Using the AIS Step Test Protocol
I estimated that my current 2k time would be around 7:00, or a 1:45 pace. The protocol provides the following step powers

Getting myself setup was a bit complicated.
- HR sensor, connected to PM5
- PM5 connected to painsled
- headphones on, connected to iphone
- Spotify set to my “8×500” playlist
- Drag set to 130
- Seat pad on erg
- wet towel
- dry towel
- lancet tool
- lactate meter battery OK
- lay out 10 lactate test strips neatly
- fan on
- PM set to 4 minute intervals/1 minute rests, display to power
Then I did a baseline lactate test, just to make sure that I remembered the process.
- wipe hands with dry towel
- wipe sample finger with wet towel
- dry sample finger
- lance fingertip
- pick up meter with non-sample hand
- pick up strip with sample hand and insert strip
- make sure meter has woken up
- visually judge size of droplet
- If it is too small, put hand down low for 5 seconds
- dab droplet with strip, starts 13 sec count down on meter
- wipe finger
- get a drink
- get feet set for next interval
- write down reading
When executed well, this whole process takes about 45 seconds. If there are any hiccups, it takes the whole 1 minute rest time.
The baseline was 1.4mmol/ml which is a perfectly reasonable reading.
Then I did the test. The steps start ridiculously easy. The key thing in my mind is to try to hit the power exactly on every stroke. I ended up bouncing a few watts up and down, but staying focused on the target kept me occupied while it was easy.
The target for the last step is “max”. I changed the display from watts to pace per 500. I wasn’t quite sure how hard I could hold for 4 minutes. I decided to try to make sure I did it faster than 1:45. As it turned out, I was able to accelerate in the last minute, so I probably could have pushed it a bit harder, but my HR was above 180 at the end so I was working pretty hard.



OK, so I did the test. What’s the point?
The main reason to do the test was to have a baseline to compare after training blocks are completed.
Beyond that, the curve can be used to estimate my current Lactate Threshold. There are multiple ways to extract this from the curve. The simplest is the power at which the lactate level exceeds 4.0mmol/ml. This gives an estimate of 243W (1:52.9). A more conservative method is to draw a best fit line for the sub-2.0mmol points and another for the steep part of the curve, and the point where they meet is the LT.

This method yields about 238. So, about 240 is a good guess. How does this compare with the 20′ test? The test measured 244W. So there!
The other useful output is a first cut at appropriate steady state training power. The goal here is to train at the highest power that results in a steady state lactate level at or below 2.0mmol/ml. From this test, training power should be between 180 and 205W.
Other testing I have done in the past shows that lactate tends to drift up slowly over a long steady state session, so something towards the bottom of this probably more appropriate.