23' Prime - Gas engine starting in EV mode

I believe it’s more about the battery not being able to handle much regen when it has lots of charge and particularly when it is cold. The car can’t switch to mechanical brakes so it starts the engine and the engine then needs to run long enough to warm up.

Try coasting down the hill in neutral and the engine will not start.

Also pre-warming the car for 20 minutes, using the app, helps prevent the engine from starting.

 

I

I don't think that is right. I will pay more attention to the contributions of both battery power, battery pack remaining charge and engine power next time I take that route. If I remember correctly even when the engine cycles on it is still not making a significant contribution to driving the car down the road. I am driving in EV mode and using Predictive Efficient Drive when I have made these observations. Unfortunately there isn't a PID I have found in OBD Fusion for battery temperature.

Do you have a reference for your theory regarding battery not being able to handle regeneration current from MG2? The gauges I have in OBD Fusion are capable of showing regenerative power and I have never seen any indication in regards to the hybrid system not being able to handle regenerative power when going down hill.

Here is an excerpt from ChatGPT regarding what Toyota has referred to as "Engine Exercise."



I believe the algorithms on the hybrid control ECU are trade secrets. That partially explains why it is hard to find information in regards to what is going on with these short run engine cycles.

 

Hi Otatrant,

Since I’m just a Prius Plug-in user myself, I should probably correct my wording: I “suspect” the system’s purpose is to burn off the excess energy from regeneration. This idea came from a discussion I saw in another Facebook forum, and so far it’s the most logical explanation that fits what I’ve observed. I’m not sure about the background of the person who originally gave this explanation, but here’s what he wrote:

“The ICE is kicking in when in B or D because the traction battery is full (around 80%), and therefore it has to expend the extra incoming energy by letting MG1 spin the ICE at a faster RPM than the usual B-mode engine-braking range (when the battery isn’t full yet). At that point, it’s sending the surplus energy through the spinning pistons—so they act as an air pump to slow the car down while releasing the extra energy without using fuel. But when you press the accelerator, MG1 has no choice but to stop using that extra energy to spin the ICE and instead use it to drive the wheels. As a result, the ICE drops to 0 RPM, B-mode temporarily disengages, and you no longer get engine braking—causing you to ‘free-fall’ or speed up downhill. To counter this and keep slowing down, you can use left-foot braking to engage the mechanical brakes, dissipating the energy as heat through friction at the rotors. Once you’re on flat ground (in B or D), the car will try to use as much stored energy as possible and delay starting the ICE until the state of charge drops back into the normal operating range. Of course, you also need to switch back to D if you’re still in B.”


Few months ago, I actually recorded the same phenomenon—when the gas engine was forced to start on a downhill:

In the video, you can see that after about 1:45 of regeneration, then about three seconds of peak charging caused the ICE to start. So, at this point, I’m quite convinced that this “burn-off-the-energy-from-regeneration” explanation fits the system logic best.

You might try this on the downhill section where your ICE usually kicks in: occasionally press the accelerator slightly to interrupt continuous regeneration and see if that prevents the ICE from starting. Based on my own commute experience, avoiding long, uninterrupted regen almost always keeps the gas engine from being triggered.

In addition, based on my current driving experience, whether or not the car is pre-warmed doesn’t have a direct correlation with preventing the gas engine from starting.

 

How do you know that was the cause? The charge meter on the MID still shows the car is charging the battery as the car slows down coming to a stop and the ICE is running.

Even if that worked at preventing the ICE from starting on its own that doesn't mean the rest of the theory about the battery turning the cylinders to burn off energy is correct. In addition I have never come across any information indicating that the battery rotates the ICE which is the reverse of what every explanation of how the hybrid system works. The consensus I am referring too is that the ICE can generate electricity to charge the battery but the battery does not drive the ICE. Do you have any references supporting this theory beyond facebook?

In addition if the car was using electricity to rotate the cylinders of the ICE wouldn't you expect the EV range to drop. In the video it remains at 41 km as the engine runs for 5-10 seconds.

 

Burning of energy wouldn’t be the reason, as the engine produces energy, does not burn it.

