Using the jurassictest range estimate tool. How accurate...?

[edatoakrun]

http://www.jurassictest.ch/GR/" onclick="window.open(this.href);return false;

The tool above was posted recently by GRA here:

http://www.mynissanleaf.com/viewtopic.php?f=31&t=5425&start=50" onclick="window.open(this.href);return false;

Only a few, mixed reports of it's accuracy so far. Since it looks promising to me, I thought a dedicated thread for anyone using it,to discuss how you use it, and how well it works, might be worthwhile.

I've just been looking back at my past driving by CW records and comparing to this tool, and there is a close correlation in the results.

But, it is impossible for me to exactly reproduce my past CW driving records with this tool, for two reasons.

The first is that Google maps does not recognize the first few hundred yards, and 150 ft. descent on my driveway as a "road" so to more accurately test, I will have to stop and restart at some point “on the map”, to get this segment of my driveway "off the record" of the "trip” reported by CW, on the "electric rate simulation".

The tool also chooses it's own default rout, and doesn't seem to let me alter it. So it has me take a parallel rout to my most frequent destinations,

However, since the Ascent and descent totals, grades, and miles driven, are very similar, and am able to specify the same speed, the two routes should have similar kWh use results, from the two sources.

They do. In fact, all results seem to be within the uncertainty I'd expect, simply due to my inability on any drive, due to road and traffic conditions, to maintain a anything close to a constant speed, or estimate the average with great accuracy.

Not only does it seem to me,to be, at least fairly accurate, for the entire days driving consisting of the two-way trip, but the ascent and descent directions, which are recorded as separate trips, on CW on the "electric rate simulation" page, also seem to be reproduced by this tool, accurately, always within the percentage of variation, I’d expect from speed variation and uncertainty

It will be difficult for me to verify just how close the tool matches real-world results.

From both directions from my home, the tool’s embedded Google map directs me to highway 299, rather than the shorter routes with less speed variation I normally choose to drive.

So for me to use it, on all trips from my home, I will have to match my route to that the tool chooses, routes with large speed variations, putting a large uncertainty into the full test, due to the difficulty estimating an "average" speed. And since I will have to drive much of the trip at under 30 mph, and much at over 50 mph. So I will only be able to test a short segment of a larger trip, at near-constant speed, never over close to a 100% charge range.

Obviously, the predictions could be much more accurately tested by others, who could do as close to a steady-speed drive, over a long distance, with significant ascent/descent, and repeat it, to get a larger sample for real world average results.

Some other factors and uncertainties you may want to consider, and resolve.

The first, DO NOT use the LEAF estimate of average speed, without adjusting for your actual speed while driving. The car includes every stop light, and the speed you back out of your garage, as part of this average, though they have little effect on the driving efficiency of your tip, overall.

Small differences in your estimate of average speed have large result, both in real life, and in the range prediction the tool provides.

I don’t know exactly how accurate CW kWh consumption reports are. I use them, since they seem to consistently and accurately replicate range results for me, and no one with accurate SOC info, has reported either constant or variable errors in them. You can of course, check the tool against any other method you currently use, to estimate your range.

Of course, the best test, is to use the tool to plan a near 100% battery capacity future trip, and see what your results are.

My car does have a 2.5% constant under-reporting error in CW miles driven. Most others who have checked, seem to report the same thing, but you should definitely check CW against your odometer as well as mile as reported on the tool site. The “Help” page there also suggests it’s own inaccuracies for “tortuous” routes, so I’d suggest an independent check of trip miles, from another source, especially if your route has a lot of winding roads.

The tool has a default value of 21 KWh (which you can adjust) I suspect ABC would be closer to 20 kWh at 100% charge, at my current battery temperatures. No accurate account of available battery capacity, in kWh, adjusted for battery temp, is yet available. So you should never expect range predictions from this tool, any more accurate, than your estimate of ABC, before you begin your trip.

 

[TonyWilliams]

There was another web page like this, and I cannot find it. It was very detailed, with things like variables for rolling resistance, etc.

