AZ Leaf Capacity Kerfuffle:
Much Ado About Nothing?

September 21, 2012

Mark D Larsen

[Note: I have since edited the following analysis with
updated calculations as specified below in Addendum 3.]

* Extrapolated from Nissan Technical Bulletin NTB11-076a:
LEAF Range Estimates at 4 Miles-per-kWh

For the last several months, some early adopters of the Nissan Leaf in Phoenix have been complaining about losing capacity bars prematurely, supposedly because of the extremely hot temperatures in Arizona. What originally started out as a worrisome concern has now snowballed (fireballed?) into well over 400 pages of discussion in the “My Nissan Leaf” forum.

Because their local dealers had told them that the bar loss was normal, a few of those Leaf owners decided to contact the news media, and the Arizona CBS affiliate, KPHO, aired a segment with their complaints on July 18, 2012. Although toward the beginning of the broadcast two owners praised the Leaf, they then lamented that their excitement had “shriveled up” with “disappointment” and “frustration.” One claimed that, when his Leaf was new, he could drive a 90-mile commute on a single charge, but now he was only getting 44 miles —less that half the original range. The other owner predicted that “Soon I'll be left with a very expensive paper weight in the garage.”

Two days later, KPHO ran another segment, in which they showed numerous owners gathering together to commiserate about the problem. In this broadcast, the reporter said that owners had told her they were “losing nearly 30% of their driving range after only one year,” and one individual stated that Nissan failed to mention at the time of purchase that they “were going to lose capacity at 3 times the rate of anybody else.”

It will undoubtedly irk those owners to say this, but I had to groan upon viewing these broadcasts. Such news segments can turn their concern into another “runaway Prius” news blitz in front of the public eye. I totally agree that Nissan should have addressed these concerns much earlier. However, according to these broadcasts the company had, in fact, already arranged to analyze several of the affected Leafs. So why air these stories? Whatever the motives, you can bet that any potential car buyers who viewed these segments have now crossed the Leaf off their list of vehicles to consider.

A few days after the second broadcast, Carla Bailo, Senior Vice President of Research and Development, released a letter to reiterate that engineers at the Arizona Testing Center would analyze some of the affected Leafs. I was glad to see that someone from Nissan finally issued a public statement, even if the tone and substance conveyed more anticipatory disclaimers than heartfelt concern.

Nissan eventually completed its investigation, but —most lamentably, and true to form— a company spokesperson has yet to release the actual data to the public. Nonetheless, Executive Vice President Andy Palmer hinted at the results in a statement to the Australian automotive website Drive on September 4, when he claimed that the hot weather in Arizona is causing a “faulty battery level display” rather than an abnormally accelerated drop in capacity.

Rather than feeling reassured by his statement, the owners in question reacted with skepticism, in essence asserting that Palmer was not telling the truth. Consequently, several of them decided to conduct their own experiment with several of the affected Leafs to prove to Nissan, once and for all, that their complaints were legitimate. Indeed, they openly stated in Green Car Reports on September 12 that “Nissan appears to be in complete denial at the highest levels,” and even that “it appears on the surface that an outright fraud may have been committed.” Whoa! Now those are fighting words for Nissan.

Last weekend, with a truly admirable and collossal effort, they carried out their tests. They then posted the results in the “My Nissan Leaf” forum on Tuesday. Do the data gathered prove that the battery packs in those Leafs have indeed lost more capacity than normal? Or do they show, conversely, that the problem really does have more to do with faulty gauge readings than actual capacity?

On the one hand, the testers admit that “Andy Palmer was right... they have poor instruments.” Yet they double-down with their accusations by also stating that “he was wrong about the batteries. It was sheer stupidity to tell this group of owners that the batteries are ok” [my emphasis]. I do not understand how they draw this conclusion from the data.

For example, as a benchmark they used Nissan’s own range estimates according to its Technical Bulletin NTB11-076a. That bulletin states that, at 4 miles-per-kWh, in 70° temperatures, a fully charged, brand new Leaf should achieve between 76-to-84 miles of range. Ergo, the testers attempted to duplicate these same parameters in their experiment: 4 miles-per-kWh speeds in Leafs charged to 100%.

