WEC Energy Group furnished its September investor deck under Regulation FD at 20:15 UTC on 4 September. Its headline is a weather-normalised forecast to "add 3.9 GW (~45%) of electric demand" between 2026 and 2030, with Wisconsin segment electric sales growth of 6.0%–8.0% in 2028–2030. The deck also shows, on its own customer slides, exactly where all of it comes from.

Two customers, and the arithmetic closes

Microsoft at Mount Pleasant: "2.6 GW of Demand Forecasted through 2030", against announced investment of "$20+ billion" — $7.3bn for phases 1 and 2, $13bn additional, 15 more data centres beyond phase 2, more than 2,200 acres, phase 1 operating since April 2026. Vantage at Port Washington, the Stargate site: "1.3 GW of Demand Forecasted through 2030", four buildings under construction on 670 acres, "$15+ billion" expected. 2.6 plus 1.3 is 3.9. There is no third contributor.

The marketed gigawatts and the booked ones

The same Vantage slide carries both numbers: "Site potential up to 3.5 GW over time" across approximately 1,900 acres — and the 1.3 GW the utility that would have to serve it actually forecasts through 2030. That is the clearest public gap this year between a developer's marketed capacity and the serving utility's plan for the same address, and both figures are on one slide.

What is signed and what is not

Under "What's New?", the deck says WEC "entered into an agreement with NextEra to extend the Point Beach PPA for ~1GW of carbon-free capacity", Unit 1 from October 2030 to 2050 and Unit 2 from March 2033 to 2053 — and, in the next line, that "the Company is seeking PSCW approval of the PPA". Signed, not approved. Wisconsin's Very Large Customer tariff, ordered by the PSCW on 21 May 2026, applies from 100 MW of forecast new load, with a fixed 10.48%–10.98% ROE band and early termination at remaining net book value. Total capital plan: $37.5bn, 2026–2030.

What the received framing gets wrong

"45% demand growth" reads as regional economic expansion driven by AI. It is two hyperscale campuses on one corridor, which is a concentration risk rather than a growth story: if either customer slips, the forecast does not degrade gracefully, it halves. The deck itself notes Oracle "has met the financial security requirements" for the Vantage site — an acknowledgement that the risk is real enough to collateralise.