Is This The Best Solar Chart Yet?
Originally published on Sustainnovate.
By Henry Lindon
Solar energy has been in a bit of a boom phase as of late (one that will arguably continue for a long while). Why exactly, one might ask, has the technology been finding more and more of a foothold in the global marketplace? Well, the graph above (courtesy of the Earth Policy Institute/Bloomberg) puts it more succinctly than I could, so have a gander.
That really says quite a lot, doesn’t it? Solar energy technology has been on quite a journey as of late, and seems set to continue to do so for at least another decade or so. (Diminishing returns on research investment will likely begin to hit in a significant way by that point, in my opinion.)
As one can see, the price of solar photovoltaic (PV) was near $100 per watt back in 1975, and total global installations totaled only around 2 megawatts (MW). Things have changed quite a bit since then, with the price per watt in 2015 being around 61¢, and total global installations totaling around 65,000 MW (65 gigawatts). I don’t know what’s a more astounding way to look at it: with the chart or just the numbers.
Of course, the cost of the promising technology fell rapidly right from 1975, but global solar installations – the blue bit – barely registered until that magic moment somewhere between 2000 and 2005, when price per watt reached a tipping point and the blue bit quickly soared to a total of just under 65,000 MW in 2015.
As TreeHugger’s Michael Graham Richard notes, “the beauty of having exponential growth on your side is that very quickly, even the current blue spike will look tiny. In 2020 or 2030 we’ll look back on 2015 and it’ll barely register as the beginning of the curve on the chart.”
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
CleanTechnica Holiday Wish Book
Our Latest EVObsession Video
CleanTechnica uses affiliate links. See our policy here.
The right hand axis is a ‘per year’ value. Cumulative total is much higher, about 240 GW (240,000 MW).
Good. I was just thinking even with exponential growth we’re screwed if that’s a cumulative value.
Ross, PV is currently at a 44% CAGR (last 10 years). This would be 9,000+ GW by the year 2025 (9 TW or 9,000,000 MW)! We might need grid storage by then 🙂
Or a bigger grid 🙂
Or a lot of electric car charging, which i guess is effectively storage.
Do you have a link for that CAGR figure? I was looking for the worldwide CAGR for PV over the last several and the CAGR numbers I found were a lot smaller than 44%. Even 21% for 10 years would get us to where we need to be.
It was 44% up to 2014. One or two poor years will have lowered it a little, but not much. Sources and spreadsheet attached to this old blog post of mine, towards the end (link)
It is just possible that the last few bumpy years mark a downwards tipping point to a lower “mature” rate of growth.But beware of the cognitive bias of conservatism: higher than trend years become exceptions, below trend years reversions to the normal.
Ross: >> PV is the fastest growing segment with 44% compounded annual growth (CAGR) in installations between 2000 and 2014. <<
https://www.bbvaresearch.com/wp-content/uploads/2015/09/U.S-Industry-Analysis_Bright-Prospects-for-Solar-Energy-Report1.pdf
Agreed, wind is 20% CAGR but solar is at least twice that.
I was wondering why on earth someone uninstalled a pretty huge amount of capacity in 2011 😛
A reduction in the number of panels installed per year. The cumulative number of panels installed is the area in blue.
Which was it, a tipping point or exponential growth? In the purest form, an exponential increase is the same at every point on the curve, as in the classic fable of the chessboard and the grains of wheat or rice. A tipping point is a change in the rate of growth: as with the recent plateau in carbon emissions. The solar growth curve is pretty smooth, and the exponential story fits better.
A very few hardy souls have predicted a tipping point in solar once it reaches grid parity, and an acceleration in installations. I can’t bring myself to believe this – defenders of historic growth rates like me are already outliers. But the theoretical case is pretty strong.
Ideally a log plot would be used, then exponential change shows up as a straight line.
Except that those are read incorrectly by the majority of people, who don’t understand such matters well enough.
