How long would the grid be down after an EMP or solar storm?
What did EPRI find?
EPRI, the Electric Power Research Institute, studied the effect of a high-altitude pulse on the transmission system. Its journal summary (2019) says a moderate E1 could disrupt or damage about 5% of transmission-terminal digital relays, and a severe E1 about 15%. EPRI says the findings do not support nationwide blackouts lasting years.
That is a limited finding. EPRI says power plants, distribution systems and renewables were not covered and still need study. We read EPRI's own summary of the work, not the full report.
Why do transformers matter?
Large power transformers are slow to replace. The US Department of Energy's 2024 Large Power Transformer Resilience Report says:
- Lead times of 36 months are commonly quoted, and up to 60 months at the maximum. Before COVID they were under a year.
- In 2019, 82% of large power transformers (617 of 754) were imported.
- Sandia tests of E1 on transformer winding insulation found it robust.
- E2 is typically ignored for these transformers.
- E3 is similar to a geomagnetic disturbance but can produce larger fields.
Long lead times matter only if many large transformers are damaged at once. How many would be damaged by a solar storm or a pulse is not stated in these sources. NOAA says only that at G5 transformers may experience damage.
What range does NOAA cite?
NOAA's satellite service, in its article on solar storms and infrastructure, cites a National Academies estimate of nearly $2 trillion in costs and 20 to 40 million Americans facing outages of 16 days to 1 to 2 years. The NOAA page does not say how those figures were derived, so treat the range as a stated estimate and not as a forecast. We did not open the National Academies report itself.
For scale, NOAA also reports the March 1989 Quebec storm left six million people without power for nine hours.
How the sources compare
| Source | Event | What it says | What it leaves out |
|---|---|---|---|
| EPRI (2019) | Nuclear high-altitude EMP | About 5% (moderate) to 15% (severe) of transmission-terminal digital relays affected by E1. Does not support multi-year nationwide blackouts | Power plants, distribution systems, renewables |
| DOE (2024) | Neither; lead times apply to any cause of transformer loss | Transformer lead times of 36 months commonly quoted, up to 60. 82% of large transformers imported | How many would fail in any event |
| NOAA citing National Academies | Not stated on the NOAA page, which is about solar storms | 16 days to 1 to 2 years for 20 to 40 million people | How the range was derived |
| NOAA (1989 Quebec) | Solar storm | 6 million people, 9 hours | Not a nuclear event, not a worst case |
These numbers measure different things, for different events, and cannot be combined into one answer.
What we do not know
- Unknown: how long any real outage would last. No source we read gives a tested figure.
- Unknown: how EPRI's relay percentages translate into restoration time. EPRI's finding is about not supporting multi-year blackouts, not a day count.
- Unknown: how many large transformers would be damaged, which determines whether the DOE lead times matter.
- Unknown: whether the National Academies range rests on solar or nuclear assumptions. NOAA's page does not say.
- Different events: a solar storm and a nuclear pulse are not the same, so use EMP vs. solar storm before borrowing one event's evidence for the other.
What to do about it
Since the length is unknown, plan in steps, cheapest first.
- Water. The CDC advises at least a 3-day supply and, if possible, 2 weeks. Ready.gov says several days and gives no number. See how much water to store per person.
- If you are on a well, read well water without power.
- Learn how to make water safe without power.
- Food. Ready.gov says a refrigerator holds about 4 hours and a full freezer about 48 hours; discard food at 40 F or above for 2 hours or more.
- Decide on power for the specific things you need: backup power: what to look for. If you have a generator, read how to protect a generator from EMP and run it outdoors, at least 20 feet from the house, with carbon monoxide detectors.
- Before spending on whole-house measures, read whole-house EMP protection: is it worth it?
Sources
- EPRI Journal, A Deliberate Approach to Electromagnetic Pulses, 2019. eprijournal.com
- US Department of Energy, Large Power Transformer Resilience Report, July 2024. energy.gov
- NOAA NESDIS, When Solar Storms Attack: Space Weather and Our Infrastructure. nesdis.noaa.gov
- NOAA Space Weather Prediction Center, Space Weather Scales. spaceweather.gov
- Ready.gov, Power Outages. ready.gov
- Ready.gov, Water. ready.gov
- CDC, How to create and store an emergency water supply. cdc.gov