# Solar power: a looming environmental disaster?

Had not seen much discussion of this, but of course PV cells are semiconductors with a finite life and will have to be disposed of within a decade or two

China will have the world’s worst problem with ageing solar panels in less than two decades, according to a recent industry estimate.

Lu Fang, secretary general of the photovoltaics decision in the China Renewable Energy Society, wrote in an article circulating on mainland social media this month that the country’s cumulative capacity of retired panels would reach up to 70 gigawatts (GW) by 2034.

That is three times the scale of the Three Gorges Dam, the world’s largest hydropower project, by power production.
https://www.scmp.com/news/china/soc...panels-are-going-be-big-environmental-problem

Related General Discussion News on Phys.org
russ_watters
Mentor
I don't see where the issue is, and expressing it in terms of discarded power output seems odd. Are they saying there is a waste issue? A resource issue? Neither seem likely to me, even if the panels are not recycled.

Bandersnatch and fresh_42
I took it as meaning the cost of recycling was uneconomic and they were likely to just be dumped somewhere, analogous to the current problem with PCs and other electronics

“We can sell them to Middle East,” said the manager who requested not to be named.
These panels are mostly ageing, not breaking down. On mid-term I'm also about getting some cheap, used panels since they are cheap and despite having ~15 years in them they are still producing ~80% of their nominal power. Efficiency would be ~12% instead of ~15% - I just can't see why should I care.

Ps.: We'll have to rebuild the roof of our house within a few years anyway. If the price of used solar panels can make it cheaper than roof tiles... Well

tweedledum, BillTre and berkeman
expressing it in terms of discarded power output seems odd
I agree, it would make more sense to say "XX square meters" or "WW kilograms" of panels will need to be disposed / recycled / dumped somewhere...

russ_watters
berkeman
Mentor
I skimmed the article, but didn't quickly see what the degradation mechanisms were for the silicon (erosion of the clear envelope material is straightforward, but not too hard to fix). Is it an increase in discontinuities in the silicon from heat or UV causing the reduction in efficiency?

It does seem like there are several opportunities for patents here, with the goal of extending lifetimes and improving recycling efficiency...

fresh_42
Mentor
... they were likely to just be dumped somewhere ...
If asked whether to live next to a pile of solar panels or a depony for nuclear waste, I'll surely vote for the panels.

BillTre
russ_watters
Mentor
If asked whether to live next to a pile of solar panels or a depony for nuclear waste, I'll surely vote for the panels.
I'm not sure I would, if the nuclear waste was in a small building I couldn't even see from the street and the solar panel dump site was a mountain.
A good 360W solar panel has dimensions of 1m x 2m x .04m. It will generate an average of about 90 watts in good conditions. A typical nuclear reactor is 1,000 MW, so you need 11 million of the panels to equal the output of a nuclear reactor. After you discard them, if you arrange them neatly on a football field, the stack will be 4,400 panels or 176m high - about 10m taller than the cooling towers of a nuclear plant....or two towers (cooling and solar panel), one for each reactor.

If the hypothetical solar panel towers are just inert solar panels, I don't think I'd mind much living near them, but I'd certainly rather live next to my nuclear plant than basically any other type of waste dump you can think of - landfill, sewage treatment plant, coal ash pond, etc. I'll take my nuclear waste over any of that.

Strictly speaking, I *do* live just a few miles from a nuclear waste storage facility, on the grounds of the Limerick nuclear plant.

Last edited:
tweedledum, gmax137 and BWV
After you discard them, if you arrange them neatly on a football field, the stack will be 4,400 panels or 176m high - about 10m taller than the cooling towers of a nuclear plant....or two towers (cooling and solar panel), one for each reactor.
I'm new here and I'm enjoying the company of folks who *think*. That being said, Russ points out a deficiency in John Q Public"s thinking which needs to be addressed, namely, <b>Innumeracy </b>. I've noticed many teachers, tutors, and professors here and were *I* teaching science I'd make this book required reading:

https://www.amazon.com/dp/0809058405/?tag=pfamazon01-20&tag=pfamazon01-20

One of my pet peeves about solar power is twofold: the absolute amount of power which goes into the production of a single solar panel, and the environmental effect of said manufacture vs. that of say, nuclear power.

russ_watters
russ_watters
Mentor
Expanding:

FYI, the waste storage situation at Limerick and in general in an old article from 1997:
https://www.baltimoresun.com/news/bs-xpm-1997-08-15-1997227147-story.html

In reading the article I had to double-check several times to make sure it really was a 1997 article; it reads like a 2017 article. All of the issues ar exaclty the same.

