Electricity - an ongoing technical discussion

[whole diary]

[broadgage]

I say chrome plate them and adorn with fairy lights. 

Would require a rather expensive transformer to step down 25KV to 230 or 115 volts to power the fairy lights.
And they would dim whenever a train went past, because the nominal 25KV supply varies a lot more than does the normal domestic mains.
A regulated power supply could be used. Transformer intended for 33KV input, so as not to be overstressed when the nominal 25KV approaches 33KV. Nominal 240 volt output, which might average nearer 200 volts actual. Then a switched mode power supply that accepts from 90 volts up to 270 volts AC input, regulated output of 12 volts DC^▸ (Direct Current) to the lights.

[Tim]

I say chrome plate them and adorn with fairy lights. 

Would require a rather expensive transformer to step down 25KV to 230 or 115 volts to power the fairy lights.
And they would dim whenever a train went past, because the nominal 25KV supply varies a lot more than does the normal domestic mains.
A regulated power supply could be used. Transformer intended for 33KV input, so as not to be overstressed when the nominal 25KV approaches 33KV. Nominal 240 volt output, which might average nearer 200 volts actual. Then a switched mode power supply that accepts from 90 volts up to 270 volts AC input, regulated output of 12 volts DC^▸ (Direct Current) to the lights.

Surely you would want the fairy lights to dim as a train went past so as to not distract the driver.  If such a safety feature has been cleverly designed into the electrification system then I take back any criticism of NR^» (Network Rail - home page) I have ever made.  Presumably, diesel traction can be banned on safety grounds.

[Western Enterprise]

Would require a rather expensive transformer to step down 25KV to 230 or 115 volts to power the fairy lights.
And they would dim whenever a train went past, because the nominal 25KV supply varies a lot more than does the normal domestic mains.
A regulated power supply could be used. Transformer intended for 33KV input, so as not to be overstressed when the nominal 25KV approaches 33KV. Nominal 240 volt output, which might average nearer 200 volts actual. Then a switched mode power supply that accepts from 90 volts up to 270 volts AC input, regulated output of 12 volts DC^▸ (Direct Current) to the lights.

I hadn't realised that 25 Kv varied by too much around a 25Kv level.
How much does it wary by?
I presume there would be some sort of voltage thingy on board loco's to regulate the voltage back to a constant level for traction motors.
The National Grid tries to keep the voltage constant, but varies the frequency, does something similar happens here ?
W.E

[Electric train]

Would require a rather expensive transformer to step down 25KV to 230 or 115 volts to power the fairy lights.
And they would dim whenever a train went past, because the nominal 25KV supply varies a lot more than does the normal domestic mains.
A regulated power supply could be used. Transformer intended for 33KV input, so as not to be overstressed when the nominal 25KV approaches 33KV. Nominal 240 volt output, which might average nearer 200 volts actual. Then a switched mode power supply that accepts from 90 volts up to 270 volts AC input, regulated output of 12 volts DC^▸ (Direct Current) to the lights.

I hadn't realised that 25 Kv varied by too much around a 25Kv level.
How much does it wary by?
I presume there would be some sort of voltage thingy on board loco's to regulate the voltage back to a constant level for traction motors.
The National Grid tries to keep the voltage constant, but varies the frequency, does something similar happens here ?
W.E

The nominal voltage rail to contact wire is 25kV or for AT (auto transformer 50kV between contact wire and auto transformer feeder wire). There is a minimum voltage that trains can still run which is around the 16 or 18kV mark although the ECR will get an alarm from the (Grid) Feeder Station if it drops below 22 kV after 15 mins there is also a high threshold of around 28kV again after 15 mins

These values are from my memory so might be different but its the general principle

Traction units are specified to be able to operate at a fairly wide voltage range, this is important for instance on GWML^▸ (Great Western Main Line) if say Didcot ATFS was to drop off line the railway should still be able to operate a full timetable feeding from Kensal Green and Bathampton, traction power systems are designed to N-1, the N bit is the cheap bit the -1 is the expensive bit 

[TonyK]

The National Grid tries to keep the voltage constant, but varies the frequency, does something similar happens here ?
W.E

It's the other way around - National Grid has an obligation within its licence to control the frequency at 50 Hz ^ 1% - in other words, between 49.5 and 50.5 Hz. In practice, the operational parameters are set at between 49.8 and 50.2 Hz, a mere ^ 0.4%.

