So, only 750Mw to replace perhaps, if the Bayswater upgrade doesn't cover it. Or, 5 of those big batteries.
Power versus energy.
For those not familiar think of it in terms of power being equivalent to how fast you are traveling in a car and think of energy as being equivalent to how far you have travelled.
Speed x time = distance travelled. Likewise power x time = energy.
Now the original design specifications for Liddell were 2000 MW power, with an availability of 75%.
That is, each of the generating units should on average be available for use 75% of the time and not available 25% of the time. That may sound unreasonable to those not familiar, 3 months downtime each year, but if you've ever seen the sheer complexity and the number of parts involved, and the length of high pressure pipe that's subject to failures and so on, then you'll understand why so much downtime.
Bearing in mind this is 1960's technology and the engineers were erring on the side of caution there - if it works better than expected, if it's more reliable, then that's not a problem. Hence the conservatism.
Bearing in mind there's 4 generating units, that is 4 complete sets of everything which operate independently. So 4 x 500 MW and take any one of them offline and the rest keep running. The only real common aspects are the stacks (chimneys) of which there's 1 stack per two units, and some sharing of conveyors but they do have some redundancy. Noting that a stack is a very unlikely cause of a problem so no real risk there.
Over time, given it's essentially worn out, capacity of the generating units has been de-rated to reduce what had become an excessively high incidence of failure.
The first reduction a few years ago was to 460MW per machine as the upper limit, later reduced to 420MW which remains the limit to the end. In practice however, in real world usage they're running at 320MW each which has avoided problems in recent times.
One of the units was closed just over 12 months ago, so in practice it's a 3 unit station at present with an upper limit of 1260MW and a normal running output of 960MW. So it's basically half shut and has been for a year now.
Here's a chart of Liddell's output for April 2023 up to the present moment:
Black line is total station output and from the 5th onwards it's sitting pretty constantly at 960MW with minor variation. What you see on the 5th is a normal, orderly startup (and the last ever in practice) of unit 2 following an outage. Individual unit outputs being the line at the bottom, generally sitting at 320MW each apart from that return to service and ramp up.
So it's been running at a stable output, but well down from original capacity, as above.
So what's it all mean?
In its present form, Liddell can meet about 8% of NSW peak demand. To the extent that poses a problem, and it could, it's one that's really only an issue during particularly hot or cold weather since, when it's mild, demand doesn't reach anywhere near the peak.
What's more significant however is the energy output and this is a detail widely overlooked by those (media, politicians etc) commenting on the issue.
Liddell is a capacity constrained generation source, not an energy constrained source. In simple terms that means the constraint on its output is the generating capacity of the machines not the ability to run them. That is, there's plenty of coal and water so provided it's not broken or undergoing maintenance then it can run constantly at full capacity if required.
The opposite of that, an energy constrained generation source, simply means the plant cannot be operated constantly due to lack of sufficient fuel, water or other inputs. That isn't necessarily a design flaw but simply making best use of a resource. For example hydro facilities are commonly built with more generating capacity than the available water simply because the intent is to operate them intermittently, storing the incoming river flow in the reservoir when they're not running. Underlying reason = it's a lot easier technically, and also cheaper, to stop and start hydro plant than to stop and start coal plant hence that design approach.
There are exceptions to that. For example Tasmania and the south island of New Zealand both do have baseload hydro operations simply because they've got enough of it to use hydro as the basis of the whole system with anything else being supplementary. Most places don't have that luxury.
NSW however does have quite a few "gotchas" in the system, a relatively large portion of energy constrained plant, and this is where any failure is likely to arise from. Excluding intermittent sources and looking only at those which are centrally dispatched, NSW generating plant as follows with categorisation as follows:
Not energy constrained:
Bayswater (coal) = 2740 MW
Eraring (coal) = 2880 MW
Liddell (coal) = 1260 MW
Vales Point B (coal) = 1320 MW
TOTAL not energy constrained excluding Liddell = 6940 MW
Possibly energy constrained:
Smithfield (gas) = 126 MW
Tallawarra A (gas) = 425 MW
Tallawarra B (under construction, gas) = 318 MW
Uranquinty (gas) = 692 MW
TOTAL possibly energy constrained = 1551 MW
Energy constrained:
Guthega (hydro) = 68 MW
Tumut 1 & 2 (hydro) = 674 MW
Tumut 3 (hydro / pumped storage hybrid) = 1800 MW
Jounama (hydro) = 14 MW
Shoalhaven (pumped storage) = 240 MW
Blowering (hydro) = 80 MW
Hume (hydro) = 58 MW
Colongra (gas) = 724 MW
Kurri Kurri (under construction, gas) = 660 MW
Eraring gas turbine (diesel) = 40 MW
Mt Piper (coal) = 1430 MW
Wallgrove (battery) = 50 MW
TOTAL energy constrained = 5838 MW
So 41% of the NSW dispatchable generation fleet is definitely energy-constrained and 51.5% is either definitely or possibly energy constrained. That's where any serious problem will arise. With running out of fuel (of whatever form, including water or battery charge) rather than simply running out of generating capacity.
Some of this does need some explanation, since you're probably wondering why I've listed coal plant as energy constrained. Going through them:
Bayswater, Eraring, Liddell and Vales Point B are all unconstrained coal plants under normal circumstances. Coal supply problems have occurred in the past, notably at Eraring, but under normal circumstances they've got all the coal and water they need to simply operate.
Smithfield, Tallawarra A, Tallawarra B and Uranquinty aren't subject to any specific gas supply constraint although
their operation does depend on NSW as a whole having adequate gas available and that's at risk. There's some doubt as to the ability to operate these facilities extensively, noting that apart from Tallawarra B none of the others have the ability to use any alternative fuel other than gas.
