Phoenix Range Test Results, September 15, 2012

[DaveEV]

Interesting though when faced with a battery that is 70%, 80% or 90% capacity knowing that at best, LBW occurs with about 4 kWh remaining in the pack and for the sake of the argument, range when new is 80 miles.

We all know that the average user doesn't like to go below LBW - which means leaving 4 kWh (out of 22.5 kWh assuming 281GID and 1GID=80Wh) on the table. We'll call 100% - LBW "usable".

100% capacity = 22.5 kWh - 4 kWh = 18.5 kWh, 66 mi before LBW.
90% capacity = 20.3 kWh - 4 kWh = 16.3 kWh, 58 mi before LBW, 12% reduction in usable range.
85% capacity = 19.1 kWh - 4 kWh = 15.1 kWh, 54 mi before LBW, 19% reduction in usable range.
80% capacity = 18.0 kWh - 4 kWh = 14.0 kWh, 50 mi before LBW, 25% reduction in usable range.
70% capacity = 15.8 kWh - 4 kWh = 11.8 kWh, 42 mi before LBW, 36% reduction in usable range.

So for most people (who generally try to avoid LBW and below), the higher the capacity loss - the worse the reduction in range feels 20% worse because of the fixed LBW setting. This could be even worse since it appears that the BMS appears to bury even more of the battery below LBW once you've lost a bar or more...

Psychologically - a 15% loss in capacity sure feels huge - going from 66 mi to LBW to 54 mi to LBW sure feels a lot worse than only being able to go 68 miles / before turtle compared to 80 miles before.

This feels even worse if one tries to charge to 80% mostly and they've lost a bar. Then they go from 52 mi before LBW to 42 mi before LBW - a 20% reduction in usable range.

[surfingslovak]

So... in the case of our car, the GOM now shows a number in the high 60s every morning after a 100% charge, and that's realistically what we can get without dragging it into LBW and VLBW, provided we're rather conservative with speed and air conditioner use. Nissan says this is where you look for a charger. So I don't care too much if there are 2,784 driving miles available below LBW. That is where Nissan is suggesting that the car be recharged. If I drag the pack down below this point (on a regular basis), I run the risk of having Nissan point the finger at me and blame me for capacity loss.

That's a good point, and the GOM apparently behaves the same way in Randy's Leaf as well. He realistically has about 50 miles before hitting the low battery warning, and the GOM was showing 47 yesterday.

But back to Volusiano's question. What other instrument or control is out of whack or behaving inconsistently? I could think of the obvious one: battery capacity bars. If I recall correctly, they were not showing the correct remaining capacity in a number of cars. Several of us have the 2011 shop manual and could make the relevant section available.

Then there is the less obvious one: energy economy. It should have been about 4 m/kWh for all of the cars, but some of them were not even close. Randy's car ended the test at 4.4 m/kWh. This defect might be hard to prove without proper instrumentation or a dyno, but it would also be nearly impossible to fix.

Is there anything else we could be missing?

[RegGuheert]

turtle voltages had a SIXTY volt spread. I'll suggest that batteries at 350 volts could have gone much farther, maybe as far as my car went with 290 volts remaining. This reeks of a BMS problem; hardware of software, or both.

There's only two things that might cause one car to turtle at 350V and another to turtle at 290V.

1. A software problem. No further explanation needed here.
2. At least one cell-pair with significantly lower capacity than the rest - one cell-pair hit the low-voltage limit and the BMS shut the party down. Should be easy to check with a Consult by taking this car down near turtle.

Agreed, except I think a software problem is a very long shot here. I would expect software problems to arise in more than a couple cars and in more than just one climate.

The BMS in the LEAF terminates both charging and discharging based on voltages at the cell and the pack level. Phil has told us that his LEAF's voltmeters appear to be accurate to within about 0.1% and that the low-voltage meters cross-check with the high-voltage meters and throw codes if there is a discrepancy. So, no, I do not think we are losing any range due to instrumentation issues.

