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

[gadgetrants]

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.

Love that! And I'm gushing...the forum-famous @insightman responded to my second post! Awesome!

And yes I'd be all over that. I think a worthy exercise, as a start, may just be a brainstorming session on what the display itself might look like. We subleased a 2017 Bolt for about a year, and I want to say it had something like a min/max display, which to me is far more logical and user-friendly. Basically, "if you keep doing stop-and-go traffic you'll get 100 miles," but "if you try to climb a steep hill you'll get 30 at most." I honestly think a format that communicates what recent driving history looks like and then projects out to various (2 or 3) future scenarios would be very cool.

Of course, I'm the guy who LOVES things like a "30% chance of rain," where co-workers are quick to say, "Matt, NOBODY understands what '30% chance of rain' means!!!"

EDIT: I also realize, in my rant about GOMs, that I assumed that Honda is throwing in all the logical variables/features for predicting range. That's actually a questionable assumption. A MUCH MORE parsimonious model would simply be to put in a basic gauge in everyone's car that captures the current charge and gas levels, and then computes how much they go down (say after a short distance, like 1 or 5 miles) and then building a simple model from those time-series data alone (e.g., computer laptops tend to work on such models for battery life estimates). I might also want to pool data over lots of cars, but the basic idea is to compute a "time-to-failure" (i.e., no more battery/gas) model that extrapolates to 0 based on your current levels. It's a highly naïve model but my guess is, all things being equal, that it will capture more than 50% of each driver's real-world experience (i.e., knowing that I'm at 70% charge predicts, on average, 30 miles left, 50% predicts 20 miles, etc.). A nice feature of such a model is that it "naturally knows" that percent changes in capacity/fuel don't have to be linear, e.g., a 5% drop when the battery is full might occur with 10 miles, but the same 5% drop when the battery is near-empty might happen after driving just 1 mile).

-Matt

 

[18ClarityPLUG]

I had to learn this recently also, thinking that these numbers would be consistent each time I charged or filled up the car. They will change with your driving habits. I fill up about once a week, more than most I suppose. I've already noticed my HV guess 0'meter adjusting. Today it was 320 after my fill up.

 

[gadgetrants]

Summer so far is a bit irregular for us, but in August my wife goes back to work and will drive the car in a very consistent way (day, time, route). I'm super curious to see our range estimates settle in to a more predictable pattern. But just filled the tank and the app is reporting 381 (50 EV + 331 HV).

-Matt

 

[MrFixit]

I look at it this way:

• The gas tank is 7 gallons.
• In HV mode, on normal terrain, with moderate speeds (60 mph or less), you can expect mileage in the mid to low 40's
• At 43 mpg, 7 gallons will yield 300 miles.

When I fill up, the GOM typically says 350+ for HV range. Add your ~50 for EV and your total range is ~400.

In my experience (and as reported by others), the GOM is optimistic (maybe 15%). But being off some is not so important when you have a full tank of gas. What I care about more is when the gas gets "low". I don't believe the vehicle will ever run out of gas while still indicating that HV miles remain. Thus, I think the GOM becomes closer to reality as the gas is depleted. I have found the GOM's to be quite adequate... I am not annoyed by them as some are.

You can alter this either way based on your driving circumstances (you may drive faster or slower, or have a lot of mountainous terrain)... I think the GOM will adapt some for you if you operate with consistent conditions, but still probably be optimistic.

Remember the days when all you had was an analog fuel gauge with tick marks that may or may not have accurately represented fuel level in 1/8'ths of a tank?

 

[gadgetrants]

@MrFixit in reading this thread and others about range estimation in the Clarity, your approach:

I look at it this way:

• The gas tank is 7 gallons.
• In HV mode, on normal terrain, with moderate speeds (60 mph or less), you can expect mileage in the mid to low 40's
• At 43 mpg, 7 gallons will yield 300 miles.

When I fill up, the GOM typically says 350+ for HV range. Add your ~50 for EV and your total range is ~400.

