Strange ICE behavior

[M.M.]

There has to be some engineering sense to this for both Honda and Toyota engineers to design the system this way. As mentioned by @bfd this behavior was commonly observed in the plug-in Prius. See this thread and this one from PriusChat for examples.

I confess to not fully understanding the Prius drivetrain, but in that vehicle, being a form of series-parallel hybrid, it makes at least some degree of sense. Basically, since the Prius is capable of running entirely on the ICE, since it includes a full transmission with a mechanically linked motor/generator, I can see how you could mechanically or electrically decouple the larger EV motor and just let the ICE (presumably allowing assist from the smaller hybrid motor?) provide all power to the wheels until the battery had reached a safe level.

It doesn't exactly make sense to me from a basic energy flow standpoint, but because of the ICE-centric design I kinda get how that might be the way you'd end up doing it.

There doesn't seem to be much documentation on the PHEV Clarity's powertrain available at this point, but the Honda description is of a different system, where the ICE is normally only connected to a generator, and the mechanical transmission only is engaged when the car is at cruising speeds, which is much closer to a series hybrid with a little bolt-on parallel hybrid action for highway-speed only.

In this case, I just don't see how it would make any sense to have the ICE running when the battery was full. When you're accelerating, you want the electric motors doing the work to draw down the battery, so any energy being produced by the ICE at idle is a disadvantage (and that's if the ICE transmission is even set up to be able to accelerate the car from a stop). At cruising speeds, same. The only use of the ICE would be to provide mild compression resistance on a gentle downhill gradient, but once you touch the brakes it's going to either have to regen or just use the calipers.

But an ICE can dissipate energy as heat, even if no energy is going into the generator or the wheels, and that seems to be how the system is designed.

I'm not sure I follow. Yes, an ICE can use compressive heat generation and the exhaust system to engine brake, but are you suggesting that it's actually putting reverse load on the generator from the battery pack in order to force a charge level reduction? It's mechanically possible, but sounds nuts to me.

Going downhill with a full charge is exactly the situation that several of us have described for when the ICE comes on for no other apparent reason. You do ask the question that's in my mind as well: why can't the system use the friction brakes instead of rigging up the ICE for engine braking? One guess on my part is due to safety and liability reasons. You can have the situation where someone lives at the top of a long hill (some Prius owners in Colorado esp. describe this as a common scenario). If the car only uses the friction brakes to burn off the excess energy, it's possible that the brakes could overheat and thus create a safety hazard.

Every vehicle on the road prior to ten years ago had friction brakes as the prime means of slowing down, so this isn't exactly an unusual situation, but they do have the ability to use compression braking to spare the calipers, so your theory is the only thing remotely resembling logical I can come up with either.

If the regen has maxed out the battery pack, the car only has two means of burning off excess energy: brake calipers or compression braking in the engine. In order to keep the brakes from overheating on a long downhill, it might be firing up the ICE entirely with the intent of using it with the mechanical transmission to do compression braking. It needs to be running for this, which would explain why it would have to stay on. I would wonder if the little "you're using the transmission" circle would appear all the time in this situation, if it exists.

However... the original post in this thread is not describing this situation:

2. Last week, with early morning temps in the high 30s (daytime temps in the low 60s), my ICE came on when I pulled out of my garage. My battery was fully charged as usual (see image below).

If the ICE started up as soon as the car left the driveway, the only way the battery could be overcharged is if it was overcharged when it pulled in, which would require the car was fully charged elsewhere then driven downhill all the way home.

Rajiv: I don't suppose you live at the bottom of a long hill and charge at work? If not, it seems more likely this particular case is either something to do with temperature management or flawed firmware relating to same.

I'm trying to imagine how the Clarity can use the engine to "burn off excess energy." Could there be an energy-wasting mode where the motor/generator reverses it's torque and uses battery power to resist the torque from the engine it just started?

That would be mechanically and electrically possible, but sounds insane to me. It seems more likely, if anything (as I note above) that it might use the mechanical transmission to do engine braking in lieu of regenerative braking. Although that sounds odd to me as well.


I have a coworker who is very knowledgeable about hybrid powertrains, so I'll try to corner him at some point and see what a professional's insight into it is. I'll report back if I learn anything illuminating.

