3 that has been modified to behave like its schnitzel eating cousin.
I drive the BMW i3 from my home in Mountain View, California to Donner Memorial Park in Truckee.  The state park is placed at one of the sites where the ill-fated Donner Party settled for the winter in 1846.  The snow that winter was as high as the memorial behind the i3.

A Quick Summary of Part 1

The California Air Resources Board (CARB) developed a class of car called the Battery extended range Electric Vehicle (BEVx). Some say the BEVx was never intended to be a car with mass appeal that can be driven like any ICE-mobile.  But I ask, why not? Actually, what I usually say is “Why the hell not!” while shouting and pounding the table with my fist. I digress.
I believe that the BEVx class of cars represents the bridge from plain ol' Internal Combustion Engine (ICE) cars to pure electric that will finally allow the public to embrace EVs without looking back. Except.

There is one teensy exception and it is what engineers call a "corner case." In this situation the “corner case” refers to those people who require a car to maintain freeway speeds over sustained elevation gains. That’s what this post is all about -- to test how big that corner case actually is.

California's San Francisco Bay Area lies at sea level and the drive east to Lake Tahoe follows the Sacramento river, never gaining significant altitude for about 50 to 100 miles, depending on one's starting location. Continuing east past the capital of Sacramento begins what is at first a gentle climb into Gold Country. Assuming the route is along I-80, the slope increases significantly past Gold Country until Donner Summit (elevation 7,228 feet) is reached 95 miles east of Sacramento.

The i3's APU is sized such that it can maintain freeway speeds, but not to maintain freeway speeds and simultaneously gain significant altitude.  It’s simply not possible to drive from the SF Bay Area to Tahoe in a reasonable amount of time with the US spec’d i3. Of course if you have the patience to charge every 60 to 80 miles, you can drive your i3 from the Bay Area to Tahoe or anywhere else for that matter. But that is impractical, even with with so-called fast chargers.

Since this post comes in two parts, and the test drive to Tahoe also comes in two parts, potential for confusion exists when referring to them.  Let’s dispatch any confusion and call the first test The Apple Pie Test and the second test The Lederhosen Test.

The purpose of these tests isn’t to prove you can drive an i3 to Tahoe by taking logical opportunities to charge. You can. It’s been done. The purpose is to prove the assertions made in Part 1. First, that the US spec’d i3 REx is hobbled as compared to its European counterpart and second (and more importantly) that an i3 REx is more than a great EV; it has potential to be the only car you need.

Oh yeah. No math in this post. I promise.

The Apple Pie Test

The Apple Pie Test is simple: try to “REx it” to Tahoe and see how far you get. (Oh, I’ve made REx a verb, but the Oxford dictionary hasn't caught up yet.) Since this is my test, I get to make up the rules. The rule is simply to take a BMW i3 as CARB intended it to be delivered to the public and drive it along I-80 until the car becomes speed limited, then compare the observed results with the predicted results from Part 1.

To do this test I left the Benicia, California, CCS fast-charger with 90% SOC and a predicted range of 60 miles.  The drive toward the Tahoe region is essentially flat for about 63 miles along I-80, then the road climbs into the Sierra Nevada mountains. I planned this section of the drive to be all electric until such time I hit the foothills. The goal was to set the cruise control to the posted speed limit (65 MPH) and simply keep driving powered by the REx until the car became speed limited.

In Part 1 I calculated that the car would become speed limited at about 725 feet elevation gain and by using the elevation profile in Google Earth, I estimated that would occur about 12 miles east of Sacramento.

The actual drive didn’t work out exactly like that, but close enough for the rough assumptions that were made. At first the speed limitation was subtle. I started to suspect the car was speed limited at around 800 feet elevation (750 gained), “flooring it" to coerce an increase from 65 MPH with the cruise control set, I achieved about 67 or 68 but no more. But by 950 feet elevation gain the effect was no longer subtle. Not only could I no longer keep pace with traffic, but was feeling very vulnerable and was searching for an exit in earnest. On some of the steeper portions of that section I was under 55 MPH indicated with traffic whizzing past at 70 MPH and above.
The BMW i3 was clearly speed limited on this section of road after leaving Sacramento powered solely on the REx

This photo was snapped moments after the previous photo

Anyone who has owned a BMW for very long can tell you that the speedometers are optimistic by at least 5%, if not 7%. So, that 58 MPH in the photo is closer to an actual speed of 55 MPH.  In Part 1 of this post I made a table of predicted top speed as a function of the grade of the road. Using the GPS coordinates of the road and Google Earth, I found the grade of the road at the precise point is 3%; the table from Part 1 predicts a top speed of 60 MPH on a 3% grade; close, but some refinement of that table is in order.