However, to interpret the results, they then compared them to the very highest range of Nissan’s estimate (84 miles). I deem this a mistake. Nissan gives a range of the range for a reason, i.e., your-mileage-may-vary. Such variations depend upon several factors that cannot be strictly controlled in any such experiment, such as driving technique when accelerating to the target speed, tire pressure, road surfaces, traffic conditions, etc. Given the possible differences, I purport that it would be better to use the lowest range of Nissan’s estimate (76 miles), because this is the minimum number for which the company can be held officially accountable according to its bulletin. Nevertheless, for the sake of argument, I’ll use the average 80 miles in the middle of Nissan’s scale as a benchmark, but will also apply the full scale to each Leaf tested to underscore that there is a spread behind the projected range.

There are 12 capacity bars in the gauge, and if we divide 100% by that number, the result is 8.34% per bar. However, the owners informed Green Car Reports that “11 capacity bars showing instead of the full 12 capacity bars” indicates “approximately 15 percent loss in battery capacity” —nearly twice my estimate. They then claim that the remaining bars each represent 6.25%. Oddly, my calculator tells me that (11 bars x 6.25%) + (1 bar x 15%) = 83.75% —thus falling significantly short of 100% for all 12 bars. Obviously something isn’t right with those percentages, but to avoid further argument, I will use them regardless. By the way, to bolster the claim that there is a capacity problem, they also stated in the article that “One woman, who just bought the car a month ago, couldn’t come close to 100% capacity. When fully charged, her car was in reality only holding 91 percent of its original ‘new’ capacity.”

My biggest concern is that their report does not effectively factor in the mileage of the vehicles tested. Nissan has made it clear from the beginning that battery capacity will diminish with time and miles driven. They have repeatedly informed owners that, if they drive an average of 12,500 miles per year under normal conditions, the capacity will drop to around 80% after 5 years, and about 70% after 10 years. It is therefore possible to extrapolate from the declared benchmarks of 100% (0 miles), 80% (62,500 miles), and 70% (125,000 miles) the corresponding drop in capacity from month-to-month over time. The following graph plots the polynomial curve that intersects those three benchmarks::

The long table down the right side shows the results when applying over time the above polynomial curve to the average 80 mile range in the middle of the scale projected by Nissan at 4 miles-per-kWh. The graph at the top of the page also illustrates that same diminishing capacity. We can now compare the actual range achieved in the test vehicles with their corresponding extrapolated percentages according to mileage from the long table on the right. The following table lists the results side-by-side, from the lowest mileage vehicle to the highest.

Here is a scatter plot that illustrates those results along the polynomial curve (in green) for the anticipated average range loss according to mileage. We can see (in blue) that 7 of the 12 Leafs achieved ranges within their projected scales (the grey variance bars). Of those, one is actually one mile higher than its spread, three are in the middle, and three toward the bottom. Of the remaining 5 Leafs, three (in purple) are in such close proximity to the low end that the differences are practically negligible: 0.8, 2.6, and 2.2 miles short. However, two of those vehicles (in red) fell notably below the predicted spread: Blue494 (4.8 miles short) and White272 (8.6 miles short). Oddly, this latter vehicle had the same mileage on its odometer as White 626 which fell well within its projected scale. In point of fact, assuming that the woman’s car purchased a month ago is Blue842 (the lowest mileage, with 2,500 miles), its range was also well within its corresponding scale.

In general, the results for these Leafs seem to show that, whatever capacity losses they are experiencing, the majority of them fall within “normal” parameters according to their mileage. As for the two glaring exceptions to the rule, I would hypothesize that there must be other factors involved. For example, as I have indicated in the last column of the table, it would be informative to know those vehicles’ usual driving speeds, D vs. ECO mode, temperatures when charging, the number of charges per day, and the use of climate control. Other crucial factors could be 100% vs. 80% charging and/or the number of quick charges in those Leafs, since Nissan explicitly warns owners that such factors can have a detrimental effect upon the battery pack’s longevity over time. The company must have such tallies in the CarWings database, and I would be most interested in seeing them.

Now let’s assess how accurate the capacity bars are in those Leafs. Here is a comparison between what the remaining bars might lead one to believe and what range they actually achieved:

The two newer Leafs (Black782 and Blue842) still had all 12 capacity bars visible, so there was no need to include them in the above table. We can see that all of the other Leafs with missing capacity bars drove farther than the gauges indicated would be possible. Evidently, the Arizona heat is affecting what data the capacity gauges are gleaning from the pack, but not necessarily the capacities themselves according to miles driven. It would appear that Andy Palmer might be right: “The problem is a faulty battery level display. We don’t have a battery problem.”

