My response to Bob Ireland from yesterday’s thread about the pricing and use of available generating assets. Numbers are rough guesses based on past experience. I am not privy to recent data....
“”I am not sure that I understand that; there is no way to store large amounts of available power for later higher demand - is there? {batteries, flywheels... ?} Or is it just a matter of not using their own local power at night so that they get better rates at times of peak demand?”
~~~~~~~~~~
“Let’s take a typical early summer day - maybe in the 60s overnight, and around 90 in mid-afternoon. Generation dispatching is set up to utilize the lowest cost of generation that is available for the next incremental increase in demand.
I’ll use TVA as an example because it’s most familiar to me. At 3 AM on a very mild night, system demand might be 14,000 megawatts. The 3 nuclear plants, if they’re on, will be running at full output. They don’t reduce output even though it’s very possible. There are reasons for that and they’re pretty important, but suffice to say that outside of a grid emergency, a nuclear plant won’t drop down its generation at night.
It just so happens that nuclear plants are among the lowest cost, in terms of fuel, of all resources so if you need something on line with the throttle stuck wide open, the nukes are a good choice.
Every plant, every generator or resource has an incremental cost/megawatt. So at low load periods, the generation that is cheapest will be utilized while more expensive generation will be at its minimum output or off line. Hydro generation is generally run according to weekly schedules, based on rainfall, navigation of the waterways, and providing adequate cooler water downstream for plants that need river water for that function (of which few are left). There isn’t enough water available to run hydro plants 24/7, so the schedule allows to run them when the energy is most needed, and allow adequate river flow in between.
Anyway, as far as price, in terms of dispatching generation, the cost/mw is the most important though there are grid reliability concerns as well. The cost spread is pretty wide - around 8-10 dollars per megawatt hour for the nukes, combined cycle gas runs around, say, 15-25 while coal can run from around 20-50 depending on the unit. Simple-cycle gas might run in the $40-80 range and combined cycle plants usually have a feature called duct burners that will add a little extra heat to the exhaust to allow a little extra output from the associated steam turbine. Duct burners are around the same price as simple cycle gas turbines. A few gas turbines, usually much older units, may cost $80-100 per megawatt hour so you use those sparingly and they sit idle most of the time (and often in disrepair).
So you want to bring up the hydro as available, based on the schedule as demand increases, then bring the combined cycle gas plants up to their normal full load, coal, depends on the unit, as does gas. As the load starts coming up in the morning, you start raising output on the combined cycle plants and coal plants. You bring on hydro as it is scheduled. When you start reaching peak, then the simple cycle gas turbines will be brought into play along with duct burners. Each incremental step costs more than the last, and on a day with a very wide spread between minimum and maximum demand, you can have a 10:1 cost ratio.
There is another piece of the pie - purchasing power from other entities. Price varies according to on peak vs off peak, and the purchase period contract. You can buy from others for economic reasons or for high demand, or for emergency situations. And there is a market price for power, and it can get quite high at peak periods, much more expensive than even those old worn out gas turbines.
As for energy storage, there are some moderately small battery systems out there, and TVA (and others) has a very nice pumped hydro storage facility that is generally pumped at night, and run during the day to meet demand and to provide some contingency supply in the event a large generator trips off line (which they occasionally do). Raccoon Mountain pumped hydro is a pretty large facility, and people that actually operate the generation system would really like about 3 more just like it. Because sometimes, the peak demand is double what the minimum load is in a 24 hour period, and its hard to bring generation down low enough to follow the minimum without taking units off line.””
For Bob and any others interested, you can look at “marginal” electricity pricing variations by consulting the web sites of PJM ( pjm.com ) or MISO ( misoenergy.org ) that track electric demand, grid conditions, and marginal pricing in their respective reliability footprints.
Note that the marginal pricing in these cases includes not only the cost of generation but the cost of transmission plus a factor based on congestion, or loading of the transmission system in a given area. So, for example, a purchaser in, say, Pittsburgh, may be buying generation from an entity in central Virginia for $22 per megawatt hour, but since the lines between Pittsburgh and the plant are heavily loaded and constrained, there may be a large adder to that price, which might make it more economical to buy $30 power from a more local source because the transmission adder might be so high that the $22 price is no longer the most economical resource.
It’s both a simple and a complicated issue. TVA didn’t use marginal pricing based on constraints for transmission access, but that may change when they finally get their new energy management system computer into operation. The old computer system, reliable and durable, was from the 90s and didn’t have the capability to deal with that much additional data and calculations.
Incidentally, this pricing model does well to deal with transmission system constraints and loading, but it also drives the construction of new lines, particularly between the heavily wind-powered midwest states and the midwest and eastern cities to get that wind power form the source to the customer. States in between, naturally, don’t want these transmission lines across their states if they are not benefitting so it’s a big mess. I mean, if you’re Missouri, do you want a transmission line strewn across your state to provide wind power from Kansas and Oklahoma to be transmitted to, say, Louisville or Columbus? Probably not.
“no way to store large amounts of available power for later higher demand”
Coal, oil, gas, nuclear fusion are exactly that. If nuclear regulation were based on 2026 expertise and not 1947 fears it would be competitive.