Bishop Hill Constraining generators
Apr 14, 2014 Energy: wind There was a story doing the rounds a week or so ago about how much windfarms were receiving to switch themselves off. The levels of these "constraint payments" has now apparently reached £8.7m in a single month.
When the story appeared in the Times (£), there was a response in the Guardian which noted that constraint payments to windfarms are dwarfed by those to conventional generators.
National Grid made special payments of £300m over the last 12 months to big energy companies – sometimes for switching off their power stations in an attempt to "balance" the system.
The huge payout dwarfs the £37m paid to windfarms to remain offline over the same period to the end of February – a figure used by critics to question the advisability of supporting renewable energy.
The wording is interesting -big generators are paid, "sometimes for switching off".
Clarification comes in the form of a blog post at the Renewable Energy Foundation blog which explains that conventional generators are usually being paid to switch on when they have planned to be switched off (for example for maintenance), or compensating them for fast startups. And the need for conventional plant to make fast startups is, of course, to compensate for the fluctuating output of all those windfarms on the grid.
And we should not forget that the compensation received by the windfarms exceeds even the bloated subsidies that they have lost by not generating any power.
The technical term for this is "a racket".
Reader Comments (62)
Chandra, demand due to cups of tea and TV are catered for and handled. The NG knows well in advance of such TV linked events, or anything else that is likely to cause a significant energy demand. No fast startups needed as they know hours and days in advance of the requirements.
Clip from Bang Goes The Theory on the NG and how it copes with EastEnders. Watch the rest of the programme to see how it all works.
Apr 14, 2014 at 8:30 PM Chandra | Apr 14, 2014 at 8:30 PM
"And yet, despite the grid requiring fast startups with those 'fluctuations', you say that 1600MW can just drop off the grid in an instant and yet no fast startup is needed. That seems mighty odd to me so I'll have presume that you, the Bishop or I, or some combination thereof, don't understand fast startups."
You appear not to understand the electricity grid system. A grid system is run differently from a non-grid system - the rules are to maintain maximum grid utility and safety.
1. The grid supply of power is consumer demand mediated. The base load required for the UK is about 30GWatts. This base load must at all times be maintained, failure to do so will result in significant outage and possible grid failure.
2. All generators on the grid must, at all times, maintain the nominal frequency, and voltage. Failure to do so will lead to undue equipment stress, instability, and possible fault outage.
3. As the demand has to be predicted ahead of use generator are connected to/removed from grid before they are needed, they may or may not actually supply power to the grid, even when running and connected to it.
4. The biggest 'consumer' on the grid are the generators themselves in their role as providing active load balancing, and power factor adjustment(phase control). This necessary requirement means that more generators are running and on the grid that the consumer load would appear to demand.
5. Connecting a generator to a grid system does not automatically mean that it is supplying or dissipating power (power balancing or load shedding). By altering the generators output phase while on the grid the generators output could be a supply, load, or niether. The correct adjustment of the generator is determined by the grid's requirements (not necessarily the consumers).
Please note that as Gridwatch shows fast start-up and shut-down is currently being used. If you look at the gas power output it fluctuates to keep both generation variation from wind, and the 'load demand' in order. Gas power generation is easily configured for fast start-up and shut-down when compared to other conventional power generators (coal and nuclear).
http://www.gridwatch.templar.co.uk/
Oops I forgot to stop the bold - sorry. [Done. BH]
Thanks for the description tom0mason. Can you perhaps explain why the fluctuating output of wind farms causes "fast starts" yet 1600MW dropping out in an instant doesn't.
Too bloody ill at the mo to get into a looong winded discussion about minutiae but I must say that Chandra does seem to be making a good point.
P.S. Chandra, I think large scale windfarms are NOT fit for purpose in the main though...Hideous waste of money on the large scale. Fine to have one on your house though with a bank of batteries in the basement I feel.
There....!
Go gentle please...ill....
Chandra| Apr 15, 2014 at 12:09 AM |
Thanks for the description tom0mason. Can you perhaps explain why the fluctuating output of wind farms causes "fast starts" yet 1600MW dropping out in an instant doesn't.
The simple answer is mechanical inertia!
As I pointed out in 4. above there is a lot of generators online just to even out the load.
Normally (pre wind mills days) this was fairly easy as the grid was less reactive and fewer but bigger generators could cope. As more and more renewables get thrown on the grid, the heavy generators get what to them looks like very reactive loads.
