Some background: My turbo (Tacx Flow Smart) has a fairly limited range of resistances it can generate. As a consequence, when TrainerRoad wants me to warm up or recover at 80…100W at 90rpm, I need to be in a low gear or ERG mode can’t get the resistance low enough and I end up generating more than the target power. Similarly, when target power is 200+W it can’t provide enough resistance in that gear, so I need to be sure to throw in 2 or 3 or 4 up-shifts to keep it in the range it can perform effective ERG adjustments within.
Which brings me to my most recent FTP… I find watching the power creep upwards and the seconds tick by is not helpful to a good performance, so once the workload gets higher, I’ll either look straight ahead or close my eyes and visualise effortlessly dancing up a favourite climb or something. That does mean that for this test I didn’t notice that ERG was in need of another up-shift so for the last two complete steps and the partial one I stopped in, I wasn’t actually generating the demanded power. (img link).
Also, I didn’t hit pause until the stars stopped spinning. But I did get a great FTP increase! (23W!) Or did I…?
(For both this and my previous test, I didn’t start the test until two consecutive calibrations had given the same middle-of-the-target-zone result)
On my next session, I had to dial the first interval all the way down to 90% to even stand a chance of completing it, and even then I was destroyed after only two intervals.
So, my question is: does TrainerRoad do FTP calculations on generated power, or target power?
It feels like it must have credited me with two full minutes at target power and the final 20s after I’d actually stopped.
As far as I know, TR simply uses your highest 60 second power output and then takes 75% of it as FTP. So that’s actual power output.
If anything I would assume that the calculated FTP now is a bit lower than what you could have achieved, if you did the upshift, because your highest 1 min power probably would have been a tat higher.
So if the workouts are now too hard, I think there is probably a different reason for it 
It uses the power measured, not target.
The “break even” point in the test is something like minute 19 (I.e. if you match the target to that point and stop you get the same FTP that you went in with). It looks like you were mostly matching to ~21 minutes which implies a notable increase.
Break even is 19:30, assuming you are “matching” target power. It will vary by any deviation of your actual power being high or low at that point (and the 60 seconds before it).
I had a tacx wheel on trainer and faced similar issues. A trainer tire can help.
Tossing that aside I would use resistance vs erg for vo2 workouts and at that time I also did the 8 min ftp vs the ramp test because of this issue. So maybe consider trying resistance mode in some workouts and gauge if it could work.
Last point. Next workouts after a ramp test and an increase are normally harder. Dialing it down isn’t the worst thing
The workout I couldn’t get near to completing was “Washington +4” (7 minutes at 105%, x5).
This was within 12 hours of my first vaccine dose, so maybe that wasn’t the best timing…
Since posting, I’ve now done “Dade+1” (3x 2.5 minutes at 120%, x3) with my new FTP (3 days after Washington). I just barely made it to the end of the second interval without dropping power though, so dialled it back to 97%, which got me through the next two intervals. I felt the need to dial it back a little further (95%) and that was just enough to make it to the very end (revs dropped by about 15rpm in the last interval, but kept the power going, barely).
I’m going to dial my ftp back to 97% and then increase it mid-session when I feel I can, rather than the demotivating alternative.
Thanks all, for the info and suggestions.