In summary, the Apple Pie Test demonstrated that all that analysis, the calculations, graphs and so forth from Part 1 were within the margin of error that could be expected for the rough assumptions that were made.  

More importantly, It proves that you can’t just REx it to Lake Tahoe in an i3. Luckily, there is a CCS charger in Sacramento, so moments after the above photos were taken I turned around and headed straight for it. With the miracle of regeneration the i3 got its SOC back up to a respectable level and I REx’d it all the way back to that CCS charger with no issues.

Hold Mode and Coding

The Lederhosen Test requires the use of a feature known as “Hold Mode”, which is on all Euro-spec i3 equipped with the REx; perhaps even all such cars destined for anywhere in the world outside of North America. What Hold Mode does is engage the REx (or more specifically in CARB-speak the APU) to maintain the battery State of Charge (SOC). Sounds a bit boring and perhaps it is.

The fact of the matter is, US spec’d cars have Hold Mode; the car’s onboard intelligence switches it on automatically when the battery SOC reaches 6.5%. The European version of the car also will switch on Hold Mode automatically when the battery SOC reaches 6.5%, but the European version also allows the driver to manually engage Hold Mode whenever the battery SOC is 75% or less.

The difference in the US spec’d car and its European counterpart is perhaps subtle, but as we shall see, the difference means everything if you require a car to maintain freeways speeds and gain significant elevation simultaneously.

What is important here is to understand that the US-spec cars do in fact have the European-spec Hold Mode programmed into the car.  The menu option that allows the driver to engage Hold Mode manually is simply hidden from the i3’s iDrive menu. For someone skilled in the seedy underbelly of the BMW tuner world known as “coding,” enabling this hidden feature in the iDrive menu is trivial.  To be clear, this practice is most likely frowned upon by both BMW NA and CARB.

To satisfy scientific curiosity, I “coded” my i3 to enable Hold Mode, Euro-style. On to the Lederhosen Test!    (click through this link to read about how to code your i3: Code your i3)

The Lederhosen Test

As noted in the last paragraph of the Apple Pie Test, as soon as I became speed-limited near Auburn, I turned around and returned to Sacramento and specifically to the CCS fast charger there. After plugging in and after i3’s SOC had reached 90%, I once again set out along the same route toward my final destination in Truckee, California, near Lake Tahoe. Hold Mode is only available if the SOC is 75% or less, so after leaving the CCS charger I drove the first 12 or 13 miles all electric.

The CCS fast-charger in Sacramento in Sacramento is at an elevation of 50 feet.

The only difference in the two drives was the SOC at the bottom of the hill and manually engaging Hold Mode. This simply means the REx was used in the Apple Pie Test to “hold” a 6.5% SOC but on the Lederhosen Test, it was used to “hold” a 75% SOC.
Engaging Hold Mode at 75%.  Note there are 88 miles to my destination, with 39 miles of all-electric range available.  

In Part 1 of this post I calculated that by engaging Hold Mode at 75% SOC the i3 should be able to climb essentially any mountain pass in North America, so long as one keeps the gas tank filled. What isn’t visible in the photo above is that Donner Pass, a 7,228 foot climb, is between me and my destination; it is time to put my hypothesis from Part 1 to the test.

With Hold Mode engaged, as one drives the i3 the REx keeps the battery SOC constant at the level set.  If driving conditions are such that the REx (due to its limited power output) cannot keep the battery SOC maintained, then energy from the battery makes up the difference and the battery SOC falls commensurately.

Soon after leaving the CCS charger in Sacramento and engaging Hold Mode at 75% SOC I found myself once again in Auburn near where I had turned around just 90 minutes earlier during the Apple Pie Test.  It was time for a lunch stop.

