The Carrington Event

If global warming plus the possibility of a major asteroid strike aren't enough to worry about, there's another threat to our civilisation reported in the New Scientist magazine of 23rd March 2009. No doubt some SF writers and film-makers are beavering away at disaster stories based on this already – or maybe this one is a bit too grim.

It concerns storms on the surface of the sun which throw out plasma balls, a process known as coronal mass ejection (CME). These fly through space at high velocity and occasionally connect with the Earth. They vary a lot in size and small ones are common, but if a giant one hits the Earth, we are in trouble deep. This is no idle threat; the outcome of such a major geomagnetic storm is described in a report funded by NASA and issued by the US National Academy of Sciences (NAS) in January 2009. Furthermore, such an event has already happened, in 1859, as reported by the British astronomer Richard Carrington. This caused stunning auroras even at equatorial latitudes and severely disrupted telegraph networks. The consequences of an event on a similar scale today would be far worse.

Our problem is that in the 150 years since the Carrington event we have become much more vulnerable to its effects. Satellite communication and navigation systems in the CME's path would be fried. Much worse, the long cable lines of our electricity grids would act as aerials, capturing the plasma and focusing it on the transformers which convert the high-voltage grid supply to lower voltage domestic supplies. The massive flow of DC current would overheat and melt the transformers' copper wiring, effectively destroying them. Power supplies in the area hit by the plasma ball would fail. A small-scale version of this happened in Quebec province in March 1989, and six million people were without electricity for nine hours. A strike the size of the Carrington event would be orders of magnitude worse.

The NAS report outlines the consequences if a Carrington event hit the USA. Within 90 seconds, 300 key transformers would be knocked out, cutting off power to 130 million people. All kinds of electronic communications would fail. Within a few hours, water taps would run dry as there would be no power to pump the supply. All electrically-powered transport would grind to a halt. So would petrol and diesel vehicles as their tanks ran dry, because there would be no power to pump fuel at the filling stations. With no transport, supplies of food in urban areas would rapidly run out; typically, cities only have about three days' supply of food (and much of that is in freezers or refrigerators, so would soon spoil). Even establishments with backup generators, such as hospitals, could only keep going for as long as their fuel lasted – probably three days. Medicines would soon begin to run out, as the factories would have no power to make them and the vehicles no fuel to transport them.

Worst of all, it would take a very long time to put matters right. The wrecked transformers would have to be replaced, a job which takes a skilled crew at least a week for each one – assuming they have a spare one handy. There are very few spare ones lying around; they are usually made to order, a process which can take a year. And the factories which can make them will probably have no power – or, if they are outside the affected zone, problems in transporting them to where they're needed. Even with the transformers repaired, there would be a kind of Catch-22 because almost all the natural gas and fuel pipelines which supply power stations require electricity to operate. No electricity = no fuel = no electricity. Coal fired power stations may have 30 days of fuel, but nuclear ones would automatically shut down when the grid fails.

Given the difficulties and delays in responding to and recovering from Hurricane Katrina, an event which affected only a very small percentage of the USA, it is easy to see that rescue and recovery services would be completely overwhelmed by a national disaster on such a scale. The net result of all this, according to the NAS report, is that the recovery time would be four to ten years – and the USA may never be the same again. The New Scientist article quotes an estimate of the death toll of "tens of millions of lives". The rest of the developed world is just as vulnerable to a major geomagnetic storm as the USA. Ironically, it is the poorest and most rural societies which would be least affected.

Can anything be done to guard against this? Precautions to protect the grid could be taken given enough warning, such as adjusting voltages and loads and restricting energy transfers. However, this process takes at least 15 minutes – which is about as long as it can take for a CME to reach Earth from the nearest existing solar satellite. Fortunately, a follow-up report in the 11th April issue of the New Scientist describes a new technique for predicting CMEs using NASA's STEREO (Solar Terrestrial Relations Observatory) pair of spacecraft which follow the same orbit as the Earth. A check following a small CME event of 16 December 2008 discovered that STEREO spotted changes to the sun which presaged the event. Given some improvements to the software to speed up data analysis, up to 24 hours warning could be provided of a CME about to head our way.

Let's hope that someone is working on that software and setting up a system for automatic warnings to be sent to power grid organisations worldwide, and that those organisations are compelled to put in place and rehearse the precautionary procedures. This can be done quite easily and cheaply, and the consequences of failing to do so could be catastrophic.