National Grid Frequency

[a_c_skinner]

This may be off topic but I doubt I am the only person wondering this, it is rail related of course.
To my latterly untutored A level physics (got an A in 1973) I'd have expected a shortfall in generating capacity would "show" as falling voltage. Plainly this is not the case, it comes out in the wash as falling frequency. This seems to be the cause of modern rolling stock "tripping" out as in the recent Class 700 events. I know this isn't solely rail related but it clearly mucked up Thameslink and I would like to understand the physics, which seems counter intuitive.

Andrew

 

[Elecman]

It’s to do with the Grid system being considered a Infinite Busbar which will hold its voltage as constant but loss of capacity shows as a lowering of frequency against the ‘constant’ load.

 

[a_c_skinner]

Thank you, but I'm not much wiser. I had imagined the AC frequency as determined by how fast the moving parts in the big alternators at power stations were turning. There must be something physical going on to generate the A in the AC.

 

[Indigo Soup]

Thank you, but I'm not much wiser. I had imagined the AC frequency as determined by how fast the moving parts in the big alternators at power stations were turning. There must be something physical going on to generate the A in the AC.

The principle is, as I understand it, that when there's a shortfall of generating capacity the grid voltage is maintained by robbing kinetic energy from the rotating machinery, presumably more-or-less in unison, if the load is well balanced across the grid. Since the machinery is slowing down, the grid frequency reduces.

 

[Elecman]

The principle is, as I understand it, that when there's a shortfall of generating capacity the grid voltage is maintained by robbing kinetic energy from the rotating machinery, presumably more-or-less in unison, if the load is well balanced across the grid. Since the machinery is slowing down, the grid frequency reduces.

More or less yes that is how it functions

 

[HSTEd]

Thank you, but I'm not much wiser. I had imagined the AC frequency as determined by how fast the moving parts in the big alternators at power stations were turning. There must be something physical going on to generate the A in the AC.

The alternators will produce an output with a frequency that is tied directly to their rotating speed.
If you remove more energy from the alternators than is provided to them by whatever they are coupled to, they will slow down and thus the output frequency will drop.

This is traditionally the method used to control grids, the governors on steam turbines would be set to monitor grid frequency and increase turbine power if the frequency falls, or reduce power if it rises.
This allows the grid to be controlled without any communication between generating plants.

 

[a_c_skinner]

Thank you all for an obvious explanation which would never have crossed my mind.

Andrew

 

[coppercapped]

As a postscript, as I understand it the control of the frequency of the grid has become more difficult following the demise of the large coal fired power stations which had biblical quantities of rotary inertia in their turbo-alternators. In most cases short term load increases could be absorbed as the frequency would only fall slightly. Modern generating plant, gas and wind turbines, do not have such massive inertias so frequency falls will be more marked. This is allowed for by running such plant at about 90% of its rated capacity so up to 10% extra load can be supplied almost immediately.

 

[GRALISTAIR]

As a postscript, as I understand it the control of the frequency of the grid has become more difficult following the demise of the large coal fired power stations which had biblical quantities of rotary inertia in their turbo-alternators. In most cases short term load increases could be absorbed as the frequency would only fall slightly. Modern generating plant, gas and wind turbines, do not have such massive inertias so frequency falls will be more marked. This is allowed for by running such plant at about 90% of its rated capacity so up to 10% extra load can be supplied almost immediately.

And of course Solar panels use inverters so not much there at all!!

 

[a_c_skinner]

Both those had crossed my mind...

 

[InTheEastMids]

As noted above, the 50Hz comes from having lots of spinning metal in (usually large) power stations turning at 3000rpm (50Hz). This is called synchronous generation. All that momentum gives the grid inertia, which in turn gives stability as demand changes

One of the increasing challenges in system operation is that this is a declining share of electricity generation.

Solar has no inertia (no moving parts). Wind turbines use power electronics and are asynchronous so do not contribute to system inertia. This is where batteries have made the most significant impact on the system because they can respond very quickly to frequency changes to support the system.

Given these changes there is currently a debate involving National Grid, Ofgem, BEIS and others about whether what is currently done provides adequate resilience, and if not, what needs to change.

 

[GRALISTAIR]

Nuclear, nuclear and nuclear. I know we are getting into speculative territory and risk drifting off topic, but If we are going to decarbonize the economy - they must feature somewhere in the future one would think. This would introduce huge resilience to the grid.

