Is Snowy 2.0’s call to can coal an own-goal?

Snowy Hydro Limited has declared wind and solar are 'clearly cheaper' power options than coal. This has been reported by Ben Packham in The Australian (Friday, 18 May 2018, page 6). Snowy Hydro is reported to have said its modelling showed the following costs:

• New wind power: $70-$80/MWh, including price premiums for energy storage.
• New solar power: $77-$99/MWh, including price premiums for energy storage.
• New high efficiency, low emissions (HELE) coal generators: $78-$120/MWh.

What is the basis for these findings? One possibility is that these comparisons are between 'plate ratings' of new wind turbines, new solar panels (plus some allowance for storage in both cases) and new HELE coal plant. All new wind, solar and coal generation equipment has such a 'plate rating' or maximum generation capacity. These are expressed in MW, or some multiple or fraction of these. Similarly, batteries have rated maximum capacities, measured in MWh, or some multiple or fraction of these.

For intermittent energy, 'plate ratings' may be convenient for broad capacity comparability. Such measures avoid the need to deal with varying intermittency and uncertainty (eg, by season, latitude, weather cycle and specific location). For new HELE coal generators, this is sensible anyway because such plant can operate continuously. Over time, for all, depreciation will reduce maximum capacity as a % of 'plate rated' capacity. However, intermittency and uncertainty can't be ignored if the plan is to replace reliable fossil fuel power with intermittent renewables like wind, solar and even hydro. Continuous reliable power is needed in all cases.

Assume Snowy Hydro has allowed for intermittency and uncertainty. That means it has allowed for the multiplied generation and storage capacity needed for wind and solar to match reliability of, say, HELE coal. If so, we can estimate the cost per MWh for required wind and solar generation capacity, and the cost per MWh of required storage capacity.

Assume 2016-17 AEMO 'efficiency' (SA) of 29% capacity for wind power and 15% for solar power capacity. Assume a fixed cost of about $7.13 per MWh over a ten-year life for batteries, based on the announced cost estimates for the Victorian regional 'big battery' project. (For both, see my 17.04.18 OLO opinion piece).

Using $7.13/MWh as the battery cost, and allowing for generation and storage capacity multipliers reflecting AEMO 'efficiency' ratings for wind and solar in 2016-17, we can work back from Snowy Hydro's costs cited above to estimate costs for wind and solar generation capacity, plus storage capacity, per MWh, needed for reliable supply to offset intermittency:

• Wind: $15.24-$18.14/MWh (or $20.30-$23.20/MWh, adding storage @ 71% of generation capacity).
• Solar: $5.49-$8.79/MWh (or $11.55-$14.85/MWh, adding storage @ 85% of generation capacity).

These capacity costs look too low to be commercial. Yet they are required to ensure enough reliable generation and storage capacity is available at the $/MWh costs for wind and solar cited by Snowy Hydro. Are the Snowy Hydro wind and solar costings simply 'plate rated' capacity plus 'something' for storage?

Some might suggest Snowy Hydro is 'talking its book' by calling time on coal. Could that be an own-goal? Will 'pumped hydro' be economic, if pumping water from lower storage dams to upper dams can't rely on fossil fuel off-peak power to do the job? There are significant efficiency losses pumping water uphill to fuel the next cycle of hydro power generation. These can be 20%-30%, including pipe friction and turbine efficiency losses. These must be outweighed by bigger price differences (arbitrage) between peak power prices (when hydro power is sold) and off-peak prices (when energy is purchased to pump water uphill).

Snowy Hydro has provided an answer to this arbitrage question, according to a report by Ben Potter (Coal no match for renewables – Snowy boss) on page 9 of the Australian Financial Review for 23 May 2018. He says Snowy Hydro has modelled Snowy 2.0 on the assumption that it would pay about $50/MWh for the power it uses to pump water into the upper storages 'at times of low demand', and charge about $80-$90/MWh wholesale when it sells the energy back into the grid 'at times of high demand'. (The printed AFR version reads '$90-$90MWh', but I assume that's just poor editing, and the lower estimate should be $80/MWh.) Snowy Hydro also apparently assumes a 25% efficiency loss pumping water uphill. So Snowy Hydro is expecting an arbitrage margin of 60-80%, outweighing an efficiency loss of 25%.

