Updated: Solar Photovoltaics (PV) is Cost-Competitive Now
I hear that PV really costs about 40 c/kWh, at least that’s what so many people who have their 2 cents to add to the energy debate have to say about it. And then all of Chicago cringes, and with them, the Obama administration.
I would quit if PV cost 40 c/kWh. After 30 years of working in PV, I would quit.
It’s true that it’s hard to understand what PV costs, since we don’t know what dollars per watt means in cents per kWh, and we don’t know what it means in different locations.
Put simply, there are some locations where PV costs 40 c/kWh; and there are some where it costs a third of that. There is no one price for PV, because sunlight varies, and system costs vary with size and design. Large systems are cheaper than small ones.
So some nudnik from the oil or coal industries can stand up and say, PV is 40 c/kWh and not be lying. And I can say it is 13 c/kWh and not be lying, and all without a cent of incentives, not even traditional depreciation.
But I want to prove this to you, because it is important that we get this straight. The debate must move another notch. Simply put, there are places and PV systems today that can sell electricity at 13 c/kWh, or even 10 c/kWh, and make an adequate return. They are cost-effective at those prices without a cent of incentives, no carbon price, and not even traditional depreciation. And there is a potential for billions of watts of these systems and, as the years go by, a diffusion of their locations from the sunniest to less sunny places.
Assuming the simplest system, a flat-plate CdTe system like First Solar makes, we can expect about the following properties:
- Cost less than $3/W installed (stated publicly by First Solar VP, Maja Wessels, at our GW Solar Institute annual symposium in April)
- Annual AC output in the US SW of about 1.9 kWh/W (DC) installed (including all losses)
- O&M about $15/kW-yr, including insurance and inverter replacement
- Thus one can calculate annual revenue per watt, at
- 13 c/kWh as 24.7 c/W-yr (multiply 1.9 kWh/W-yr times 13 c/kWh)
- 10 c/kWh as 19 c/W-yr
- Calculate annual gross profit (revenue after subtracting operating costs of $0.015/W) of
- 24.7 minus 1.5 c/W-yr = 23.2 c/W-yr (at 13 c/kWh)
- 17.5 c/W-yr (at 10 c/kWh)
- And calculate first-year return on investment as annual gross profit divided by initial investment (300 c/W)
- 23.2 c/300 c = 7.73% (at 13 c/kWh)
- 17.5/300 = 5.83% (at 10 c/kWh)
Present PV systems have warranties for 25-30 years at under 1% per year degradation, so worst case, this cash flow drops 1% per year. Most PV systems are actually expected to do better, and lose less than 0.5% per year. One could argue for a power purchase agreement to sell the electricity for 10 c/kWh and a 0.5% per year inflation clause and get the 5.83% return indefinitely, like a perpetual bond or an annuity with no end date. Right now, I can’t get 5.83% without selling my soul to Goldman Sachs. How about you?
This picture of PV deployment (large systems in the desert) may not be your cup of tea if you want a PV system on your residential roof. But for the US and world it is great news, because it means that PV is cost-effective in the best sunlight. We now have two sturdy, cost-effective weapons (PV and wind in the best resources) against climate change and for energy self-sufficiency. And over the years, the price of PV will probably follow its 20% reduction for every doubling in volume (as it has for 40 years) and get to be cost-effective everywhere, not just in Arizona. But that’s icing on the cake. There are enough sunny places to do the trick now, if we want to use them.
Ken Zweibel
Obviously, present PV systems are NOT “waranteed” for 30 years at under 1% degradation, and the annual AC output in the US SW is nowhere near 1.9 kWh/W (DC) installed. So, no, Solar Photovoltaics (PV) is NOT cost-competitive now, and if it were, why does it require subsidies?
One would have expected that the founding Director of The George Washington University Solar Institute and the former President of PrimeStar Solar would be a bit more informed and/or truthful.
Sorry, I meant 1% per year. The original post stated “Present PV systems are waranteed for 30 years at under 1% degradation, so worst case, this cash flow drops 1% per year. i.e., it included the 1% per year and then proceeded to speak about 0.5% per year increase in revenue from an inflation clause. So, fortunately, the typo didn’t perpetuate to the logic.
The 1.9 kWh/W is simple to get – 2400 kWh/m2-yr divided by 1000 W of power is 2.4 kWh/W-yr. This would be the maximum if there were no losses. Assuming 20% losses including operating temperature, DC-AC conversion and etc, this would be 1.92 kWh/W-yr. I am talking about the best sunlight in the Southwest.
