[angainor]

Hi all,
I saw different numbers regarding the efficiency of charging EVs in general, and the Outlander in particular. People usually mention 10-15% loss in the charging process. I'm trying to see how my car does, but can't make any sense of the numbers I'm seeing. I'd appreciate any insights!

Today in the morning the PHEV Watchdog reported 12Ah in the battery. I charged the car partially (not to the top) using the supplied charger, from a socket that monitors the power draw. My socket reported a total of 3.65kWh used. This was the same as what the car reported in the MMCS (charging cost set to 1). Given the mains voltage of 238V, this gives 15.3Ah charge.

After the charging the PHEV Watchdog reported 22.6Ah in the battery. I waited a few hours and measured again - no change. That is 10.6Ah added, out of 15.3Ah measured on the socket. If that is true, the loss was 31%. I guess this can't be. Am I making a mistake somewhere? Or am I missing something fundamental about how batteries work?

Thanks for any clues!

 

[richi]

I think you're confusing the line AC current with the battery DC current. Because they're at different Voltages, they reflect different amounts of power (and therefore energy, when integrated over time).

 

[richi]

10.6 Ah into the battery, at a nominal battery Voltage of 300V, is 3.18 kWh. I make that a perfectly reasonable 13% loss.

 

[angainor]

10.6 Ah into the battery, at a nominal battery Voltage of 300V, is 3.18 kWh. I make that a perfectly reasonable 13% loss.

Thanks, richi! I was comparing Coulombs, while I should be comparing energy (Joules) instead. Should you then take the nominal battery Voltage, or the actual one? If I take the actual voltage (324V) at the end of charging, then I get 3.43kWh (94% efficiency). Since the voltage increases over the charging period, I guess the correct way would be the average between the end and the beginning? That would then be 92% efficiency. I guess that sounds much better

 

[richi]

Well using the nominal will under-estimate the energy "stored," and using the final Voltage will over-estimate it. Nominal should indicate 50% SoC, but you're not charging 0–100% (probably more like 30–100%.

More to the point, are you sure your battery's only reading 22.6 Ah when full? Assuming it's a pre-MY2017 car, that's less than 60% SoH!

 

[angainor]

Well using the nominal will under-estimate the energy "stored," and using the final Voltage will over-estimate it. Nominal should indicate 50% SoC, but you're not charging 0–100% (probably more like 30–100%.

More to the point, are you sure your battery's only reading 22.6 Ah when full? Assuming it's a pre-MY2017 car, that's less than 60% SoH!

This is a 2017 model, and this was not a full charge. The reported capacity is 31Ah, so the above is 73% SOC. SOH is 77%, which I think is on the low side. Could you say how this looks from your experience for a 4yr car with 53kkm? To me it looks low, but it's hard to find any real solid numbers around..

If you're saying that at 300V the battery is actually at 50% SOC, in my car I see 300V at 20% SOC (reported by the dog, not by the car). Do you know why that could be?

 

[richi]

Ah I see. Forgive my misunderstanding. For the mileage, 31 Ah is bang on the trend line that Daniel graphs at phevwatchdog.net. But you might find that your car is under-estimating the battery health, and you could force it to recalculate it with the tips in the FAQ section A.

As for the 300V question, I might have misremembered. In any case, the estimate of overhead looks reasonable.

 

[angainor]

Ah I see. Forgive my misunderstanding. For the mileage, 31 Ah is bang on the trend line that Daniel graphs at phevwatchdog.net. But you might find that your car is under-estimating the battery health, and you could force it to recalculate it with the tips in the FAQ section A.

As for the 300V question, I might have misremembered. In any case, the estimate of overhead looks reasonable.

Thanks a lot richi! Of course, I already did read your FAQ many times I also did follow the procedure in A1 a few times. As a result, the car reported a SOC of 32.7Ah right after the charge, with 4.1V per cell. But that goes down after a rest. What I suspect is happening is that during the charging the BMU does Coulomb counting to compute the added energy. But then, after a rest, it seems to project everything onto it's current model of rest Voltage vs. SOC. The exact same thing happens after driving: initially, after I stop the voltage can be anything. Then the voltage 'rests', and the car goes back from the Coulomb counting estimate to the V-SOC model. This is always connected with SOC drop, so it seems like the car cannot do correct counting while driving.

