At utility scale unsubsidized solar is cheaper than coal kWh for kWh, albeit with the caveat that solar is not dispatchable, and the coal is not being penalized for the externalities and deaths it causes in the general population. However, except for the health damage, it's difficult to quantify the rest of the damage that coal does without lots of arguments.
That is the strangest analysis of energy costs I've ever seen. Things that are not accounted for in that analysis are (from pg. 1):
- capacity value vs. energy value (i.e. they're just assuming peak utilisation at all times, which is nuts)
- costs related to distributed generation, congestion costs (? land?)
- waste disposal (this includes nuclear)
- intermittency and back-up generation costs (e.g. costs of energy storage, costs of intermittent gas peaker usage)
- transmission costs (which get substantially larger in distributed and intermittent scenarios)
- environmental externalities (e.g. air pollution, resource mining etc.)
- system balancing costs (i.e. not having constant random brownouts)
- PV generation assumed to all be in US Southwest
And so on.... It's hard to draw any sensible conclusions from that analysis, given the assumptions made. They're even assuming piston engines will be used for gas and diesel plants that are sized to produce ~5000 GWh/yr. For reference, ~80% of the world's electricity is produced by steam turbines (not piston engines).
EDIT: Putting all of that aside, if they're anywhere close to the mark with these numbers then wind looks like the clear and unambiguous winner. Also very surprising that solar PV on residential rooftops is claimed to be significantly more expensive than any option... I wonder what's driving the cost?
It's difficult to get single numbers for these, due to the difficulties that you're listing. I'm not sure that it's stranger than any other analysis out there, such as EIA. I believe it's slightly more focused on where to invest if you wanted to build one more of something, which fits with Lazard's other focuses.
It's not surprising that residential PV is so expensive. First off, it's all small custom jobs which means that labor is high, which is one of the larger costs of solar. Second, residential PV is all behind the meter, so it can still be a cost win for consumers since they pay transmission costs from the utility.
Good point. From Lazard's perspective it also probably makes sense not to account for some of these things (e.g. infrastructure cost, site remediation etc.) as it's unlikely these kinds of costs will be borne by a private investor.
I don't have stats for subsidy-less coal, but I'll point you to [1], which gives the unsubsidized installed price of grid-level solar (defined as >500kW) power as being about $2.30/Wdc in 2015. Page 14 has a summary graph of this parameter since 1998.
The report gives pre-incentive prices for residential (<10kW) and non-residential (<500kW) as well.