 

[bobcubsfan]

Our Leaf which of course was an EV, no ICE, behaved differently. Even fully charged, it would regen when coasting or when brakes were applied.

 

[Kendalf]

I'm trying to imagine how the Clarity can use the engine to "burn off excess energy." Could there be an energy-wasting mode where the motor/generator reverses it's torque and uses battery power to resist the torque from the engine it just started?

Yes, an ICE can use compressive heat generation and the exhaust system to engine brake, but are you suggesting that it's actually putting reverse load on the generator from the battery pack in order to force a charge level reduction? It's mechanically possible, but sounds nuts to me.

I do think that it is doing something like what the two of you describe. Here's my guess (without any actual documentation, it's only a guess) at how it works based on this diagram I found from a C&D article of the Accord Hybrid two-motor system--which is essentially the same as the Clarity hybrid system with different specs for the battery and motors.

I believe that the arrow from the blue generator to the green battery can also go the other way, such that the excess energy from regen can be shunted to the generator, which can then spin up the ICE through the mechanical connection pictured. The clutch is disengaged such that the ICE does not drive the wheels nor provide power to the main electric motor. The engine then dumps that excess energy from the generator (which is now acting as a motor for the ICE--what a concept!) via engine braking.

 

[bobcubsfan]

Wow! Great find. So basically the generator can "bleed off" excess energy by turning the ICE which is not using fuel, or is it?

 

[rodeknyt]

Yes, it uses gas. This has happened to me a handful of times recently and sitting at our loooooong red lights in SoCal, I've seen the HV range drop by a mile or two.

 

[bobcubsfan]

An analogy. When a mobile phone is charging, it will stop when the battery is full. So, why can't Honda have the electronics stop the battery pack from accepting "juice" when it is full, but let the regen circuits and the generator work? In other words, have the system behave like a mobile phone.

 

[lordsutch]

A full mobile phone would be charging from an AC adapter connected to the power grid, which could either switch off (in which case the dropped load would be redistributed across the power grid) or release the excess energy as heat (what older, non-switching power bricks did). Since a car is a closed system, there's nowhere else for the excess energy from regen to go in the "grid"; it has to be dissipated somewhere in the car. The last place you want this energy dissipated is as heat in the wheels, since that'll overheat the friction brakes (which you will need later to actually stop the car). Using the gasoline engine to bleed off energy is probably the least bad option available, since it's already designed to safely dispose of excess heat anyway through the engine coolant and cabin climate control systems.

Presumably the engineers figured that it would be relatively rare for the car to be over "100% charge" (i.e. beyond the maximum permitted state of charge of the pack, which is quite a bit below 100% of the nominal cell capacity), since most people would precondition and/or they'd have enough of a heating/AC load while driving, or at least use up enough of the battery before going down a hill, that there'd almost always be some excess battery capacity for regen to dump the power into. Given the choice of either overcharging the battery (and potentially reducing its lifespan) or burning a small amount of gas in those relatively rare circumstances with an early downhill start from 100% SOC, they chose the latter option rather than the former.

Of course, the Leaf's engineers couldn't dump excess regen energy into a gas engine, so they really had no choice but to dump it into the battery pack, even if that meant overcharging the pack and hurting battery life as a result.

I suppose Honda could have designed the car to not use regen when the battery is full, but that would probably require some fairly complex engineering to ensure it would have the same "feel" and work safely and consistently, since drivers presumably expect pressing the brake pedal to have more-or-less the same effect regardless of how charged the battery is.

 

[jdonalds]

Maybe this is completely different, but our Nissan Leaf would go into regen mode as soon as the car was "coasting" or if brakes were applied even with a full charge.

Same with the Clarity

 

[M.M.]

An analogy. When a mobile phone is charging, it will stop when the battery is full. So, why can't Honda have the electronics stop the battery pack from accepting "juice" when it is full, but let the regen circuits and the generator work? In other words, have the system behave like a mobile phone.

lordsutch answers the question thoroughly, but I want to add something:

In a PHEV (certainly the Volt, and almost certainly the Clarity and every other PHEV unless they end up having very poor battery longevity), 100% visible SOC on the battery pack (that is, what you have when the car finishes charging, and what you see in the mobile app) is not 100% of even nominal “full”. It is probably around 85% of “full”, with the extra reserved to increase battery longevity (as discussed at great length in another thread here). This is different from most BEVs, which usually give you the option to go to pretty much 100% of nominal SOC, since unlike a PHEV you would get stranded if you ran out, so for some drivers and/or some trips, the decreased life of the battery is worth being able to drive somewhere you otherwise could not get, or the reduction in range anxiety.