The sixty-volt difference in discharge voltage is *precisely* what I would expect to see between a new pack and a severely degraded pack. The cells in the new pack should all have similar capacities, so none should go below the knee prematurely. But in a pack without a TMS driven in a very hot climate some cells will live in an even HOTTER environment than others. Those cells will deteriorate faster and will therefore have higher internal electrical resistance, possible greatly higher. As a result those cells will get even hotter and will quickly go downhill.

This condition is exactly what Nissan's CELL VOLTAGE LOSS INSPECTION test is designed to identify. If Scott's car is the one that terminated at 350V, then I find it irresponsible that Nissan has not reported the amount of variation found in the cells in his pack.

Again, I will say that I would find a histogram of cell capacities from a degraded LEAF to be quite informative right about now. I don't think we can expect to get that from Nissan. Anyone want to purchase a battery and give the old one to Phil to document?

[edatoakrun]

Below, in my opinion, is an accurate short summary of the test results, and significant conclusions, based on the partial release of test data:

After selecting LEAFs nationwide, whose drivers believed them to have some of the largest range losses, a recent range test in Phoenix showed less than the range loss as had been expected, by relying on what the test showed were inaccurate capacity bars displays, and "gid" counts.

One outlier on the low side got only 59.3 miles. The unknown conditions experienced by this LEAF during over 29,000 miles of use, make it impossible to determine if any factor or factors of use contributed to this car's relative under-performance.

While the other eleven of the twelve did all have close to the highway range (and many, even more) that Nissan had estimated for new LEAFs in its promotional materials, many, but not all, seemed to show a significant reduction from the higher "new" LEAF range, as estimated by another source, Nissan Technical Bulletin NTB11-076a.

There was large variability between the LEAFs individual ranges, of between 66.1 and 79.6 miles. Inadequate test protocols could only seem to explain a small part of the large range disparities, between all twelve cars.

We can now conclude, in all likelihood, that many or all LEAFs have a a significant flaw or flaws in their energy use reports, that make it difficult to determine with great precision what capacity or range loss has been experienced by any LEAF, either from new, or from an assumed standard range.

Alternate means of testing of the battery capacity, such as by measuring the charge accepted, might allow more accurate battery capacity results, from which standardized ranges at m/kWh use levels, could be calculated.

However, all data indicating accuracy or inaccuracy of all m/kWh reports from the test LEAFs, has so far been withheld, by the promoter of the range test.

Stoaty
Ed, we already know your opinion. Please give it a rest.

No, I don't think I will.

I wrote that summary above because of the deficiencies of Tony’s own effort. But I would appreciate your future comments, and those of any others, related to the accuracy of the summary that I posted above. My intent was to post a short and accurate summary of the most significant findings from the data reported from the range test. I’m sure it can be improved. I may have left out more important conclusions, made errors, in presentation or emphasis, and I could even have facts wrong. If so, I hope these errors will be pointed out, to me, so that they can be corrected.

Any substantive comment, of course requires you to review the data and much longer summary written by Tony, and posted on page one of this thread, and review that for accuracy. I think you will find it is significantly flawed, both in data presentation, and in the conclusions made. I wrote the alternate summary because of the significant deficiencies of Tony’s own effort, and my belief that all present and future LEAF drivers deserve an accurate presentation of the test results. Those who participated in the test, giving much effort , and some incurring significant expense and damage to their Leafs, even more so.

Much of the most important data (IMO) apparently was not collected during the test. Other data, which Tony has previously characterized as “bad”, is still being withheld by Tony, but is trickling out in the comments of others, or selectively released by Tony, as in the m/kWh results from one LEAF in his own comment, below.

I hope other test participants will also continue to post more of the “bad’ data, as well as continue to correct the factual data errors in Tony’s comments.

There is actually much more to write about many other interesting aspects of the test results, as some of the comments of others have shown.

I’m sure I will have a few more points to make myself, if no one else does first, (hopefully) after more complete data from the test becomes available.

Yes, we already know that an 84 mile car is missing from the data...