When the battery is full, the motors are freewheeling. For some reason, Toyota wants some engine motive power is used when that happens so that the car is not barrelling down in neutral gear.

 

Engine doesn’t need to be combusting for engine braking. Also, combustion produces energy, doesn't burn it.

Try the same at 50% SOC and see what happens.

Why is your battery SOC not showing? The Gen 5 display is worse than the Gen 4 display. Gen 5 seems to be actually a Gen 4 refresh with some upgrades to power but then some downgrades to other things. Gen 6 will be a major redesign.

Eco score only 23%?? It is supposed to be close to 100%. Going well over the speed limit? LOL

 

I was also confused at first by the idea that the ICE might start up just to burn off excess regen energy.

Even now, I can’t say for sure that’s the exact reason.

But — I’m 100% certain that regeneration directly triggers the ICE to start.


I’ve become very familiar with this behavior（skill?) after driving the same two long downhill routes near my home many times.

As long as I carefully control the level of regeneration — especially by avoiding long, continuous maximum regen — the ICE never starts.

I’m also quite sure that the ICE activation has nothing to do with the traction battery’s charge level, warm-up status, or even ambient temperature.


To me, the strangest part of this logic is this:

If there’s too much regen energy, why can’t the system simply reduce engine braking and let the mechanical brakes handle the excess instead?

Starting the ICE — and even potentially burning some fuel — just to deal with over-regeneration feels counterintuitive.

From a driver’s point of view, it doesn’t make sense that the system would rather start the gas engine than use the friction brakes.

 

Haha, I knew someone would have something to say about the speed and that Eco score

No worries — next time I’ll make sure not to upload any clips unless they hit 100%!

Anyway, my main feedback here is still about the ICE kicking in during regen — that’s the part I’m trying to figure out.

 

Beyond a post on facebook and a correlation between long periods of regeneration the ICE cycling on how do you jump to causation?

 

I have some evidence that the amount of charge effects the engine startup:

The only time I experience the ICE starting with EV range remaining has been leaving home with a full battery and going downhill right away. It doesn’t start during summer months and doesn’t typically start if I prewarm the car.

I have also experienced the engine starting on a long, extended, steep hill after driving 13 km and with about 60 km of EV range remaining. Going down this same hill with a near empty battery with about 10 km of EV range never causes the engine to start regardless of how much regeneration I apply.

 

If you are really curious you could get a bluetooth OBD adapter and one of the many OBD cell phone apps or even just look at you charge meter on the MID display.

First I am pretty certain that the car is not starting the ICE engine to burn off excess regeneration energy because the battery is too full or too hot.

For the last couple months I have not been driving one route where I observe similar behavior i.e. descend a couple hundred feet of elevation followed by an ascent of a couple hundred feet of elevation. When I do travel this route the ICE engine cycles on in between the short segment in between the descent and ascent almost like clockwork.

In the last couple of months I have also been experiment with an old OBD iOS app OBD Fusion and the Carista bluetooth dongle. Here are some of the PIDs I selected to display and what I would expect to see on them during one of these brief ICE run cycles.

First. Hybrid battery remaining charge. The car completes charging at home somewhere around 89-90%. After regenerative braking during the couple hundred feet of descent I have noticed maybe a ~1-2% increase in remaining charge. Here is a screenshot to give you an idea of what I would see. Since I am currently not connected to the dongle it reads 0% but when I unplug from the charger it will read something like 88-89%. At the bottom of the hill it will read like 89-90%.


A couple of other PIDs are traction battery or hybrid/EV battery system current, hybrid/ev battery power and trip fuel. I use predictive efficient drive which means that for the most part on short trips the only gas I use is when the ICE cycles on to do whatever it has to do. During these brief run cycles I have noticed trip fuel consumed to be somewhere around a tenth of a gallon.



On the hybrid/ev battery power I see a positive number indicating that I am driving primarily on electricity and similarly a positive number of the gauge for hybrid/ev battery current. Here are two screenshots to give an idea of what it looks like. Negative numbers indicate energy or current going into the battery and conversely positive numbers indicate energy/current leaving the battery.