Has been by far the most cogent, the most logical, and for my needs, the most effective strategy. My rants about not "second-guessing the GOM" have not been to discourage any owners from trying (because it's fun and challenging!), but rather just to note that it's a bit like chasing a horizon that's always receding. As an engineering exercise we don't have enough information to do it -- so broad-scale approaches like yours make the most sense.

I don't know if someone else said this, but it's a bit like: how do I know if it will rain today?...I look and see whether I'm carrying an umbrella!

Your point about the GOM (my words) "converging to reality as fuel goes down" is a nice property of time-series predictive systems: predictions are highly variable (high variance) the farther out from the predicted event. In other words: as "time-to-failure" (no fuel) approaches, there are fewer and fewer alternatives available (e.g., with 80% charge there's a wide range of possible distances you could drive, but with 10% charge that range narrows to a much smaller number).

Remember the days when all you had was an analog fuel gauge with tick marks that may or may not have accurately represented fuel level in 1/8'ths of a tank?

We have a 2007 CRV, and (TRUE STORY!) after years and years of driving it, one day I was toggling through the odometer settings and said, "OH HEY!!!! It has a range estimate!!!" My first thought was "how did we survive without it?" and my second thought was "because all the prior cars we've ever owned!" Sometimes I repeat that mantra wrt: electric seats, heated seats, leather seats, bluetooth, ACC, sunroofs, touchscreens, navigation...the list goes on!

-Matt

 

[insightman]

I wonder, was the 89-mile range Clarity Electric's GOM optimistic or pessimistic? An optimistic GOM in a PHEV isn't a real problem like it could be in a BEV. Also, PHEV drivers get to measure their true EV range over and over while BEV drivers discover their true range only in dire circumstances.

My MINI BEV's GOM is very pessimistic, predicting ~110 miles when the battery can usually take the car more than 130 miles (mountains, 95-mph driving, gale-force headwinds and cold weather excepted). Yeah, I know that's not a lot of miles, but I'm happy to exchange the range for a ton of fun. We have our Clarity for the long-distance runs.

 

[Dan Albrich]

So my own Clarity has always performed at the low end of the spectrum. I don't know why, but there ya go. I typically see about 265 miles on a full-tank of gas. In the summer time, combined with EV, lets call it 300 miles total range. So I start looking for a gas station if I'm in the rural west at about 180-200 miles traveled. I definitely start looking for gas before the low fuel light comes on which seems to only yield something less than 40 miles worth of warning in my case. It's actually a weak point of the Clarity-- if I could of given the Clarity PHEV a 10 gallon tank I'd be perfectly happy. But the 7 gallon tank is what we get. In the summer, I don't think much of driving 3000-5000 miles in the rural west when on vacation so it matters where you go and if you like long road trips. If you tend to be in populated areas where access to gas is easy, then it's a non-issue.

I'd add that pre-pandemic, my daily routine was to drive 15-25 miles total, all EV. I have a simple non-highway commute to work. There are times when I go about two months without getting gas because I simply don't need gas. Love it.

I'm still plenty happy with the car and its fuel efficiency. So just something to be aware of. Not a deal killer.

 

[gadgetrants]

Dan your comments about "rural west" had me thinking that your trips are mostly long-distance, but then you also mentioned the "pre-pandemic" routine. I'm probably confused but IIUC I'd want to untangle those two before asking the GOM to give a consistent answer. Just speculating (wildly) but I'd wager that with the 15-25 mile routine you'd see something more like high 300's, and with lots of long-distance trips (where the battery is mostly charged by the ICE?), you'd have a mostly HV-type experience with mileage more like lower 300's. Keep in mind that I'm a total Clarity noob...

-Matt

 

[Dan Albrich]

Hi Gadgetrants-

So for usage, like 10 months a year I'm home and primarily travel around town. I live in a smallish town (Eugene, OR). My commute is almost exactly 5 miles so to/from work is 10 miles. After work, I often go out again "to town" and add another 10-15 miles. So a normal day for me is at or under 25 miles of driving. Then and separately, usually for about two weeks in the summer, I take vacation. For a few years I drove (and camped) in the southwest and a typically 9 day trip would exceed 4000 miles of driving. This was pre-Clarity, but I don't think much about 10 or more hours of driving during the day. To me, being on the road is total freedom. In several of my long trips, I start out with no destination in mind. More recently and with the Clarity, my wifes dad moved south of Tucson, so that's a couple 10 hour days of driving if we go visit by car. I've done that trip with Clarity a couple of times, but my intent was to indicate that some people like to drive including long distances (which I realize is not the case for everyone).