I'm not quite sure why there's fixation on a "84 miles" when the gauges in the car can't be relied upon. To assume that a new car should have gotten 84 miles on the range test would mean that you'd have to assume that the cars accurately reported 4.0 mi/kWh (which you've said we can't) and that 84 miles is typical for a new car (which Nissan has not claimed - they have only claimed a range of 76-84 miles is typical at an efficiency of 4.0 mi/kWh).

84 miles for a new car is probable. "Black782" would have banged that out 2-3 months ago, and did PLENTY of times on my BC2BC trip. I used Nissan's data of 84 miles because I didn't have a car in the test that did it. Like I've said several times now, if Nissan wants challenge that, I will be more than happy to find one or more new cars right here in SoCal to bang out 84 miles.

My car did show exactly 4.0 miles/kWh on the outbound leg, and 3.8 inbound (I reset at the halfway point, like I've outlined previously to determine wind effects). If you look at the wind chart for 0251Z, the wind was reported from the east at 6 knots. The inbound leg was east (except the last 6 miles that is north/south.

So, again (I've said this several times), my car's data hit the predictions very well. My average economy was 3.9, and this matched all the expectations before I ever turned the car on that morning. The only unexpected surprise was not having 100% available to start. But it did in the very recent past.

So, to Nissan 76-84 mile thing. Sure, they can argue all kinds of BS, but I went 76, so I guess that meets the normal range. But, sure, and also matched my economy by 0.1 miles/kWh. But, if you don't think a car will go 84 miles, then you are not believing Nissan's own data.

A new LEAF will do it, and Nissan publishes that it can do 84 at that 4 mile economy, which is the economy that my car did (within 0.1.. the center Nav screen would shown 4.0).

[KJD]

Interesting though when faced with a battery that is 70%, 80% or 90% capacity knowing that at best, LBW occurs with about 4 kWh remaining in the pack and for the sake of the argument, range when new is 80 miles.

We all know that the average user doesn't like to go below LBW - which means leaving 4 kWh (out of 22.5 kWh assuming 281GID and 1GID=80Wh) on the table. We'll call 100% - LBW "usable".

OK correct me if I am wrong here but I seem to remember reading somewhere else that 21 kWh was what was usable not 22.5

[edatoakrun]

turtle voltages had a SIXTY volt spread. I'll suggest that batteries at 350 volts could have gone much farther, maybe as far as my car went with 290 volts remaining. This reeks of a BMS problem; hardware of software, or both.

There's only two things that might cause one car to turtle at 350V and another to turtle at 290V.

1. A software problem. No further explanation needed here.
2. At least one cell-pair with significantly lower capacity than the rest - one cell-pair hit the low-voltage limit and the BMS shut the party down. Should be easy to check with a Consult by taking this car down near turtle.

Agreed, except I think a software problem is a very long shot here. I would expect software problems to arise in more than a couple cars and in more than just one climate.

The BMS in the LEAF terminates both charging and discharging based on voltages at the cell and the pack level. Phil has told us that his LEAF's voltmeters appear to be accurate to within about 0.1% and that the low-voltage meters cross-check with the high-voltage meters and throw codes if there is a discrepancy. So, no, I do not think we are losing any range due to instrumentation issues.

The sixty-volt difference in discharge voltage is *precisely* what I would expect to see between a new pack and a severely degraded pack. The cells in the new pack should all have similar capacities, so none should go below the knee prematurely. But in a pack without a TMS driven in a very hot climate some cells will live in an even HOTTER environment than others. Those cells will deteriorate faster and will therefore have higher internal electrical resistance, possible greatly higher. As a result those cells will get even hotter and will quickly go downhill.

This condition is exactly what Nissan's CELL VOLTAGE LOSS INSPECTION test is designed to identify. If Scott's car is the one that terminated at 350V, then I find it irresponsible that Nissan has not reported the amount of variation found in the cells in his pack.

Again, I will say that I would find a histogram of cell capacities from a degraded LEAF to be quite informative right about now. I don't think we can expect to get that from Nissan. Anyone want to purchase a battery and give the old one to Phil to document?