Interesting thought and now that you say it, I realize that I agree. Having said that, these graphs with linear axes serve a useful purpose for dramatic effect and promotion and are useless for seeing into the future. A log axis makes it crystal clear to see into the future. (Apparently the BP analysts who ran that report recently don’t understand how to use them either….)
The guys at BP don’t want to understand, because it’s such bad news for them.
Sadly, this is too true… but for those of us who do, they’re incredibly useful! (Could we possibly use both?)
True, but for the same reason a chart or cumulative install would have been better. As you can tell from the comments about. 😉
Yearly exponential growth actually looks slow in a neighborhood. You should see it Nationally in 2 years.
First you see 1 house in year 1, then 2 in year 2, then 4 in year 3, then 8 in year 4, then 16 in year 5, then 32 in year 6, then 64 in year 7, then 128 in year 8.
So, suddenly in 8 years the energy revolution is over.
“A very few hardy souls have predicted a tipping point in solar once it reaches grid parity, and an acceleration in installations.”
Doesn’t require a particularly hardy soul. Especially for business properties, this tipping point is the most likely scenario. If another landlord is able to supply tenants with cheaper electricity than what I can get from the utility, I either have to sell electricity at a loss, or reduce my rentals, or go ahead and install solar.
Or you could think of it another way. Assume I have a million bucks sitting in the bank getting a return of 5%. I could invest in another building, for a 9% return, or I could install some solar on my existing buildings, for a 12% return.
Hmmm….
Once it makes sense, and once there are a few “guinea pig” installations up and running in an area, and suppliers/installers have sorted out the teething problems, pretty much everybody has to get on board, or get left behind.
Just wondering what the prices will be in say 2020, less than 50 c/watt and in 2025 30 c/w?
And what a third graph would look like total money invested, as in 2 MW at $ 100 W in 1975 to 65 GW at. 61 c/w in 2015,
I did a piece on that for a publication of The Economist Group, reposted (the original version) here: http://cleantechnica.com/2015/12/05/solar-power-prices-low-future/
Here’s a key bit from that:
In fact, [BNEF Manager of Solar Insight Jenny Chase] projects that the price per watt of a crystalline silicon solar panel will drop from 62 cents today to 21 cents in 2040, when it will still lead the market. The cost drop will come from “incremental improvements in crystalline silicon technology (thinner wafers, better-shaped busbars, better AR coating, more targeted doping, better contact technology).”
Thank you. 🙂
Am I correct that in 1975 at $ 100/W investment was $ 200 million for 2 MW installed and 65 GW at 61 c/W investment was $ 3.8 billion or 38 billion?
Total world wide investment was, with panels about 40 % of the system, $ 38.4 billion, about $100 billion.
Not bad for a few hippies!
Growth from zero will look impressive. But the real question is how far have we come in relation to where we need to be – and are we on schedule.
Then things don’t seem so rosy.
It’s called cherry picking facts, you just pick the one curve or statistic that proves your point and ignore the rest. People dont need to know the whole picture.
There is no “ignore the rest” to ignore. Those are the growth rates.
Progress is uneven. California is way ahead of the rest of the nation in solar. Arizona, Texas, Florida, Nevada, New Mexico, and many others could do much more. China is adding much more wind than the US also. In general, US lags many other areas of the globe, despite being a tech leader.
To be fair, though, China is 4.3 times larger than us…
(Not that we have an excuse… but our per capita renewables is still way ahead. Now if you want to talk about Europe or some Canadian provinces… )
To limit climate change on any reasonable proposal, the world needs about 5TW of solar and about 5TW of wind by 2035 or 2040. If annualized growth of 30-50% per year for annual installations continues uninterrupted, we’d hit it. If there are more 15% growth years then no.
Thanks – that’s good information and a sobering calculation about growth rates! 🙂
“To limit climate change on any reasonable proposal”
What does this mean? 2 degrees? 3?
I’m not sure that $.61/watt realy means much. The price that matters is the total cost of a system, including soft-costs. Thats what people/organizations look at when they make
a decision to use solar. So while panels have gotten cheap, the total cost of even a large ground mount is still over a dollar per watt, and residential is still much higher than that.
