In case anyone's interested I think I may have started to get to the bottom of the differences I've been seeing between EVTP's forecasts and the actual figures I've had from the car, compared with the accuracy others have reported.
I hadn't fully appreciated the way EVTP estimates and adjusts speeds on each segment of a route, and once I dug in to that detail (at least for one of the trips I took) I realised that the average speed EVTP was using for some major segments was significantly below what I actually achieved. EVTP claims to use average speed for a segment at the time you run the route plan, so clearly this can and would change at different times, which I hadn't appreciated.
Bottom line was I needed a speed multiplier of 1.1 in EVTP to get segment estimates closer to actual driving speeds, and that brought the kWh used much closer to what the car actually reported. Still erred on the optimistic side, so not a complete explanation, but at least it seems to improve my ability to predict usage and adjust for likely speed variations on major route segments.
I hadn't fully appreciated the way EVTP estimates and adjusts speeds on each segment of a route, and once I dug in to that detail (at least for one of the trips I took) I realised that the average speed EVTP was using for some major segments was significantly below what I actually achieved. EVTP claims to use average speed for a segment at the time you run the route plan, so clearly this can and would change at different times, which I hadn't appreciated.
Bottom line was I needed a speed multiplier of 1.1 in EVTP to get segment estimates closer to actual driving speeds, and that brought the kWh used much closer to what the car actually reported. Still erred on the optimistic side, so not a complete explanation, but at least it seems to improve my ability to predict usage and adjust for likely speed variations on major route segments.