Eli5 how could an electromagnetic impulse, say from the sun for example, disable a whole power grid?

228 views

Eli5 how could an electromagnetic impulse, say from the sun for example, disable a whole power grid?

In: Physics

Electromagnetic radiation is made up of photons. There’s photons interact with pieces of metal in electrical components. These pieces of metal convert the electromagnetic radiation to electrical energy, same way an antenna works. There is too much energy for the electronic components inside the devices, the smoke escapes and you can’t put the smoke back in. If these components as the vital to the power grid, like transformers, no more power grid

Lets say you have a high voltage transmission line going between a power station and a city. These can be hundreds and even thousands of miles long. So if you get a powerful electromagnetic wave hitting it then it will act as an antenna and the wave will create a current in the line. Since the line is so long you can get a huge current in it from this electromagnetic wave and more then likely some part of the line will not handle the current and melt. The power grid is made up of these transmission lines all connected together.

But even if these transmission lines will fail, hopefully it will just blow all the fuses, it will be an easy fix that would take no more then a few days for linesmen to go around fixing all the issues. The real problem is with the more sensitive equipment like switches, transformers and generators. In addition to being much more complex and therefore much harder to fix they also tend to be enclosed in an oil bath to prevent short circuits in the extremely high voltages they operate in and also to cool the equipment better then air. So not only do you need to unwind a huge coil to get to the part of the wire that failed and then wind the coil again but most of the coil will not be there any more as the oil have burned it away. If you get any of these high current events in just a fraction of the transformers it will cause huge issues for the grid as you can not transfer power from the power stations to the cities any more. And it is not like you can just order a new transformer station to be delivered the next month because there are very few factories making them in the world as the demand for new power grids is relatively low.

The current plan for dealing with such an event is to be able to detect it early using space probes that have been installed between the Sun and Earth. When we have detected a wave of charged particles from the Sun strong enough to damage our power grids we will shut it all down. Turn off all the power stations and disconnect all the switches. It will probably require linesmen going out and manually disconnect huge disconnected or even cutting some of the lines intentionally. This means that when the electromagnetic pulse arrives we will not have one huge grid but only small pieces of lose wires around and these will not build up as much current as the longer lines would. Most importantly sensitive equipment like power stations and transformers will not be connected to the grid.

The most vulnerable item would be the transformers that are used to step up the voltage from the generator voltages to transmission voltages for delivery to the loads, where they are stepped down again. These units are huge and lead times for buying them are can be in terms of years. Many utilities will have spares for critical locations (imagine having a multimillion dollar “spare” sitting there idle for years on end) but even then, there is considerable time involved in making the changes. Repairs usually involve removing the unit back to the factory for de-tanking and complete rewinds, although some could be done on site depending on severity. Even then, they can be time consuming.

A transformer is designed for a specific frequency and voltage. When these change from the induced currents of a magnetic pulse, the impedance of the unit changes drastically resulting in overheating, insulation failure from voltage spikes, core saturation, etc.

Just to add another aspect to the other good explainations:

As well as an electromagnetic pulse causing damage to various individual bits of equipment in the grid, one of the biggest problems is that it affects a wide area all at once.

Most grids are designed so that if any single piece of equipment fails, there will be a short blackout when the broken equipment is automatically disconnected. A blackout is how the grid protects itself from being further damaged by the faulty equipment (by making too much electricity flow through it, for example). Some grids are really clever and can very quickly restore power by sending it through a different route that avoids the broken equipment.

But, if there’s suddenly lots of equipment everywhere that’s broken simultaneously, that process doesn’t work as well. For example, the equipment that automatically disconnects failed equipment (a circuit breaker) might also itself be broken. This can lead to further damage to the transmission equipment that sends the power around the grid.

As an engineer, you can design systems to be able to handle failures and damage and still continue to work (often called ‘resillience’). The more failures and damage your system can deal with, the more difficult and expensive it becomes to build it. You have to decide that the system will be able to deal with a certain number of failures at once before admitting that it will stop working – if you don’t, the system becomes prohibitively complex and expensive and is not seen as good value – and probably won’t be built.