Frequency falls when demand exceeds supply and rises when the opposite occurs. It is managed second by second by adjusting supply and / or demand. It is for this reason that someone at National Grid watches TV in the main control room, ready to hit switches the minute the shouting dies down at the end of Eastenders. As a million kettles are filled, Dinorwig can be booted up to provide 1.8GW extra power within 15 seconds, before said kettles have been switched on. Or combined cycle gas turbine generators can be ramped up. Prayers can be said to the gods of wind to blow a little harder into the turbine blades of wind farms, although so far this has proven counter-productive. These gods tend to wait until demand is low before kicking off, necessitating constraint payments to shut down windmills. In extremis, a big electricity user such as a chemical works will be asked to shut temporarily, and /or small diesel generators will be started up around the country.

The domestic voltage supplied in the UK^▸ (United Kingdom) is nominally 240V AC, but is declared to be 230V AC +10% or -6%, so 216.2V to 253V is acceptable. Most modern equipment is designed to work perfectly well within these limits and beyond. The actual average is around 242V, more than is needed for most purposes, so wasting energy in use. In supply, however, it is more economic to supply electricity at higher voltages, as less power is lost in transmission. Doubtless similar principles apply to 25KV equipment, perhaps even with bigger margins for variation.

As I write, the frequency is 50.085 Hz according to gridwatch. That gives a real-time summary of the mix of the power supply. Demand is low at 28.98GW, it being after 11pm on a fairly warm night. The frequency is the driver, and is constant across the National Grid. Voltage varies locally across the network according to local use and supply, but within the +10% / -6% limits.

Or so I understand -  I am not a professional, and am not entirely sure why frequency is so hugely important. Someone will explain!

[ellendune]

Unless frequency is kept in close tolerance interconnection with other grids is impossible.  Try and connect supplies that are out of phase and it will not go well. 

I was once told about a power station which was refurbished and the phase meter was incorrect.  When he generator was connected to he grid it was 180 degrees out of phase.  The result was a big bang and the very heavy armature broke through the generator casing and out through the roof of the generator hall. 

[JayMac]

It's the other way around - National Grid has an obligation within its licence to control the frequency at 50 Hz ^ 1% - in other words, between 49.5 and 50.5 Hz. In practice, the operational parameters are set at between 49.8 and 50.2 Hz, a mere ^ 0.4%.

Frequency falls when demand exceeds supply and rises when the opposite occurs. It is managed second by second by adjusting supply and / or demand. It is for this reason that someone at National Grid watches TV in the main control room, ready to hit switches the minute the shouting dies down at the end of Eastenders. As a million kettles are filled, Dinorwig can be booted up to provide 1.8GW extra power within 15 seconds, before said kettles have been switched on. Or combined cycle gas turbine generators can be ramped up. Prayers can be said to the gods of wind to blow a little harder into the turbine blades of wind farms, although so far this has proven counter-productive. These gods tend to wait until demand is low before kicking off, necessitating constraint payments to shut down windmills. In extremis, a big electricity user such as a chemical works will be asked to shut temporarily, and /or small diesel generators will be started up around the country.

The domestic voltage supplied in the UK^▸ (United Kingdom) is nominally 240V AC, but is declared to be 230V AC +10% or -6%, so 216.2V to 253V is acceptable. Most modern equipment is designed to work perfectly well within these limits and beyond. The actual average is around 242V, more than is needed for most purposes, so wasting energy in use. In supply, however, it is more economic to supply electricity at higher voltages, as less power is lost in transmission. Doubtless similar principles apply to 25KV equipment, perhaps even with bigger margins for variation.

As I write, the frequency is 50.085 Hz according to gridwatch. That gives a real-time summary of the mix of the power supply. Demand is low at 28.98GW, it being after 11pm on a fairly warm night. The frequency is the driver, and is constant across the National Grid. Voltage varies locally across the network according to local use and supply, but within the +10% / -6% limits.

Or so I understand -  I am not a professional, and am not entirely sure why frequency is so hugely important. Someone will explain!

You should have been a Physics teacher FT,N! That's pretty much how my excellent Physics teacher explained it many moons ago. Dinorwig's 'Electric Mountain' swinging into action to cover the 'TV pickup' is an impressive feat of science and engineering.

https://www.youtube.com/watch?v=SkIzKGot0Ss

It is somewhat disappointing though that plans for a second 'Electric Mountain' large pumped storage power station on Exmoor were never realised. That would be most handy today.