It’s all relative to the cost of grid electricity, the local solar resource and whatever incentives are available. Take New Mexico for example. Plenty of sun, cheap labor costs, but not a lot of solar energy because electricity rates are a lot lower than in neighboring states. Arizona and Nevada had much bigger utility and residential markets until they pulled the plug with added fees and bait-and-switch net metering changes.
California utilities have embraced solar in a big way. Apparently the costs are so low, that California has integrated large amounts of economic solar.
“The reports in sum show that in 2015, utility-scale solar power plants produced 15,591,964 megawatt-hours of electricity for CAISO. That’s 6.7 percent of the system’s total of 231,965,326 MWh. Wind came in at 5.3 percent. Hydro contributed 5.9 percent, with the portion that the state considers renewable, “small hydro,” at 0.6 percent of generation.”
http://ww2.kqed.org/news/2016/01/11/solar-power-california-top-source-of-renewable-energy
California has more solar than all the other states combined.
“When it comes to solar power in the U.S., California’s in a league of its own.
The home of Hollywood, Yosemite, and the Golden Gate Bridge also boasts almost half of the nation’s roughly 20,000 megawatts of overall solar capacity, new data from the Energy Information Administration show. Even more striking, in both types of utility-scale solar—photovoltaic and thermal—California operates more capacity than every other state combined. The next runner up is Arizona, at one-fifth of California’s solar might.”
https://cdn.theatlantic.com/assets/media/img/posts/2016/02/California_Solar/979e8e508.png
http://i1.wp.com/ww2.kqed.org/news/wp-content/uploads/sites/10/2016/01/2012-2015-CAISO-Wind-and-Solar-Line-Chart.png
Utility solar averaged just under $1.50/watt installed in the US. Other countries are ahead of the US, with lower BOS costs.
Utility solar has reached parity with natural gas in some states.
http://www.greentechmedia.com/articles/read/Utility-Scale-Solar-Reaches-Cost-Parity-With-Natural-Gas-Throughout-America
They produce around half their power with gas and coal, for which I suspect they import all of the fuel. On top of that, I think they import some electricity directly. They have some of the best solar resources in North America. Putting up a bunch of solar is just good business.
Very little coal. Nearly all that solar is less than 5 years old and new installations continue at a fast pace. Hydro is way down with the drought.
And at peak times, well over 25% solar. February 12th shows 6.4 GW solar at peak out of 30 something GW in February. And there’s s couple GW that’s invisible to caiso because it’s residential rooftop. Later in May peak get higher. Every day, day after day.
http://www.caiso.com/outlook/SystemStatus.html
Note the giant crater in the middle of the day as traditional generation is curtailed.
But I would argue it’s not just solar that matters – look at total renewables without hydro (as, while still very beneficial, Hydro is legacy).
From that, Maine and Iowa are the places to beat.. and South Dakota isn’t too shabby either!
Fair enough. However as solar panel efficiency increases installation costs drop as fewer mounts and less installation time needed to produce same amount of electric.
True but efficiency gains have been minor compared to cost reductions. PV panels have been in the high teens, low twenties for many years.
Fine enough point. But I imagine $/watt was much easier to track globally, and it is a critical component to the cost of a system.
According to one source, cost of the panels are about 40 % of the whole system.
And BOS in the US is excellent for utility scale and poor for residential.
Exactly. Labour as a proportion of the total cost is rising, as the prices of panels and inverters come down.
Expect to see: Innovations in racking systems, systems for lifting panels onto roofs, wiring harnesses/trunking/conduit etc
Also: lots of systems starting to go up in the third world, where labour is cheap (and “old” electricity can be expensive, and is often unreliable).
But it’s not actually the installation labor that’s the problem – it’s the man-hours that go into permitting/connections that really drive up our costs here in the US…
We need a streamlined permitting process. Even just exempting any solar-system with battery backup that doesn’t put any power onto the grid from these permitting requirements would drop the costs enough to make home storage economical (just from the permitting savings).
