Wednesday, October 28, 2015

Love letter to Nissan Leaf DCFC users

This love letter is for those who care how Leaf is charging using DCFC. I hope to save money for those who pay to DCFC, and have non-paying Leaf drivers be aware how expensive it can be and hopefully become better members of the EV community. This is as romantic as I get, right to the point!

Previously, I've harped on Leaf's "No charge to have other people wait while Leaf is slow charging" program as well Leaf's slow DCFC speed. I call waiting for Leaf DCFC when L2 would be cheaper as getting "Leafed" and waiting for Leaf DCFC slower than L2 speed as getting "Leafracked". I also call Leaf drivers pulling into and using dual head DCFC when perfectly working Chademo is sitting empty as "Leafrackers".

There's not much I can do about Leafrackers. They are truly mouse frackers, and I don't have much love for them. Don't be a Leafracker!

I don't have much love for getting Leafracked, either. If they're willing to take up a DCFC spot and have others wait while charging slower than L2, they are frackers. Especially bad are those who plug in after their first DCFC when they already have 90% SoC. But there could be some that don't really understand that they're charging so slowly.

While I don't like getting Leafed, those who let others get Leafed may not have much choice. If Leaf battery has deteriorated so much that they cannot get much further than 50 or 60 miles, they may need all the help they can get. Leaf at 60% may only charge at 36kW (instead of 45kW). Even if they pay for charging, they may have to stick around and pay at higher rate until there's enough charge, though switching to L2 at 10kW (80%) instead of 6.6kW (86%) would be nice.

Below are two tables of MPG equivalent to gas car if Leaf has to pay to charge using eVgo OTG plan. It's similar to my MPGe table in that rows are $/gal of gas at local gas station. It uses Leaf's 114 MPGe EPA (3.38 mi/kWh EPA). Reds represent worse than 50 MPG, though many are masked by light yellow row markers.

First table is useful for Nissan Chademo only charger. It is more informative charger, because it shows the current in amps. The voltage is roughly constant at 400 volts, so the second column header (second row) is corresponding power in kW (multiply current by 400 divide by 1000), just for FYI. To find equivalent MPG to gas car, simply look up the closest gas price and current.

For example, if it's 25 amps (about 80%), local gas prices are $2.60/gal (Oct. 2015), you'd be paying equivalent to 14.6 MPG gas car.

Second table is useful for dual head ABB charger. It's not as good, because it does not show power nor current. As such, we have to "eye-ball" and infer the data we need. One can deduce power from energy (kWh) and elapsed time. 0.01kWh in 1 second would correspond to 36kW, 2 second would be 18kW, 3 seconds 12kW, 4 seconds 9kW, and so on. In general, if it takes more than 2 seconds for 0.01kWh (18kW), it's better to disconnect and use L2 instead. That's only with regard to time if others are waiting. If money is your objective, it's better to move to L2 when it takes more than 1 second per 0.01kWh (about 60%)

One can deduce more accurate kW from ABB by doing time averaging. For example, one can count how many seconds it took to get 0.1 kWh, and divide  the time by 10 to get more accurate time for 0.01kWh. Because Leaf slows down so quickly, much longer averaging many not be beneficial, although if you're trying to charge at 95% (hopefully, no one's waiting for you), you may have to wait for about 4 minutes to get 0.1 kWh.

For example, if it's taking 3.5 seconds for 0.01kWh and local gas prices are $2.60/gal (Oct. 2015), you'd be paying equivalent to gas car between 13.2 MPG and 17.6 MPG.

Edit Oct 29, 2015

A commenter by the name of Todd was gracious enough to make measurements of Leaf DCFC, and made a plot of Power vs elapsed time. Following is a quote of his comment.

"The battery was at 24% at start and ended at 88% in 30 minutes. The temperature here in Mira Mesa was 70 degrees @ 10PM. The battery started at 70 degrees and ended at 87 degrees. The rate was 41Kw, increasing to 44Kw until around 58% charge and then it was a somewhat linear drop to about 9Kw over the last 21 minutes. That works out to 34% in the first 9 minutes and another 30% in the last 21 minutes. The last image is showing the voltage of each cell and SOH%."