Oh and don't write of wind generation completely FT,N! Grid wide storage solutions for wind generation are feasible and there is proven technology. It just needs both government and the generation industries to realise that generation isn't the be-all and end-all. If we are to continue building wind generation (and we should, along with other generation methods which use a variable source of energy - sun, tide) then we need to research, fund and build storage too. Compressed air storage is likely the best candidate. The UK has a lot of abandoned mines that could be converted into such facilities.

[Worcester_Passenger]

Is Dinorwig really 1.8GW? Surely 1.8MW?

[ellendune]

!.8 MW would not have much effect.  It is 1.8 GW^▸ (Great Western).

[Worcester_Passenger]

!.8 MW would not have much effect.  It is 1.8 GW^▸ (Great Western).

Sorry - not really awake at 03:30!

[broadgage]

Frequency control of the national grid is important for a number of reasons, though arguably not AS important as in years gone by.

Firstly some types of clock rely on the grid frequency for timekeeping and will become inaccurate if the frequency varies. Not just wall clocks etc. but also time controls for outdoor lighting or for heating controls. This is still important, but more and more clocks use an electronic timekeeping circuit and not the mains frequency.

Secondly a lot of industrial machinery is driven by induction motors the speed of which is closely related to the mains frequency. If the frequency is low then a whole factory is in effect slowed down and may produce say 0.5% less goods but still have the same labour and other costs. 0.5% was said to be many millions of pounds an hour in lost production over the whole country.
Less important these days as we have less manufacturing, and also a lot of modern machines use variable speed drives whereby the motor speed is controlled by an inverter and is not locked to the mains frequency.

Thirdly, a great deal of non-domestic lighting uses fluorescent or discharge lamps on copper/iron ballasts. If the frequency is low extra current is used, significant over the whole country, and still significant despite new installations tending to use electronic ballasts that are immune to frequency variations.

Fourthly, some specialist equipment including older types of audio and video recorders and cinema projectors  rely on the mains frequency to control to playback speed.
A 2 hour recording might run for 119 minutes or 121 minutes depending on frequency, and the pitch of music would be slightly wrong. Less of a concern with the spread of digital technology.

Finally the actual generation and transmission equipment is optimised for 50 cycles and may be less efficient or less reliable at other frequencies.

There is also a certain amount of national pride involved "my national grid has better frequency control than yours" The continental grid has much tighter frequency control than ours, a fact of which the French are inordinately proud.

[stuving]

Broadgage has listed the main uses that were sensitive to mains frequency. The increasing use of inverter and other electronic drives at all power levels might eventually eliminate all of those, but there are still reasons to control grid frequency. National Grid has a licence obligation to keep within that 1% limit "save in abnormal or exceptional circumstances", and I suspect the real importance lies with limiting those exceptionally low - or, worse, high - excursions. I can imaging some factory machinery being very embarrassed at being asked to run more than 10% overspeed.

I can have one more go at getting closer to the way the grid is managed, if you like.

If you have a big generator - say 500MW alternator - in your garden, just to run the Christmas lights, its frequency depends on the rate you turn it and it voltage on the excitation (the current magetising the rotor). If you connect it to the mains, both of these have to match the local grid, or else power flows. But not in quite the way you expect.

If your alternator tries to run too fast, it can't ever get more than a fraction of a turn ahead of the grid, and power flows to the grid depending on its "lead angle". Similarly, if you don't pedal hard enough, power flows back to keep it up to speed but lagging the grid by a few degrees. If the whole grid has more power going in (mechanical or purely electric, adding all generators) than going out (summing everything) then everything spinning locked to the grid speeds up together.

So rate of change of frequency is a sensitive measure of power balance, and it can be monitored anywhere on the grid. NG^▸ (Natural Gas) use it to control power balance for that reason. Note that the size of this spinning inertia keeps the system stable - if it was a lot smaller, the grid would accelerate a lot faster, and the feedback control would be more difficult - get it wrong and serious mayhem ensues.

So, what is the effect of varying the excitation? The generator tries to alter its terminal voltage, which is forced to be the same as the grid's local voltage. Well, no real power flows in or out - but reactive power (measured as MVAr) does. That's a bit mystical, but basically it's a necessary evil in AC systems, but too much of it is wasteful of resources and energy and, ultimately, it's dangerous. One if its effects to to cause local voltage variations, and NG control those by getting the big generators to vary their excitation so as to import or export the stuff (paying per MVArh of it).

