How much Gas-only range should I expect from a 2018 PHEV full tank?

Discussion in 'Clarity' started by Jacobsk, Jul 4, 2019.

To remove this ad click here.

  1. KClark

    KClark Active Member

    On a trip from LA to southern Utah, about 750 miles total using HV mode the entire trip with 3 adults and luggage and driving 75 mph most of the way we averaged 40 mpg.
     
  2. To remove this ad click here.

  3. TomL

    TomL Active Member

     
  4. TomL

    TomL Active Member

    fffffff
     
  5. TomL

    TomL Active Member

    I'm a new owner with 1000 miles on a Touring model bought June 2019 and am pretty confused about range estimates. I have two questions. First, I started out with HV estimates around 330, but over the past 200 miles it has risen to 370 and not changed. Is this a spuriously high estimate and do I need to get the software fix to reduce high estimates? Second, my EV charge using a Level 2 Morec 32 amp EVSE, has consistently been 46-47 miles but today it was 53.8. That's a great number but is it also unreliable?
     
  6. Easy to find out. Set a trip meter to zero, and track actual EV miles obtained vs the estimate. I find they’re fairly close, mainly depending on highway miles driven.
     
  7. To remove this ad click here.

  8. MrFixit

    MrFixit Well-Known Member

    An HV estimate of 370 miles is 'normal' for full tank of gas. You will likely not actually get that many miles, but that is what is consistently reported when the tank is filled. I would be shocked if you didn't have the range update fix. My car was manufactured in August 2018 and it came with the fix from the factory.
     
    TomL likes this.
  9. TomL

    TomL Active Member

    Thanks for the prompt reply. I did as you suggested. This afternoon's charge was even higher at 55.3 for EV miles. I drove 29.7 miles in EV mode under ideal driving conditions on country roads mostly at 50+ mph. So, 55.3-29.7= indicates 25.6 EV miles should be left. The actual EV reading upon return was somewhat higher than 25.6 at 27.9. This is good news, right? 50-plus EV miles is better than the advertised 47.
     
  10. Yep, good news. I tend to do “mental math” almost every drive to see if I’m gradually “getting ahead” or “falling behind” the starting EV range estimate. Helps to see what techniques help and hurt EV range. And it’s a fun game!
     
    insightman likes this.
  11. This morning I found we had charged overnight to a disappointing - for this time of year - 47 mile EV range. Part was the need yesterday to jump on I-40 for a stretch, in the rain no less.

    But I think a bigger factor was that on several occasions, I sat in the car with A/C running while Karen ran into stores to shop and at the vet while checking out. I think when one is forced to do that, EV miles both actual and predicted take a big hit - which is entirely logical.
     
  12. To remove this ad click here.

  13. TomL

    TomL Active Member

    And it is getting funner (not funnier?). My EV charge has increased for the third day in a row from 53.8 to 55.3 to 57.5. I will test this out by going further on EV mode as Fast Eddie B suggested.
     
  14. 4sallypat

    4sallypat Active Member

    With all these higher than the 47 miles per charge stated by Honda during the warm weather months, can we stop the complaints when we are in winter ?

    The 47 number is an AVERAGE number of miles per charge - year round!

    My summer months are averaging 60+ miles per charge while the winter months are averaging around 50 miles per charge....
    But then my winter months are way warmer than the rest of the nation....
     
  15. TomL

    TomL Active Member

    I followed up with a complete litmus test that Fast Eddie B suggested. My overnight HV charge was up to 58.3 and I drove until that charge was exhausted under ideal conditions in HV mode on surface roads going 40-60 mph. The trip miles went from 0 to 60.8, about 4% higher than the starting estimate. While I am 73 and normally have a taciturn demeanor, my wife thinks I am downright giddy over this marvelous new car. Kinda makes me want to thumb my nose and splutter a raspberry in the general direction of Volt owners.
     
  16. MajorAward

    MajorAward Active Member

    For the record, actual readings from my last HV trip gave me this: 41.37mpg X 7 Gal = 289.59 HV Range. This was all interstate except for the first 11 miles, 74-85mph on the interstate, 0-55 getting to the interstate.

    Over the past two days I measured EV range in Economy mode ( I usually drive in Normal). Started with a 52 on the GOM for electric range, and ran my battery all the way to the point where HV came on. No engine operation at all during this period. Actual reading, assuming the trip meter is correct: 51.4. This was local roads (no interstate), speeds 0- 60mph, but mostly 0-50. It has been 88-91 degrees the past few days (lower at night, but all my driving was in daylight. I had the AC set to auto at 75 degrees for the duration.

    I'm still not getting the super high GOM readings, but very happy with the actual results. I haven't been through a winter yet, but will try not to complain when those numbers drop... but probably will anyway:)
     
  17. insightman

    insightman Well-Known Member Subscriber

    I just want to check. Driving 60.8 miles in HV mode (when the ICE helps out) isn't a big deal. Getting 60.8 miles in all EV (no ICE) travel is pretty good for a car rated at just 47 miles of EPA EV range.
     
