11

I heard once that when a wind turbine power plant doesn't produce enough electricity the power companies are sometimes forced to turn on a couple of jet engines in order to compensate for the loss, is there any truth to that? I imagine stability is a key factor in keeping the production static and efficient, so what would the power company do?

Transistor
  • 168,990
  • 12
  • 186
  • 385
Rob
  • 229
  • 2
  • 6
  • 14
    "Peaker plants" and "load following plants" (see Andrey Akhmetov's answer, below) would exist even if wind turbines had never been invented. They're needed to match the on-line generating capacity to the demand for electric power, and the _demand_ can change just as quickly as the wind can change. – Solomon Slow Mar 11 '19 at 01:32
  • 1
    Jet engines are actually an overlapping set of the type of powerplant these plants use. They are properly gas turbine engines which are also used on the Abhrams tank and New York busses. Why are jet engines not a subset or superset? Because there are some types of jet engines that have nothing to do with turbines such as ramjets. – slebetman Mar 11 '19 at 03:58
  • Supercaps are now being touted for commercial use in peaking supplies. This [Kilowatt LABS SIrius Capacitor Module](https://arvioshop.com.au/supercapacitor) is rated at 3.55 kWh storage and costs $Australian 4500 on this page. Say $US3000 estimate for 'a few'. That's about 5 x the cost of LiIon batteries - which makes it an utter bargain if the specs are true. 1,000,000 estimated cycles (probably at 100% DOD) [!!!!], 45 year capacitor life, 99%+ round trip storage efficiency, [Brief specs here](https://arvioshop.com.au/supercapacitor) . Astounding. – Russell McMahon Mar 11 '19 at 05:19
  • Not enough for an answer, but gives me an excuse to plug my favourite website: http://gridwatch.co.uk/ it will show you how Andrey Akhmetov's answer works in practice, and shows how different plants vary their outputs as required over the day. (This is for the UK's power gird, there are similar other ones for other girds) – Puffafish Mar 11 '19 at 13:41
  • 4
    When I read "turn on a couple of jet engines..." I instantly imagined jet engines blowing into the wind turbine to make it turn faster. – IMil Mar 12 '19 at 01:28
  • It's important to note that sudden dips in supply look a like sudden spikes in demand. Power companies have always had to deal with the latter, for instance because a couple million Brits decided to make a cup of tea (using electric kettles to boil water) at the same time each day (that story may be apocryphal, but it's illustrative). – MBraedley Mar 12 '19 at 19:07
  • @IMil that's actually what happens. The heated output of a jet engine blowing onto an *internal* turbine makes it turn. That turbine is then on a shaft connected to a generator, or in a modern turbofan, a bigger fan which makes most of the thrust by moving bypass air, or else geared to drive the propeller blades of a turboprop, or the rotor of a helicopter... – Chris Stratton Sep 16 '20 at 18:58

2 Answers2

33

This is correct. When the demand exceeds supply, voltage will sag and frequency will drop (which can risk equipment failure and is certainly an undesirable situation). The operators of power grids will turn on alternative sources of generation in order to correct the imbalance as soon as it is noticed (often under the coordination of a regional transmission organization such as CAISO).

Grid operators are very careful to ensure that the grid frequency is properly maintained (source); even a few seconds of drift (i.e. a few hundred cycles ahead or behind) require RTOs and related agencies to take corrective action where safe. Most of these measures work the same whether demand increases or supply decreases (and thus are relevant whether we are speaking about an increase in consumer load or a decrease in supply from wind or other renewable sources).

In order to understand the mix of energy a bit more thoroughly, it's necessary to take into account the types of generation, which include base-load plants, load-following plants, intermittent sources, and peaker plants:

  • Base-load plants are designed to operate at high cost efficiency (not necessarily environmental efficiency or any other measure of efficiency, unless dictated by local laws and priorities), but cannot be adjusted quickly. Examples of these may include large coal and nuclear base load.
  • Load-following plants can adjust if they have capacity (e.g. hydroelectric or smaller fuel-burning plants)
  • Peaker plants are agile and can be brought online quickly (e.g. gas turbines), but are inefficient. When the base-load plants are insufficient, load-following plants increase their load; if this capacity is exhausted or the grid is experiencing rapid swings in load that the load-following plants cannot keep up with, then peakers will come online and begin burning fuel to achieve enough supply to balance the demand.

Another factor to consider is planning: If an area has consistent winds and enough wind turbines, the wind can be considered part of base-load: It cannot be adjusted, but is relatively predictable and consistent day-to-day. Gaps in the wind are treated the same way as any other shortfall of base-load: first via load-following plants if possible and then with the help of the peakers.

