Anaerobic Capacity (W'): Why Your 30-Second Power Matters for a 4-Hour Race

You blew up 80 km into a 90 km bike leg. Pace held. Power looked fine. Then one punchy climb at 120% of your critical power and your legs just stopped talking to you. That’s W’ (W prime), your finite anaerobic reserve, and you spent it without knowing you had a budget.

Most cyclists know their FTP. Far fewer know their W’. That gap costs time on race day.

What Anaerobic Capacity (W’) Actually Is

Think of W’ as a battery pack that only runs above your critical power (CP), the highest power you can sustain aerobically for a long effort.

Below CP, your aerobic system can meet the demand. You can hold that power more or less indefinitely (glycogen aside). Above CP, your body draws on a finite reserve to make up the gap. That reserve is W’. Drain it to zero and you stop. Not because you’re mentally weak, but because the chemistry locks up.

Trained male endurance cyclists average 12.7 kJ of W’, compared to 9.6 kJ for untrained riders (Chorley et al. 2020, n=20). Punchy criterium riders can exceed 20–22 kJ. Ironman specialists often sit at 9–13 kJ, and that’s fine. More on why shortly.

The depletion formula is clean:

T_lim (seconds to exhaustion) = W' / (Power − CP)

Translation: if your CP is 280 W and W’ is 11 kJ, riding at 308 W (+10% above CP) exhausts your reserve in exactly 1,100 seconds (about 18 minutes). Ride at 336 W (+20% above CP) and it’s gone in 366 seconds, under 6 minutes.

Every watt above CP burns through your battery faster. The math is unforgiving.

The CP–FTP Gap: Why the Wrong Threshold Breaks Your Zones

CP and FTP (Functional Threshold Power) sound interchangeable. They’re not quite.

A 2021 Frontiers in Physiology study (Karsten et al., n=21) found CP averaged 256 W vs. FTP 249 W, a mean gap of just 7 W. The correlation was strong (r = 0.969). But individual limits of agreement spanned −19 W to +33 W. For roughly 1 in 10 riders, the gap exceeded 20 W.

That matters for zone setting. If your CP is 20 W above your FTP, your “threshold” zone is miscalibrated by nearly a full zone. You’d be training in Zone 4 while thinking you’re in Zone 3.

For long-course pacing, use CP as your ceiling, not FTP. They might be close. They might not be. See how FTP is calculated and tested if you’re not sure which number you’re actually using.

How Long a Single Match Actually Burns

Here’s the depletion table for three real athlete profiles. All times are calculated from T_lim = W' / (P − CP).

The Ironman specialist has less W’ to burn, but they don’t need it. Their CP is high enough that they never cross the threshold in a paced race. The criterium rider carries more W’ but hits those surges far more often.

Your match budget is shaped by both numbers together.

W’ Reconstitution: The Battery That Recharges Slowly

Here’s what almost no pacing guide explains well. W’ doesn’t snap back the moment you ease up.

It recharges along an exponential curve. Chorley et al. (2020) found these recovery milestones at near-zero recovery power:

• 2 minutes: 37% recharged
• 6 minutes: 65% recharged
• 15 minutes: 86% recharged

Six minutes of easy spinning to get back 65% of your reserve. That’s a long time mid-race.

And it gets slower during a sub-CP cruise. The recovery time constant (τ) depends on how far below CP you ride. Skiba et al. (2012) showed:

τ = 546 × e^(−0.01 × D_CP) + 316

Where D_CP = CP minus your recovery power. At 20 W recovery (coasting), τ ≈ 350 s. At 250 W recovery for a CP-280 athlete (40 W below CP), τ climbs toward 580 s. You’re barely moving the needle.

Steady is faster than fast-then-slow. The math proves it.

The Bi-Exponential Problem: Each Match Costs More Than the Last

Chorley, Marwood & Lamb (2022, n=10) found something the standard Skiba model misses. W’ reconstitution isn’t one simple curve. It has two components:

• Fast phase: time constant ~21.5 seconds, handles about half the initial recovery
• Slow phase: time constant ~388 seconds, handles the other half

After a second exhausting effort with the same 2-minute recovery window, the slow-phase time constant degraded from 388 seconds to 716 seconds. That’s 85% slower than the first recovery.

The second match you burn is almost twice as expensive to recharge as the first.

The bi-exponential model fit the data with R² = 0.999. The simpler Skiba model: R² = 0.614.

This compounds across a 4-hour race. Three surges above CP and your recovery between them progressively lengthens. You arrive at the run already short-changed.

Why a Larger W’ Isn’t Always Better

This is the part most W’ articles get wrong.

