Key Points You Need to Know When Talking About Energy: Alternating Current and Grid Instability

In this installment, I'll discuss a major element of grid instability that is not commonly understood, which is frequency instability.  Frequency is a much more critical aspect of the electric grid than most people realize.  To understand why, you need to understand roughly how AC power is transmitted.

AC Power transmission: No electricity is produced

One aspect in which the standard "water analogy" of electricity fails is that by this analogy, one would expect that power plants pump electrons into transmission wires, which then flow down to substations, are split up and then pour into homes when needed.  In fact, however, AC power generation results in a net-zero motion of electrons.  The alternating current does move electrons back and forth a bit, but there is no round trip of electrons moving around a circuit.  This is what the "alternating" part of "Alternating Current" means: charge flows one way very briefly (1/120th of a second in North America), but then flows right back the other direction, cancelling out the previous flow of charge.  The total electrical charge delivered from all power stations to the grid over time is exactly zero.

If power stations don't deliver electricity to the grid, how do they deliver power?  The answer is that they deliver cyclical changes to the electrical potential of the grid components.  You can think of this as a constant back-and-forth motion in the electrical field that can be harnessed by any electrical tool that keys off of this cyclical motion.  Incandescent light bulbs, for example, run simply off of resistance, which is electrical friction.  Just as when you are attempting to start a fire by rubbing two sticks together, the heat that is generated doesn't care what direction the sticks are going, so too you can hook an incandescent light bulb straight up to either an AC power source or a DC power source and it will work exactly the same way.  On the other hand, things like computers and cell phones rely on actual directional electricity flow because they make use of electrical logic gates, which do very much care about the direction of energy flow.  This is why such devices need the AC adaptor--that bulky "wall wart" to which the charge cord going into the phone or laptop; this is the electrical equivalent of a "worm gear," changing cyclical energy into linear energy.

My current favorite explanation of this phenomenon is this YouTube video by Veritasium: The Big Misconception About Electricity.  His analogy of AC power transmission as being a chain inside a tube being pushed back and forth is an excellent way of understanding things on a basic level (though he himself qualifies this analogy pretty severely in the above video.)  It's a cool video and I recommend you watch the whole thing, but for efficiency's sake I have linked to the precise moment in the video where he begins to talk about AC power transmission, and you only need to watch about 2 minutes of the video from that point.  (The details about wire transmission versus electrical field transmission are interesting from a physics perspective, but not really relevant to energy policy.)

AC Power transmission: No electricity is stored

To the above point, I will add also the fact that there is, currently, no real utility scale storage in the power grid.  This is no longer *strictly* true, as there are (as mentioned before) now "utility scale" battery storage plants in the world which do store energy and release it to the grid.  However, all such power storage plants combined have a negligible impact on the grid.  They currently amount to less than a rounding error in the total energy scheme of things.

One interesting consequence of this is that as you look at anything around you that is consuming electrical energy, you can know that (since electrical power travels over the grid at about the speed of light) the power that is lighting that lightbulb or driving that computer monitor was--less than a millisecond ago!--a scalding hot bit of steam pushing a steam turbine, or a photon hitting a solar cell, or a puff of wind pushing a wind turbine.  Electricity delivery from power plant to home happens quasi-immediately--that is, at the speed of light.

Frequency synchronization

So AC power transmission happens (basically) immediately and also cyclically; in North America, the back-and-forth of alternating current on the grid happens at 60 Hertz (cycles per second).  These two facts together means that all power plants putting energy onto the grid must be synchronized.  Every input into the grid *must* be at a frequency matched with the grid frequency.  As the electrical fields are going back and forth on the grid, if some power source attempted to add energy into the system but was pushing while everyone else was pulling, it would instead *remove* energy from the system: it would cancel out instead of adding.

Actually, "cancelling out" is not a good description of what would happen in such a case, since the actual result would be much more violent.  If you have ever driven a stick shift, you have probably at least once accidentally put the vehicle into reverse when you meant to put it into a forward gear.  Remember the gear grinding?  That's getting closer to what would happen if a power plant tried to put electricity onto a grid with the wrong frequency.  Except, that's not a violent enough image for what would happen.  There are *massive* amounts of energy involved here.  Let me come up with a better analogy:

Suppose you were to take a lawn mower, turn it on, and then keep it running while upside down.  Then if  you took another lawn mower, turn it on and put it on top of that lawn mower--imagine the wreck and metallic carnage that would ensue. 

Then imagine doing that, but instead of using lawn mowers, use two of those massive turbines that are in power plants:

Two turbines connected to the same grid but operating at different frequencies would result in enormous destruction.  Massive, expensive pieces of machinery connected to one or the other or both would tear themselves apart due to the conflicting magnetic forces that situation would create.  Explosions, sparks flying, massive crank shafts breaking apart . . . well, it would be bad, let's say that much.