From what little research I did, GIDS (named after a guy who made Leaf tool?) is actual battery capacity that is used whereas SOC (state of charge) is absolute battery capacity. Like all EV, the battery does not discharge to 0 and does not charge to 100%. When 0% is indicated by the car, there's still lots of energy in the battery. GIDS is what's indicated by the car. As such, SOC is pretty meaningless with regard to driving; only GIDS will be discussed. In fact, when I use SOC in my blog, what I really mean is GIDS.

The plot was made at ambient temperature of 70F, so it's hard to know how it'll translate at different temperatures. It may not be absolutely accurate for all conditions, but it's a guide to make some rule-of-thumb observations to help Leaf drivers.

Left axis is power in kW for Green plot. Right axis is % for GIDS in magenta, battery state of charge in % in red, not sure what black is; maybe temperature?

1. If you're paying eVgo OTG plan to charge, and you have 6.6kW L2, it's best to switch to L2 at 40kW (100A current) to minimize cost. That occurs at about 60%. If you have 3.3kW L2 (2012 and earlier), it's best to switch to L2 at 20kW (50A current), which occurs at about 75%.

2. If you want to optimize for time, the best place to disconnect is when it starts to dip below 40kW, which is about 60%.

3. If you want to optimize for time, and you need more than 60%, it's up to the individual when to stop, but it seems the "knee" occurs around 70%.

4. If you need more than 70%, time taken will be less than optimal, but it will be quicker than L2 up to about 88 85%, taking 15 minutes to gain 15% (averaging about 1% per minute, VERY SLOW!). This would be the maximum charge Leaf should get from DCFC. It's easy to remember: back to the future! Get it? Time optimized, back to the future 88? Yeah, go see the movie! (based on Tom Saxton's data, 88% might not be best; stop at 85%. See below)

5. Based on his charging from 24% to 88% in 30 minutes, you should prorate your charging accordingly. Best would be to plug in at 10% or less to get to 70% in 30 minutes (roughly). But if you're already at 50%, you should only take about 10 minutes to 70%, not full 30 minutes. No-charge-to-charge or not, waiting around to charge wastes time, and I'm sure your life is worth more than $1/hr.

Edit Oct. 30, 2015

As I was researching "GID", I came across another blog (by Tom Saxton?) that deal with Leaf DCFC.

As a test, he charged from 21% to 80%, then plugged in again to get 95%. Hopefully, others weren't waiting while he was doing this! I think he got clever with axis labels. He states he charged to 80%, yet the light blue is "pack kWh" in legend. I suspect he cleverly chose the axis so that %GID and pack kWh line up. As such, pack kWh plot will be used as %GID.

For 21% to 80%, he found it to take 13.2kWh in 26:40 min (0.44 hours). That's 30kW on average, which is roughly similar to eye-balling from Todd's plot. That's also 44MPGe$ from first table above when gas prices are $2.60/gal.

For 80% to 95%, he found it to take 3.2kWh in 36:35 min (0.61 hours). That's 5.25kW on average, far less than Leaf's L2. That's about 7.5MPGe$ from second table 5.1kW column when gas prices are $2.60/gal.

Far more troubling is what happens at 88%. Basically, it stops charging for 8 minutes while you're paying by time; $0.80 gone to dollar heaven! In addition, first 3 minutes of charging adds more energy than 33:35 min after. This is probably what I saw some Leaf at >90% charging less than 2kW with DCFC. Talk about wasting time!

Based on this data, maximum recommended DCFC level would be 85% (3% margin from 88%). Otherwise, you risk being in charging limbo at 88% for 8 minutes. But you should still disconnect at 60% (40kW) or 70% (~20kW "knee" region of slowing) if you can live with it. After all, why waste time and money needlessly?