As expected the battery SOC falls as elevation is gained
The photo above was taken at my lunch stop in Auburn. Note that the SOC has fallen 4% to 71% at 1210 feet elevation (1160 feet of gain).  If I had stopped the car and let the REx run sufficiently long, the SOC would have returned to 75%. But that would have both taken time I didn’t want to spend and defeated the purpose of the Lederhosen Test. So, after a quick bite to eat I got back in the car and re-engaged Hold Mode at 71% SOC.

Leaving Auburn, I resumed toward my destination of Donner Memorial State Park 65 miles away in Truckee, California. The only thing between me and my destination was Donner Pass at 7228 feet, one more stop for gasoline, and the potential to run out of battery. But I had done my homework and was confident that I had plenty of energy left in the battery to complete my drive.

It was the perfect day for such a drive; the sky was a beautiful blue, the temperature was in the mid 70’s, the traffic light and SiriusXM’s Classic Vinyl accompanied me. During the drive I took pictures of the i3’s displays every 1,000 feet of elevation gain, but suffice it to say that the battery SOC slowly dropped in an expected and predictable fashion as I glided up the mountain's slope. After 45 minutes or so I once again stopped to top off the fuel tank.

The i3’s other “fast charge” port.  I don’t like to use this method of adding energy, but sometimes a guy’s gotta do what a guy’s gotta do.

The battery SOC has fallen from 75% to 54% after climbing 5300 feet.

After refueling, Donner Summit was less than 30 minutes away. I found myself so absorbed in monitoring the progress of the battery SOC prediction that I nearly blew past the sign marking the summit! 
The i3, with Hold Mode engaged, used a mere 31% (75% at the bottom of the hill less 44% at the summit) of its SOC to gain nearly 7,200 feet of elevation.  In simple terms, one can think of it as if the REx’s power output is used to propel the car forward, the battery’s power output is used to climb the hill.

By using less than a third of its battery to gain those 7,228 feet, the i3 REx is obviously capable of much more. In Part 1 I asserted that the i3 with the European-style Hold Mode was probably capable of summiting any road in North America. After making the drive over the Sierra Nevada’s I-80, I believe that point has been verified.


The i3 REx with the European-style Hold Mode is more than capable of conquering Donner Summit simply by engaging the feature at the beginning of the climb and keeping the tank filled.  The US spec i3 REx is not.  But the implications are far greater than this.

The entire thesis of this post and the previous one is much larger in scope than “can BMW’s i3 make the drive to Lake Tahoe.” The thesis is much more than the car or the corporation. It’s about an idea.  A brilliant idea.

It’s about a transitional electric vehicle that the public can embrace without looking back, without asterisks and without range anxiety. The embodiment of that transitional electric vehicle is the BEVx class; to date only one car is made to that standard. It’s a brilliant piece of engineering.  Yet that brilliant piece of engineering is emasculated by regulations imposed by a governing body that should be championing it.

I’m surprised that Sir Isaac Newton hasn’t leapt from his grave and set his hair on fire.

The use case I have been passionately trying to demonstrate, that the i3 is fully capable of, may be an inconsequential corner case for the majority of owners worldwide.  But it is a legitimate use case and one that the many buyers consider. And people buy to the corner case, especially if it is their only means of transportation.

Until such time that adding energy to an EV takes as much thought and effort as adding energy to an ICE-mobile, technologies like the BEVx are going to be required to get the public to embrace electric mobility.

If removing the restriction on the operation of the APU is not made, the genius of the BEVx classification will never bear fruit.  That’s because even though the average driver does less than 40 miles a day, they also want the flexibility to take their car wherever they want, whenever they want. For this reason, PHEVs are about as “electric” as the general public is willing to go.

Once the current limitation of the APU software managing the SOC is understood by the public, the public will eschew the BEVx classification for PHEVs, such as a Volt. While that may be a better choice for the environment than, say, a Camry, the Volt driver will not be able to drive as much on electricity as if he bought a BEVx, such as an i3.

That’s why I’m writing; to beg CARB to Unleash the REx. It’s been said that the PHEV is the gateway drug to a pure BEV. If that is so, the BEVx has the potential to be crack -- instantly addictive. Make it so.