These big generators are there to maintain the base load of the grid (some 30 GWatts) and are not originally designed to cope with the highly reactive loads - in essence they are big rotating machines and do not change their rotational speed quickly. On the other hand if a mere 1.6GWatts was to be suddenly called on from these brutes they can, in a few seconds, deliver it and the grid would recover.
See the Electricity generation section of http://en.wikipedia.org/wiki/Drax_Power_Station to get some idea of the physical masses involved here.
NOTE -
Maintaining the integrity of the grid is primary function of the whole system, so if some consumers are to be dumped to maintain it, this will happen.
see http://en.wikipedia.org/wiki/National_Grid_(UK)#Cost_per_kWh_of_transmission
Adding Renewables -
Adding renewables to grid regardless of load requirements can not be done, that would compromise grid integrity. So renewable (if available) are usually only put on the grid during periods of predictable demand but they make the grid overall look more active (they 'look' like a variable generator and/or load) to the rest of the grid. To obtain useful power from these generators other generators on the grid must react to 'fill in the dips' of their output.
All this extra reactive loading from renewables demands the addition of generator sets that can react rapidly to the variation - hence the installation of smaller gas generators that are more load agile while delivering the peak power demanded from the consumer.
Remember that this system has to maintain this while balancing the national 3 phase load at the very tight limits of voltage and frequency.
A related theme is here at http://chiefio.wordpress.com/2013/04/14/is-the-uk-grid-approaching-instability/
Please note that since April 2013 the UK has somehow reduced it's demand by nearly 10GWatts - how?
Hope this helps.
Thanks Tom0, you've saved me a lot of effort.
I recall an incident on 27th May 2008 in which two generators were lost in 2 minutes (1582MW). Grid frequency fell, but a further unexplained loss of about 250MW meant that 581MW of demand had to be disconnected for up to about an hour. There would have been no disconnections had the 250MW loss not occurred. The 250MW was put down to embedded renewable generators, for which the grid code does not apply. Basically, embedded renewable generators are a huge liability to the operation of the grid. Embedded renewable generators do not maintain the required information, so that NG could not determine exactly what happened. It will happen again, with more drastic consequences next time.
Actually, the way it works with 'big' generators makes the £300M paid to them an underestimate. Generators which can actually control their output place bids and offers to produce more or less than their planned generation. The price they offer to turn up will always be at above cost, sometimes quite a lot above (depends on the market and the time of day/day of week etc). If such an offer is accepted, then they turn up and the generation above the original plan is sold to the Grid at the offer price. This generates a payment to the Generator.
At the same time, they also make bids to buy power from the grid instead of producing it them selves. If a bid is accepted, the generator turns down - generates less - and buys the difference at the bid price from the grid. This generates a payment from the generator to the grid. (The Bid price will generally be below the generator's cost to produce the same power.)
Thus if the net payment is £300M, it would be interesting to know what the gross in both directions is.
I suspect that Chandra may not appreciate that "reactive" refers to current and voltage not being in phase, rather than speed of reaction to changes in network load.
Thanks tom0mason for the explanation. I am still unclear as to why variation in wind generation should be more serious than dropouts of conventional generators or unexpected variation in demand, but the subject is clearly more complicated than I had understood.
That the system is resilient in the face of huge losses of generation capacity is to be expected. Dropouts are very common and the grid has been engineered to survive with intermittent conventional plant. See figure 15 in the report report on the May 27 2008 incident Philip mentioned for a scatter plot of hundreds of trip events (2001-2009) and figure 14 for 60 or so over 500MW (1996-2009).
Philip Bratby, you said
> There would have been no disconnections had the
> 250MW loss not occurred. The 250MW was put
> down to embedded renewable generators, for
> which the grid code does not apply.
That might be what you want to be the reason, but your memory is wrong about that. The final report I linked to above says that "Although this embedded loss did not cause the frequency to fall to 48.8Hz, it only hastened it". The wording is a bit odd, but clearly it was the fall to 48.8Hz that caused disconnections and that fall was caused by conventional plant failing. The next sentence says that, "In slightly different circumstances, it may do so and initiate operation of the National Low Frequency Disconnection Scheme". For this reason it recommended reviewing the rules that apply to embedded generators with a view to preventing them from disconnecting during low frequency events (which they were, oddly, allowed to do). The report is from 2009, so one might hope that new rules now apply.
Chandra:
You have just demonstrated that you don't understand the impact of reactive power on network stability. Time to wrap your head in a hot towel and learn some science.
You will have to pay for outside work like fuel Pump, radiator cleaning, Lathe work welding and Boring etc.