 

[HSTEd]

As noted above, the 50Hz comes from having lots of spinning metal in (usually large) power stations turning at 3000rpm (50Hz). This is called synchronous generation. All that momentum gives the grid inertia, which in turn gives stability as demand changes

It is however worth noting that when Hinkley Point C arrives, it will have huge masses of metal but it will be spinning at 1500rpm, which is traditional for nuclear abroad, rather than at 3000rpm.
That is for technical reasons created by the limitations on materials used to make turbine blades (if they get too long the centripetal acceleration will snap them).

They have four pole generators instead of two pole generators to produce 50Hz.

 

[InTheEastMids]

Nuclear, nuclear and nuclear. I know we are getting into speculative territory and risk drifting off topic, but If we are going to decarbonize the economy - they must feature somewhere in the future one would think. This would introduce huge resilience to the grid.

Having to withstand the instantaneous loss of 3.2GW of generation will require changes so as not to reduce the resilience of the electricity system.

And resilience from what? Nearly all loss of supply arises from failures in distribution networks. I'm not saying that a Black Start isn't important, but any discussion about electricity system resilience needs to think a lot more holistically.

It is however worth noting that *if* Hinkley Point C arrives

There, fixed that for you
Of course my serious point is the uncertainty in the 'when' and if/when another project will happen at all.
Thanks for the detail on HPC. Interesting.

 

[Mike Hodgson]

My probably outdated simple layman's understanding of the big benefits of Dinorwic Pumped Storage power station (where the Festiniog railway was diverted) was that it was excellent for filling shortfall in demand at very short notice (compared with coal or gas fired stations which take quite a while to get their kettle boiling), and that it could continue to do so as long as there was water at the top of the hill and that it could equally well absorb surplus power generation by pumping it back again. It was also supposed to be useful for smoothing out demand between high evening loads (generate) and slack overnight demand (pump). One problem was that it could only do this for that part of the country. Other similar systems aound the country were desirable but the geography isn't as convenient although there are some others elsewhere.

 

[pj334]

As noted above, the 50Hz comes from having lots of spinning metal in (usually large) power stations turning at 3000rpm (50Hz). This is called synchronous generation. All that momentum gives the grid inertia, which in turn gives stability as demand changes

One of the increasing challenges in system operation is that this is a declining share of electricity generation.

Solar has no inertia (no moving parts). Wind turbines use power electronics and are asynchronous so do not contribute to system inertia. This is where batteries have made the most significant impact on the system because they can respond very quickly to frequency changes to support the system.

Given these changes there is currently a debate involving National Grid, Ofgem, BEIS and others about whether what is currently done provides adequate resilience, and if not, what needs to change.

Reading the Ofgem report into the national power outages that occurred on August 9th (https://www.ofgem.gov.uk/system/files/docs/2019/08/incident_report_lfdd_-_summary_-_final.pdf, what I found surprising was that there was a knock-on affect on local small scale generation.
Hornsea One offshore wind generation disconnection lost a total of 737MW supply to the grid.
Little Barford Gas and Steam Turbine generation disconnection lost a total of 641MW supply to the grid.
There was then a further 500MW approx supply lost from local distributed generation, typically solar and some small scale gas and diesel fired generation, due to the operation of their Loss of Mains Protection.
I had previously assumed that solar panels installed on the roof of domestic premises would continue to provide power into the house even if there was a power cut, but this would seem to suggest their inverters require an incoming mains supply to maintain synchronism (and presumably to stop them trying to feed the whole street).
As IntheEastMidlands rightly said, the national grid needs large physical rotating machinery to provide inertia and thus stability, which electronics alone does not provide. Whilst the trend towards wind and solar sources for electricity generation is commendable, they can not provide the whole answer as both sources provide their output power via electronic inverters.
Inevitably, the country will need a significant amount of power from conventional turbine generators, and the only non-carbon source of energy currently is nuclear, and this form of generation is ideally suited to providing steady base load generation.

Mike Hodgson mentioned Dinorwic Pumped Storage power station, which indeed is used to provide support during peaks. There is an interesting near-realtime chart showing how and when it is used at
https://www.gridwatch.templar.co.uk/
Regards, Paul.