As of 2012 Limerick had 1,143 metric tons of waste. Let's call it an even 2,000, 7 years later. Uranium's density is 19g/cc, so that's 105 cubic meters. Let's say with the containers they finally get stored in they get packed and stacked at a 10% efficiency. That would cover the football field at an odd 21cm high. Or more realistic, a room of dimensions 5x5x4.5m; A decent-sized living room.

That solar panel has a mass of 26kg, so each tower of panels has a mass of 286,000 metric tons; roughly equal to 3 aircraft carriers.

russ_watters
Mentor
One of my pet peeves about solar power is twofold: the absolute amount of power which goes into the production of a single solar panel, and the environmental effect of said manufacture vs. that of say, nuclear power.
It is unfortunately difficult to find reliable information on that. Most of it comes from "advocates" who focus on the downside of one without giving the equivalent downside of the other. I tend to believe the downside for solar there is under-reported based mostly on the fact that nuclear is big so it's easy to pick on; typically you find comparisons between a pretty/small solar or wind farm and a big ugly nuclear plant, but no mention of the fact that the nuclear plant is producing 100x as much power.

russ_watters
Mentor
Nuclear is more expensive than Solar to build, if you look at table 1b in the link below, the Levelized Cost of Electricity is in the $40-$50 per MWH range for Solar, Onshore wind and Combined Cycle Gas but $77.5 for nuclear... Those numbers, I would think, do not include provisions for dealing with the intermittency. The numbers for solar would look a lot less attractive if the purchase price of an MWH of solar power included the 3MWH (guess) of nuclear or natrual gas power required to back it up. The USA can ignore this brick wall for a few more years, but "high" (~20%) countries like Germany have already hit it. According to https://cleantechnica.com/2018/03/25/solar-power-energy-payback-time-now-super-short/ Energy payback for Solar is now 1-4 years, depending on the technology and location When I was taking the CEM course, the instructor said that if people recognized how good a 4 year payback was, they'd mortage their children to invest in energy conservation. Besides the previous objection (which small scale adopters don't and probably won't ever need to deal with), I'd like to see the fine print of what level of subsidies this includes. In some cases, the solar subsidies are extreme, but that can't last much longer (and in many cases has already ended).  Oh; that's "energy payback", not financial payback. You switched metrics on me. I'm not very interested in "energy payback". Those numbers, I would think, do not include provisions for dealing with the intermittency. The numbers for solar would look a lot less attractive if the purchase price of an MWH of solar power included the 3MWH (guess) of nuclear or natrual gas power required to back it up. The USA can ignore this brick wall for a few more years, but "high" (~20%) countries like Germany have already hit it. But that is not really a cost - just that Solar cannot be a sole source of electricity. The current mix of solar, wind and CC Natgas solves these problems. When I was taking the CEM course, the instructor said that if people recognized how good a 4 year payback was, they'd mortage their children to invest in energy conservation. Besides the previous objection (which small scale adopters don't and probably won't ever need to deal with), I'd like to see the fine print of what level of subsidies this includes. In some cases, the solar subsidies are extreme, but that can't last much longer (and in many cases has already ended). this is just referring to the payback of energy used to make the solar panel vs. energy generated by it, not money spent, so subsidies would not enter into the calculation. russ_watters Mentor But that is not really a cost - just that Solar cannot be a sole source of electricity. That depends on who you ask. Many "renewables" advocates claim that "renewables" can be the sole source of electrical energy. Or the other side of the coin, that "clean" can be accomplished without nuclear. The reality is that the fraction that intermittent renewables can do without backup is so low that they cannot be a very high fraction of the energy mix without it.  Actually, even if we assume a hard limit at 20%, you still have to include the back-up in part, if you are doing the overall analysis. The reason is that a natural gas, coal, nuclear, waterver (usually natural gas today) plant is built with a certain assumption about how much energy it is going to produce, and therefore how much it costs. If you, after the fact, add another power producer to the mix, you have to include in its cost the adjustment you just made to all the other power producers' costs. Otherwise the total doesn't add up to 100%. this is just referring to the payback of energy used to make the solar panel vs. energy generated by it, not money spent, so subsidies would not enter into the calculation. Yeah, I caught that later (see edit). I'm not sure I see the value in this metric. Last edited: But nuclear is not a complete solution either, as it only provides base power and cannot be ramped up or down in response to demand fluctuations. Solar and Wind are base generators that complement eachother as wind tends have higher output at night, then gas filling the gaps The energy breakeven is valuable in that is shows why, say, Hydrogen is not a power source or how biofuels are a boondoggle Nuclear is more expensive than Solar to build, if you look at table 1b in the link below, the Levelized Cost of Electricity is in the$40-$50 per MWH range for Solar, Onshore wind and Combined Cycle Gas but$77.5 for nuclear
I don't really know how those LCoE numbers are calculated. Do they account for the actual lifetime of the plant? Nuclear units originally licensed for 40 years have (almost all) extended to 60 years and we are now working on 80 year licenses. A plant that started up 1978, that is still operating in 2058 will look like a real bargain, even though the 4 billion original capital cost was seen as stunningly high in 1978.

