Water tank and water pipe analogies fail pretty fast when it comes to thinking about electricity. Keeping grids energized by real-time management of supply and demand is not much like keeping water or gas flowing through pipes.
If you want to examine smart grids, it's strange to ignore what's been developed in Germany and NW Europe over several decades. A lot of this revolves around fast communication strategies and supply/demand prediction algorithms:
> "To be able to operate this complex solution infrastructure, Netze BW has applied a so called “traffic light concept”. The green light indicates that no congestion is predicted, while the yellow light is a sign of a potential bottleneck in the grid that might require certain restrictive measures by the market players. For example, a Virtual Power Plant operator would adjust the operating mode of its storage and generation assets to avoid predicted transformer overload. However, despite these actions taken during the yellow phase, the actual technical limits of the electricity network might still be violated in real time due to unforeseen events. In this case, the red light would call for immediate mitigation measures enabled automatically by the REMS system."
The fact that the AC power grid is synchronous is very unlike any water/gas system:
> "In a synchronous grid, all the generators naturally lock together electrically and run at the same frequency, and stay very nearly in phase with each other...Small deviations from the nominal system frequency are very important in regulating individual generators and assessing the equilibrium of the grid as a whole. When the grid is heavily loaded, the frequency slows, and governors adjust their generators so that more power is output (droop speed control). When the grid is lightly loaded the grid frequency runs above the nominal frequency, and this is taken as an indication by Automatic Generation Control systems across the network that generators should reduce their output."
The water pipe - electrical circuit analogy is not necessarily horrible for DC circuits, but I think it becomes an impediment to learning pretty quickly. There's no water model that works for things like p-n silicon junctions for example.
There's also the fact that transmission of energy by fluid in a pipe (sound wave speed I think) isn't anywhere near as fast as transmission of energy in a wire by electric fields, even though the electrons themselves aren't really moving that fast at all in bulk (*drift velocity is low). There's a 'pipe full of marbles' analogy for that effect, but the actual energy is carried by the electric field, not by a flow of electrons as with a flow of water through a water turbine. There's also the Drude model which treats the electrons as something like an ideal gas with electrons scattering off the positive metal ions:
However if you get into band theory of solids as a means of explaining conductors, insulators and semiconductors, anything related to a water pipe analogy falls apart:
A specific analogy doesn't have to be used forever. You can drop it when it starts being problematic.
Asked a different way, what didn't work for the purposes of this article? It's at a high enough level that "becomes an impediment" isn't really an issue. Nor are "p-n silicon junctions". Nor the speed of transmission.
That's the concern being raised. It's indeed problematic to use a water bucket/pipe analogy in an AC grid when talking about transmission and generation of power, and related things like spinning reserves. (It's less problematic when discussing energy.)
The water analogy is perfectly fine to use to talk about the bulk AC transmission grid (which is low-frequency alternating current).
photochemsyn speaks about how the hydraulic phenomena is not adequate to describe some electronic aspects (e.g., modeling p-n silicon junctions). These issues are irrelevant to how well the hydraulic analogy works to describe the workings of the AC transmission grid.
Yep, exactly my point. For the purposes of this explanation in the article the analogy works fine, hence why I'm asking what exactly didn't work.
That it doesn't work in all possible cases talking about A/C transmission doesn't really impact its use as an analogy in a high-level view of the grid.
There are many aspects where the analogy is flawed, when talking about the electrical grid, e.g. you will have different types of loads (capacitive, resistive, inductive) and those loads will change the way your grid behaves. This is not something that you have in fluid dynamics at all (to my knowledge) yet any grid operator who would ignore such things would have their gear just explode.
The grid is a heavy rope that you are trying to keep tension on so that it doesn't touch the ground. One side of the rope is the power station, the other end splits and frays into individual strands that connect to every meter on that grid.
Strands are continually, almost randomly, disconnected (let go), reconnected (picked up), and the amount of "pull" on the rope each strand creates will vary (usage fluctuations). This means the power station needs to continually be adjusting how much it "pulls" on the rope to maintain tension and keep it from dropping.
I like this one because it not only avoids water and pipes and cuts to the core of the balance issue that is going on. It's not "thing A flows from point B to point C" but rather "this is an active system and careful balancing act".
That being said I recognize that the water pipe models are well studied, even in how they successfully model some aspects (and fail to capture others) of electrical grids.
Well, for one, if you get low pressure in one set of water pipes the entire water grid doesn't attempt to collapse because it goes out of sync.
The grid is a gigantic clock that at least in the US is running at 60 ticks per second. This is pretty easy to manage if you have some massive power source on it like gigantic turbines at a nuclear power plant. All your small clocks aren't going to push that around so much.
The problems come in when all you have is small clocks, who sets the phase of the grid?
> “We find that replacing conventional generators with inverter-based resources, including wind, solar PV, and certain types of energy storage, has two counterbalancing effects,” said Paul Denholm, NREL principal energy analyst and lead author of the guide. “First, it’s true that these resources decrease the amount of inertia available on the system. But second, these resources can reduce the amount of inertia actually needed
> “Ultimately, although growth in inverter-based resources will reduce the amount of inertia on the grid, there are multiple existing or possible solutions for maintaining or improving system reliability,” Denholm said. “So, declines in inertia do not pose significant technical or economic barriers to significant growth in wind, solar, and storage to well beyond today’s levels for most of the United States.”