For the hydro stations:
Guthega has an extremely limited water storage capacity, it stores a mere 0.53% of its annual water use so two days' worth basically, since it has never been fully developed upstream. If it's not raining then its ability to operate is very rapidly constrained to whatever water comes down the river.
Tumut 1 & 2 draw on a very large water storage, Lake Eucumbene, and are in series. Water runs in order through Tumut 1 > Tumut 2 > Tumut 3 > Jounama > Blowering. Long term output, annual, from Tumut 1 & 2 is limited to an average of about 190 MW but in the short term they can and from time to time do run constantly due to that large upstream storage.
Tumut 3 is the big one in every sense of the word. It's the largest hydro station by far at 1800 MW but it's also the largest risk of running dry. Reason being it discharges water about 10 times as quickly as the upstream Tumut 2 releases water. So that makes it extremely possible to drain out the Talbingo Reservoir into which T2 discharges and from which T3 draws water.
Under normal circumstances it works simply because Tumut 3 is run for peak loads and also incorporates some pumping capability to put its own discharged water back into Talbingo Reservoir for re-use. So long as it's only run for the peaks, and can pump during the off-peak, then it works. But in a shortage situation, if the required hours of generation increase and the available hours of pumping decrease, then it's a pretty rapid descent toward draining out the relatively limited Talbingo Reservoir which only holds 36 hours' worth of water at maximum output (based on T2 at maximum discharge into Talbingo, and T3 at maximum draw from it).
Tumut 3's design being one brought about by politics at the time. Long story but it was done instead of fully developing water storage above Guthega and building 3 other power stations in the Snowy scheme. Instead of doing that, it was decided to greatly enlarge Tumut 3 and add pumped storage to it on the basis that it would work using the surplus capacity of coal-fired generation to pump the water back up overnight. Doing so was cheaper and avoided an environmental conflict at the time associated with building the other storages and power stations. Downside = it comes with the risk as above both with T3 itself and with the lack of water stored at Guthega (it also reduced the amount of water stored for the Murray 1 & 2 stations which are electrically in Victoria).
Jounama is immediately below Tumut 3 and simply re-uses typical discharge from T3. It's output depends on Tumut 3's.
Blowering and Hume are power stations as such but they're not operated based on the need for electricity but rather, the need for irrigation water. Blowering simply re-regulates discharge from the Tumut side of the Snowy scheme and stores a full year's worth of water. Hume does the same for discharge from the southern end, the Murray stations, of the Snowy scheme. Operation is driven by farming requirements not electricity requirements - that they do generate is just a bonus. With an added complexity that Hume can be connected electrically to either NSW or Victoria and is physically on the border. So they're not really dependable as a generation source, they're in a similar category to wind and solar in that they're intermittent due to that mode of operation. I guess if push came to shove, the relevant Minister might make a decision to let the water out....
Wallgrove Battery and Shoalhaven pumped hydro are self-explanatory. They are storage systems for peak power, requiring external recharging. They serve a purpose but can and will be drained out if not charged.
So why have I listed Colongra, Kurri Kurri, Eraring gas turbine and Mt Piper as energy constrained? Don't we have enough gas and coal? Well it's complex.....
Starting with Mt Piper well quite simply no, it really hasn't had enough coal, that's been an ongoing problem for years now. 2013-14 was the last year it actually ran to its unconstrained capacity and since then it's been hit and miss with fuel supply. For the short term it can run to capacity so long as there's coal in the stockpile but on an annual basis not really.
Eraring gas turbine is easily explained. Nothing physically limits it but the NSW government won't allow it to operate more than 200 hours a year due to air pollution. It's a regulatory limit, nothing other than that.
Now for Colongra there's more detail. In short it has a limited connection to the main gas system and operation of the power station simply drains out the gas pipeline. It's good for a bit over 5 hours a day and that's it. Anything beyond that requires burning diesel fuel. Trouble is, for constant operation on a 24 hours / day basis the required volume of diesel fuel is about 3.5 million litres per day, all of which needs to be delivered to site by road tankers of the same sort which deliver fuel to service stations. I don't think anyone's ever put an exact figure on how much fuel could actually be supplied in practice but suffice to say it seems pretty unlikely that 3.5 million litres per day could actually be delivered and that being so, the tank would run dry in due course.
Kurri Kurri has a bit more gas but overall same constraint as Colongra at the technical level. Plus the added problem of being capped by legislation to 876 hours a year burning gas + 175 hours a year on diesel.
So if we put all that together then what it means is that to the extent production from wind + solar + Bayswater + Eraring +
Liddell + Vales Point B + the actual operation of Blowering and Hume falls short of total consumption, that requires the gap to be filled by an assortment of facilities all of which have actual or at least potential constraints on their operation. Hydro with limited water, coal plant with limited coal, gas turbines with limited gas, diesel with a capped number of operating hours and storage systems which have to be recharged.
To the extent there's any risk to NSW electricity supply, that risk is of draining out those energy constrained facilities. Ending up with no water left in the dams, no gas pressure left in the pipes and so on. They simply can't run constantly, their operating hours are limited, and that point is overlooked by most who focus on peak power.
That's not saying a failure will occur, but I could certainly present a plausible scenario where it would happen. It's not out of the question, mathematically it's possible. Closing Liddell doesn't guarantee that outcome, but it does increase the chances. It's akin to knocking a few more supports out from under the building or taking a few more rivets out of the aircraft wing - it's weaker now but that doesn't guarantee it actually fails.