What I cannot see in these data is evidence that a Leaf is only achieving half its original range after one year, nor that another Leaf has lost 30% of its capacity in the same amount of time, nor that yet another is losing its capacity at 3 times the rate as anybody else, nor that a relatively new Leaf has lost 9% of its range in only one month. Have those owners consequently modified their claims, thanks to these test results? Au contraire. KPHO aired yet another segment in which one of the owners assert that “We have records online of over 80 cars now complaining of significant capacity loss and they all seem to be focused in the hotter climates” [my emphasis]. You’ll have to forgive me, but I wonder if these owners have examined the test data, for the results do not appear to substantiate such claims.

In my humble opinion, I think apologies to Nissan are in order for the accusations of being "in complete denial," committing "outright fraud," and “sheer stupidity.” And I would hope those concessions could also be given air time to the general public, because the news coverage to date has likely sullied Nissan’s reputation and contributed to the recent slow down in Leaf sales. Consider, for example, that Time Magazine recently published an article questioning if the Leaf is a flop, despite the fact that it won the 2011 World Car of the Year award. When I read stories like that, I can well imagine the petrolpushers and their sockpuppet naysayers rubbing their hands with glee, because public witch hunts like these not only harm the particular vehicle in question, but also the EV movement in general.

Of course, to be fair, I also have to decry that Nissan continues to withhold its own results rather than openly communicating with both those owners and the public in general. They would only issue a prepared statement to KPHO for yesterday’s story, which reads:

Nissan has been working hard to understand some LEAF customers’ concerns in the desert southwest. We've tested a number of individual vehicles and will be contacting those owners to discuss their individual results in the near term. We also anticipate having more information to release to the wider Arizona customer base soon. We are taking Phoenix customer concerns seriously and are working hard to ensure their full satisfaction.

As far as I am concerned, that “near term” couldn’t get here soon enough, and should have occurred many months ago to nip this fiasco in the bud before it morphed into such a public relations nightmare. At this point I perceive that they might have to convey some “tough love,” since the owners’ own test data appears to undermine their complaints.

Addendum 1
September 25, 2012

I was pleasantly surprised to discover that, only one day after posting the above analysis, Nissan finally provided their promised feedback: the “near term” came sooner than I thought it would! The Director for Product Planning and Advanced Technology, Mark Perry, gave an exclusive interview to John Voelker of Green Car Reports. Among the many issues discussed, the salient points of Perry’s message were that:

Later that afternoon, “My Nissan Leaf” forum member Michael Walsh posted a follow up letter from Carla Bailo. She elaborated further upon Perry’s statements, providing more details about the analyses Nissan conducted on the seven Leafs:

As you can imagine, these conclusions are pretty much what I anticipated, given my own reading of the owners’ test data above. What is the most encouraging to me, however, is that Bailo expressed the commitment that “we stand by our customers.” In point of fact, recognizing that Nissan has fallen short when it comes to open communication with Leaf customers, Bailo announced that they have asked none other than Chelsea Sexton to convene an independent advisory board. Their hope is that said board will “help us to be more open and approachable in our communication and to advise us on our strategy.”

I am encouraged by this move. Chelsea Sexton is not only the EV Poster Girl, but one of the most experienced advocates in the movement. As she herself has often described her, Chelsea strives to serve as the “grout” between the various EV “tiles”: manufacturers, consumers, government officials, utilities, journalists, advocacy groups, venture capitalists, retailers, advertisers. I sincerely hope that she can help smooth the waters and facilitate frank, constructive dialogue between Nissan and its consumers from this point forward. I wish her the very best in the effort!

Addendum 2
September 27, 2012

This morning I discovered that KPHO had aired a fourth broadcast on this issue. I watched the segment with the hope that they would report Nissan’s findings, and that the company was now offering to resolve the complaints via individual, face-to-face meetings with the owners of the seven vehicles analyzed. Most unfortunately, this was not the case. I had to groan even louder. Here are some key quotes:

Reporter Heather Moore: “The Leaf owners were Nissan’s guinea pigs, if you will.”

