regen/charging efficiency

[JRP3]

Actually rolling resistance, coefficient of drag, surface area, motor and controller efficiency, and charge efficiency all play a part in vehicle efficiency. Still missing the entire point. I'm not arguing which vehicle is more efficient overall, I'm talking about one specific aspect. No I would not say a vehicle with 95% charge efficiency and 1 mi/kwh was good, I'd say it's got great charge efficiency and obviously has problems everywhere else. I'm capable of discussing specific parts of a vehicle independent of the whole. I really don't know how to simplify it more for you.

 

[Doug_G]

Please provide data to explain your contention of poor charge efficiency. (I assume you have posted it previously but it could take a while to dig for it.)

 

[JRP3]

Roadster owners claiming 25% or higher loss during charging. Lithium charges at 97% or better efficiency, with charger losses maybe around 90% or so. My conversion and others charge around 92-93%.

 

[Doug_G]

Is that a "worst case" number under high heat when the cooling system has to run full tilt to preserve battery life? If so simply reducing the charge current would increase efficiency.

Or is that all the time?

 

[JRP3]

I don't know the conditions during the reported charge events. I assume there will be a range, depending on temperature and current.

 

[stopcrazypp]

Has anyone even really measured charging efficiency of both (accurately) under normal/standard conditions (the same for both, meaning same charging rate, temperature, SOC range etc)? It's also hard to guess the Leaf's charging efficiency since they haven't been totally upfront about what exactly is the capacity of the battery (have heard 24kWh 0-100%, but also heard 24kWh "usable"). For the Roadster I have heard two numbers too (53kWh and 56kWh).

I think in the end the EPA sticker numbers for the Model S will tell all (assuming they also include charging losses, which they do AFAIK). But currently judging by EPA sticker (city/hwy kWh/100mi) for Leaf (32/37) and Roadster (29/32), the Roadster is better despite any supposed increased charging losses.

I confirmed it, EPA testing procedures do include charging losses, that means the Roadster is still significantly more efficient (at least on the EPA cycle) than the Leaf even given supposedly high charging losses:
http://www.smidgeindustriesltd.com/leaf/EPA/EPA_test_procedure_for_EVs-PHEVs-1-13-2011.pdf

That article says it measures consumption using "a normal AC source" which is quite vague, but for the Leaf there's really only 120V or 240V 15amp. Any comparisons should be based on those two.

 

[JRP3]

Here's one report:

Due to cooling and other losses in charging, filling from empty takes about 68 kWh, or 26% more than 54 kWh the battery holds.

http://evworld.com/article.cfm?storyid=1733&first=2423&end=2422

 

[stopcrazypp]

Here's one report:
http://evworld.com/article.cfm?storyid=1733&first=2423&end=2422

Sourced from here originally:
http://www.teslamotors.com/blog/pvev-we%E2%80%99re-getting-72-miles-day-sunlight-or-72-mps

No details on how they got that number.

 

[Doug_G]

I don't know the conditions during the reported charge events. I assume there will be a range, depending on temperature and current.

I'm not convinced that you can draw conclusions about regen from that. I expect cooling will require more power when stationary, since there is less air flow. More air flow means the radiator coolant is closer to ambient, so the compressor won't have to work as hard. Also under cooler conditions you won't need the compressor at all. So I don't believe you can make a blanket statement like that.

 

[dpeilow]

I've measured 59.5 kWh to do a range mode charge on a car that was in the 0 miles range state. This was a 30A charge.

I've also seen 2 miles estimated range lost over a ten mile climb to the highest point on the english motorways, which was regained when dropping down the other side.

 

[JRP3]

I'm not convinced that you can draw conclusions about regen from that. I expect cooling will require more power when stationary, since there is less air flow. More air flow means the radiator coolant is closer to ambient, so the compressor won't have to work as hard. Also under cooler conditions you won't need the compressor at all. So I don't believe you can make a blanket statement like that.

Part of the problem is we have been mixing two different topics, the charge efficiency of the Roadster and regen efficiency in general. However, since the pack is already warm from the higher loads of driving, and regen current can be higher than charge current, it's likely the Roadster pack would need even more cooling when driving than when charging. As you say this is compensated by the better airflow when moving, but it's hard to say to what degree. But again, my point is that in any vehicle regen has to involve conversion losses and there has to be a hit taken when driving hills compared to the same distance on the flat.

 

[JRP3]

I've measured 59.5 kWh to do a range mode charge on a car that was in the 0 miles range state. This was a 30A charge.

That's good, though without knowing how many kWh were actually left in the battery it's hard to say exactly what the efficiency is.

I've also seen 2 miles estimated range lost over a ten mile climb to the highest point on the english motorways, which was regained when dropping down the other side.

Since we know there have to be losses, otherwise it's perpetual motion, something wasn't the same on the way down as on the way up, or the range estimation was off. My guess is that range estimation saw all the regen on the way down and projected range from that input, just as it did when it saw the higher draw on the way up and subtracted some range. Obviously when you hit the flat you won't be getting that regen input. The estimating software can only do so much. The only way to really know is with kWh or amp hour data.

 

[Nik]

The estimating software can only do so much.

A 'range remaining at destination' estimate could use GPS and map data to make allowances for significant elevation changes. One could also imaging having metadata in a sat-nav that described the road surface characteristics if this signficantly affectd range.

Satnavs already calcualte 'economical route' based on ICE-specific algorithms, we just need an app to do it for EVs.

Thought: Would EV users share anonymised power-usage data to allow the rolling resistance of each part of the road netowrk to be determined? I would.