Known gaps and shortfalls can also be handled through trading. For example, Washington State, US has abundant hydroelectric power, and exports energy (as of 2019) to fourteen other states. Its overproduction of energy (which can itself be as harmful as underproduction when it causes overvoltage and overfrequency) is usefully diverted to help make up some of the supply of neighboring states such as California (source). This export includes base-load if the local demand is dropping too quickly for the operating power plants to adjust.

Stored energy also makes a contribution. The sources for such extra energy may be storage sites such as pumped energy storage, batteries (e.g. this), or they may be generation (not necessarily burning fuel).

Lastly, load-shedding is a last-resort. If conditions are adverse (very high demand such as air-conditioning on a hot day, transmission line failures, loss of base-load, etc) then the grid operator may increase the real-time price of industrial energy, or even require that industrial grid users curtail their demand to avoid grid instability. If this is insufficient then blackouts and brownouts will occur, to prevent the total loss of the grid and its most critical users (hospitals, emergency services, communications).

nanofarad
  • 18,062
  • 2
  • 47
  • 73
  • That's all fairly accurate except I don't think coal is very efficient. Wind generation doesn't really count for much in the big picture yet. The NG gas turbines are expensive to operate but can load balance very quickly. Base line plant adjust so slowly that when demand drops too quickly electrity has to be dumped elsewhere. Which means selling it at significantly less then the cost of producing it. I know that our price in Canada changes with the American dollar. Excess power goes back and forth across the border and makes a mess of the price. The whole grid is interconnected. – Joe Fala Mar 11 '19 at 01:09
  • 15
    @JoeFala Coal is not efficient relative to its environmental effect, but it is efficient relative to its financial cost in many parts of the world, to the best of my knowledge. – nanofarad Mar 11 '19 at 01:10
  • Oh yeah it's cheap but from a combustion efficiency point of view I don't think it's very good. I believe that many of the plants are being upgraded but because the cost is still low enough it not financially sustainable to run the higher efficiency plants. I haven't brushed up on this for several years, so I'm not familiar with the current technology. I'm pretty sure nuclear is the cheapest to run and the cleanest overall but expensive to set up and people are terrified of it. Nuclear is actually cleaner then solar panels if you factor in the production of the material in the panels. – Joe Fala Mar 11 '19 at 01:16
  • 1
    J-Power's unit 2 ultrasupercritical(mouthfull) in Japan has 45% efficiency which is pretty damn good. Nuclear power is is like 55% I think more of those ultrasupercritical plants are coming online soon. – Joe Fala Mar 11 '19 at 01:23
  • 1
    @JoeFala I've edited the answer to mention cost-efficiency in particular to avoid any confusion. Thank you for letting me know about the imprecise wording. – nanofarad Mar 11 '19 at 01:25
  • If demand exceeds supply, won't it rather be that **either** voltage **or** frequency will drop, which of course will be immediately sought remedied by addition of power from a responsive source? I mean, you regulate either voltage of frequency to keep stable the other, and the target is ofc to have both stable. – Stian Mar 11 '19 at 12:16
  • 1
    Note that peakers (gas turbines) are historiclaly inefficient and expensive, but really really cheap gas in the USA due to fraking has changed the math a bit. – Yakk Mar 11 '19 at 14:29
  • 1
    @StianYttervik the utility may drop both depending on how regulation is implemented throughout the grid. Either may drop, or both may drop by a lesser amount. – nanofarad Mar 11 '19 at 15:12
  • 1
    Note that the total frequency adjusts are very small, there is a website that monitors the current grid frequency in Europe with a precision of 3 decimals: http://www.mainsfrequency.com/ – Ferrybig Mar 11 '19 at 15:35
  • No way is nuclear 55% efficiency, unless the waste heat is used for useful purposes that are taken into account for the efficiency calculation. – juhist Mar 11 '19 at 17:41
  • 3
    In the Pacific Northwest we get a huge amount of our power from the dams. The dams have the added benefit that they quickly increase/decrease power in order to compensate for wind changes. We actually make so much power in the spring that the wholesale electric price occasionally goes negative and we have to pay people to take our power/ask other plants to shut down which REALLY annoyed the wind plants who had federal matching funds that wouldn't pay if they weren't generating. – Bill K Mar 11 '19 at 19:57
  • @JoeFala No, coal power plants have great efficiency at combustion - they convert about 40% of the fuel's energy to usable electricity, which is superior to most forms of combustion. Their waste heat is also often used productively, which can raise the effective efficiency as high as 90%. Most gas turbines are similar, though the most modern plants can manage as high as 60% (for electricity). In contrast, nuclear plants have substantially less than 1% efficiency - but they don't consume trainloads of coal every day, just a truck of fuel once or twice a year. Still very much worth it :) – Luaan Mar 12 '19 at 08:29
  • @juhist It's complicated. The thermal efficiency of an NPP is essentially the same as of a coal plant - unsurprisingly, since they use pretty much the same steam turbines. Most of the energy in the fuel is lost long before the steam is generated - so if you compare the idealized energy released in a nuclear reaction to the actual amount of electricity generated, you get a total efficiency of less than 1%. In the end, comparing such vastly different generation schemes is silly - coal can in no way be turned into nuclear fuel and vice versa. The same for dams - 90% efficiency is nice, but... – Luaan Mar 12 '19 at 08:35
8