Is this plausible? Maybe:

• With current timing of peak power demand, and fossil-fuel base-load delivering supply during current off-peak periods (mainly at night), such arbitrage margins seem quite plausible, and enough to outweigh efficiency losses pumping water uphill (as they apparently are today).
• If we assume current peak- and off-peak user demand timing is unchanged, but supply shifts from coal to gas, what then? Gas has been thought of as a 'peaker' generation back-up energy supply. If coal disappears, do we move to gas 'peakers' to 'firm up' renewables reliability, including pumping water uphill? At Snowy Hydro's $50/MWh, I don't think so. Neither does Snowy Hydro.
• Suppose we forget fossil fuels and move to renewables to re-charge Snowy 2.0's 'big battery'. Suppose also that the current user demand cycles continue. When is 'off-peak', given renewables are re-charging batteries, including Snowy 2.0? For solar, it's no longer at night: that's when solar-charged batteries are being discharged. For wind, it could be any time, depending on the weather. With renewables, we don't control when power supply occurs: the solar cycle and the weather do.
• Today, 'peak' and 'off-peak' still are mainly demand-side phenomena. Fossil fuel supply smooths out price effects of demand peaks and troughs (a bit).
• In a renewables world, power market 'peak' and 'off-peak' periods depend both on demand and supply peaks and troughs. Will these coincide, or be 'out of sync'? The answer will help determine practical arbitrage margins on which the viability of Snowy 2.0 and its ilk will depend.

The solar cycle dictates the middle hours of the day as the best time to re-charge batteries from solar power. In a renewables world, this is a time of peak solar power demand. It's also the time when multiplied PV generation will be operating at full capacity, with most power (85% in the example above) going to re-charge multiplied power storage capacity (batteries). Solar energy is at a maximum, but a lot of expensive generation and storage 'kit' is needed to collect and store it. It's not clear that this 'off peak' period is cheap.

Outside the middle hours of the day, increasingly, solar-powered storage will be discharging. This is solar power's supply-side 'peak' period, when grid-scale battery owners hope for higher prices to cover their re-charge/storage costs. This supply-side 'peak' and 'off-peak' cycle must be overlaid on the existing demand cycle. For demand in a fossil fuel world, 'off peak' was in the wee small hours of the morning. For solar, this is a battery discharge time, not a traditional 'off-peak' time. The night is also a time when the increasing penetration of household and small-scale business storage batteries for PV systems will be discharging most. These will take the top off 'peak' demand for grid power from industrial-scale batteries like Snowy 2.0.

What's the net effect of these supply and demand cycles on highest and lowest power prices? Power prices in the middle of the day generally are significantly higher than in the small hours of the morning. Grid power prices in the small hours of the morning will no longer exploit cheap fossil fuel base-load prices. With renewables, this period becomes part of the 'peak' period, when batteries are being discharged.

It seems likely that arbitrage gaps between highest and lowest prices will narrow quite a lot, and probably around a higher cycle-average price.

Similar considerations apply for wind power, but wind weather cycles are not diurnal/nocturnal. Wind cycles can be longer or shorter than a day. Again, 'peak' and 'off-peak' price cycles will change and, especially for wind, their timing becomes less certain. What happens to the amplitude of wind peak/off-peak price cycles?

Without fossil fuel power, when will 'pumped hydro's' water be pumped uphill? That's its 'battery re-charging' phase. It won't be when the wind doesn't blow or the sun doesn't shine. Such periods will be when discharging hydro power (and other battery storage) is needed for reliability. Are these the new 'peak' periods? Snowy Hydro seems to be planning on that for its viability. In the recharging phase, the wind must be blowing and/or the sun must be shining. Are these the new 'off-peak' periods, albeit when all renewables' storage will be re-charging, in a 100% renewables world?

Is a 60-80% arbitrage margin likely in this new renewables world? If yes, can punters be told, in detail, how?

Ben Potter reports Snowy Hydro claims 'coal plants couldn't respond quickly to sudden changes in the energy market when wind and solar energy come and go in the grid'. I agree. Why has this happened?

• The inherent intermittency of renewables like wind and solar, and the way renewable energy targets (RETs) are applied through the NEM bidding process, means the intermittency features of renewables are also imposed on fossil fuel energy sources – especially base-load. When renewables supply is 'on', fossil fuel supply power bids are 'off'. The intermittency of renewables is mirrored in fossil fuels. That increases their costs (and thus those faced by power users) a lot. They're less competitive.
• Traditional fossil-fuel base-load power provided reliability through 'spinning reserve' and grid stabilisation services because it wasn't stop-start supply. Renewables and RET policies effectively have made fossil fuels intermittent as well, and made them increasingly uncompetitive. This is the 'other blade' of the RET 'scissors', killing coal. (See my opinion piece of 30.01.18, Power failure: some inconvenient renewable energy realities and the links there to my booklet with the same title).
• Increasingly, Australia's total power supply is intermittent and either unreliable or more expensive, or both. We'll need much more (intermittent) generation capacity, and nearly as much power storage capacity, to do the job if we want the same reliability. That means we'll pay more for power, not less.

There are lots of questions about 'pumped hydro' and other renewables. For wind, solar and hydro, even if reliable, (i) unsubsidised economic viability and (ii) user affordability in a world dominated by renewables, are two big questions. I don't think politicians' answers to date are persuasive.