 

[Elecman]

Reading the Ofgem report into the national power outages that occurred on August 9th (https://www.ofgem.gov.uk/system/files/docs/2019/08/incident_report_lfdd_-_summary_-_final.pdf, what I found surprising was that there was a knock-on affect on local small scale generation.
Hornsea One offshore wind generation disconnection lost a total of 737MW supply to the grid.
Little Barford Gas and Steam Turbine generation disconnection lost a total of 641MW supply to the grid.
There was then a further 500MW approx supply lost from local distributed generation, typically solar and some small scale gas and diesel fired generation, due to the operation of their Loss of Mains Protection.
I had previously assumed that solar panels installed on the roof of domestic premises would continue to provide power into the house even if there was a power cut, but this would seem to suggest their inverters require an incoming mains supply to maintain synchronism (and presumably to stop them trying to feed the whole street).
As IntheEastMidlands rightly said, the national grid needs large physical rotating machinery to provide inertia and thus stability, which electronics alone does not provide. Whilst the trend towards wind and solar sources for electricity generation is commendable, they can not provide the whole answer as both sources provide their output power via electronic inverters.
Inevitably, the country will need a significant amount of power from conventional turbine generators, and the only non-carbon source of energy currently is nuclear, and this form of generation is ideally suited to providing steady base load generation.

Mike Hodgson mentioned Dinorwic Pumped Storage power station, which indeed is used to provide support during peaks. There is an interesting near-realtime chart showing how and when it is used at
https://www.gridwatch.templar.co.uk/
Regards, Paul.

You are quite correct all domestic generation must disconnect from the mains in the event of mains power-loss to provide protection to the DNO staff carrying out any work on the disconnected mains

 

[GRALISTAIR]

My probably outdated simple layman's understanding of the big benefits of Dinorwic Pumped Storage power station (where the Festiniog railway was diverted) was that it was excellent for filling shortfall in demand at very short notice (compared with coal or gas fired stations which take quite a while to get their kettle boiling), and that it could continue to do so as long as there was water at the top of the hill and that it could equally well absorb surplus power generation by pumping it back again. It was also supposed to be useful for smoothing out demand between high evening loads (generate) and slack overnight demand (pump). One problem was that it could only do this for that part of the country. Other similar systems aound the country were desirable but the geography isn't as convenient although there are some others elsewhere.

Allegedly, it was claimed (I will have to dig out the reference) a peak demand was half time during Coronation Street where people put the kettle on and creating a measurable huge surge in demand very quickly

 

[John Webb]

...........As IntheEastMidlands rightly said, the national grid needs large physical rotating machinery to provide inertia and thus stability, which electronics alone does not provide. Whilst the trend towards wind and solar sources for electricity generation is commendable, they can not provide the whole answer as both sources provide their output power via electronic inverters.
Inevitably, the country will need a significant amount of power from conventional turbine generators, and the only non-carbon source of energy currently is nuclear, and this form of generation is ideally suited to providing steady base load generation...….

A boss of mine some 40 years ago, who worked previously at the Royal Aeronautical Establishment at Farnborough, told me he was asked to investigate there when the mains was most stable for operating a very large wind-tunnel. To his surprise he found the grid was most stable when they were operating their wind-tunnel! (It took several 100MW of power to run this equipment!)

One green power generation source that seems to be woefully ignored is tide power - it occurs all round the country twice a day and over a wide range of times and is not subject to the variability of wind or sun.

 

[PartyOperator]

Voltage does vary with time and physical location, and generally this is seen as preferable to allowing frequency to reduce, given that lots of things out there are very sensitive to frequency while most are not that sensitive to voltage. Depending on the flow of reactive power (which is kind of hard to explain without complex numbers but basically measures the lag between current and voltage), voltage varies across the network and has to be compensated by large generators adjusting the voltage sent out by their transformers.

The frequency of the power you get at home should always be within 1% of 50Hz, but the voltage could be 10% over or 6% under 230V.

 

[Bald Rick]

Allegedly, it was claimed (I will have to dig out the reference) a peak demand was half time during Coronation Street where people put the kettle on and creating a measurable huge surge in demand very quickly

Many years ago I was in the National Grid Control centre, and I saw the data and graphs. There were indeed surges of demand when East Enders / Corry finished, although the size and time of the surge depended on the date and weather. For example around March with daylight getting progressively longer, on one night people would be (on average) switching their home lights on at 1929, the next night it would have been 1931, but because 25 million people are watching the telly from 1930 they don’t switch them on until the programme finished at 2000. Of course this was in the days when 25million people were quite happy watching soap operas.