I was going to scold you for not doing a search -- then couldn't find a decent answer! So -- here's a short answer:

First, jet engines -- no. You're thinking of gas turbines, but they are not jet engines (try a web search on "Gas Turbine").

Second, there's not a lot of energy storage on the electrical grid, aside from tanks of gas, piles of coal, uranium rods, and water behind dams. Batteries are starting to look like maybe they'll be practical, eventually. But by and large, when "alternative" energy sources poop out, there needs to be a "traditional" energy source that kicks in. Gas turbines are good for this because they can be brought on line quickly.

This wiki article goes into the grid storage issue.

TimWescott
  • 44,867
  • 1
  • 41
  • 104
  • 1
    The statement on the gas turbine is imprecise but not incorrect. An aeroderivative gas turbine is basically a jet engine, do a web search on this. Peaker plants are usually aeroderivative gas turbines because they can start up in ~15 minutes. The alternative are called industrial gas turbines which are much larger and more efficient. Industrial gas turbines, especially combined cycle units, take hours to start up and shut down and so are inappropriate for peaking use. – user71659 Mar 11 '19 at 01:39
  • 1
    Supercaps are now being touted for commercial use in peaking supplies. This [Kilowatt LABS SIrius Capacitor Module](https://arvioshop.com.au/supercapacitor) is rated at 3.55 kWh storage and costs $Australian 4500 on this page. Say $US3000 estimate for 'a few'. That's about 5 x the cost of LiIon batteries - which makes it an utter bargain if the specs are true. 1,000,000 estimated cycles (probably at 100% DOD) [!!!!], 45 year capacitor life, 99%+ round trip storage efficiency, [Brief specs here](https://arvioshop.com.au/supercapacitor) . Astounding. – Russell McMahon Mar 11 '19 at 05:18
  • 13
    @user71659: When I read "jet engine" in the question, I wondered if someone was imagining using a turbofan to generate *wind* in a wind farm, literally pointing it at existing wind turbines. Totally implausible, but the kind of distortion / misunderstanding I'd believe someone have. – Peter Cordes Mar 11 '19 at 08:01
  • @user71659: Those industrial gas turbines are perfectly capable of acting as a backup for wind turbines; weather predictions are reliable enough to predict power generation 24h in advance. The fast plants are needed to deal with demand variation, but that wasn't the subject of this question. – MSalters Mar 11 '19 at 15:16
  • @RussellMcMahon Panasonic batteries as used in Tesla cars have 28000 cycles at 80% DOD and cost way less than supercaps. How many cycles do you need? I think 365 per year is enough, with the Panasonic batteries certainly achieve. – juhist Mar 11 '19 at 17:45
  • @juhist Serendipitously, for the last few days I have been wading through web claims re battery specs and costs for leading edge good quality LiIon cells. Tesla battery performance is good but also somewhat variable. That 28,000 c lycles at 80% DOD sounds like one of the very luckier users. You'd not want to be in the bottom 5-10% of user experiences. The latest 'thing' are the 21700 cells typically 4 Ah (LG, Samsung, Molicell, ...) with costs perhaps $US5 in good volume. ... – Russell McMahon Mar 11 '19 at 18:18
  • Tesla predict $200/kWh by 2020. | More than a cycle a day is desirable for eg homes with good solar PV capability and very much more in power system load balancing plants. Series to parallel switch: Series FETS off, Parallel FETs on. – Russell McMahon Mar 11 '19 at 18:18
  • Note that “water behind dams” _can_ be a practical way to store generated electricity for later use, as at the Dinorwig pumped storage plant in the UK. – Mike Scott Mar 12 '19 at 09:56
  • @MikeScott Yes, I'm aware of pumped energy storage. If the terrain is favorable I understand it can be a good approach. – TimWescott Mar 12 '19 at 15:26