So using hand-wavy numbers, a Leaf at 100% that has nowhere to dump excess energy other than the brake pads might allow some overcharge, despite a significant hit to the battery longevity, figuring that it won’t be common. A Clarity that is at 85% at “full” actually has another 15% SOC to put into the battery before it even gets to the nominal maximum charge. This does have an effect on overall longevity if you do it a bunch of times, so it’s possible that the engineers decided it was worth either using the transmission or the motor generator run backwards as a means of dumping excess energy, as lordsutch says, although if I was designing it I’d probably have let it go up at least a few % before I resorted to desperate measures, and it would be interesting to hear an estimate of how much regeneration you have to do on a “full” pack before it turns on the ICE.

As noted earlier, though, this does not seem to explain the behavior that the thread-starter is describing. The description seems to be of an ICE that starts only some of the time, and right away, before it would even have the chance to over-regen going downhill.

 

[bobcubsfan]

But, my point is this. When any EV is being charged externally, at some point the onboard circuitry will stop the charging process. when the battery pack is fully charged. The cable is still connected, but nothing further happens. So, why not let the regen function work at all times, but have a software switch disconnect the generator from the battery pack when it is full?

 

[jerry50902]

I was going to post about this problem, lucky I found the thread first.

I got my Clarity PHEV on Dec 15, 2017. It now has just under 10k KM on it.

It went through the entire winter perfectly fine. ALWAYS used electricity first and ICE never turned on at all (all except two days in Toronto where temperature dropped to -30C). I charge the Clarity to full every night on a 240V charger.

My commute is about 50 km round trip. During the winter, the EV range is barely enough to get me back home. Last month, when temperature is 20+, I can get home with about 30% of charge left, which is GREAT!

However, when the ICE-turn-on situation started last week, the ICE won’t shut off until I am almost to my office. I will return home with almost 60-80% of charge left. That simply means I am wasting gas for no reason when I should have been on pure electricity only.

I thought it was me stepping to hard on the gas on the way, but I have been observing very carefully and that is not the case.

I hope I can find an answer on here first before going to the dealer.

 

[ab13]

But, my point is this. When any EV is being charged externally, at some point the onboard circuitry will stop the charging process. when the battery pack is fully charged. The cable is still connected, but nothing further happens. So, why not let the regen function work at all times, but have a software switch disconnect the generator from the battery pack when it is full?

In order for regen to work the electricity has to go somewhere, as long as the generator turns it makes electricity. The simplest design is to stop regen and go full friction brake, but that is not desirable. The other way is to generate heat, which is not practical.

A BEV will also have this issue if the battery is full and you travel downhill.

 

[bobcubsfan]

But this happens on a pretty level road, and not always. Do we know for sure how a BEV works?

 

[jerry50902]

I have observed that the situation only happens when the battery charge is full ( during the last two weeks).

If I parked the car for a while with the battery not at full. The ICE does not turn on.

If this indeed IS the reason, then we need an option to charge the clarity to only 99% full. I would much rather lose that 1% of EV range than having the gas engine turn on. Hell, even 10% is fine with me.

But having the ICE turn on (basically the entire trip) for me defeats the purpose of a PHEV. Might as well get an Accord hybrid that is cheaper to run and buy. The point of the Clarity was so that I can do all my daily driving with EV only and only rarely use the ICE.

 

[bobcubsfan]

When I was a teenager I had a bicycle with a small generator that was connected to a headlight. I could engage the generator so the rear wheel would generate power. As I rode, the light would come on. So, what if I disconnected the wires to the headlight? The generator would still be operating, wouldn't it? It would still be a drag on the rear wheel. That is my whole point. If the battery pack is full, let the generator do its thing, but disconnect it from the battery pack.

 

[Kendalf]

If this indeed IS the reason, then we need an option to charge the clarity to only 99% full. I would much rather lose that 1% of EV range than having the gas engine turn on. Hell, even 10% is fine with me.

