The 300 MW enrolled vs 34.3 MW measured gap is the same pattern I keep seeing in data-center power announcements — headline capacity vs what is actually scheduled to be delivered. Generac's Amazon deal was announced as $8B, but the SEC filing schedules only $2.4B of deliveries across 2027-28; the rest is a ceiling, with warrants vesting as Amazon actually spends. Your Enersponse case makes the same point from the grid side: the number that matters is the measured, dispatched one, not the enrolled one. Is there a standard disclosure you'd want — measured average-event-hour alongside every enrolled MW claim — or would that just get gamed into a different headline?
The Generac comparison is useful, particularly the distinction between expected deliveries and the much larger ceiling governing warrant vesting.
One correction on Enersponse: the 300+ MW was underlying customer electricity load, not enrolled flexibility capacity. The 34.3 MW was estimated average response across dispatched event hours, not a simultaneous test of the entire portfolio.
I wouldn't mandate one measured average alongside every enrolment claim. That could be gamed through event selection and baselines, and would penalize contingency capacity that was available but never called.
I'd require a standard, service-specific disclosure distinguishing underlying load, connected and controllable capacity, enrolled and qualified MW, available MW during defined windows, activated MW, and verified response against actual instructions. Include the measurement method, duration, delivery distribution and repeatability evidence.
The critical distinction is between capacity that was unavailable, capacity that wasn't needed, and capacity that failed when called. A single average obscures all three.
You're right on both numbers, and the correction matters: 300+ MW was the underlying customer load behind Enersponse, not enrolled flexibility, and the 34.3 MW was an estimated average response across dispatched hours — not a simultaneous test of the whole portfolio. I compressed both into a headline-versus-measured contrast the figures don't quite carry.
It actually sharpens the point your piece is making. The distance between load, enrolled capacity, and verified response is exactly where these markets get oversold — Generac's $8B headline against $2.4B actually scheduled for 2027–28 is the same shape in a different market: the biggest number is the one that travels.
On the disclosure standard, I take your objection to a single mandated average — event selection and baselines make it gameable, and it punishes capacity that was available but never called. Your split (unavailable / not needed / failed when called) is the distinction that matters. One question back: for dispatched-hour averages like the 34.3, should the disclosure also have to state how many hours were dispatched against how many were available? That's where 'not needed' and 'not capable' can still hide inside each other.
The 300 MW enrolled vs 34.3 MW measured gap is the same pattern I keep seeing in data-center power announcements — headline capacity vs what is actually scheduled to be delivered. Generac's Amazon deal was announced as $8B, but the SEC filing schedules only $2.4B of deliveries across 2027-28; the rest is a ceiling, with warrants vesting as Amazon actually spends. Your Enersponse case makes the same point from the grid side: the number that matters is the measured, dispatched one, not the enrolled one. Is there a standard disclosure you'd want — measured average-event-hour alongside every enrolled MW claim — or would that just get gamed into a different headline?
The Generac comparison is useful, particularly the distinction between expected deliveries and the much larger ceiling governing warrant vesting.
One correction on Enersponse: the 300+ MW was underlying customer electricity load, not enrolled flexibility capacity. The 34.3 MW was estimated average response across dispatched event hours, not a simultaneous test of the entire portfolio.
I wouldn't mandate one measured average alongside every enrolment claim. That could be gamed through event selection and baselines, and would penalize contingency capacity that was available but never called.
I'd require a standard, service-specific disclosure distinguishing underlying load, connected and controllable capacity, enrolled and qualified MW, available MW during defined windows, activated MW, and verified response against actual instructions. Include the measurement method, duration, delivery distribution and repeatability evidence.
The critical distinction is between capacity that was unavailable, capacity that wasn't needed, and capacity that failed when called. A single average obscures all three.
You're right on both numbers, and the correction matters: 300+ MW was the underlying customer load behind Enersponse, not enrolled flexibility, and the 34.3 MW was an estimated average response across dispatched hours — not a simultaneous test of the whole portfolio. I compressed both into a headline-versus-measured contrast the figures don't quite carry.
It actually sharpens the point your piece is making. The distance between load, enrolled capacity, and verified response is exactly where these markets get oversold — Generac's $8B headline against $2.4B actually scheduled for 2027–28 is the same shape in a different market: the biggest number is the one that travels.
On the disclosure standard, I take your objection to a single mandated average — event selection and baselines make it gameable, and it punishes capacity that was available but never called. Your split (unavailable / not needed / failed when called) is the distinction that matters. One question back: for dispatched-hour averages like the 34.3, should the disclosure also have to state how many hours were dispatched against how many were available? That's where 'not needed' and 'not capable' can still hide inside each other.