Regarding the Clarity GOM, when the car was new I'd measure against actual, and for my driving it has always been pretty accurate, but if anything just a tad optimistic. i.e. actual miles for me, are generally slightly lower than estimated. I used to keep logs and do a bunch of detailed tracking when I got the car in 2018. But time passed, I've intentionally decided not to track. I notice that the GOM in my case, always "trues up" within a couple days of similar driving conditions. But yes, in both cases, my driving is predictable. When I'm home, there's no highway driving ever, and average speed of 25. There just isn't a lot of places I need to go when home that aren't easy to get to. When driving on a long trip, my drives are hundreds of miles per day, and I typically go about 70-80mph on the freeway whenever practical to do so. But again, very predictable pattern of driving within the same time period -- and hence my GOMs historically have been pretty close to actual.

One more thing- I never charge clarity on a long trip. It's all "hybrid" mode and burning gas for me. My combined GOM has *never* displayed more than 340. Even brand new, just never happened. With a full tank and a full charge, I usually see 280-320.

Then once you learn that say for your usage, there might be a 5-7% fudge factor, you know the readings are good enough.

 

[gadgetrants]

Thanks @Dan Albrich!

I notice that the GOM in my case, always "trues up" within a couple days of similar driving conditions.

I'm gonna get that printed on a t-shirt, it reinforces my experience with predictive-algorithm development on the computational side.

I also want to think deeper about the fact that you've mixed in some long-range driving, but your shorter local trips are the norm, AND your GOM lives around 340. That "smells" to me as though (1) GOM estimates are miles-weighted (which makes sense in general) but also (2) possibly, the sliding window for computing estimates seems to include long-ago drives (there are a couple ways to do this, both lookup-table based, which is unlikely, and standard times-series models). It's as though your GOM has incorporated your infrequent, long-distance drives as potentially possible in the near-term, and adjusted your estimates accordingly (i.e., esp. since, as you noted, you go "pure" HV on the long trips).

Also:

To me, being on the road is total freedom. In several of my long trips, I start out with no destination in mind.

I'm not gonna touch that because I have a healthy case of FOMO and it makes me crave a red sports car and blonde girl-friend (and neither fit my profile!)...and now I hear Springsteen saying he "took a wrong turn and just kept going..."

But on a more serious note: we just drove through Eugene (in a different car) on the way to Portland (HOT HOT HOT...could see the smoke on Mt. Shasta!). How does the Clarity feel about those hills in Southern OR? We did the socal Grapevine in the Clarity a while back and now I know what the "angry bees" are.

-Matt

 

[Landshark]

Y’all sure can make a mountain out of a mole hill. It’s just like any ICE vehicle.

Tank capacity x MPG.

The range estimate is stooooooopid. Don’t waste a minute of your life trying to figure out stoooooooopid.

There is a fuel level gauge. Use that.

 

[MrFixit]

I don't know why the GOM seems to be a "love it or hate it" kind of thing.

They both provide useful information. If all you had was a fuel gauge like the old days, you would be estimating remaining miles in your head. The GOM simply does this for you. It is not perfect, but neither is the mental calculation based on the fuel gauge !

I think there is a tenancy when someone sees a digital readout to expect it to provide an exact answer.

 

[gadgetrants]

tl;dr It's a shame that our GOMs don't talk more to our phones (or the car navigation) and use planned route info to reason more intelligently about range estimates.

I think there is a tenancy when someone sees a digital readout to expect it to provide an exact answer.

I keep thinking the future is just around the corner where we say, "Hey [given name for personal car assistant], do I have enough fuel to get home?" At this point, I know I can ask the Google Assistant how far away [Location X] is in time, e.g., "When will I arrive home?" But maybe asking "will I run out of gas?" is not so trivial (I actually think it should be straightforward to compute, if you know the planned route).