If you're interested, Figure attached. Yesterdays charging was done in smaller steps (~1kWh). Those are the red asterix in the figure. The top ones are measurements right after I stopped charging, the lower ones are after 1h and 2h rest. You sometimes see that the Coulomb counting and the model do coincide (pure vertical updates, no change to capacity after rest change to Voltage). But in two cases the values are again re-projected. Again, seems like either the car has hard time doing integration, or the model is wrong

I have loads of fun with this car

 

[Nick Heenan]

Hi all,
I saw different numbers regarding the efficiency of charging EVs in general, and the Outlander in particular. People usually mention 10-15% loss in the charging process. I'm trying to see how my car does, but can't make any sense of the numbers I'm seeing. I'd appreciate any insights!

Today in the morning the PHEV Watchdog reported 12Ah in the battery. I charged the car partially (not to the top) using the supplied charger, from a socket that monitors the power draw. My socket reported a total of 3.65kWh used. This was the same as what the car reported in the MMCS (charging cost set to 1). Given the mains voltage of 238V, this gives 15.3Ah charge.

After the charging the PHEV Watchdog reported 22.6Ah in the battery. I waited a few hours and measured again - no change. That is 10.6Ah added, out of 15.3Ah measured on the socket. If that is true, the loss was 31%. I guess this can't be. Am I making a mistake somewhere? Or am I missing something fundamental about how batteries work?

Thanks for any clues!

I also run a PHEV, but am rather baffled by what you have been doing. What do you hope to gain by doing all this analysis?
The car's battery gives a maximum range of 15-25 miles, depending on the time of year and how it is driven. Beyond this distance it is "just" like any other hybrid ICE, and provided you keep the petrol tank topped up you will never ever experience any range anxiety. You are certainly not having to plan your journeys to leave an adequate safety margin of range between all the rapid chargers en-route.
Running costs of the PHEV are far more influenced by its relatively high fuel consumption (of petrol) at motorway speeds, than the very low cost of putting another 12kWh into its battery (at around £1 for a full charge). Yes the charging process is not 100% efficient, but the real cost of this is miniscule when compared to anything else.
Mitsubishi are not going to be interested in marginal variations in charging efficiency, certainly not for any warranty claim! Has ANYONE ever had to have a new battery pack fitted to this PHEV?
Yes, battery technology is slowly evolving, and I guess that most of us here are already dreaming about what to buy as our next vehicle (funds permitting) once the next-generation of BEVs with far longer-range batteries hit the market.
Perhaps that is also when some real charging efficiency gains will be seen?

 

[angainor]

I also run a PHEV, but am rather baffled by what you have been doing. What do you hope to gain by doing all this analysis?

NP. I'm an engineer, although different field. Just curious and hope to learn something new. I definitely don't expect to save any money on plotting all this: I could surely use this time in a more economically productive way

 

[richi]

What I suspect is happening is that during the charging the BMU does Coulomb counting to compute the added energy. But then, after a rest, it seems to project everything onto it's current model of rest Voltage vs. SOC. … seems like either the car has hard time doing integration, or the model is wrong.

Yes, exactly that AFAICS. It's a very simple BMS design, dating from around 2008. It's coded to be ultra-cautious.

 

[Nick Heenan]

NP. I'm an engineer, although different field. Just curious and hope to learn something new. I definitely don't expect to save any money on plotting all this: I could surely use this time in a more economically productive way

I'm a (retired) electrical engineer too. In planning my 2nd home charger installation, I spent / wasted far too much time exploring the cost-benefits of installing solar panels and a power wall, and the half-hidden tarrifs of various electricity suppliers, in an attempt to "optimise" the cost of increased electricity consumption.
We should all just blame the lockdown then, for having too much time on our hands

 

[angainor]

We should all just blame the lockdown then, for having too much time on our hands

Precisely! 🤣 I can't really drive very far, but I still enjoy sitting in this bloody car! what do do?!

 

[richi]

Just look forward to the next heatwave, when you can sit smugly in the car with the electric aircon running full blast.

 

[angainor]

Just look forward to the next heatwave, when you can sit smugly in the car with the electric aircon running full blast.

Oh, not my case. I live in Norway. Just getting out of the -10C season with electric heater running full blast! Don't suspect too much summer aircon needs in these parts of the world during my (nor car) lifetime