Can you rule out the possibility of adaptive battery management by the BMS?

IMO, Nissan must have designed the LEAF BMS with battery life, rather than convenience to owners, as the primary goal.

How successful Nissan was, is now the larger question.

But can you conclusively rule out variable battery warning levels, and shut down, as evidence of a the BMS functioning as designed, rather than a malfunction?

Would it not make a lot of sense for Nissan to include measures intended to protect of the battery from premature capacity loss (or, much worse from Nissan's point of view, a "failure" warranty claim) based on the prior conditions of use, into the BMS?

As one example, the BMS may have noticed a Nissan driver believes they can make daily 60 mile freeway commute in Phoenix temperatures on one charge, (or a 80 mile one, with a quickie at the local DC) in their LEAF and now, the BMS is trying to tell the Driver, that he or she cannot, and still expect to maintain (some unknown level of) capacity for (some unknown) period of time.

[Stoaty]

Stoaty
Ed, we already know your opinion. Please give it a rest.

No, I don't think I will.

OK, added to ignore list. I hope you aren't channeling Orientexpress.

[TonyWilliams]

Stoaty
Ed, we already know your opinion. Please give it a rest.

No, I don't think I will.

I wrote that summary above because of the deficiencies of Tony’s own effort.

Ed, YOU'RE A WINNER !!!!!

My test is bad. Now, you can rest easy tonight. Night, night.



PS: OrientExpress might want to snuggle tonight, so you two should get together for a Tony bashing.

[DaveEV]

Interesting though when faced with a battery that is 70%, 80% or 90% capacity knowing that at best, LBW occurs with about 4 kWh remaining in the pack and for the sake of the argument, range when new is 80 miles.

We all know that the average user doesn't like to go below LBW - which means leaving 4 kWh (out of 22.5 kWh assuming 281GID and 1GID=80Wh) on the table. We'll call 100% - LBW "usable".

OK correct me if I am wrong here but I seem to remember reading somewhere else that 21 kWh was what was usable not 22.5

22.5 kWh assumes 1 GID = 80 Wh and 281 GID = max charge.

Realistically, even a new car will not always hit 281 GIDs on a full charge - ~276 appears to be common, for example - perhaps if your battery isn't perfectly balanced. And turtle comes on around 7 GIDs, so 276-7 = 21.5 kWh usable. Pretty close to 21.5 kWh "usable".

That post is really just showing how a fixed LBW set point makes capacity loss look worse than it is for most people since most people only look at range from 80-100% to LBW as "usable". The actual numbers don't matter all that much.

[TonyWilliams]

Interesting though when faced with a battery that is 70%, 80% or 90% capacity knowing that at best, LBW occurs with about 4 kWh remaining in the pack and for the sake of the argument, range when new is 80 miles.

We all know that the average user doesn't like to go below LBW - which means leaving 4 kWh (out of 22.5 kWh assuming 281GID and 1GID=80Wh) on the table. We'll call 100% - LBW "usable".

OK correct me if I am wrong here but I seem to remember reading somewhere else that 21 kWh was what was usable not 22.5

22.5 is the official stored capacity, as shown above (281*80). But, there are losses to extract that stored power, and my rule of thumb is 75*281 of extracted power, AS MEASURED BY THE DASH INSTRUMENTS. That means that if you drive 4 miles/kWh as the dash displays, the car will go 84 miles (84/4=21), so I call that "useable energy", as opposed to 22.5 "stored energy". Edit: yes, like drees said, we leave a tiny bit of power at the bottom, between turtle and the point that the battery shuts down.

The car absolutely will NOT go 22.5*4 = 90 miles. This comes up a lot. The battery's rated capacity is 24kWh, but it is never charged above 95% SOC, and never allowed to discharge below 2% SOC of the rated capacity of the battery. 22.5/24=93.75% of the rated battery capacity is stored for use.

You'll note that I used 4 miles/kWh for all the examples, and that was exactly the target economy we intended to have at 100kmh speed.