Facts about my trip from Mtn. View -> Truckee -> Mtn. View
Left home with 100% SOC
528.2 miles round trip
246 miles on REx
6.6 gallons of gas purchased
Ended trip with about ½ gallon more gas in the tank then when I left
4.1 mi/kWh
4 CCS charging sessions totaling 62.8 kWh
0 Level 2 charging sessions
Arrived home on the REx (6.5% SOC)


  1. Point made. So how would the Lederhosen trip have been different if you'd started it completely anew from home? I think one thing the CARB guys are concerned about is people, being nervous about what terrain or need for extra power might lie ahead, will stick their hold mode on at 50% or higher, maybe never change it back, and end up using a lot of gas.

    1. That is a possibility Rob. However I believe that would be the exception, not the norm.

      The people that buy and drive electric cars go way out of their way to drive as many electric mails as possible. If you talk to Volt owners, they will tell you how important it is to them that they drive as much on electric as possible and some go way out of their way to do so.

      I don't think many people will shell out $50,000 for an i3 REx and then activate the range extender all the time, even when it's not needed. I believe the vast majority of people would only use it when they believed they absolutely needed it so they can drive safely to their destination.

    2. I agree it’s the exception, but some people do treat a plug-in hybrid like a regular hybrid and just run it on gas most the time. You hear stories of fleet Volts that almost never get plugged in. This may be the case of other incentives like tax breaks or carpool lane stickers causing someone to buy a plug-in when they don’t really intend to plug it in.

      Obviously the CARB is trying to avoid this with the BEVx restrictions. They don’t trust that owners will choose electricity if gas works just as well. So they cripple the use of gas with a small tank and the inability to turn on the Rex when desired. And you know that if it had a large tank and no REx restrictions that there would be some jokers that wouldn’t bother to plug in much if at all.

      We need to convince CARB that the small tank and a default of using nearly all the electricity first is enough to ensure the goal of 90% electric miles. The ability to engage the REX on demand won’t lead to significantly greater gas use, just additional peace of mind that will lead to greater adoption of this mode of transportation.

    3. Yes Kurt I too have heard of this, but as you mention, this was usually only done with fleet vehicles. Also, sinct the Volt can be driven hundreds of miles on a tankful of gas, it wasn't very inconvenient.

      If they were trying to drive an i3 Rex like that, first, they would have to enable hold mode every time they got in the car - something you don't need to do with the Volt. Then, they would need to stop for gas every 40 or 50 miles, again, a major inconvenience which would prevent nearly anyone from using the vehicle in that manner. I think the 1.9 gallon gas tank is fine, I just want the Rex to turn on at a higher SOC when necessary - like the trip John took.

  2. This is a CARB case where the desire for perfect has got in the way of being great. Perfect would be fine if we had affordable 300 mile battery cars, until then they should be promoting great. As my Lederhosen attired, apple pie eating friend has noted, the i3 rex is great, and has been marginalized to simply average by our bureaucracy. That works counter to California's goal of getting millions of drivers to plug in.
    California needs to adjust it's CARB.

    1. Just to add to what Tom said, the statistics from show that the community taken as a whole, drive 80% of their miles on electricity. One guy has done 30,000 miles on a single tank of gas (8 gallon tank).

      Once experienced, people love to drive electric.

      That and the fact that the tank is 1.9 gallons is all you need to ensure that people aren't just going to drive the thing on gas.

  3. And... if you had topped up at the CCS station in Sacramento before attempting the climb? How far would you have gotten up the mountain if you hit the base of the mountain with a full charge?

    You're presenting two scenarios: a) BEV that's empty at the base of the mountain, and relying on APU... b) BEV that's full at the base of the mountain, and a reconfigured APU .... That leaves out the third scenario, full BEV at the bottom of the mountain with the original APU configuration.

    1. Yes but the test was to see if the hold mode as implemented on Euro-spec i3s could allow the i3 REx to climb the mountain without going into reduced power mode. The point wasn't if a fully loaded BEV could climb the mountain. It wouldn't be often that someone can charge their EV at the base of a long mountain climb (it just so happens you can in this instance) so the real point as far as I can see is can an unrestricted REx climb any mountain in the US using the hold mode as set for European customers and the answer is yes, it definitely can.

      With the US-spec REx restricted to turning on at 6.5% it can only climb around 750ft before the 6.5% buffer is exhausted and the car slows down.

      As we've covered here, BMW is indeed working on a solution (Short of allowing Euro hold mode though) and I'm anxious to see how it works in real world applications like this one.