There were however bigger surges - notably during England World Cup games, specifically just after we lost on penalties. The biggest power demand change, though, was Princess Diana’s funeral. Because more or less all shops and places of work were shut that Saturdays morning, and everyone was in front of the TV, demand was actually lower than a normal night time. Once the funeral cortège left London, people started ‘going about their business’ and IIRC demand more than doubled in about half an hour.

 

[a_c_skinner]

One green power generation source that seems to be woefully ignored is tide power

And indeed can incorporate pumped storage. Apparently

 

[Elecman]

Voltage does vary with time and physical location, and generally this is seen as preferable to allowing frequency to reduce, given that lots of things out there are very sensitive to frequency while most are not that sensitive to voltage. Depending on the flow of reactive power (which is kind of hard to explain without complex numbers but basically measures the lag between current and voltage), voltage varies across the network and has to be compensated by large generators adjusting the voltage sent out by their transformers.

The frequency of the power you get at home should always be within 1% of 50Hz, but the voltage could be 10% over or 6% under 230V.

Most often at the higher end due to the DNOs not normally altering the previously set tapping 240volt of thier local distribution transformers

 

[edwin_m]

I guess it would be possible in principle for producers of power that feed the grid via inverters to modify their voltage and frequency to simulate the behavior of a rotating generator as discussed above, very much as a VVVF inverter on a traction package does but with a much smaller degree of variability. However I guess nobody thought to specify this so the inverters weren't designed with that capability.

And indeed can incorporate pumped storage. Apparently

A tidal barrage such as the one proposed and abandoned for Swansea Bay could have the ability to shut off all flow with some sort of sluice gate and open it later so the flow inwards or outwards is deferred but that much larger. This would give a capability for pumped storage but only at certain states of the tide. This might be useful if there was another tidal station which happened to have similar tidal flows at the same times, but for wider use it would only be of use if the demand was foreseen and the tide happened to be at the right state at the time it occurred. Hence I'd guess they probably don't have those sluices, as they wouldn't have any other use I can think of. More recently tidal power seems to be concentrating on undersea turbines that catch passing currents rather than barrages, and I can't see how they could provide any pumped storage function.

 

[AM9]

... More recently tidal power seems to be concentrating on undersea turbines that catch passing currents rather than barrages, and I can't see how they could provide any pumped storage function.

They would have to create artificial tsunamis.

 

[Belperpete]

My probably outdated simple layman's understanding of the big benefits of Dinorwic Pumped Storage power station (where the Festiniog railway was diverted) was that it was excellent for filling shortfall in demand at very short notice (compared with coal or gas fired stations which take quite a while to get their kettle boiling), and that it could continue to do so as long as there was water at the top of the hill and that it could equally well absorb surplus power generation by pumping it back again. It was also supposed to be useful for smoothing out demand between high evening loads (generate) and slack overnight demand (pump). One problem was that it could only do this for that part of the country. Other similar systems aound the country were desirable but the geography isn't as convenient although there are some others elsewhere.

The pumped-storage scheme that needed the Ffestiniog Railway diverting is TanyGrisiau. Dinorwic is its subsequent, bigger brother, and is at Llanberis. As I understand it, Tanygrisiau was primarily built to smooth-out demand between peak and off-peak loads, using cheap electricity to pump the water up, and then letting the water back down to generate expensive electricity at peak times. However, once it was built, its ability to cope with short-term demand was realised and became more important.

As far as I am aware, neither TanyGrisiau nor Dinorwic can be brought on-line instantly, both need notice, as it takes time to open the massive valves, get the water flowing, and get the turbines up to speed. So they are useful in coping with anticipated peak demands, such as the Corry tea-break where they can have the turbines up and running ready, but not for unexpected short-falls. Coal and gas-fired power stations can also be kept on "spinning reserve" in much the same way.

Whilst the grid does allow electricity to be distributed around the country, there are limits. So while the two pumped-storage schemes in North Wales could deal with loss of generating capacity anywhere in the country, they are constrained by the capacity of the grid to get it there.

I seem to recall being told that following electricity privatisation, National Grid is not allowed to generate power, only distribute it. So Tanygrisiau and Dinorwic are no longer allowed to do their original role of generating significant power at peak times - they are now only used to stabilise the Grid by covering for short-term spikes in demand.

