furcifer said:
I have to ask, do you think the costs were calculated properly? I mean if you look at solar, a lot of the "savings" has come from China's production methods, which aren't exactly "eco friendly", or safe.
Agreed. And coal power plants built in China are horribly dirty things, as anyone who has been to Beijing can see for themselves.
There are a few ways to tally costs. One is plain construction cost. That's easy to calculate because people keep careful records of cost (because that's how they get paid.) And here you see numbers like $1000/kw for combined cycle gas plants and $1800/kw for solar-PV.
Operating costs are harder. Natural gas power plants, for example, require natural gas - and the price of that will vary. Solar-PV needs no fuel, ever, and maintenance costs are very low. But they also have "capacity factors" because the sun isn't out all the time, so that reduces the energy they can generate. Nuclear power plants have decommissioning costs, insurance costs and their physical plant takes a huge amount of power - which is power you can't sell to consumers. In addition, there is a cost of money associated with any capital expenditure that has to be taken into account.
To take all these into account, a "levelized cost of energy" is calculated. (That's what the Lazard and BNEF studies concentrate on.) If the owner can sell power for more than the LCOE, then his plant makes money. If they cannot, then it will not succeed. Potential purchasers take this into account when requesting bids on new power plants.
The LCOE isn't 100% accurate of course; it can't be. No one can know what the future cost of natural gas will be, so if a generator banks on natural gas being about as cheap as it is now, and it gets expensive, they may fail even though the project looked good on paper. In addition it does not take externalities into account like health and economic damage from pollution. But it's a good basic estimate of how much the power from a given plant will cost.
Lazard, for example, gives LCOE ranges based on where power plants are built (Niger is cheaper to build in than the US) estimated fuel costs and estimated economic performance in the future (determines how hard loans are to pay back.) From their report in late 2018:
Solar PV rooftop residential $160-$267 ($/MWH)
Solar PV rooftop C+I $81-$170
Solar PV utility scale $36-$44
Onshore wind $29-$56
Gas peaker $152-$206
Nuclear $112-$189
Coal $60-$143
CCGT $41-$74
So far their costs, based on actual bid prices, have been pretty accurate. Utility bids on solar, for example, are going from about $25 to $40 per megawatt-hour, indicating that people are willing to build utility scale solar for that price.