I’m thinking that this is the sort of thing that the DOE could instigate.
One set of standardised protocols that local authorities could simply adopt with a stroke of the pen.
The DOE could produce some guidelines for simplifying the process but building regulations in the US are largely a county decision. It might be the case that individual states could impose a standardize approach but there’s not much chance the federal government could. Installation is activity that stays within the state boundaries.
While that would be great theoretically, the reason many places have ‘permitting’ that is so difficult is because they want to discourage residential solar… so they would be unlikely to adopt the best practices identified.
I’d suggest the more effective route would be to require that the agency instituting the ‘permit requirement’ reimburse all relevant expenses for the permitting in normal circumstances. Only allow them to collect fees from the ‘abnormal’ difficult cases where they can justify why the expenses were higher than normal.
What really strikes me, is the long near plateau in price from roughly 1987 to roughly 2000, it barely budged. If what the technology really needed was enough volume to drive it down the cost curve, it looks like we lost a decade and a half. If instead it had to wait until technology in general advanced to the point where real progress was possible, then perhaps we didn’y lose too much.
Thanks, Ronald Reagan!
We gutted federal spending for solar programs right at the critical juncture where we could have hit the knee in the curve a lot quicker. All the while, fossil fuel subsidies and their ability to push external costs onto everybody else while pocketing the profits remained intact.
Only the big stall was roughly 1988:2000 which mostly corresponds to Bush-1 and Clinton. It started taking off during Bush-2. Now maybe the lag between politics and response is that long, but maybe other things were at work here.
I suspect the logjam was broken by the German push for PV, and not US politics.
Exactly, Germany (and a few other markets eg Spain) subsidising solar started to cause costs to fall and then things just took over with ever greater numbers of installs and manufacturing in China.
I big part of it had to do with affordable production of solar grade silicon. Prior to cheap large scale production solar used silicon scrap from the semiconductor industry, and the amount of this was limited. So the creation of affordable solar silicon capacity probably had a lot to do with the scale of the market.
Yep. We owe Germany a massive thankyou…
If human civilization survives, the German people will be entitled to a large part of the credit.
Indeed. I make a point of shaking the hand of every single German person I meet, and thanking them for subsidising the cheap solar panels on my roof.
And not just figuratively. I actually do this. When I hear the accent, I ask “Are you German”? Those who are brave enough to admit it, get a hearty handshake, a broad smile and a sincere thanks. Sometimes gets a bit weird, if they’ve never heard of the Energiewende. Luckily most of them have…
But in 1986 Reagan and the King of Saudi Arabia secretly agreed to crash oil prices to bankrupt the Soviet Union. That didn’t directly affect solar, but generally natural gas prices move with oil and natural gas grabbed center stage as the cheap energy alternative after that. Generally, fossil fuels appeared cheap to the investor establishment until the Bush-2 crisis and alternatives were relegated to a concern for treehuggers.
When a technology is first being developed, the individual actions of a few researchers or promoters can have a dramatic impact. Once that technology has come to the attention of a wider cross-section of the political and scientific community, any advance that can be made has a better chance of being made by someone, so progress should become more predictable. (Not necessarily faster, as each successive improvement costs more, but more predictable because the costs are being spread over successively larger populations.)
What seems like a plateau is a linear graph of a log response. Thats why it “looks” that way. Looking more closely with a log graph we can see that solar cost reduction continues according to Swanson’s law unabated.
http://blogs.scientificamerican.com/media/inline/blog/Image/naam-solar-moore_s-law-2.jpg
http://hammerandhand.com/wp-content/uploads/2015/07/Swansons-law.png
I was going to say that. The chart in the article would look far better with log scales. Here’s my chart of the prices, using a log scale. It has lower recent panel prices than the one in the article. I got the latest data from from here ( http://pv.energytrend.com/ )
Also the capacity chart shows the annual increase (new capacity) not the cumulative total capacity. That chart is also shown below. Data cobbled together from various sources.
