I've used Lazard's LCOE (levelised cost of electricity) estimates for wind, utility-scale solar, gas and coal unchanged. Note that these estimates exclude both subsidies and the implicit costs imposed on society by pollution, including emissions of carbon dioxide. Lazard assume for all their calculations an 8% cost of debt and a 12% cost of equity. If we were to assume half these rates, the cost of renewables would fall proportionately more than the costs of gas, coal and nuclear because in the case of renewables almost all the cost is upfront, whereas other technologies have maintenance and fuel costs, which being further into the future have less impact on the LCOE the higher the discount rate. The corollary is that in countries with higher interest rates, perhaps because they have greater political risk, higher inflation, or less access to capital, nuclear and coal would appear relatively more attractive.
For CSP (concentrated solar power), I've used public announcements about Solar Reserve's PPAs for new CSP plants around the world. Lazard gives higher figures, but they don't seem credible in the context of actual signed PPAs over the last year to 18 months. Note that the most recent signings (Chile & Australia) have been at the lower end of that range, and that the PPA (subsidised) of their first CSP plant in Nevada was US$135/MWh, which compares with average recent PPAs of US$67/MWh. Big decline over 5 years.
For wind plus power-to-gas (storage) I've assumed that the power-to-gas part costs 3 times the cost of wind alone, but is only needed for 20% of the time (10 weeks a year), based on previous work I did. That is, the costs assume there is no wind at all for 20% of the days each year, or half average wind for 40% of the year. They also assume there is already a natural gas grid and peaking power gas plants, which is true in most developed countries as well as many developing countries.
Lazard does not include the cost of de-commissioning in its estimates for the LCOE of nuclear, so I've added an arbitrary 10% to their estimates. I think Lazard's costings are based on the new generation IV Westinghouse AP1000 reactors that were being built at Vogtle in Georgia, USA. Westinghouse has gone bankrupt because of the huge cost overruns at Vogtle, so I suspect Lazard's estimates for nuclear are too low. By comparison, the new reactors at Hinkley Point in the UK (10 years behind schedule!) have a PPA of about $140/MWh (£92.50 in 2012 prices), but that's not fixed. It rises in line with yearly CPI inflation. (All the costs of renewables in the chart above are fixed in nominal terms.) Adding inflation indexing to the PPA significantly increases the effective LCOE. Also, the PPA for Hinkley Point does not include the cost of de-commissioning, nor the benefit of government loan guarantees.
Natural gas prices in the US are lower than in other countries, so outside the USA, gas would be more expensive than shown. Even though natural gas produces half the CO2 emissions of coal when burnt, unburnt methane is 28-36 times as potent a greenhouse gas as CO2 is*. Leakages from gas extraction and distribution can in principle be reduced to close to zero, but that will clearly up the cost of gas. Also, the momentum against fracking is building. It's very clear that it's environmentally very costly. Expect the price of gas to keep on rising.
I have included the cost of batteries to "firm" the power from wind or solar (solar only shown in chart). Extrapolating from the costs of the Tesla battery bank in South Australia, the cost of batteries is at best US$72/MWh, assuming a 15 year life. So for 8 hours, that would add at least $24/MWh to wind or gas. However, the net cost would be less, because batteries, like CSP, could store power when it is abundant and cheap and release it when it's expensive as well as providing frequency/oscillation control and current flow. I've assumed a net $20/MWh.
I won't say my calculations are definitive, but the chart is telling, isn't it? It's abundantly clear what the shape of the electricity grid will be--all the technologies on the left of the chart, except for gas, because gas still adds to global warming, and anyway is more expensive outside the USA than is depicted in the chart. Even in the USA, the prominence of gas in the grid is unlikely to last.
As we replace aging coal and nuclear power stations, it is obvious it will be with wind, solar, CSP, batteries and power-to-gas. Not just because we want lower emissions and cleaner air but because they will be cheaper. This gives the lie to the oft-repeated claim that switching to renewables would cause the economy and our living standards to collapse.
* Nitrous oxide (N2O) is 265-298 times as potent a greenhouse gas as CO2. N2O is produced in agriculture (67%) and by burning fossil fuels or bushlands and forests. Don't confuse it with NO2, or nitrogen dioxide, which is a potent pollutant produced in car engines by burning petrol or diesel.
No comments:
Post a Comment