What I find with these two that I've reviewed are a lack of real world experience, which means they carefully factor something as minuscule as differences in rolling resistance, and miss HUGE variables like battery capacity with temperature.

 

[GRA]

There was another web page like this, and I cannot find it. It was very detailed, with things like variables for rolling resistance, etc.

What I find with these two that I've reviewed are a lack of real world experience, which means they carefully factor something as minuscule as differences in rolling resistance, and miss HUGE variables like battery capacity with temperature.

It would be nice if they automatically factored battery temp, but for now when playing with it I'd just enter a manual capacity correction based on your data. That's what I've been doing for calculating 80% range, too, changing '21' to '16.8'.

BTW, I just copied the link, someone else (I forget who) posted it first in a different thread (which one I don't remember).

 

[edatoakrun]

I did find a route from the tool that I can drive, to somewhere I’d like to go. I was going to try it today, but it was just too cold for the trip to be pleasant, without using the heater. I hope to try it later in the week, in warmer weather.

I’m planning a trip to Clikapudi trail, on the shore of Lake Shasta, 27 miles each way, by Google maps. The tool slightly underestimates the distance (as mentioned on the “help” page) at 26.2 miles, from the Google “road end” about a 1/4 mile from my house. I will record trip times and speeds, and keep as close to a constant 70 kph, which is what I entered in the tool, as possible.

The tool predicts 5.8 kWh use for the drive there (including a 221 meters net descent) and that 8.1 kWh will be used to return.

So, the round trip should take a little over 14 kWh, allowing for the .8 mile distance underestimate error.

Normally I would top-off to 100% for this trip, but if I charge to 100%, I lose a lot of regen on the 800 ft. descent, in the first 3 miles of this route, which the tool would not, I expect, account for.

10 bars would be cutting it uncomfortably close, and I expect I can still keep disk brake use minimal, starting from 11 bars SOC.

So, I’m planning on charging from 10 bars, the timer “80%”, about 30 minutes before I leave, to about 11 full bars, probably close to 18 kWh of Available battery capacity, with a forty-something degree f battery. The exact SOC or ABC shouldn't matter, for this less-than–maximum-range test, as both the tool predicts, and CW reports, in kWh use.

So, I figure the tool has predicted (in bar display) that I should be about 7 full bars at the halfway point, the Clikapudi trailhead. And may be as low as one bar, and may even get the LBW as I climb back up the final 800 ft., on the 2 mile grade up to where Oak Run rd. crests Bullskin ridge, about a mile from my home, which I should reach, with about 3 kWh ABC remaining, after the final, short, “off-Google map” climb up my driveway.

And Hopefully, only a few days, later, CW will report kWh used on the trip, for comparison.

 

[GRA]

I did find a route from the tool that I can drive, to somewhere I’d like to go. I was going to try it today, but it was just too cold for the trip to be pleasant, without using the heater. I hope to try it later in the week, in warmer weather.

<snip>

Thanks for testing this out. I look forward to your report.

 

[edatoakrun]

I did find a route from the tool that I can drive, to somewhere I’d like to go. I was going to try it today, but it was just too cold for the trip to be pleasant, without using the heater. I hope to try it later in the week, in warmer weather.

<snip>

Thanks for testing this out. I look forward to your report.

I just got home from the test route today, and the short story, is that the tool works very well, IMO.

I suggest others try it, especially those who have a longer-route, and more-constant-speed option than I do, to reduce the uncertainties I had.

I lost the 8th bar 1.6 miles before the halfway point, as predicted, and still had 7 when I restarted.

I did not lose the second bar till .2 miles before Google map’s “end of the road”, or get the LBW until I actually parked, in my carport, stopped and restarted, both about a mile (at 15 mph, with 100 ft. of net descent) after predicted.

But for all I know when I adjust for actual speed variations from plan, the tool might have done better, or worse, than those results indicate.

It was very close, in any case, and probably within the uncertainty that "average speed" creates, on any route where a constant speed is impossible.