The NREL inertia video explainer felt a little like it was begging the question - "inertia protects the grid because it has inertia and keeps spinning" - it doesn't quite feel like it explains where the extra energy comes from or goes, just that the mass keeps spinning. (I also haven't had a physics class in a long long time so some of this is not obvious to me, except that I understand from just common sense that if something's spinning you had to put a bunch of energy into getting it going in the first place and it's going to keep going if left on its own)
Anyway, I was hoping someone could fill in some details for me. Imagine a simplified grid: a dam that sends water through a penstock past a turbine/generator and into an electrical circuit, and a couple of resistance heaters on the other side of the circuit. The energy comes from water flowing through the dam - the dam operator opens up the sluice gate to let water flow through, the generator extracts the mechanical energy and turns it into electrical energy and it goes down the wire to the resistance heater where it gets turned into heat energy. Everything is balanced - the right amount of water is flowing through the dam to turn the turbine at the right speed to balance out all of the energy flowing through the wires and into the resistance heaters (and lost along the way, like losses in the transmission lines, etc). In this setup, there's some measure of pressure that turbine pushes back against the water flowing through the penstock of the dam, which is balanced out by how much pressure is coming from the water behind the dam and the pressure being put on the surface area of the penstock in the dam and the pressure being relieved by the water leaving the dam.
I get that thanks to inertia, if the sluice gate accidentally slams shut and all water stops flowing through the dam, the turbine is going to keep spinning for a bit and energy is going to keep going out onto the grid, though it will start to slow down due to friction at the turbine and energy being extracted from the system by the resistance heaters on the other end of the grid.
What I'm less clear about is how does inertia help when the water keeps flowing at the regular speed but when demand drops from the grid load. Let's say one of the resistance heaters turns off in a home somewhere - what happens to the energy from the water that was previously flowing into the grid via the turbine? Does the inertia in the spinning of the turbine somehow push back against the water flowing the dam, slowing the water down a bit/building pressure up in the penstock and behind the dam - with that pressure buildup being exactly equal to the energy that used to be going into the resistance heater? And that pressure either stays built up from the turbine until someone lowers the sluice gate a bit to cut back on the waterflow through the dam? Or does nothing involving inertia happen here - if the resistance heater gets turned off the overall load is reduced and the turbine spins a bit faster because there's less pushing back on it, and the water can move through the dam a bit faster, and the turbine just spins faster until someone notices it's going a bit too fast and the gate needs to be lowered so it drops back to rotating at 60hz?
Similarly, if someone turns on another resistance heater and now more energy is needed on the grid, but the sluice gate isn't opened up immediately, is inertia involved here somehow? If the turbine has to push harder on the grid side because of extra load, presumably the turbine slows down? Or does the turbine get pulled along by the new load somehow (more inertia?), and so more water can push past the turbine, giving it the extra energy it needs (and presumably dropping the water pressure in the penstock in the dam? And the pressure stays low until the sluice gate is opened up a bit more and more water can flow through the dam?)
I am using water pressure from a dam here, but I assume this would be equally true in a gas plant generating steam - if more energy is needed, the pressure in the steam drops until someone turns up the burner and creates more steam, etc, or if less steam is needed the pressure just builds up until someone notices and turns down the burner?
If anyone can explain how inertia and the grid translates into changes in the actual source of energy, I'd much appreciate it!
I think the very simplified idea is that inertia slows down how rapidly the generators slow down respectively speed up in response to load changes, thereby giving you a better chance of adjusting power generation to match the new load.
If your ratio of inertia to power demand variability is too low, then any sudden change in power demand will lead to your generator rapidly spinning up/slowing down before (in the case of your imagined dam) you have any chance of de-/increasing the water supply to the turbines as required. If it slows down or speeds up too much, then whoops you're hitting the grid frequency limits, power trips, and you've got a blackout.
If it spins up too much and too rapidly it might even exceed the mechanical limits of the generator, though normally real-life systems should of course always be designed to be capable of safely handling even a complete load trip, because there's always the possibility of a tree suddenly falling onto a power line or whatnot…
With sufficient inertia on the other hand, any mismatch between power generation and power demand now only manifests itself as your generators gradually starting to spin faster/slower, giving you enough time to adjust the water supply valves as required.
In the pipes analogy, pressure corresponds to potential energy (voltage). Low voltage conditions does not necessarily trigger frequency stability issues.
Grid synchronization is a different issue to the electrical-hydraulic analogy (which, by the way, was the one that Maxwell used and its pretty useful). Grid synchronization comes from the fact that electrical quantities in the network (current, voltage) are alternating periodic. Generators, which are usually mechanical, oscillate and induce such periodic signals. The system must remain synchronized. But this has nothing to do with the network being electrical.
For one the "water" moves near light speed effectively and tanks andand reservoirs are far more expensive, and gravity doesn't meaningfully apply. And there is some leakage with distance. And you cannot use a turbine to generate water from torque. Even if many equation structures are the same as fluid mechanics the assumptions break from differences.
This is not how the hydraulic analogy works. The "speed" of the water is irrelevant. The analogy is used to understand losses, energy balance, energy conservation, reactive power, etc.
If you want to examine smart grids, it's strange to ignore what's been developed in Germany and NW Europe over several decades. A lot of this revolves around fast communication strategies and supply/demand prediction algorithms:
https://eu.landisgyr.com/blog/grid-control-the-future-of-the...
> "To be able to operate this complex solution infrastructure, Netze BW has applied a so called “traffic light concept”. The green light indicates that no congestion is predicted, while the yellow light is a sign of a potential bottleneck in the grid that might require certain restrictive measures by the market players. For example, a Virtual Power Plant operator would adjust the operating mode of its storage and generation assets to avoid predicted transformer overload. However, despite these actions taken during the yellow phase, the actual technical limits of the electricity network might still be violated in real time due to unforeseen events. In this case, the red light would call for immediate mitigation measures enabled automatically by the REMS system."