  18. Robert_Alabama

    Robert_Alabama Well-Known Member

    @TomL , I agree with you that this car is superior to the Volt (and I own a first generation Volt and would say it was the best car I had ever owned prior to the Clarity). The Clarity blows away the first generation Volt. More EV range, better HV mpg, and the first generation Volt requires premium gasoline. For the Second generation Volt (most recent model), I believe the Clarity still has generally equal EV range, and slightly better mpg in HV, but at least the second generation Volt does burn regular gasoline. Add to that the size of the car and what I believe to be a nicer interior for the Clarity, and for me it is hard to even compare the two as equal alternatives. The Volt has better data available to the driver, but I'll reluctantly sacrifice that for all the other upsides. By the way, my father-in-law also is downright giddy over his Clarity too, and he is 86. He strongly considered buying a Volt a couple of years ago and is very happy he waited to buy the Clarity.
     
  19. TomL

    TomL Active Member

    The giddiness continues. EV charge is up to 60.1 and tripmeter confirms 60 miles driven. To further pander to my delight I stopped at my local BP station to “fill up.” After a month’s driving I couldn’t get more than
    2.5 gallons in the tank.


    Sent from my iPad using Inside EVs
     
    Teslawannabe likes this.
  20. neal adkins

    neal adkins Active Member

    You are correct. As i previously stated you will not get close to that estimate. The system is assuming an additional charge in the estimate. Actual range is about 280 miles hv + 53miles ev = 333. Thats about 37 miles short. So dont get yourself stranded relying on displayed range.
     
  21. gadgetrants

    gadgetrants Member

    Howdy everyone! I realize there's a 2-year gap between the last post and this one, but that's what happens when you buy a 2018 Touring in 2021!!! I have a lot of catching up to do!

    Really enjoying this thread and have a lot of thoughts and feelings about GOMs but wanted to respond to this point, which gave me a very strong, visceral reaction:

    I've worked on the "data" teams at several tech companies, so I fit neatly into the description of those "software designer" guys, though my role is more the guy who creates the tables, does the math, trains the models, etc. And then a software engineer takes my model and ports it into the device.

    Anyway, I was APPALLED at @2002's story, though in all honesty, I have had many colleagues who treat their internal "customers" that way. In their defense, most teams have a long list of critical features, and a manager or product-owner who prioritizes those features -- so it's not unusual for one feature request (or many) to get lip-service from the software team (the programmers typically have zero control over the priorities).

    Where I have wiggle room, I try to do things like organize brown-bag lunches and educational meetings so the sales people, the service people, and the product-development people can create a shared language and align their priorities. In the case of @2002's regrettable story, I would have invited his team to a a planning meeting (or three) to discuss the data-table updates, and then written a project proposal that highlighted (a) the expected improvements, and (b) the impact on sales, customer satisfaction, turnover/retention, etc. It's that easy.

    When anyone in your organization has a good idea it deserves to be heard and thought through. Doesn't always happen but that's life.

    Regarding GOMs I have a very strong, and probably unpopular opinion on them. I'll put that in the next post.

    -Matt
     
  22. gadgetrants

    gadgetrants Member

    And now, in response to the dreaded guess-o-meter aka GOM:

    In reverse engineering the range estimates, what we're trying to do is make sense of the output of a device that probably has many complex inputs. The problem is that it's a "black box" -- there are an unlimited number of ways (theories) to explain its behavior, all of which may be equally good at estimating the GOMs output.

    A more academic way of framing it is in terms of "the law of uphill analysis and downhill invention" (we're doing uphill analysis, by the way):

    http://web.eecs.utk.edu/~bmaclenn/Classes/102-S11/robot_labs/Lab5.html#:~:text=In his work, Braitenberg describes,it is easier (a downhill
    So maybe what I'm saying is I wholeheartedly agree that the G in GOM is a guess and that it isn't fruitful to look for straight-line patterns between "I drove ABC miles or terrain today" and "My estimated range was XYZ." Probably the most important and critical piece of data we lack is the "sliding window of time" that the range-estimators use (and how much consumption and distance are weighted by time) -- and my experience in the world of data, sensors, and real-world products is that without that as a bare minimum, it's literally impossible to reconstruct how the estimates are generated.

    One of the reasons this is an impossible task is that with black boxes, we typically assume "as X goes up Y goes up (or Y goes down)," that is, a linear relationship. But in fact good models can be highly non-linear, which means something like "when X is low and goes up, Y also goes up, but if X is kind of moderate, then Y actually goes straight and flat for a while..." Or worse, "When X is low and Y is low, Z goes down but if Y is high..." You get the picture. Even if we succeeded in naming the 10 most important features that go into the GOM estimates, figuring out how they are mathematically related is pointless.

    This isn't intended to squash the conversation. I think, in the absence of the algorithm itself, Clarity owners sharing their real-world experiences is huge. And as we've potentially seen, it can be a valuable way for each of us to gauge our own experience (but as was already noted, without highly detailed driving logs, your car is an apple and mine is an orange, but at least they're both fruits!).

    OK, phew, got two long posts out of my system! Hopefully future ones will be more concise!

    -Matt
     
    Last edited: Jul 29, 2021
    insightman likes this.
  23. insightman

    insightman Well-Known Member Subscriber

    Wouldn't it be fun if Honda provided an SDK (software developer's kit) that let us have a go at creating our own GOM algorithms? But I'm forgetting the Clarity's lack of a future.
     
    gadgetrants likes this.

Share This Page