So maybe: just looking at the dash and seeing a single number is a very impoverished data point. Talking to a...er, forgive my choice of words...digital assistant who is programmed to say things like, "Based on your current speed and road conditions, you have an 84% chance of reaching home" (you see, there I am again with my probability estimates) is a bit more informative. And imagine the [insert favorite scifi movie name here] scenario where you can discuss at length that 84% estimate with your assistant, e.g., "Can I also stop at [favorite restaurant] to pick up [favorite food], or is that cutting it too close?"

-Matt

 

[Landshark]

I keep thinking the future is just around the corner where we say, "Hey [given name for personal car assistant], do I have enough fuel to get home?"

I’m so thrilled to not be afflicted with Digital Assistant Disorder (DAD).

Unless you have a mostly charged battery, get gas when the gauge is down to 2-4 bars. Figure 15 miles per bar. This only requires basic math skills.

Otherwise, a likely outcome may go thusly:

Matt, I may have made a miscalculation. There must be a 15mph headwind, or the fuel is the special winter blend, or the road surface has become rougher than anticipated. Is the road wet? You haven’t used the wipers, so I don’t believe it is raining. I also did not anticipate getting stopped at the last 3 traffic signals. Those unexpected accelerations have consumed more fuel than predicted. Is that 1 gallon can of gas still in the trunk?

Sometimes humans can display levels of intelligence that is superior to the artificial kind. Sometimes.

 

[gadgetrants]

Matt, I may have made a miscalculation. There must be a 15mph headwind, or the fuel is the special winter blend, or the road surface has become rougher than anticipated. Is the road wet? You haven’t used the wipers, so I don’t believe it is raining. I also did not anticipate getting stopped at the last 3 traffic signals. Those unexpected accelerations have consumed more fuel than predicted. Is that 1 gallon can of gas still in the trunk?

OMG BEST LAUGH TODAY ^^^

I’m so thrilled to not be afflicted with Digital Assistant Disorder (DAD).

I have good reason to believe I'm patient zero.

-Matt

 

[Dan Albrich]

I used to get the "angry bees" sound pretty often. Members of this forums convinced me not to worry about. I literally just turn up the tunes and don't worry about it. The car is designed to do what it does, so I just trust that. Mine was unusually pervasive (angry bees) so I reset my system by unplugging the 12v battery, then re-attached. Maybe placebo, but I do think that helped. I've only needed to do that once. [I also always have to warn, if you follow this advice, just know you get tons of SCARY warnings next drive. just drive 5-15 miles and all the warnings go away. You did not break your car.]

The formula I now keep is to, not matter what, keep at least 10 miles of EV range, even when driving long distances in hybrid mode. If for any reason the EV mile estimate goes below 10, I do the hold HV button (HV Charge) and that gets me to say 18 miles of EV in the summer time. I then run in HV mode entirely on long trips. This combo doesn't eliminate angry bees under some hilly conditions, but it seems to keep the bees mostly happy.

The other thing I do, and I did this so long ago I forget it isn't default, I change the center reading to the one that has HV and EV miles listed all of the time. This might be why I've become such a "fan" of the GOM. And perhaps unlike others, I find the GOM to be very good for my uses. For me, it has never been wildly inaccurate, and yes it does adjust over time. I also find that setting my A and B trip meters (again non-default) to last time I did EV charge and last time I filled tank respectively gives me some insights too. i.e. for EV range, if the estimate is say 17 EV miles, and say I've already traveled 20, then I know my ending estimate (next time I re-charge) will likely be 37 (or 37 ish). This gives me a realtime way to notice how my GOM estimate will change (up or down) based on driving conditions.

Lanshark and Matt, I really enjoy both of your comments. I especially like the comment not to pay too much attention to stupid. Agree with that, but like I say, the GOM in my case has been quite predictable and reasonable for my uses.