[Bmblbzzz]

AIUI^▸ (as I understand it) (which is far less than other people who have already posted) 50Hz is the frequency used throughout Europe, including Russia, and North Africa, regardless of local distribution voltages, precisely so that the voles, anbarons and electric toads can be shunted across borders.

[broadgage]

50 cycles is indeed the standard virtually throughout Europe, and does to an extent facilitate interconnection between neighbouring countries. There are however practical problems in interconnecting too large a geographical area and also problems in the operation of AC transmission lines of more than a few hundred miles.

Therefore many international interconnectors are in fact DC^▸ (Direct Current). The UK^▸ (United Kingdom) has interconnectors to France and to Holland, and others are planned. These all use DC thereby avoiding any need to synchronise the UK and continental systems despite both working at 50 cycles.

UK grid data may found via the link below, showing frequency, total load, import/export flows and from what sources electricity is being generated.

http://www.gridwatch.templar.co.uk/

[TonyK]

You should have been a Physics teacher FT,N! That's pretty much how my excellent Physics teacher explained it many moons ago.

Thank you BNM, for those kind words. My physics teacher hated me, with some justification, and banned me from his lessons for the whole of the term prior to my O-levels. I got a Grade 1 just to spite him, and because I found the subject fascinating.

It is somewhat disappointing though that plans for a second 'Electric Mountain' large pumped storage power station on Exmoor were never realised. That would be most handy today.

There was a pumped storage system in Lynmouth at the end of the 19th century. It worked well, but demand eventually outstripped it, and the Lynmouth flood finished it off. Plans for a much larger scheme were dropped because of cost. There are plans to build a new pumped storage facility in Snowdonia, although it will be a fraction of the size of Dinorwig.

Oh and don't write of wind generation completely FT,N! Grid wide storage solutions for wind generation are feasible and there is proven technology. It just needs both government and the generation industries to realise that generation isn't the be-all and end-all. If we are to continue building wind generation (and we should, along with other generation methods which use a variable source of energy - sun, tide) then we need to research, fund and build storage too. Compressed air storage is likely the best candidate. The UK^▸ (United Kingdom) has a lot of abandoned mines that could be converted into such facilities.

I think storage of surplus wind energy is on the list, right after we have cracked fusion power (under research since 1920, jam promised tomorrow many times). At present, there is seldom any surplus wind power, and when there is, it tends to be because of a lack of transmission capacity. It gets stuck in Scotland, where routes out are in short supply. As the wind power available to the grid increases, the practical effect is a decrease in the use of the backup Combined Cycle Gas Turbine plants, such as Seabank and Didcot. Nuclear and coal plants operate best at flat-out top power, and gas largely compensates for the fluctuations in other sources, including wind. As coal plants are shut down, the amount of power generated by gas increases, being mitigated slightly by wind. As of this moment, almost two thirds of our demand is being generated by fossil fuels, with less than 5% coming from wind. I agree we need renewable energy and, when it becomes necessary, storage, but I reckon we backed the wrong horse and spent the money badly. A lot of people would disagree with me on that, usually at conferences in Brazil or Tokyo, or in the queue for subsidies.

And thank you broadgage, for the very clear explanation, and for helping me to understand for the first time why the interconnectors with other countries are HVDC, not AC. Iceland next! I have friends who work in providing generated power for big rock groups, and other things that fill a stadium, in places where the local supply frequency cannot be guaranteed. That can have massive effects on tuning of electronic gear. I didn't realise that wider industry had similar issues.

This may all seem off-thread - indeed it is - but I think it serves to provide an insight into the complexities involved in electrification of the railway. It is much more than stringing wires above the tracks, and plugging them into the supply. We now know that the frequency of 50 Hz is as good as guaranteed, but the voltage available at any given moment will vary considerably. Two trains on the same electrical supply may each need to draw - what? - upwards of 10 Mw each, causing fluctuations in voltage locally. If regenerative braking is involved, the flow of current will go the other way, and the voltage will have to increase to cope. The infrastructure and the vehicles involved will all have to be capable of managing such extremes, and often in less than ideal ambient conditions, without risking blowing an armature through a roof or similar.

Or so I understand.