So if the theory that these short run cycles burn off excess electricity what should I expect to see? In my experience the car is already using electricity when the engine cycles on for a brief period. I also don't think I have ever seen a jump in battery current or battery power during one of these brief run cycles.

 

Your observations totally make sense

But I’ve been wondering if maybe the energy flow in this situation goes something like this:
MG2 (regen) → DC bus → HV ECU → MG1 → ICE (spinning as an air pump).
If that’s the case, the energy might not actually pass through the battery, so the OBD data would still show stable battery power even though the system is dissipating energy internally.

Technically, that process shouldn’t need any fuel at all — the ICE could just spin freely to create pumping losses. But from my own experience, every time the ICE kicks in, it still seems to burn a tiny bit of fuel.
I could be wrong, but maybe that’s part of Toyota’s logic for system maintenance — keeping the oil pressure, catalyst, and temperatures in check??

 

Just that the other explanations don’t seem to fit as well.

I’ve tried disabling Predictive Efficient Drive, driving the same route in both hot and cold weather, and with the battery at different charge levels — none of those factors made much difference.

What does make a noticeable difference is how much regen I apply — basically, how my right foot manages the braking strength.
That’s been the one variable that consistently changes whether the ICE decides to kick in or not, at least in my experience.

 

Energy can never actually pass through a battery, for the simple reason that it doesn't have separate input and output wires.

You will have a net inflow, net outflow, or net balance.

If the car is going downhill with energy flowing into the battery, and the car is choosing to dissipate energy elsewhere, then what you will see is your energy inflow getting smaller.

Unless it's triggered by you increasing the regen by letting off the accelerator or applying the brake more - in that case maybe the battery flow remains the same, and you now have energy dissipated elsewhere.

The OBD variables should show engine power output, and I think that should include any dissipation as negative.

Looking at this video, you can see negative engine values from when it's started.

Another thing to observe there is how dynamic the regen limit is - the green marker at the right. That would be the most useful thing to locate if your OBD display can show it.

The hypothesis is that as well as being based on all the factors you can observe in that video (capacity, speed, temperature), it will be limited based on how much regen has occurred recently. (I guess that would feed into temperature even if not specially handled).

So as you're going downhill, that regen limit marker will be edging up.

At the point that you release the accelerator enough to drop the requested drag below the green marker, the car will have to start the engine for more drag.

When applying the brake pedal, it can get more deceleration from the physical brakes when you cross the max-regen marker, but afaik in normal driving, the "coasting" drag without the brake pedal is dependent on the HSD. So it's relying on that coasting drag normally being above the green marker.

 

Are you agreeing with the theory that the these short engine run cycles are to burn off excess energy from regeneration?

I took a short drive in a hurry tonight. 4 miles one way with the OBD Fusion app running. Started at an elevation of ~525 ft, ascended to about 720 ft, than descended down to ~50 ft. I was driving in EV, Sport mode with a destination set in the Toyota navigation and in a hurry to get to the grocery store before it closed.

The engine started after descending ~500 feet at an elevation of about 200 feet in a relatively flat area with stop signs every quarter mile or so. The start of the engine coincides with an initial drop in engine oil temperature from about 70° F to 66° F. The engine oil temperature than rises to 107° F over the remaining 3.5 minutes of the trip while using roughly 0.04 gallons of gas. In this same time period EV battery power ranged from -30 kW to +50 kW. I can not discern any patter between regeneration power and the internal combustion cycling on.

Another detail that stands out is that the fuel rate jumps up to a rate of 1 gallon/hr 1 second before the oil temperature drops to the low of 66°F. For next ~40 seconds the fuel rate remains in the range of 1-2 gallons/hr as the engine oil temperature rises to 85° F.

From the data I see a much stronger correlation between engine oil temperature, fuel rate and the start of the internal combustion engine. I still can't see any data to suggest that the internal combustion engine starts to burn off excess charge.

Here is a screen recording I uploaded to youtube of the OBD Fusion app replaying the drive. I had to switch screens to show the various data fields. The spreadsheet OBD Fusion generated is much more helpful in looking for correlations.

And here is .csv file on google drive of the in case anyone wants to look at the data.

 

I think it's one of the best hypotheses. No clear evidence though. Clearest evidence for or against it would be that battery charge limit value, but we don't have that recorded.