To avoid the risk of creating too much of a tangent, I'll repost this thought as a new thread.

Update: http://www.teslamotorsclub.com/showthread.php/5348-EV-Specific-Satnav-Technology?p=62305#post62305

 

[dpeilow]

That's good, though without knowing how many kWh were actually left in the battery it's hard to say exactly what the efficiency is.

I'd guess at 19 miles as it had only just flipped over. It was in range mode.

Since we know there have to be losses, otherwise it's perpetual motion, something wasn't the same on the way down as on the way up, or the range estimation was off. My guess is that range estimation saw all the regen on the way down and projected range from that input, just as it did when it saw the higher draw on the way up and subtracted some range. Obviously when you hit the flat you won't be getting that regen input. The estimating software can only do so much. The only way to really know is with kWh or amp hour data.

Well, yes, I realise that. But it was under the granularity of the range algorithm was kind of my point. If I got 51% back would it still appear as 2 miles?

 

[Doug_G]

Part of the problem is we have been mixing two different topics, the charge efficiency of the Roadster and regen efficiency in general. However, since the pack is already warm from the higher loads of driving, and regen current can be higher than charge current, it's likely the Roadster pack would need even more cooling when driving than when charging. As you say this is compensated by the better airflow when moving, but it's hard to say to what degree. But again, my point is that in any vehicle regen has to involve conversion losses and there has to be a hit taken when driving hills compared to the same distance on the flat.

Well, what limited data I have seen suggests charging efficiency on the order of 90% or so, which I would deem entirely acceptable.

I have some experience with cooling systems, albeit in a very different context. But what I have learned about them is that the biggest loss is pretty much always from the radiator to the air. Getting sufficient air flow makes a huge difference.

I've noticed that the fans in the Roadster tend to run when you're driving slowly, but speed up a bit and they switch off.

In any case, if you're getting back 90% of the energy going back down the hill, that's wonderful. Perhaps in theory you could save more energy by coasting down the hill, but in the real world you have to regulate your downhill speed. Even if you were to end up with 25% losses its still far better to put that energy back into the battery then to have to slow the car with 100% lossy friction brakes. Especially when the alternative is either (a) being stopped by the cops for speeding or (b) flying off a cliff.

 

[JRP3]

Well, what limited data I have seen suggests charging efficiency on the order of 90% or so, which I would deem entirely acceptable.

Absolutely. I'd be interested in your average air temps when charging, that may be the difference.

In any case, if you're getting back 90% of the energy going back down the hill, that's wonderful. Perhaps in theory you could save more energy by coasting down the hill, but in the real world you have to regulate your downhill speed. Even if you were to end up with 25% losses its still far better to put that energy back into the battery then to have to slow the car with 100% lossy friction brakes. Especially when the alternative is either (a) being stopped by the cops for speeding or (b) flying off a cliff.

I don't think you'll get 90% back from regen. Even if your wall charge efficiency is 90%, regen incurs losses in the motor and inverter, as well as your losses from the drive line, tires, and air resistance. Regen is an excellent replacement for brakes, no question. The problem comes when people, including myself at times, drive in a way that maximizes regen instead of trying to maximize coasting. Jack Rickard of EVTV did some testing with and without regen on the same route and both times came out with almost the same energy use by maximizing his coasting. He had others do the same routes in the same car and they benefited more from regen but never got better numbers than he did. Certainly steeper longer hills and a lot of stop and go would increase the difference. I'd love to see some solid numbers on regen but there are so many variables it's like trying to pin down range numbers. Best we can probably do is say for X vehicle in Y conditions you might get Z for regen, or an average number you can expect overall. It is going to be less than your best charging numbers.

 

[Doug_G]

Absolutely. I'd be interested in your average air temps when charging, that may be the difference.

Temperature in my garage usually in the range between -15C to +30C. Varies a lot over the year.

I don't think you'll get 90% back from regen. Even if your wall charge efficiency is 90%, regen incurs losses in the motor and inverter, as well as your losses from the drive line, tires, and air resistance.

I don't think you can include things like tire and air resistance, because you would have had that loss travelling the distance regardless. That's more a factor of how fast you are driving.

 

[tomsax]

I can't speak to what's theoretically possible, but my model is that when driving between 55 and 60 mph in good weather, I lose about 7 ideal miles for every 1,000 feet of elevation gain and get about half of that back for each 1,000 feet of drop. This model fits the data I've collected going over Snoqualmie Pass several times (3,00 feet) and Mt. Rainier once (6,500 feet).

I also drove back and forth over Stephen's pass (4,000 feet) in 23-degree F; that trip fits my model if I reduce range by about 10% to 15% due to the cold weather and wet roads.

It's going to be a challenge to do those same drives in a Leaf as there isn't currently any helpful charging along those routes.

 

[JRP3]

I don't think you can include things like tire and air resistance, because you would have had that loss travelling the distance regardless. That's more a factor of how fast you are driving.

Yes I was just trying to point out that there are other forces slowing the vehicle besides regen, and that on some mild hills speed would drop even with no regen at all. In that case you would recoup none of the energy used going up the hill. A theoretical loop with steep uphills and long gentle downhills might result in little to no regen. Does the Roadster have an amp gauge? You could drive up a hill and see the amps required to hold a steady speed then see what the regen amps are on the way back down at the same speed, hopefully on a day without wind.

 

[Doug_G]

Does the Roadster have an amp gauge? You could drive up a hill and see the amps required to hold a steady speed then see what the regen amps are on the way back down at the same speed, hopefully on a day without wind.

Yes it does. My daily commute doesn't take me over hills, but I'll try that when I have a chance.