High-intensity interval training reliably increases CP, but the change in CP is inversely correlated with the change in W’. As your aerobic threshold rises, your anaerobic reservoir often shrinks proportionally.

Jones and Vanhatalo (2017) put it plainly: “Endurance training consistently increases CP but also tends to reduce W’.”

For Ironman, that’s exactly what you want. The optimal long-course profile is:

• High CP (65–80% of which is your Ironman bike target)
• Modest W’ (enough for tactical surges, not a sprint race)

Jenkins & Quigley (1992) showed CP rising 31% (from 196 W to 255 W) after just 8 weeks of endurance training. That’s the adaptation to chase.

A big W’ with a low CP means you’re built for short punchy efforts. That’s a criterium profile, not a 5-hour bike leg.

W’ in Action: A Real Example

Take a rider I’ll call Kenji, 41, training for his second Ironman, CP 275 W, W’ 13 kJ, targeting a 5:15 bike split.

In his previous race, Kenji held an average of 195 W but let his power spike to 340 W (124% CP) twice on a climb, each surge lasting about 90 seconds. Using the depletion formula, each spike cost roughly 5.1 kJ of his 13 kJ reserve. Those two spikes burned 78% of his W’.

At 195 W race pace, D_CP = 275 − 195 = 80 W, giving τ ≈ 430 s. He had about 30–40 km of sub-CP riding before T2, which recovered roughly 75% of his W’. He arrived at the run with about 3 kJ left.

His marathon fell apart at kilometer 18. Not fitness. Not fueling. Match management.

In training, AthleteOS flagged his W’ balance dropping below 2 kJ on tempo rides, the same pattern showing up weeks before race day. He capped his hill surges to +8% above CP. His next race: 4:58 bike, the strongest marathon of his life.

How to Test Your CP and W’ (The 3-Minute All-Out Method)

Vanhatalo, Doust & Burnley (2007, n=10) validated a single 3-minute all-out test that measures both parameters in one ride.

The protocol:

1. Full warmup, then go absolutely all-out for exactly 3 minutes.
2. Your CP estimate = mean power over the final 30 seconds (EP). In the study, EP = 287 ± 55 W vs. independently measured CP = 287 ± 56 W (r = 0.99, no significant difference).
3. Your W’ = total work done above EP across the full 3 minutes.

Power stabilizes to a plateau around 135 seconds in. Everything above that line in the first two minutes is W’.

A multi-duration alternative: use best efforts at 3, 5, 10, and 20 minutes from your training history. Plot them on a power-duration curve and fit the hyperbolic model. AthleteOS pulls these automatically from your workout history so you don’t need to run a dedicated test.

What This Means for Anaerobic Capacity and Ironman Pacing

Race power targets by distance (Goeringer, Working Triathlete 2023):

• 70.3: 75–90% CP (mid-pack 80–83%)
• Ironman 140.6: 65–80% CP (mid-pack 70–72%)

At 70% CP, your aerobic system handles everything. W’ sits untouched, unless you surge.

That’s why Variability Index matters. VI above 1.05 consistently predicts slower run splits. Each power spike above CP is a W’ withdrawal, not just a momentary intensity bump.

Athletes exhausted in field conditions had W’bal averaging 0.5 ± 1.3 kJ (essentially zero) at the point of failure (Jones & Vanhatalo 2017). Athletes who finished successfully held a minimum of 3.6 ± 2.0 kJ in reserve.

Arrive at T2 with W’ in the tank. That’s pacing.

How AthleteOS Tracks Your Matches

AthleteOS computes your CP and W’ from your power-duration curve using multi-duration best efforts pulled automatically from your training history, then calculates W’ balance in real time during and after every workout. You can see exactly how many kilojoules of anaerobic reserve you spent, how long each sub-CP recovery block would take to recharge it, and whether your CP-to-W’ ratio fits your goal distance.

The session analysis view flags any ride where you crossed CP more than three times. That’s the pattern that shows up in athletes who consistently underperform their fitness score on race day.

It replaces the guesswork of “I think I went too hard on that climb.” The data already knows.

Learn how Zone 2 builds the aerobic base that drives CP upward, and why aerobic decoupling on long rides tells you whether your base is actually holding. If you’re unsure whether to target CP or FTP for threshold work, understanding what a good FTP looks like gives you the benchmark context first.

PCr (phosphocreatine) resynthesis has a time constant of about 57 seconds (Broxterman et al. 2016). W’ reconstitution takes roughly 334 seconds, six times longer. The limiting factor isn’t PCr — it’s the clearance of inorganic phosphate, hydrogen ions, and ADP, making W’ an integrated bioenergetic reserve rather than a purely “anaerobic” one.