I remember reading somewhere that the regen efficiency is about 40 percent, but my sense from this trip is that it is considerably better than that. Has anyone else tried this kind of experiment or does anyone know what the alleged regen efficiency is, meaning from the kinetic energy of the rolling car to the kWh actually absorbed by the battery?
It's very misleading to use the range meter to do this kind of evaluation because it takes recent driving conditions into account. So if you spend 20 miles driving downhill the car thinks "gee, this guy is driving really efficiently!" and ups the range estimate.
For a better estimate of how effective regeneration is you should base it on the kilowatts consumed going uphill vs. downhill.
Have been curious and had the day off today so went up the local pass to get some numbers! After much trial and error I settled on a 2.7 mile run where I could regen all the way to a complete stop barring traffic. It was steep enough that in "D", I didn't need to lighten up the regen, but did need to add some extra regen at a couple of curves and the very end of the run. It would run up to about 63 mph top speed on the run. Started each run up by gradually accelerating up the hill to a cruise-controlled 55mph (later tried some at other speeds). I then assessed kWh used as I passed a sign post near the top at that speed. Then stopped down the road, turned around and got the car up to the same speed (55mph) and started recording kWh used from that same point until the complete stop at the bottom, at the starting location (well, other side of the road of the starting location, but negligible diff).
I did 6 55 mph runs, one of which was confounded in an interesting way. They were very reproducible, in fact, on the last run I predicted what the kWh would say at each leg including the stop and start at the top, and it was right on. Each of the 6 55 mph run used 2.0 kWh on the way up, and the 5 unconfounded runs all regenerated 0.8kWh on the way down. You could tell that if you were able to measure in hundredths some of the runs were a bit different, but not over a tenth, so pretty reproducible. Using the 5 unconfounded runs gave, interestingly, 40% efficiency (talking broad strokes here, 1 tenth diff either way gives 5% difference!).
The confounded run occurred when, on the way down, I got behind a slow poke, therefore had to regen more to keep speed lower (around 40-45 all the way down) to not run into him. This run regenerated 1.0 kWh, which would give you 50% (broad strokes, people!). I'm pretty sure this difference is real, and likely related to wind resistance stealing energy from the regen at the higher speeds. I did another run where I went 55 up, then mimicked the slow way down by using more regen to keep speeds at 40-45 and it came out the same (2.0 up, 1.0 regen on way down).
Did a few other runs to "heighten the contrasts" as they say.
Three runs with 45 mph up, and letting it go but from 45 mph on the way down. Each gave 1.8kWh up, two regenerated 0.9 kWh down, the other 0.8kWh down. So that would give you 48% if you average the runs. Makes sense for slower up being more efficient due to same factors (i.e. drag) as matter on level ground. Going down was kind of in between the 55mph starts and the really slow downhill runs above, also fits that the loss from drag might matter.
Did one last run at 70 mph up, and going even slower on the way down like 35-40 (without traffic of course!). Used 2.3 kWh going up, regenerated 1.1 kWh on the way down for about 48% again. I think these numbers are pretty good estimates of regen efficiency on a hill steep enough to require regen. But it does make you wish for hundredths of kWh on the gauge. Also, as the uphill and downhill runs are largely separate events except for the starting speed, it makes you realize this isn't a perfect test of the actual regen efficiency. This is due to how much the different speeds you use driving up the hill affects the regenerative % numbers, even though it's regenerating just the same on the way down. For example, if you took the most efficient way I tested on the way up at 45 mph, and used 1.8 kWh, then regenerated the most efficient way I tested on the way down (35-40), regenerating 1.1kWh, it would suggest 61% regenerative efficiency. If you took the least efficient way up and used 2.3kWh, and the faster, less efficient way down as well, getting 0.8 kWh regenerated, it would suggest 35% efficiency. Makes you wonder how you would do with slower traffic, gentler pedal pressure than you have to use on the hill, etc... My guess is it actually would be closer to the 60% or more in that setting, based on what we've seen with the slower runs here.
I'm sure someone will tell me where my math was wrong in here somewhere