[I work at a university, and our police captain in the past made a comment about someone being "stuck on stupid" which i dearly loved. As if "stupid" is a setting that one can flick on or off. -- and yes, try not to spend too much of my energy on stupid (or folks who are stuck on it). Good life advice.]

 

[gadgetrants]

@Dan Albrich the same goes for me, I think this is an awesome forum and your comments are very enlightening!

FWIW I also keep the same EV/HV gauge up in the center console and check it against Trip B distance, which is set to reset with each full charge (which is pretty much daily). And pretty much on all our local trips the GOM comes within a mile or two of adding up to the true distance travelled. Sometimes even, I get home with a 2- or 3-mile "surplus" (perpetual energy!!!).

Speaking of "avoiding 0 charge like the plague" last night something happened for the first time: I went out on a 20-mile grocery-run with maybe 21 or 22 miles on the EV GOM (this is a neurotic story: it's important to note that the prior day I'd filled the tank, reset the A trip meter, and noted that the HV range estimate was 331). I knew what would happen...and it did...the last 2-3 miles home are uphill and I hit 0. The ICE turned on and NOOOOOOOOOOOOO I burned 3 of those precious 331 HV miles making it home. I've read many times on the forum, and verified last night, that the battery still had 10% capacity, even though the console displayed 0. It might be the only thing at this point that I'm tempted to hack (which of course I'm completely unable to)!

[I...sadly...USED TO work at a university and miss that life sometimes. But in my newer persona as a "private industry" guy I had one of my favorite managers sometimes say that he needed "to flip his a$$#ole bit" which I'd never heard before and now dearly love...nice guy most of the time, but sometimes you need to flip "that" bit!]

EDIT: regarding stooooooooooopid when I got my Pixel 2 (backstory: I had HUGE battery problems with the Pixel 1) I tracked usage and battery data religiously for about 6 or 7 months. And then I stopped doing it and realized I was getting about 4-5 hours of SOT daily so why track? I still check a favorite battery app in case something goes awry but I think I'm on the same path with the Clarity: crazy new owner who wants to measure everything, but eventually learns to just be chill.

-Matt

 

[Landshark]

Speaking of "avoiding 0 charge like the plague" last night something happened for the first time: I went out on a 20-mile grocery-run with maybe 21 or 22 miles on the EV GOM (this is a neurotic story: it's important to note that the prior day I'd filled the tank, reset the A trip meter, and noted that the HV range estimate was 331). I knew what would happen...and it did...the last 2-3 miles home are uphill and I hit 0. The ICE turned on and NOOOOOOOOOOOOO I burned 3 of those precious 331 HV miles making it home. I've read many times on the forum, and verified last night, that the battery still had 10% capacity, even though the console displayed 0. It might be the only thing at this point that I'm tempted to hack (which of course I'm completely unable to)!

One tactic to reduce the likelihood of the engine coming on at the end of a drive is to just bite the bullet and run the engine for a few miles in the middle of the trip. That way the engine can fully warm up. Then shut it down and finish the trip on batteries. If the engine comes on just before the end of a trip, it should be allowed to run until it completes its warm up cycle and shuts off. That could mean sitting in the driveway, burning fuel. At least use the fuel to move the car forward.

And don’t try fudging with the BMS protocol. Battery manufacturers have good reason to not allow a DOD below 90%.

 

[gadgetrants]

I knew if I mentioned "hacking" I'd get a tongue-lashing and I completely earned it. Note to self: don't joke about reengineering the car!

And @Landshark thanks for the reminder about letting the engine run. I was "running" through that exact scenario in my head as I was headed back home, "If the engine starts now, will it keep running when I get home and turn off the car?..." You're absolutely right. I was trying to sneak back home under the wire but I knew in my heart that I should have pushed the HV button the moment I got on the freeway.

-Matt

 

[BrianRC]

One tactic to reduce the likelihood of the engine coming on at the end of a drive is to just bite the bullet and run the engine for a few miles in the middle of the trip. That way the engine can fully warm up. Then shut it down and finish the trip on batteries. If the engine comes on just before the end of a trip, it should be allowed to run until it completes its warm up cycle and shuts off.

So Landshark, how long would you say the warm up cycle is typically?