  4. David, I think you mean I'm presenting two REx scenarios, yes? Here is the third REx scenario that you suggest.

    I've analyzed this several different ways trying to minimize the time to get to Tahoe. Assuming an i3 REx, if you are at the bottom of the hill with 100% charge, you will get to Colfax easily, but there are no other charging opportunities until after the summit (still several thousand feet of climbing remaining). You must charge in Colfax, or use the REx to continue. Engaging the REx too late and you are screwed (will be speed limited).

    If trying to avoid using the REx (driving nearly 100% electric) the most time efficient way to get to Tahoe is to charge 100% at the CCS station in Sacramento. You will reach Colfax (52 miles away, a couple of thousand feet in elevation gain) with very little to spare but it should be possible in an i3.

    There is a L2 in Colfax and if you charge completely you *should* make it to the top of Donner Summit, but it will be very close in a REx. A BEV (lower mass) will fare better. From the summit you can regen down to the next L2 in Truckee.

    Everyone has different thresholds of what they are willing to do, etc. but for me, Tahoe is a weekend trip or a three day weekend trip. I'm not going to use several hours of that precious time window to sit at a Level 2. I'll either REx it (like in the Lederhosen Test) or I'll take an ICE-mobile.

    1. John, your argument, viz., that what the Air Resource Board regulators intended to achieve in improving air quality is actually working against that very aim with their REx restrictions, is compelling. Why not contact an environmental reporter from the San Francisco Chronicle, the Sacramento Bee or the San Jose Mercury News--actually, do all three papers--present your case, and get them to do a piece on how Air Board regulations are working against improved air quality. CARB officials are appointees and therefore do not really respond to a constituency except for those who appoint them or those who fund them.. Build some pressure. A couple of good newspaper stories on this would be a terrific way to get the attention of decision-makers. John, your posts are not just informative, they are exhilarating. Please keep writing. One last question. As you understand them, how far will the software changes BMW will provide in 2015 make the REx more user friendly, for instance, how would they impact a trip to Tahoe?

    2. Excellent idea! I can and will do just that.

      I also plan to write elected officials. To date, I've only written appointed officials from CARB. After I gather addresses, etc., I'll publish a list of who to write here and also on FB & the i3 forum.

    3. Another brilliant post John.

      Not sure how the BMW folks will feel about "schnitzel eating" and "lederhosen," but hopefully they will nonetheless be motivated (and educated) enough by this to take another crack at CARB themselves.

    4. Even the occasional Bavarian has a sense of humor. Or so I have heard :-)

  5. Bavarians eat Weisswurst, the Austrians across the border eat Schnitzel. Both wear Lederhosen.
    Aside of that, great post and great discussion. Thanks for your effort!

  6. I think there is a benefit to having such strict restrictions, like what CARB has done for the BEVx category. Sure it hobbles the current implementation of i3 in the US, but it also discourages manufactures from designing an EV with training wheels, but instead go more boldly with the electric drive train.

    1. I think I see your point, but CARB is throwing out the baby with the bathwater. A better way to reach the same goal is to require ALL plug-in cars to have a DCFC port. Those without (specifically the PiP, but also unfortunately my Volt) would not be elegible for all the goodies attached to PHEV's and BEV's such as HOV stickers and rebates. If an OEM is going to supply the DCFC port, they will probably also supply a reasonable battery, not the miniscule one found in the PiP.

  7. I believe that you have seriously discounted the Volt as just a PHEV. It is everything the i3 should be with a smaller battery. It has Hold and Mountain Mode. It can run on electricity only, although for a shorter range. It is a standard size car, not a compact. It has a real gas tank, not a tiny fuel bladder. With the Volt, you wouldn't have to recode the car, nor would you have to stop for gas. Overall it is a better car, and it sells for quite a bit less. The only advantage of the i3 is its bigger battery, and that only counts if you daily commute is more than 40 miles RT.

  8. I think I see your point, but CARB is throwing out the baby with the bathwater. A better way to reach the same goal is to require ALL plug-in cars to have a DCFC port. Those without (specifically the PiP, but also unfortunately my Volt) would not be elegible for all the goodies attached to PHEV's and BEV's such as HOV stickers and rebates. If an OEM is going to supply the DCFC port, they will probably also supply a reasonable battery, not the miniscule one found in the PiP.