My gut feeling is that my battery may have had slightly less than the 18 kWh ABC I was anticipating when I started ( 80% + 30 minutes of 16 A L2 at about 38F), and regen may have been better than the tool’s default setting of 0.4 “recup”, which I used for the estimate above. But don’t hold me to either of those statements, till I work the numbers over, and get a better picture.

I suppose I could go out and drive to VLB, but not much point, as I didn't have a either a 80% or 100% charge to begin with, this test will not give very good data on total ABC, anyway.

I have a whole page of one-hand-on-the-wheel-written-chicken-scratch-notes to decipher, and compile, and that may take me a while. I've used a written log a few times before, and IMO, having a passenger to take them down, makes the process a lot less stressful.

I tried to keep records of time, mileage, and temp, at each bar loss. And I stopped and restarted at 2 points each way, to break down the trip for average speed over different road sections. Hopefully, I remembered to “accept” each time, and CW picked it all up.

 

[GRA]

Thanks for the report. I'm currently playing with the tool to see where Qc stations need to be on 80 and 50 between Sac and Lake Tahoe.

 

[edatoakrun]

I may have spoken too soon, when I posted on Monday night:

"... just got home from the test route today, and the short story, is that the tool works very well, IMO."

Long story, but, actually, CW data indicates lower kWh use, than as predicted by the tool, when using the most likely settings.

But before I go on, and risk starting any further disagreements with the chart-watchers, I'd like to ask opinions, of what percentage of battery capacity, I should use for estimating remaining battery capacity, at the end of this drive, given the facts below:

"..I did not...get the LBW until I actually parked, in my carport, stopped and restarted..." (this was 0.4 miles after losing the second bar).

I've never before had the LBW on start-up, that had not also occurred, during the previous trip. Do I use it, as a indication of SOC, or not?

So, looking at the range chart, which, in your opinion, is the most accurate, of either of these points, LBW, or when I lost the second bar?

And where do you think I was, as % of battery capacity, at the most accurate of the two points?

 

[surfingslovak]

There was another web page like this, and I cannot find it. It was very detailed, with things like variables for rolling resistance, etc.

What I find with these two that I've reviewed are a lack of real world experience, which means they carefully factor something as minuscule as differences in rolling resistance, and miss HUGE variables like battery capacity with temperature.

Tony, you are probably thinking of a paper David Roper, a retired physics professor from Virginia, put together. On a related note, would you have time to connect sometime soon? Perhaps early next week, when you are done with local meetup.

http://bit.ly/leafrangecalculation" onclick="window.open(this.href);return false;

 

[edatoakrun]

...
Tony, you are probably thinking of a paper David Roper, a retired physics professor from Virginia, put together....

http://bit.ly/leafrangecalculation" onclick="window.open(this.href);return false;

Regeneration Fraction: G O.8

Interesting paper.

Thanks

 

[TonyWilliams]

There was another web page like this, and I cannot find it. It was very detailed, with things like variables for rolling resistance, etc.

What I find with these two that I've reviewed are a lack of real world experience, which means they carefully factor something as minuscule as differences in rolling resistance, and miss HUGE variables like battery capacity with temperature.

Tony, you are probably thinking of a paper David Roper, a retired physics professor from Virginia, put together. On a related note, would you have time to connect sometime soon? Perhaps early next week, when you are done with local meetup.

http://bit.ly/leafrangecalculation" onclick="window.open(this.href);return false;

No, not that one. It was primarily written with the Tesla roadster in mind, with links to every RV park in the country.

 

[TonyWilliams]

So, looking at the range chart, which, in your opinion, is the most accurate, of either of these points, LBW, or when I lost the second bar?

And where do you think I was, as % of battery capacity, at the most accurate of the two points?

Honestly, the fuel bars have their own limitations, and as such, I just think of them as somewhat "soft", but useable information.

Of all the crazy displayed data that Nissan has endowed upon us, the two separate battery warnings are consistently at fixed battery performance values. Any time a warning pops up, reading fuel bar data becomes history.

If the battery were 50% degraded, and you charged to 100% SOC, you'd show 12 fuel bars of a battery that has 140 Gids of capacity.