The point of the post is to clearly state that in the best sunlight for the cheapest systems (which are large) PV is competitive at 10 c/kWh in the sense described in the blog – a nearly perpetual dividend at 5.8% (if there is a half percent inflation clause in the PPA).
Thanks for pointing out the typo, and please with any further concerns.
Ken, in real life it doesn’t come so cheap. Output in the SW is more around 1.75 kWh/Wp. O&M is a notch higher. Project costs are at $3.5/Wp (Ms. Wessels $3.0 is probably the EPC price and does not include permitting and land cost)
But most importantly, the WACC in real life is above 8% post-tax, as opposed to your lower pre-tax returns. You end up somewhere around $0.18/kWh, which is still way below $0.40/kWh but still a few years from being competitive.
1. SW output. The output in the SW does vary depending on how well made the system is, especially the inverter efficiency, and also the module technology. Using a FSLR CdTe module gets you about 10% more output per installed W than a typical silicon module. Just go to the NREL Solar Adviser Model and check it out. In Daggett, CA, I can get 1.9 kWh/W-yr using SAM defaults and FSLR modules – but with a better system (less derating, better inverter, you can beat that and in a larger region of the SW).
2. “Real” O&M depends on inverter replacement for the most part. There are other, less-real O&M costs – insurance, property tax, land lease…and all of them are somewhat arbitrary to our current set of rules. This may be the “real world,” but it is also a world whose playing field was defined before the advent of solar energy.
3. Suppose you are right (I can’t read Maja’s mind, so I don’t know) – suppose $3/W is only EPC. Why are we paying much more for land and permitting solar PV? Good farm land in Illinois is $3000/acre, which is under a $1/m2? Assuming there are 3 m2 of land per m2 of module, this is $3/m2 of module area, which for a 10% module is still only $0.03/W – about 0.15 c/kWh (that’s point 15, not 15). If we were not charging for land as a form of taxation on PV, land would be a negligible cost; the same for permitting. So am I wrong to use $3/W? Indeed, in six months, that may be high, given the progress in cost reduction lately.
4. WACC. Yes, normal corporate decisions are made using this sort of WACC (altho I might argue the various rates implicit in it are too high – the thirty year bond is 4% now). But if we had the same tax treatment as Oil and Gas Royalty Trusts or REITs, there would be no corporate tax on PV, and my calculation would be correct. PV is the perfect “royalty trust,” as I have argued elsewhere.
Aside from the technical issues of what designs would give 1.9 kWh/W-yr in what locations, I challenge energy people to think newly about what we should assume the government portion of solar energy costs should be. Instead of nickel and dime-ing solar with land fees, permitting, and property taxes, the government should simply shoulder every non-hardware, non-labor cost as an externality far outweighed by the value of reduced climate change and energy diversity. Don’t load down PV with all the environmental “bribes” paid for by fossil fuels (like those now being paid for the Gulf spill).
Under these circumstances, of the best sunlight and enlightened treatment by the Federal government, 10 c/kWh in large tracts of the US SW is real right now.
Hey Ken, in the 2010 Annual Energy Outlook the Energy Disinformation Administration is forecasting 40 cts/kWh for PV in 2016 (!!!) so you have 6 years left to quit.
the 40 cts result from an assumed CAPEX of over $6000/kWp. Chances are that by 2016 we will be below $2000/kWp for utility systems.
It would be great if the top 10 influential energy agencies around the world recognized the current low CAPEX and LCOE position of PV and the potential for cost reduction. It would have a huge impact on energy policy. So I challenge you in your own country. Can you make the DOE and EIA publish much lower numbers in the next Annual Energy Outlook ? Have them call Maya for a start…
This would be useful:
http://www.eia.doe.gov/oiaf/aeo/electricity_generation.html
Thanks for posting those. Yes, those are among the issues I am concerned about. These are the kind of numbers quoted by 3rd parties, based on the expertise of our Federal government.
I take your challenge to do something about it, and will report back on my progress.
Thanks,
Hi Ken
at U of T in Toronto Canada we are being taught that large scale projects will cost around $0.60/kWh given 400 w/m2 and an efficieny of 0.1 (even though we know efficiencies are already at about 0.2)…this subject is being taught as a Carbon free energy course by a prof interested in green techs.
http://www.earthscan.co.uk/?tabid=102427
check it out…he may need some updating