I agree. I wish there was a setting in the car to limit the charge percentage. My EVSE allows me to set amount of energy to charge for a session or quickly set the timer for how long to charge, and I've been using this feature plus some fuzzy math from the % SOC of the battery to try to set each recharge session to only give me above 95 but less than 100% charge. It's been working, but definitely inconvenient.

 

[bobcubsfan]

Since the behavior is inconsistent that is a puzzlement.

 

[M.M.]

When I was a teenager I had a bicycle with a small generator that was connected to a headlight. I could engage the generator so the rear wheel would generate power. As I rode, the light would come on. So, what if I disconnected the wires to the headlight? The generator would still be operating, wouldn't it? It would still be a drag on the rear wheel. That is my whole point.

Now I understand where the confusion lies. In your bike example, no, it would not be operating. It would be spinning freely and providing virtually no drag on the bike.

A generator with no load on it is just a small amount of extra mass on an axel, and in something the size of a car provides no appreciable drag. The drag is produced when the generator is loaded--when it's converting mechanical energy of the car into electricity that is going into a battery or the motors on the wheels.

You can experience this for yourself with an electric motor from a toy. Spin the driveshaft when nothing is hooked up. There's a bit of resistance, but very little. Now touch the wires to either end of a resistor and spin it. It will provide much more resistance and the resistor will get warm.

(An aside: One thing an EV could do to address this would be to have a hefty resistive heater and fan for it somewhere that it doesn't have to blow hot air into the cabin, and just dump the excess energy into that in order to save the brakes on a long grade.)

There are two things being discussed here--one is whether the Clarity responds to a long downhill with a full battery by dumping energy somewhere other than the brakes (which appears to involve the ICE, either using the generator running backwards--pulling energy from the wheel regen motor/generators and putting it into the ICE-or using the ICE with the transmission), and the other is whether there is a completely unrelated issue forcing the ICE to run when it shouldn't, which very much looks to be the case.

If this indeed IS the reason, then we need an option to charge the clarity to only 99% full. I would much rather lose that 1% of EV range than having the gas engine turn on. Hell, even 10% is fine with me.

As I noted, in reality "full" on the Clarity is almost certainly actually around 85%, although the top 15%, give or take, is not visible to the user. I would very much expect at least 5% or so of that to be usable during an initial downhill regen, which makes me suspicious that there is a bug or poor design decision in Honda's control logic. Impossible to say for sure, though, without understanding the details of what's going on, both in terms of when this happens and what the Clarity's response to excess energy during a long downhill with more-or-less full battery is supposed to be..

But this happens on a pretty level road, and not always. Do we know for sure how a BEV works?

We do know how a BEV works. Which doesn't help, because we don't really know how the Clarity works. But to the point, if this is happening on a pretty level road, and only some of the time, it does not sound to me like there's any good reason for it to happen. It is possible Honda programmed its "dump excess energy into the ICE" routine way too aggressively, but that does not appear to be the only issue. Case in point:

My commute is about 50 km round trip. During the winter, the EV range is barely enough to get me back home. Last month, when temperature is 20+, I can get home with about 30% of charge left, which is GREAT!

However, when the ICE-turn-on situation started last week, the ICE won’t shut off until I am almost to my office. I will return home with almost 60-80% of charge left. That simply means I am wasting gas for no reason when I should have been on pure electricity only.

Question: Is your commute flat, or is it downhill most of the way to work?

The fact that it behaves differently when it's cold would imply that it doesn't need the ICE when it's cold because it can run excess energy into the heater, but does something different when it's warmer and doesn't have that load available.

But the numbers you give look really weird, unless you have like a 15% downhill grade for the entire 25km to work. If not, then the ICE is generating a significant portion of the energy for one leg of the drive, which should not be happening.

 

[jerry50902]

It’s weird but real. I will film my entire way to work tomorrow and upload it.

There are even times when the car is parked, the ICE is behaving like its in HV-Charge mode. (The graphic shows a big green lane straight from engine to battery, no, not tires to battery, but ICE)

 

[jerry50902]

I did drop by a Honda dealer today. One possibility was discussed.

Because I have my AC on Auto (set to 21C), it is “possible” that the ICE had to kick in just to run the AC. And because my car is parked outside, the interior temperature is high.

I will try it out tonight and report back tomorrow.