It does not bode well that, to set the tone for this segment, the newsperson accuses the company of experimenting on their own customers like laboratory animals.

Reporter Heather Moore: “Now Nissan has acknowledged at least two of the cars are ‘lemons.’”

Surely those owners know that this is patently false. Nissan has offered to buy the cars back simply to please them. Just because the terms of the buyback are in accordance to Arizona's “Lemon Law” does not mean the company concedes that the cars are “lemons.” Indeed, its announced position is that “The Nissan LEAFs inspected in Arizona are operating to specification and their battery capacity loss over time is consistent with their usage and operating environment.” Using the label “lemon” only further sullies the vehicle’s reputation in the public eye.

Owner Mason Convey: “We’re not sure if this is the beginning of their admission that there is an issue, or if this is just them trying to quiet a few isolated individuals.”

In other words, rather than clarify that Nissan’s analyses showed that “The cars and the battery packs are behaving as we expected,” the Conveys still maintain that there is a capacity problem and postulate that the company might be trying to just get them to shut them up about it.

Reporter Heather Moore: “Scott Yarosh’s Leaf was the worst. After only 15 months, he couldn’t even make his 45 mile one-way commute to work.”

This reveals that Yarosh has a daily commute of 90 miles round trip. I therefore have to wonder why would he even consider a Leaf in the first place, knowing that the EPA’s official range estimate for the vehicle is 73 miles per full charge. Indeed, according to Nissan’s 6 range scenarios not even a brand new Leaf would go that far unless he were driving the equivalent of “Scenario 4: LA4 test cycle,” i.e., an average speed of 19.59 mph, with temperatures between 68° and 86°, and climate control off.

I seriously doubt such was the case in Phoenix. The closest match would likely be “Scenario 2: Cross-Town Commute on a Hot Day.” If so, according to the polynomial curve, after 29,000 miles a 100% charge would only give a “normal” range of 61 miles; an 80% charge merely 49 miles. I think it logical to therefore assume that Yarosh was charging twice a day, at both home and work, and perhaps to 100% both times. He just could not have made it to work and back on a single charge under those conditions, even in a brand new Leaf (68 miles with a 100% charge; 54.4 miles with an 80% charge). Impossible.

It is also reasonable to postulate that charging at work might have been outside, in the blazing AZ sun, parked over baking asphalt. Obviously climate control was a must. Finally, with a commute that long, my guess is that it involved driving the interstate, at freeway speeds. If such guesses are right, these are all “usage” factors that, as the disclaimer we all had to sign explicitly states, “may hasten the rate of capacity loss.” I can't help but wonder what Yarosh’s CarWings data shows about such factors.

I can only conclude that, if such were Yarosh’s driving and usage patterns, it just doesn't surprise me that his Leaf experienced a battery capacity loss more accelerated than Nissan’s parameters would predict. But that is not the case for the majority of the Leafs tested —and probably not the case for most AZ Leafs. There are always a few exceptions to the general rule.

Owner Scott Yarosh: “I think they're trying to get me to shut up, to be honest: keep my mouth shut.”

Well, that sure worked!

Owner Scott Yarosh: “I couldn’t believe that Nissan was trying to pin it on me.”

So the miles, speed, charging, and conditions of Yarosh’s 90-mile daily commute were not a factor?

Reporter Heather Moore: “Most believe our high heat is baking the batteries far beyond what Nissan expected, and the company refuses to admit it.”

Owner Scott Yarosh: “I believe temperature is the number one key factor in this whole range loss.”

Again, no matter what those owners believe, the test data above do not substantiate that claim. There is no evidence that even the few exceptions to the general rule are solely due to the “high heat.” As Yarosh’s daily commute shows above, his own “unique usage cycle” likely also played a key role in accelerating the capacity loss in his Leaf.

Owner Mason Convey: “There’s a lot of bitterness left over from the way they treated us.”

Yes, this is evident, given the statements these customers just made over the airwaves —despite the way Nissan is now treating them by buying back their cars.

Reporter Heather Moore: “An inside source with the company tells me that there is talk about the company yanking the car from dealerships here.”

Since the “inside source” remains anonymous, I am skeptical of this claim. However, if indeed true, given what these segments have broadcast to the viewing public, I wouldn't blame Nissan in the least.