  1. Hmm, lots of information and potential misinformation here. While I can't agree more than No Charge to Charge is a terrible idea for the rest of the EV community, I don't agree with a lot of the conclusions you've drawn based on such a tiny sample size.
    I primarily use the NRG evGo for my quick charging lately... and I PAY for it. I pay a lot for it out of convenience and necessity to enable longer trips. Looking at it in terms of cost/kwh for that single charge doesn't tell the whole story, because I charge at both endpoints for free and only the midpoint has a cost. So the overall trip is still cheaper than $2/gal gas, and without the speed/convenience of the L3, I would have taken the ICE. Basically, no one should ever tie up an L3 unless necessary.
    Secondly, I while I don't have nice plots over time for you, I do have several datapoints that conflict with what Tom gave you. Maybe it's his particular leaf or charge, or just a bad day.
    I have charged from 17-87% in 25 min, roughly 30kwh average, peak that I noticed was 44kw, and was 12kw at the end.
    I have charged a few other times where I noted around 10kw at 90% and 3.1 kw at 98%. Unfortunately I can't fill in the gaps in between for you but what I can say is that L2 is never faster than L3 because it also ramps down near the very end. I have never noticed a weird 8 minute gap either. Barring any issues with the specific charger they were converge to an equal speed and moving out of someone's way would still make sense, but wouldn't be faster.

    1. Since I don't have Leaf, I have to rely on plots from other Leaf owners. Based on about a dozen Leaf I've spot checked, the plot by Todd seems about right. I did not pay attention to "pause" as I wasn't even expecting it. This is only meant as a rule of thumb guide. If you have better data for your Leaf, by all means, that should be used.

      L2 won't be faster than DCFC. However, when DCFC is slower than L2 charger rate (eg. 6.6kW), switching to L2 probably won't be any slower than DCFC. But it doesn't really matter; I don't recommend L2 for time savings at all. Quickest would be DCFC to 60% and disconnect, etc. as I describe in points 2, 3, 4. Even making additional DCFC stop could be faster than L2. For time savings, only stick to DCFC up certain %GID, and unplug to avoid limbo spots, if any.

      Only times to use L2 would be to save money, or in those rare times you need to top off to get to next DCFC. I mean, why pay 6X cost of L2 when you don't have to? I describe maximum money savings in point 1.

      As for tables to show MPGe$, obviously it's only applicable to what you pay at DCFC. Taking into consideration what may happen outside is impossible. But I've met few (I think 3?) Leaf drivers who pay to charge. When I pointed out how much they're paying with DCFC (about 15 MPG gas car at 80%), they were surprised. They thought they're always getting 114 MPGe, as in out of pocket cost equivalent to gas car that gets 114 MPG. One pointed out that the voltage wasn't changing much, so he thought it's charging at max rate. This post is meant to inform Leaf drivers and help them. After all knowledge is love (or is that power?)

      There's another post that deal with $/kWh + $/gal to find MPGe$. When one's charging at home or Blink, this would be helpful. Not everyone gets free charging, even L2, so I'm hoping this will help Leaf and other EV drivers understand what they're paying.

    2. Anon,

      Your numbers do appear to jive exactly with my experience regarding start and end kW rates. I find the 8 minute dwell time from toms data odd. I'm sure that was just a charger malfunction since that has never happened to me.
      I did do a charge from 80% to 98% on L2 the other day since I've never done that and never charged to 100% before. I found it interesting because the rate was 5.5kw consistently until around 95% and the. It had a steady decline to 2.2kw as it approach full charge. Forgot to save the graph though. Perhaps the next time I need full charge I'll capture it for this post.

    3. Todd, the tail end of your charging plot shows power to be less than 10kW, probably 8 kW (looks more than 7.5kW). But Anon states he's 12kw at 87%. I paid more attention to details at start and few minutues after than details at the very end. It could be that different Leaf behave differently after about 80%. Anon didn't mention the rate at 60% or 70%, but your data at those levels match well with various Leafs I've seen.

      I've also seen Leaf at 95% taking close to 20 seconds to add 0.01kWh, so 2kW rate at high %GID is also real. You tend to remember very painful memories, and that was especially painful waiting 1.5 hours at 10PM. But from some comments, it could be that was due to variant behavior of different Leafs at high %GID.

      In any case, I think the suggestions I make are reasonable; if you can live with 60%, only charge to there, etc.

      Another interesting bit of news is that I haven't been Leafed in past week or so. One car that I saw often at fast charger (always at Chademo only spot, thanks!) was now charging with L2. I don't know if it's coincidence or people read this to save money or my nagging. In any case, I hope things are improving. That'll be good for all of us.