So, it is conceivable that we may see LBW pop up with 3 fuel bars. But that LBW will still have 48 Gids of energy; the same as when it was new. Use the LBW data.

 

[edatoakrun]

So, looking at the range chart, which, in your opinion, is the most accurate, of either of these points, LBW, or when I lost the second bar?

And where do you think I was, as % of battery capacity, at the most accurate of the two points?

Honestly, the fuel bars have their own limitations, and as such, I just think of them as somewhat "soft", but useable information.

Of all the crazy displayed data that Nissan has endowed upon us, the two separate battery warnings are consistently at fixed battery performance values...

Use the LBW data.

Well, I was afraid you were going to say that, Tony.

I'd ask you to again consider whether the stop/start process, could trigger a "premature" LBW, such as the way it causes bar disappearance, before those "fixed...values", or that Professor Roper, in the article posted above is correct, is his (apparent) conclusion that LBW actually varies with "recent driving style".

In fact, my LBWs and the few VLBWs have never followed the patterns of your chart closely, coming "early" in most all cases, perhaps not coincidentally, since they also almost always follow a period of high kW use, over 30 kW over several minutes, in the final ascent to my home.

My recent experiences, are particularly hard to reconcile with the range chart.

OK, back to the CW report for the trip mentioned above.

One big surprise.

As reported earlier, the tool predicted 13.9 kWh use for this trip.

CW reported 12 .1 kWh (+ 0.3 kWh for the 0.6 miles off the map", and 4.3 m/kWh for this entire trip).

This looks like a much bigger variation than can be easily explained by errors in my speed records, or by adjusting the tool, for variations, such as regen efficiency.

However, as noted above, the tool seemed very predictive of the bar display behavior, I posted above, which poses an unexpected question, which will have to come first.

It would appears, my understanding of the accuracy of CW kWh usage reports, and the what the bar display and LBW shows, according to the range chart, are inconsistent.

If the bar display and LBW give consistent results for kWh capacity, under all conditions, it would seem to be a waste of time, to check the tool for accuracy, as I did, using CW, or to use CW at all.

I know the explanation of my reasoning below is lengthy, but I hope you will take the time to read it and offer useful comments or criticisms.

First, to examine the CW report for evidence indicating level of accuracy. 4.3 m/kWh, for the entire trip, seems entirely consistent with speed and ascent/descent for this trip. Again, had I used the near 14 kWh I predicted before the tip, using my understanding of the bar display, I would have gotten only about 3.7 m/kWh, for this trip. Had I calculated that earlier, I would have noticed the tool’s m/kWh (which the tool displays, but as kWh/100 Km) number was way too low, as my “commute” is very similar to this drive in speed, distance, and ascent/descent, and I always get very close to 4.3 m/kWh, when driven at about this speed.

Coincidentally, my dash (behind the wheel) display, also shows 4.3 m/kWh, for the 7,300 miles on my LEAF from delivery, also corresponding to my CW monthly averages, since the August 2011 update, which proves, to my satisfaction, that the the correlation between CW energy usage and dash is very close, if not exact, as I’d expect.

And My CW reports from long-range trips, from 100% of charge to V LBW, from 83 to 92 miles, at 4.9 to 5.1 m/kWh, correspond to the entire range chart (except, possibly, for those parts below VLBW, that I've never used)

But, as far as I know, no one with gid or SOC data, has actually checked out CW, so there remains the possibility, it was, for this trip, and has always been, wrong for me, or I suppose, even for everyone else using CW.

And there is also the (hopefully remote) possibility, that I have suffered a non-gradual loss in battery capacity, since my last long-range test in January. Which I will take a look for when, I get the chance, as per my advice on the TickTock thread.

As to the accuracy of the bar display, I think I had close to 11 “full” bars (80% by timer + 30 minutes of L2) when I began. the 11Th bar did not go, until 7.5 miles into the trip, and even with the regen from the descent, that seems to indicates the 11Th bar was at a significant fraction of “full”, when I began.