Reporter Heather Moore: “The company agreed to speak with me early this week, we agreed, and we have never heard back” [My emphasis].

If true, somebody sure dropped the ball at Nissan! For a news report to be “fair and balanced,” it needs to give equal time to the other side of the story. I wish the automaker would have taken advantage of the opportunity to publicly present the findings of its analyses and express its commitment to nonetheless “work with individual owners to ensure their satisfaction.”

I am deeply disheartened to see this news broadcast. I had purposely refrained from mentioning a relevant issue is my original post above, but at this point... I might as well include it. When I took delivery of my Leaf, I had to read and sign a four page disclaimer which includes this specific paragraph [my boldface]:

  • Gradual loss of battery capacity. Like all lithium ion batteries, the 2012 LEAF battery will experience a reduction in the amount of electricity or charge it can hold over time, resulting in a reduction in the vehicle’s range. This is normal and expected. The rate of reduction cannot be assured, however, the battery is expected to maintain approximately 80% of its initial capacity after 5 years of normal operation and recommended care, but this is not guaranteed. This number may be higher or lower depending upon usage and care. Factors that will affect and may hasten the rate of capacity loss include, but are not limited to: exposure to very high ambient temperatures for extended periods of time, driving habits, vehicle usage, and charging habits (Quick Charging the vehicle more than once per day).

Did the disclaimer for the 2011 model not include this stipulation? If it did, and those owners signed their acceptance of its terms, I cannot see how they have a leg to stand on with their complaints —even if they were legitimate. The warning seems absolutely clear to me. When it comes to capacity degradation, Nissan covered its bases, and in my opinion would be on firm legal ground if the owners decided to carry out their threat to file a lawsuit “in court.”

However, rather than assume a confrontational stance, even though the evidence to date appears to support the company’s position, Nissan is instead offering an olive branch which even includes buying back individuals’ cars. Yet this news story is how those owners bury the hatchet and express their appreciation...? The proverbial hand that feeds them now has a couple of fingers missing.

Addendum 3
October 12, 2012

I see that yet another news segment has aired the AZ owners’ complaints, this time broadcast by KPNX. Most unfortunately, the report basically repeats the previous stories, but the reporter did announce the discouraging news that in California a group has in fact filed a class action lawsuit against Nissan over the capacity kerfuffle. I need not likewise repeat my previous commentary: readers can view the story for themselves and form their own opinions.

On a more encouraging note, Nissan itself recently posted a video of an interview with Chelsea Sexton and Andy Palmer. The first topic discussed was the AZ capacity issue, and Andy reiterated that the capacity loss in those owners’ vehicles falls within normal specifications according to usage and environment. Nonetheless, to stand by its customers, the company is meeting with those owners individually to assure their satisfaction. He also clarified the specifications themselves by repeating that Nissan projects and average capacity of 80% after 5 years, and 70% after 10 years, with proper use, care, driving, and charging habits. Those benchmarks indicate that the capacity loss is faster at first, then slows down over time, and thus show a curve when plotted on a graph.

When I first posted the above analysis, I had used a simple linear calculation. A few AZ Leaf owners rejected, dismissed, and ignored my findings, claiming that they were invalid because of the linear extrapolation. Ironically, I tried to point out that plotting Nissan’s curve instead would actually bolster the automaker’s conclusions and undermine their complaints —but to little avail. I have consequently recalculated the capacity loss using the polynomial equation generated by Nissan’s benchmarks to plot the true curve over time. I have updated the tables and graphs above accordingly, as well as my Capacity Loss Tool and its corresponding PDF chart.

Another objection to my original analysis was that I used the minimum figure from Nissan’s range estimates at 4-miles-per-kWh: between 76-to-84 miles. My reason for doing so was to not allow Nissan any “wiggle” room to contest the results. Nonetheless, because my rationale fell on deaf ears, I have decided to just use the average figure in that range (80 miles), but then purposely include error bars and the spread for each Leaf tested to illustrate the full scale of Nissan’s range predictions, top to bottom.

Finally, since those critics insist that the capacity bars do not represent an equitable distribution, I have instead entered the percentages they prefer: 15% for the first bar, 6.25% for every bar thereafter —even though the sum total of their percentages does not equal a full 100%. Go figure!