What I think you'd see is a shift in the limit more than a shift in the other values. It's not so much "burning off excess charge" as "keeping regen below the current battery limit", with that limit having been reduced due to long continuous charging.

If you can't find the battery charge limit, just having the engine power value would be good - I think that should read negative. Although there would be some uncertainty about whether a visualisation tool would correctly display negative, or exactly what the car reports. I'd trust a display of the battery limit value more.

 

Do you have data to support this? For example a max hybrid battery pack remaining charge at the start of a trip coinciding with a large regenerative current at the start of a trip meanwhile a steady battery pack %charge remaining. I don't think it works like this. I think that since the car battery stops charging at around 90% that means you could unplug from a charger, turn off all the drains on power (ac, radio, lights etc.) and descend from the top of a hill with max regeneration and observe that % battery pack remaining charge would go above the limit the car sets for the EVSE. That is what I have observed.

I live in the middle of a hill. If I go one way I descend 400 ft and if I go the other way I can ascend 200 ft. What I have observed is that when I go the direction of descending the car will add charge to battery above the limit set for the EVSE. For the last month or so I have never seen the charge get to above the low nineties in percent charge remaining. I have no idea what would happen if I lived on a larger hill and could descend 1000 feet immediately after completing charging. It is an interesting question though.

What would happen if hybrid battery pack remaining charge reached 100%? Would the car turn off regenerative braking? I don't know. I think most people would have to go out of there way to set up this scenario. In addition I don't think it explains these short engine run cycles people are discussing on this thread. The internal combustion engine is not like the electrical system where it can regenerate fuel like the electric motors can regenerate electricity. I suppose the internal combustion engine could use excess energy to produce heat but beyond this thread and the facebook post cited I have not seen any reputable documentation indicating this is the case.

 

"This" being a reduction in the battery charge rate limit? No, because you haven't recorded it. I'm suggesting you do.

Just in case it wasn't clear - not sure you're getting this, based on the following text - the limit I'm suggesting is a charge rate limit. That's what that green marker is in the video I linked to. Not a stage of charge limit.

This is well observed and documented behaviour - it will regenerate, but it will dump the energy into the engine.

And there are lots of posts from people asking how to limit the charge to prevent this scenario:

Limit charging to 95%??? | PriusChat

Yes, it spins the engine to create drag, which ends up as heat, same as any other form of drag. That's what every Prius has done for 20 years now.

The last form of skepticism we had on this was from that "Paul" user who knew it did this but was convinced the car wouldn't do it unless "B" was selected, thus the car would run away as if in neutral. (Turning off regeneration).

He was totally impervious to documentation, but if you're not, there's plenty on this thread, and threads linked from it.

https://priuschat.com/threads/the-most-misunderstood-aspect-of-the-toyota-hybrid-synergy-drive-system.249598

Everyone knows that the car will do this energy dump when it hits an SoC limit; the fine detail being discussed on this thread is whether it can also be triggered by a charge rate limit before hitting that SoC limit.

My belief is that it absolutely could, because the condition is basically "battery can not currently accept sufficient current to produce sufficient simulated accelerator-released drag", for whatever reason.

So the remaining question is can the current limit drop low enough to do that due to continuous regen significantly before hitting the SoC limit, as it seems we're not hitting the SoC limit here.

You need an ODB reader that shows the charge rate limit, like @john1701a's does in that video.

 

Just noticed that I think we do have an engine start situation due to low charge rate limit in that video above, at 01:30. Compare that stop with a "normal" one at 03:00. Regen marker rapidly moves towards zero, and the engine starts.

Not clear what exactly the trigger was though. I'll have to watch more. SoC was high.

 

So, the regenerative charge-rate limit is around 2C, which would be 13.6 kW × 2 = 27.2 kW for the Gen 5 PHEV?

For some reason, it likes to turn on the engine when the motors are freewheeling. It must be some artifact of the power-splitting system. Tesla doesn’t have an engine; so, they have nothing to turn on downhill.

Can you check if you get a higher regenerative charge-rate limit if you have a lower battery SOC, like less than 50% for example?