But nuclear is not a complete solution either, as it only provides base power and cannot be ramped up or down in response to demand fluctuations...
Yes they can. The submarines and aircraft carriers do it continuously. The stationary electric plants can do it too, within limits (daily load cycle capability was a design requirement for some of the later units). It is, admittedly, more complicated -- with xenon building in and burning out, but it works. The power companies don't operate them that way, because once running they are the cheapest generating source in their fleets. Deregulation and subsidies have changed that equation somewhat, but the low fuel cost of nuclear helps a lot.

Astronuc, Rive and berkeman
I don't really know how those LCoE numbers are calculated. Do they account for the actual lifetime of the plant? Nuclear units originally licensed for 40 years have (almost all) extended to 60 years and we are now working on 80 year licenses. A plant that started up 1978, that is still operating in 2058 will look like a real bargain, even though the 4 billion original capital cost was seen as stunningly high in 1978.
They usually assume some cost of capital and time value, so the \$4 billion original cost is amortized at, say 7%

russ_watters
Mentor
But nuclear is not a complete solution either, as it only provides base power and cannot be ramped up or down in response to demand fluctuations. Solar and Wind are base generators that complement eachother as wind tends have higher output at night...
I'm sorry, but none of that is really true and/or relevant. Nuclear can be throttled, but even if it couldn't (I'm not sure how well/fast it can), the fuel cost is a small fraction of the energy cost, so you could just dump it if you couldn't find a use for it and it wouldn't change the cost much.

Solar and wind, even if they "tend" to compliment each other cannot be designed to compliment each other because they don't always compliment each other. There will be days where there is neither wind nor sunlight, and in developed countries we don't accept grids that black-out for several days/weeks a year. To me, that's one of the defining characteristics of a "developed country" and an utterly non-negotiable most important feature of our power grid.
...then gas filling the gaps...
I agree, but again, I'm an environmentalist, not an "environmentalist". I do not support the Green New Deal, for example, which states a goal of 100% "clean, renewable" electricity. That would prohibit nuclear and natural gas and provide us a grid that worked...most of the time.
The energy breakeven is valuable in that is shows why, say, Hydrogen is not a power source or how biofuels are a boondoggle
IMO, that's too self-evident to be useful in most cases. If someone doesn't already know that hydrogen + oxygen <=> water + energy adding a metric to quantify that 1=1 is just an extra layer of complexity. Fuel is fuel and storage is storage and frankly, anyone who doesn't know the difference shouldn't be discussing energy. We ban perpetual motion machine discussion here because the subject is too basic/silly/crackpotty to bother with. About the only place I've seen it applied, with some value, is to oil extraction. But even then I've mostly seen it used for the purpose of failed Peak Oil predictions.

To put a finer point on it, I have no idea how the EROEI of solar compares to nuclear and I can't fathom why I'd use it in a decision on which to build next/more of. Our energy source decisions are made mostly on the economic basis and the rest on the political basis.

Bystander
Astronuc
Staff Emeritus
Nuclear can be throttled, but even if it couldn't (I'm not sure how well/fast it can), the fuel cost is a small fraction of the energy cost, so you could just dump it if you couldn't find a use for it and it wouldn't change the cost much.
Typical LWR ramp rates are about 3%/hr, but some plants can do 5% to 10%/hr (typically above 40% of full power for PWRs), or higher depending on the power level and other conditions. Some French PWRs have done (and may still do) daily load follow and/or frequency control.

One German PWR had the following with regard to power ascension (from a study I co-authored 20 years ago).
• Ref.: PCI-RELEB. Limit 465 W/cm
• Ramp: 1%/min from 1 to 70% (4.6 kW/ft core average LHGR), while some others used 0.5%/hr.
• "No Limit" above 70%

It's a matter of conditioning and de-conditioning the fuel. Some new fuel designs may eliminate PCI restrictions on the fuel, then the plant can maneuver as fast as the turbine or balance of plant will allow.

gmax137 and russ_watters