The second visible bar disappeared, just before the 12.1 kwh trip ended.

So, I think could have been significantly off in my prior understanding of the bar display, thinking they each are fairly consistent , and pretty close to 1/14 of total ABC, 12 visible, and the two hidden or reserve. This is the view I had, perhaps mistakenly when I projected the bar display on this trip.

I've never had the constant speed level drives, nor gid observations, that were used to compile the range chart, so I’ve taken it for granted that the figures posted were both accurate, and consistent for the underlying kWh use. The only purpose the chart could ever serve, for me, is to use the level drive values of m/kWh usage at a given speed, to calculate maximum range, after adjusting for other factors, and for that purpose, it clearly has been very close to correct, for my long-range driving.

So, I've never had reason to question the range chart’s conclusions, of battery capacity from 11 to one bar of display, or to LBW, until this trip.

In fact, the range chart,if I read it correctly, seems to indicate these, bars, from between 10 and 11 to 1, would correspond to a slightly larger % of battery capacity than a simple 1/14Th of total battery capacity each, would indicate, more like 15 kWh, and even only 9 bars would be about significantly greater than 12.1 kWh. For example, If I am reading it correctly, the chart shows, for the same m/kWh, as I believe I got on this trip, 4.3, to reflect 63 miles of range from 11 to one bar, and 57 miles just for the 10 bars to one bar segment. I drove 54.4 miles (by odometer) and used well over 10 bars.

Is that much variation from "average” bar representations of range/battery capacity likely?

While those numbers are, perhaps, “fudge-able”, If I instead use the LBW as an indicator, it seems the range chart would show about 68 miles and 78% of total ABC, about 15.6 kWh (estimated using 10. 5 “bars ” as calculated to my my approximately 20 kWh ABC, adjusted for temp) miles at 4.3 m/kWh) between my start point and LBW.

And on this trip, the same events, seemed to have corresponded, to only 12.4 kWh use, and 62% of ABC ,as reported by CW.

This 25%+ variation, is in the “non-fudge-able” range, IMO.

One further indication of total kWh use for this drive, supporting the CW report, over the bar chart calculation, is total charge over 80% + recharge time. My “charging stopped” notification, for the recharge to 80%, after this trip, at very similar temperature, to prior the trip charge, was received after 3 hours and 30 minutes.

I had also added 30 minutes of charge, from 80% prior to the trip. Again, the total trip, was slightly longer than the 12 .1 kW and CW , showed an entirely consistent, 12.4 kWh of use, for the entire trip.

85% charge efficiency, which is about what many have reported, IIRC, using Phil’s 16 amp charger, would seem to limit the total energy use at just over 13 kWh use max, even if a full 16 amps were coming “from the wall”, over the entire 4 hours.

Which also strongly suggests to me, the 12.4 kWh use reported by CW, is likely fairly close to correct, if not exactly so.

Yesterday, the charging complete notification for my almost exactly 80% to LBW trip, was received after 3 hours and 36 minutes of charging. I will post the CW report, of kWh use for the trip, which this recharge time corresponds to, once I get it.

Edit, 5/5: 11.1 kWh, out of the potential 11.75 kWh, calculated by 85% efficiency, at 16 amps constant.

I’m not trying to say my observations above are conclusive evidence. of either CW, or range chart, accuracy.

But the same test procedures I used, can be reproduced by others, many with more sources of information, and the ability to do a more accurate constant-speed drive tests than I can.

I’d suggest you give it a try, if the facts they could show, are of interest to you.

 

[surfingslovak]

So, it is conceivable that we may see LBW pop up with 3 fuel bars. But that LBW will still have 48 Gids of energy; the same as when it was new. Use the LBW data.

Hm, I see where you are coming from, and given the data Phil has shared with us, this makes perfect sense. Looking at it from another angle however, I would be surprised if the LB and VLB warnings protected a fixed amount of energy in the battery. That wouldn't make much sense at all. You typically don't want the battery to go below a certain SOC percentage, and not a certain Wh or kWh figure.