That’s a good point. For example cumulative solar pv module growth from 1977 to 1981 was from less than 1 MWp to more than 10 MWp. Explosive growth! But only shows up as less than a blip compared to the Gigawatt deployments that we have now.
The plateau is meaningless if you don’t compare it to fossil fuel competition at the same time. If the boom in solar research had by some miracle made solar close to coal costs before 1980, then the plateau would have been a boom time for solar installation. It wasn’t even close.
If you squint, you can see a second collapse in solar prices after 2008. That’s about when installations really go nuts. Now which led to which? Probably, the oil & gas price explosion the year before the crash triggered both.
Something that surprised me is First Solar that is doing grid scale thin film (CdTe) solar installs is projecting to be at $1/watt installed by 2017 (for grid scale).
In residential solar a ballpark figure I see thrown around a lot is more like $3/watt installed.
Right now grid PV plants are beginning to crush solar thermal plants in size, installed capacity, cost etc.and that will come with huge implications itself for grid scale solar. Finally hitting high installed capacities will help dilute R&D costs too.
It isn’t thermal solar’s time. Thermal solar almost certainly can’t compete with PV solar but it might be able to compete with stored PV solar.
As time goes along wind and solar will continue to eat into natural gas’s business and gas will be forced to increase its asking price. As that price goes up room will be created for stored power to come onto the market. That’s where thermal solar (with storage) may find a role.
I suspect it would be wise to keep building thermal solar plants in order to further the technology and discover its potential.
PV cells only respond to certain solar frequencies. I don’t understand why we don’t see more development of compound solar systems, where reflectors concentrate light onto a single row of optimized PV cells, which then have their heat transferred to a liquid coolant which can run a low-temperature turbine or feed a storage system.
Inefficient. Thermal solar with all the energy being used for heat isn’t economically competitive.
Hey, Bob! Off topic, but I have no other way of communicating with you: you might find this amusing:
http://thearchdruidreport.blogspot.gr/2016/02/renewables-next-fracking.html
Wednesday, February 10, 2016
Renewables: The Next Fracking?
I didn’t make it to the bottom of that screed. I slowed when I hit this – “The first group consists of those people who believe that of course sun and wind can replace fossil fuels and enable modern industrial society to keep on going into the far future. The second group consists of people who actually live with renewable energy on a daily basis. It’s been my repeated experience for years now that people belong to one of these groups or the other, but not to both.”
Having been off grid for over 25 years with solar and understanding that solar and wind can replace fossil fuels and knowing many people with the same experience and understanding it was clear that this guy was talking through his hat
Next he went off on how lots of people held that biofuel was the answer when, again having been there, most people interested in getting off fossil fuels calculated that biofuels were, at best, a niche solution – well strike two.
And he did a bit of hero worship on Tom Murphy, the nuclear advocate who sets up arguments for renewables and storage in ways that insure their failure and declare that only nuclear can save the world. I have no idea why a stellar institution like UCSD tolerates his crap.
At that point I quit. Bullshitters are gonna bullshit….
Yeah, Mr. Greer is a major pessimist and is clearly looking forward eagerly to the collapse of civilization.
Thanks for your comments. Yes, I knew you’d find it amusing. 🙂
You need to make it clear that the blue graph is installed capacity PER ANNUM.
Otherwise that dip aroun2011-2012 would make no sense. How would total cumulative installed capacity decrease in any year?
In the case of solar it will almost certainly be a case of exponential growth at first followed by probably steep but certainly linear growth until market saturation is reached.
Installation growth requires more and more manufacturing capacity per year. New factories will have to be built in order to increase panels produced per year.
At some point panel manufacturers look ahead and see the approximate year of market saturation. Then they have to decide if it makes sense to invest in more factories in order to increase output. Will those additional factories be able to pay themselves off and return a profit?
As we reach the saturation point there will be more manufacturing capacity than can be utilized. Some factories will be closed. Growth will be over and only those factories needed for replacement (and future demand growth) will be needed.