 

[Belperpete]

A tidal barrage such as the one proposed and abandoned for Swansea Bay could have the ability to shut off all flow with some sort of sluice gate and open it later so the flow inwards or outwards is deferred but that much larger. This would give a capability for pumped storage but only at certain states of the tide. This might be useful if there was another tidal station which happened to have similar tidal flows at the same times, but for wider use it would only be of use if the demand was foreseen and the tide happened to be at the right state at the time it occurred.

Most of the new green energy generation methods, such as solar and wind-power, are erratic. They mean that the coal and gas-fired power stations don't need to burn as much coal or gas, but they still need to be there to cope with the times when the sun isn't shining and the wind isn't blowing. Tidal is little different, except that you can predict much more accurately when a tidal scheme is and isn't likely to be generating.

There is now a lot of attention being given to batteries, to reduce the number of standby power-stations needed, and also to moderate the load on the grid (so that the grid doesn't have to cope with large swings of power, but is instead fed with a smaller but constant supply).

 

[GRALISTAIR]

So applying all this to railways, I assume the demand would be fairly predictable if a huge electrification programme was announced. Morning and evening rush hours of course , but more or less predictable. I am hoping to try and calculate how many power stations would be needed if 4000 kilometres more was electrified.

 

[Bald Rick]

As far as I am aware, neither TanyGrisiau nor Dinorwic can be brought on-line instantly, both need notice, as it takes time to open the massive valves, get the water flowing, and get the turbines up to speed.

0 - 1.8GW in 16 seconds. Not instant, but prett dam(ned) close!

So applying all this to railways, I assume the demand would be fairly predictable if a huge electrification programme was announced. Morning and evening rush hours of course , but more or less predictable. I am hoping to try and calculate how many power stations would be needed if 4000 kilometres more was electrified.

Well that depends on the size of the power station....however (back of fag packet estimates follow...)

Currently GB rail takes a little less than 2% of the GB grid supply for traction electricity, or an average of (about) a constant 500MW over the course of a year. In reality this will be about a tenth of that between midnight and 0500, but up to around 800MW at peak times.

Whilst only 36% (2018 figure, it may well be 38% now) of the network is electrified, most of the electric network deals with fast and/or frequent and/or long trains, which are more power hungry than most of the non-electrified network were it to be electrified. It is therefore reasonable to assume that approximately 75% of all current vehicle miles are electrically propelled. (I’m sure there’s a stat confirming this out there somewhere, but I can’t find it).

Therefore if the entire rest of the network was electrified, it would represent an extra 33% power required assuming a like for like service frequency / length / speed. Being bold, let’s assume that electrification delivers longer / faster / more frequent trains, so we actually need an extra 50%. That’s equivalent to an average demand of 250MW over a year, peaking at around 400MW.

In context, the government recently agreed deals for three new wind farms, each with a capacity of around 1,200MW. Hinckley Point and the proposed Sizewell C are each 3,200MW. Drax is almost 4,000MW.

Therefore the answer is “as little as 1/10th of a power station”

Some useful data from the ORR here; it’s a year and a half old, and the amount o power used for electric trains will have risen notably in that time due to the roll out of the new GW fleet and further electrification in the Midlands, NW and Scotland.
https://dataportal.orr.gov.uk/media/1114/rail-infrastructure-assets-environmental-2017-18.pdf

 

[InTheEastMids]

There were however bigger surges - notably during England World Cup games, specifically just after we lost on penalties. The biggest power demand change, though, was Princess Diana’s funeral. Because more or

Remember seeing a presentation saying that the solar eclipse in August 99 was the biggest recorded short term demand change... Perhaps the Diana pickup was over a longer period.

I seem to recall being told that following electricity privatisation, National Grid is not allowed to generate power, only distribute it. So Tanygrisiau and Dinorwic are no longer allowed to do their original role of generating significant power at peak times - they are now only used to stabilise the Grid by covering for short-term spikes in demand.

The Electricity System Operator (for GB; ESO) and the Electricity Transmission Owner for England & Wales are both (separate) National Grid companies. They are forbidden from owning generation or storage of any kind.

Dinorwig is ultimately owned by Engie and I expect earns most/all of its money by contacting with the ESO to provide various services to balance the grid