Inrush currents en mass?

[emeritus]

Circa 1970 a cascade of over-current trips blacked out a large part of the electricity grid of the North-East coast of the USA, and it took some time to get it back on line. I read that, in at least one case, the power company had to borrow dozens and dozens of car batteries to provide the DC power required to excite their alternators, the DC normally being obtained by dynamos driven from the then-dead AC mains supply.

[broadgage]

Originally Posted by kalee20

All this makes me think what would happen if the National Grid ever crashed - or even separated in two. It's never happened yet although I understand the procedures exist, but obviously can't be tried!

I guess restarting would be getting one power station up and running, at the right frequency, bringing another station up, syncing it in frequency phase and voltage and switching in, then another (bearing in mind that some power stations suck a fair bit of power till they're established), connecting some users, then bit by bit adding more generation and more load.

Syncing two separate sections of Grid, if it ever split, sounds nigh-on impossible. One would have to be progressively shrunk as its generators and users were isolated and reswitched into the other.

Such events are very rare, but HAVE occurred. Following the great storm in 1987, there was a regional black start of the national grid area serving London and the South east. This ran as an "island" separated from the rest of the country, for some hours. AFAIK re-synchronising the two parts of the grid was achieved without undue problems. Simply adjust the frequencies to be very nearly equal, wait until the two parts are in phase and close the circuit breaker. Once conditions are stable, then close the circuit breakers on other grid lines, so as not to be reliant on a single connection.
Synchronising a large power station generator to rest of the grid is generally automated these days, but I would expect the duty engineer to know how to do this by hand if need be.
Synchronising two portions of the grid would probably by done manually, by a senior engineer. A very slight frequency difference between the two portions is desired, such that the two portions drift in and out of synchronisation every few seconds. Most engineers avoid synchronising on a "standing phase" that is with no difference between the two parts of the system, for fear that something has gone wrong with the instruments.

[G6Tanuki]

These days I doubt there would be any manual intervention in the re-syncing process; the old-style two-bulbs-that-you-get-to-equal-brightness, or the later oscilloscope type Synchroscopes, were only accurate to within a few degrees of phase-difference; these days I would expect there are bunches of microprocessors involved, running mathematical models of the load/rotational-speed curves of the alternators and with realtime feedback of steam-pressure/gas-feed to the turbines.

Many of the large coal/biomass/CCGT power stations also have co-located OCGT generating capacity [typically a couple of Rolls-Royce Avon gas turbine engines more commonly found in warships] that can provide site power. Similar arrangements are also used in nuclear stations to provide continuity of cooling if the site has to go off-load suddenly.

A sort-of friend works for RWE in Germany where he looks after 1.2GW of combined-cycle-gas-turbine generating plant. Next time I see him I will ask how they do it!

[kalee20]

Originally Posted by broadgage

Following the great storm in 1987, there was a regional black start of the national grid area serving London and the South east. This ran as an "island" separated from the rest of the country, for some hours. AFAIK re-synchronising the two parts of the grid was achieved without undue problems. Simply adjust the frequencies to be very nearly equal, wait until the two parts are in phase and close the circuit breaker. Once conditions are stable, then close the circuit breakers on other grid lines, so as not to be reliant on a single connection.

Simple in principle - but syncing a mini-grid to a bigger grid? How to adjust the frequency of the incoming 'island' with its maybe half a dozen power stations, so 20 generators all fighting each other against any speed change... bringing them all up or down to speed together? And having got the frequency right, then adjusting excitation, again all together, until the voltage is equal to the master grid... and then throwing the switch with any unbalance current measured in kiloamps?

I can imagine the switch-on thump...

[Wellertcp]

I used to work with a chap who's background was in power-station/sub-station/large commercial building power control and distribution.

Nowadays, loads are controlled, triggered and monitored by PLCs activating ACBs (air or oil circuit breakers) with millisecond timing. All controlled by highly-complex software subroutines that he used to write (Asian guy, he was something of a genius)

So in a sub-station, or a large commercial office building (City or Canary Wharf type thing) there is a precision controlled and timed start-up sequence from 'blackout' to prevent an 'all-on' situation.

And just to add, all the PLCs are '2N' dual-redundant, so if one fails, another takes-over instantly, and UPS and battery backed, etc. And, of course, all monitored 24/7.

[emeritus]

We never lost our mains power at all in the 1987 storm, although the phone line was dead. My mains electric bedside alarm clock had kept perfect time, while my wife's bedside digital clock (red LEDs) didn't have a backup battery fitted and its display would have been flashing and showing the wrong time if there had been a power interruption. We had slept soundly through it and in the morning were surprised to see the fallen trees blocking both ends of our road.

[trobbins]

One way to look at it is that each generator facility, whether that be a PV domestic inverter, or a large battery or gas generator of whatever facility, is either designated as a grid follower or grid former (ie. typically a grid follower). So most generator facilities have a key role of detecting if there is a grid or not, and disconnecting if not. The rules of engagement are all mature and are increasingly more onerous as power level and connection rail voltage increase. Some generators can swap between functions, of following versus forming, and this is all highly regulated and requires dedicated communications.

One striking example in Oz was when a large windfarm disconnected due to repeated grid transient disturbances (due to distribution towers going down, and lightning). The disconnect switch had been programmed to stay connected for up to 6 or 7 substantial disturbances, but then finally disconnected. On review, the programming of that switch logic should have allowed for one more disturbance (as I recall)- so a mistake from the windfarm operator, but it had obviously got past the onerous checking and inspection and approval process that has to happen during commissioning. The consequence of disconnection is the main concern, as the remainder of the grid has to take up the slack, and so some frequency slip occurs, and can cause consequential domino type trips (as per USA huge outage).

The protection relays that automate such facilities are exotic beasts. For the mundane mass market of small generators, the prescriptive connect/disconnect settings are all mandated, and equipment can't be sold in a country unless compliant. That all has a cost, especially for the low power end of the market, but has pretty much avoided islanding risks, although the anti-islanding mechanisms used by different manufacturers are not the same, and can cause quirky hassles/grief especially for weaker grid situations.