When trying to figure out the balance of how much failure to tolerate, you look at the risks to your system and how often they occur.

Individual failures happen quite routinely, and potentially cause a lot of damage if not handled. Over a reasonable amount of time, the amount of money you spend on protecting from damage is less than the damage that would have been caused had you not prevented it – making it easy to justify.

A big electromagnetic pulse from a solar flare that has a noticable effect on the surface is a very rare event – and there are relatively few examples of it occurring to study to understand the system wide effects. So it’s hard to know exactly how much damage it would cause, and what we can do to properly protect the system from it, and it’s also hard to know whether it’s likely to happen. That makes it difficult to balance how much you protect from it, compared to how much damage would occur if you didn’t. Often, this leads to engineers (and more importantly their bosses) accepting that if it is to occur, they’ll build in some ways to limit the damage such an event would cause (i.e. ‘damage limitation’), but they won’t try to protect against it completely, because it’s just too expensive and complicated to do so.

Which leads to the outcome that if such a rare system-wide event were to occur, the likely outcome is that the grid will be non-functional for a significant amount of time. The better the design is, the quicker they’ll come back online again.

Top response is incorrect.

The issue for transmission is not the directly induced current from the radiation to the line. That’s only an issue for small PCBs with undersized traces, but even then it’s the secondary concern. The issue is that the rapid changes in the earth’s magnetic field will induce a voltage in the ground itself across several miles and can last several minutes or longer. This isn’t high enough to be problematic for humans, but if you have equipment that depends on having a common reference point (literally “ground”) and they suddenly don’t, then you can have what are called Ground Induced Currents, which will take the path of lowest resistance, which means it will go from the ground into the equipment and into the system, which is literally designed to have as low a resistance as possible to save energy and money (resistance means heat, and heat is energy bought that is not sold.) Transmission lines aren’t usually (edit: earth) grounded (look up delta vs wye if you want) and can handle this increased current with more or less no issues. Transformers are the real problem because of what’s called core saturation.

Transformers work by magnetic Flux in = magnetic Flux out. Magnetic Flux is based on the product of the number of turns, voltage, and current on each side. If you are running your transformers at 80 or 90% of nominal and suddenly increase the current on one side, you’ll leave what’s called the “linear region” of behavior and enter the “saturation region” (read: non-linear) and the core will heat up very quickly, which is bad. Transformers monitor their temperature and if it gets too hot, they’ll force out of service to stop damage. Lines do this too, but like I said, they’re not grounded, and their temperature curves are more generous.

If multiple transformers do this at once, or if another disturbance happens at the same time, it can lead to what’s called a “cascading condition” which is the thing you see in movies where its multiple events bang bang bang bang, cities blacking out one block at a time, switches operating one after another, all trying to protect themselves (overload trips itself, which causes another overload, which trips to protect itself which causes another overload.) If the operators don’t pre-posture their system for that eventuality before it starts and the system isn’t designed to handle the transient behavior, then that cascading condition can black out either the entire system or a large part of it. Check out the 2003 northeast blackout to see what cascading looks like. I believe California had one as well in 2008 ish.

Source: monitoring GICs are in my job description. If you wanna look up more stuff, the NERC term for this event is a Geomagnetic Disturbance (GMD)

Power plants make electricity by quickly moving wires through a magnetic field (or the field through the coils).

Long distance transmission lines are very long wires, and solar storm/EMP is a very big moving magnetic field.

Thus, it creates electricity in places that can’t handle (that form/amount of) it.

Electro-magnetic pulses can induce voltage spikes in ungrounded/poorly shielded circuitry. These voltages can fry the junctions in transistors, killing the device. This is one of the problems during a nuclear war….all the modern circuitry would die as the electromotive forces were produced during the bomb explosion.