React FAQ
← All metros

How many nuclear reactors would it take to power Detroit?

The Detroit–Warren–Dearborn, MI metro area is home to 4,390,913 people. Below: scenario estimates of how many reactors of various sizes would be required to cover its annual electricity demand — and what it would take with other power sources.

Population
4,390,913
Annual electricity demand
43,406,017 MWh
Primary state
MI
Today's grid mix
eGRID 2023 subregion RFCM — RFC Michigan
50% Natural gas
largest source · 2023
Natural gas 49.8%
Coal 25.7%
Nuclear 10.2%
Wind 8.8%
Other 3.1%
Oil 1.5%
Solar 1.3%

Replacing the fossil portion (77% of generation) with nuclear would avoid roughly 16,220,004–25,202,337 tons of CO₂ per year for this metro's share of demand. Range uses EPA eGRID 2023 Total Output rate (low) and Non-baseload rate (high) for RFCM. See methodology.

Annual electricity demand
MI state total — all sectors
-2.2%
2001 → 2025
050M100M150M200M200120052010201520202025

Reactors needed

Each row shows how many of one reactor type would cover annual demand. Icons scale with count — fewer big reactors, many more small ones.

Advanced Microreactor (10 MWe representative)
10 MWe per unit · 90% capacity factor
551
reactors needed
+ 95 more (visualization capped at 456 icons)
Small SMR Module (77 MWe, NuScale-class)
77 MWe per unit · 93% capacity factor
70
reactors needed
Large SMR (300 MWe, BWRX-300 class)
300 MWe per unit · 93% capacity factor
18
reactors needed
Large Reactor (1,117 MWe, AP1000 class)
1,117 MWe per unit · 93% capacity factor
5
reactors needed

Or with other power sources

Counts assume same annual MWh. Variable sources (solar, wind) would need additional storage or backup to provide firm equivalent power — disclosed per card. Land use is typical farm/plant footprint; CO₂ is operational rate × annual demand.

Nuclear · Large
Large nuclear reactor
Raw count (same annual MWh)
5
units needed
Land
6.5 sq mi
Cap. factor
93%
CO₂
Zero

Firm baseload. Long build times, high upfront cost.

Nuclear · SMR
Small SMR module
Raw count (same annual MWh)
70
units needed
Land
3.8 sq mi
Cap. factor
93%
CO₂
Zero

Firm baseload. Pre-commercial in the U.S.

Solar
Utility solar farm (100 MW DC)
Raw count (same annual MWh)
199
units needed ≈ 49.8M panels (≈400W each)
Firm equivalent (with storage)
498–796
units needed for 24/7 firm power ≈ 124.5M–199.0M panels (≈400W each)
Land
187 sq mi
Cap. factor
25%
CO₂
Zero

Variable output. Raw count assumes same annual MWh; firm equivalent accounts for storage and oversizing needed for 24/7 power.

Wind
Onshore wind turbine (3 MW)
Raw count (same annual MWh)
5,720
units needed
Firm equivalent (with storage)
8,496–14,160
units needed for 24/7 firm power
Land
738 sq mi
Cap. factor
35%
CO₂
Zero

Variable output. Raw count assumes same annual MWh; firm equivalent accounts for storage and geographic diversification needed for reliable power.

Natural gas
Natural gas combined cycle (600 MW)
Raw count (same annual MWh)
16
units needed
Land
480 acres
Cap. factor
55%
CO₂
17.4M tons/yr

Firm dispatchable. Emits ~800 lb CO₂/MWh.

Coal
Coal-fired plant (600 MW)
Raw count (same annual MWh)
17
units needed
Land
5.3 sq mi
Cap. factor
50%
CO₂
47.7M tons/yr

Firm. Highest CO₂ rate; significant air pollutants.

Capital cost bands

Wide bands reflect first-of-a-kind delivery uncertainty. Mid value from NREL ATB 2024; high band reflects delivered Vogtle / NuScale history. See methodology.

Advanced Microreactor (10 MWe representative)
microreactor · 551 units
$82.7B$121.2B$165.3B
Small SMR Module (77 MWe, NuScale-class)
smr · 70 units
$48.5B$80.8B$118.6B
Large SMR (300 MWe, BWRX-300 class)
smr · 18 units
$32.4B$43.2B$64.8B
Large Reactor (1,117 MWe, AP1000 class)
large reactor · 5 units
$39.1B$55.9B$83.8B

Data provenance

Demand basis
MI state per-capita × metro population (v2 methodology, 2025 basis year)
Demand period
historical state trend: 2001–2025
Population source
U.S. Census Vintage 2025 estimates
Reactor cost basis
NREL ATB 2024, with widened bands for FOAK uncertainty
Grid mix source
EPA eGRID 2023 (subregion RFCM)
Source comparisons
Hardcoded archetypes — see methodology page

These figures are screening-level scenario estimates. They are not forecasts, project proposals, or permitting determinations.