This is why all utility scale equipment have trip-safeties built in.  We are familiar with *current* circuit breakers, which break the circuit in case too much electricity is flowing because of overload or a short circuit.  Utility equipment, on the other hand, have *frequency* circuit breakers.  In the event that a generator detects that it is producing electricity at a frequency too far off the grid frequency, the equipment will automatically trip, removing itself from the grid.

The relationship between frequency and power

There are a number of issues that can affect the frequency of a power plant, the most usual one being excessive load.  This is something that bicyclists will be familiar with, actually.  If you are cycling along a flat road, pedaling at a specific rate, but then hit an uphill slope, you will find it very difficult to maintain the same speed.  Normally, the rate at which you push the pedals around will naturally slow down.  The same thing happens to power plants when they experience higher than normal load: frequency goes down as power output is too low to meet demand.

While this is good from the standpoint of equipment safety, it does create an inherent danger in the grid of cascading failures.  Suppose, as was the case last year in Texas, that you have a systemic problem affecting power output in many power plants across the grid.  Supply cannot quite keep up with demand.  Consequently, some power plants start to have trouble keeping up with the required grid frequency.  Once some of these plants start removing themselves from the grid due to frequency issues, however, this puts even more burden on the remaining plants.  What can happen, then is a cascading failure where, all of a sudden, all power plants across the grid trip, one after another.

Most people are unaware of how close to this total failure Texas came last year.  (Here's a decent video talking about this: What Really Happened During the Texas Power Grid Outage?).  Texas was 4 minute 37 seconds away from triggering this sort of cascading failure.  If power managers had not managed demand by "shedding load" (i.e., turning off power to large segments of the grid), the entire grid in Texas would have gone black.

And "going black" is worse than you probably realize, again because of synchronization.  Turning power plants back on and getting a grid back up to fully operational is a massive task, because every power plant must be carefully brought back online *in sychronization* with every other power plant.  Such a "cold black start" in Texas has never happened, but is projected to take weeks or months to finish, during which time almost the entire State of Texas would have been without electricity.

Implications for the energy debates

Let's draw a few implications for energy policy from the above discussion:

1. Variable power sources have inherent grid stability problems.  Because of the relationship between power balance and frequency, power sources which dramatically change their power output are inherently tricky to manage on the grid.  It's not just a question of whether you have the total raw power at any moment to keep up with your demand; you have to do this *and* at the same time keep all of your separate power sources properly magnetically coupled, lest everything come crashing down.  The greater the extent your grid relies on such fluctuating power sources, the more of a challenge this becomes.
2. Some people have criticized Texas for not connecting to the wider grid, evidently thinking that the more power plants are linked to a system, the more secure the system will be.  That is not necessarily the case.  More power plants magnetically coupled does mean more total power, but it *also* means more plants that must be precisely aligned with each other.  It's been shown that more interconnections (beyond a certain limit) will tend to *de*-stabilize the grid, not increase reliability.
   
   This is why for some of the largest interconnections between grid areas, you will see high-voltage DC powerlines connecting grid to grid.  You convert from AC to DC at one end, pipe the electricity to the other end, and then convert back to AC.  This provides a power pipeline from one grid to the other without entangling the two systems with the same frequency requirements.
   
   However, such high-voltage DC power lines are a modern, specialized, high-tech, and extremely expensive solution.  Most such power lines are rather short and limited to high-density areas, because they are so expensive to create.  The most obvious consequence of this is that any infrastructure bill that doles out money to individual regions and tells them "improve your grid infrastructure" is not going to result in such power lines.  These things are always created as specific projects in order to connect grid-to-grid: no single region is going to be able to justify high voltage DC power lines *for the purpose of that region alone*.
3. Frequency instability is a major reason why the supposed "Smart Grid" is still very much a pie-in-the-sky idea.  The idea that a power grid can be made stable and usable even with widely varying power inputs by having all of the nodes in the grid be intelligently switchable is nice--but completely beyond any existing grid right now.  The extent of "Smart Grid" technology available right now is really all about dynamic load shedding--smart meters that turn down your AC on a hot summer day because the grid load is getting too high.  The ability of a Smart Grid to dynamically handle variable *power plant* loads is only possible via the addition of massive, currently non-existent machinery: intelligently connected synchronizing relays, or something to that effect.
   
   Again, a key takeaway here is that no amount of *normal* funding from an infrastructure bill which is aimed at "repairing our crumbling infrastructure" (or whatever the rhetoric is) will allow for this sort of transformation of the power grid into something that can handle huge amounts of variable power supply.  What would be needed for such a thing is a radical transformation, not just a repair.