The effects of mild hills on PIP EV efficiency

I've attempted to measure the effects of traveling over seemingly flat roads that have a mild change in elevation. My commute is 6.5 miles each way. But I get dramatically different mileage (watt-hrs per mile) in each direction. I've noticed over the years in my previous cars (2001, 2004, 2010 Prius) that I get 5-8 mpg better mpg going to work than returning home). In my plug-in Prius the effects are greatly amplified.

Therefore, I conducted some (somewhat careful) measurements. From home to work the drop is 174 ft, or 27 ft/mile, on average (varies from 13 - 37 ft/mile). It is mostly not really noticeable. Most perople would say it is mostly completely level. 27 ft/mile is ~6 inches per 100 ft. However this mild slope produces a up to a 33% difference in MPGe on EV.

I get ~225 w-hrs/mile going uphill and ~150 w-hrs/mile downhill, on average on this route. I'm driving on 35 - 50 mph roads going about 30 - 45 mph as traffic permits.

For those that don't know or care about watt-hrs/mile: The battery numbers are about 4.6 miles of indicated battery miles for 6.5 miles going slightly downhill and 7.0 miles of indicated battery miles for the 6.5 miles on the uphill return route. This is based on initial values of 14.1 - 14.2 miles estimated on the display for a full charge.

I estimated the w-hrs by assuming 3.0 kw-hrs per full charge based on numerous ChargePoint charges reporting ~2.9 - 2.95 kw-hrs for a charge starting at 0.0 miles remaining. I realize that the actual usable battery capacity is probably closer to 2.7 kw-hrs so my absolute numbers may be uniformly too high by ~10%. During the test runs I took the battery miles readings at the start, end and several midpoints and scaled them based on the initial full charge miles remaining. I repeated these readings for 5 full round trips of approx the same conditions and averaged the results. My commute conveniently consists of (about) .75 miles, followed by five 1-mile segments (each with distinct cross streets/traffic lights) followed by .75 miles... so for the graphs (below) I threw out the first and last segments, keeping just the 5 one-mile segments. This also eliminated any small warmup penalty going each direction.
I obtained the elevation data from the web ( Google Maps Find Altitude ), calculated the elevation drop/rise for each segment, and plotted the results. Each segment is continuously uphill or downhill except for two overpasses. One peaks at the end of a segment (see worst case of 323 w-hr/mile...the following segment is 185 w-hrs/mile even though without the overpass each is gaining 30 - 37 ft/mile; I guessed that the overpasses are ~20 ft high). Interestingly the overpasses don't affect the otherwise continuously downhill commute much at all. The worst case on the downhill commute is accelerating onto the 50 mph road with a ~5-8 ft uphill.

(The table shows all the averaged data. The graphs only show points B to G.)

 

That looks like an interesting tool. It doesn't have the Prius plugin as a car choice, though.

Mike

 

The weight must be in kg?

Mike