Five official tariff sheets — three in Missouri, one in Indiana, one in Michigan — run through the same load factor give an annual effective price with a median of 64.0 $/MWh. The low is I&M Indiana at 56.8; the high is I&M Michigan at 76.7: one company, two neighbouring states, 19.9 $/MWh apart. The rule was registered before the number: the first reading is the unweighted median of all full-level cells.
Updated 2026-11-12 (sixth cell): Xcel Minnesota's A97 rate sheet is now verified as full level, giving an annual effective price of 54.8 $/MWh (60.1 summer / 52.2 winter; transmission level; on-peak hours are 34.83% of the year). Run it through the same pre-registered rule and the reading holds at 64.0 $/MWh. The ranges in the body below (56.8-76.7) and the two-part split keep the 2026-11-11 five-cell basis; the current six-cell table is on the DCPI index page.
| State / utility / tariff | Energy summer / winter (¢/kWh) | Demand summer / winter ($/kW-mo) | Seasonal ($/MWh) | Annual |
|---|---|---|---|---|
| MO · Ameren Missouri 11(M) Large Primary Service · effective 2025-06-01 | 4.06 / 3.71 | 23.90 / 10.63 | 81.5 / 55.3 | 64.0 |
| MO · Evergy Metro Schedule LLPS, Sheet 77 · effective 2026-03-09 | 2.99 / 2.99 | 24.04 / 22.04 | 71.0 / 67.6 | 68.7 |
| MO · Evergy Missouri West Schedule LLPS, Sheet 187 · effective 2026-03-09 | 2.88 / 2.88 | 21.89 / 19.89 | 66.3 / 62.9 | 64.0 |
| IN · I&M (AEP) Tariff I.P., transmission level · IURC No. 20 · effective 2025-02-19 | 3.93 / 3.93 | 10.19 / 10.19 | 56.8 / 56.8 | 56.8 |
| MI · I&M (AEP) Tariff LP, transmission level · M.P.S.C. No. 18 · effective 2024-07-15 | 3.63 / 3.63 | 23.59 / 23.59 | 76.7 / 76.7 | 76.7 |
Source-sheet links and the demand build-ups (grid charges, blended two-tier TOU energy) sit on the DCPI index page and the machine-readable endpoint /api/v1/dcpi/latest.json.
The demand term is divided by load factor × 730 because demand charges are billed on peak kW per month while the index needs a cost per MWh. The lower the load factor, the more that fixed demand charge costs per kWh — so 80% is a stated basis, not a sector average. Change the load factor and the reading changes with it.
| Tier | Meaning | Examples | In the reading? |
|---|---|---|---|
| full | The official sheet discloses both energy and demand components; every line is recomputable | Ameren 11(M), Evergy LLPS ×2, I&M I.P. / LP | yes |
| partial | Only some components disclosed (e.g. demand-only, or delivery-only); no complete effective price | Dominion GS-5 (demand only), ComEd Extra Large Load (delivery only) | no |
| formula | Priced per deal under a trade-secret formula; no verifiable numbers in the rate book | Otter Tail (Minnesota) | no |
| confidential | Amounts redacted or bound by confidentiality agreements | Entergy special contract, Xcel Minnesota E002/M-26-170 | no |
This is the one-line reason the index exists: the terms layer is fully public, the price layer is almost entirely not. PUC dockets are public by statute, yet the entries that actually carry a price are either redacted wholesale as trade secrets or disclose a single demand component. The reading covers states whose tariffs can be verified, not the whole country — that sentence travels with every citation.
| Tariff | ¢/kWh | vs this site's median (6.40 ¢) |
|---|---|---|
| Oncor transmission service (TX) | 5.44 | −0.96 |
| Dominion GS-4 (VA) | 6.54 | +0.14 |
| Dominion GS-5 (VA) | 7.96 | +1.56 |
| ComEd Extra Large Load (IL) | 8.86 | +2.46 |
| Georgia Power PLL-18 (GA) | 9.43 | +3.03 |
E3's model includes customer, demand, energy and rider charges with E3's own load shape and site assumptions — it must not be conflated with this site's line-by-line recomputation from official sheets. It is here for one purpose: a cross-state order-of-magnitude reference. The same 25 MW load can cost nearly twice as much.
Riders (fuel adjustment, efficiency, renewable standards), reactive charges, customer charges and taxes are excluded — the all-in bill is higher. 64.0 is a comparable floor, not a final cost.
Real data-centre load factors run roughly 70-90%. 80% gives every tariff a common denominator; changing it moves both the energy and demand terms, so it lives in the methodology rather than in a footnote.
Negotiated PPAs, redacted contracts and commercial quotes are out of scope. This page does not forecast power prices and does not judge any state's or company's rate policy — it puts five public tariffs over one denominator.
Data endpoint: GET /api/v1/dcpi/latest.json · tariff ledger: data/dcpi/tariffs.json (source_url and docket per cell)