June 14, 2010
2010 10 Gigawatts Solar Power Install Expected

Eric Wesoff of Green Tech Media reports on their projection that well over 10 gigawatts of solar cells will be installed in 2010.

In 2010, we will cross the threshold of 10 gigawatts of photovoltaic solar installed globally in a single year -- a record-setting and once-inconceivable number.

Rewind to ten years ago: the total amount of photovoltaics installed in the year 2000 was 170 megawatts.  Since then, the solar photovoltaic industry has grown at a 51 percent annual growth rate, and 170 megawatts is now the size of a healthy utility installation or a small solar factory. 

Contrast that with 200 gigawatts of wind installation this year. Wind continues to far surpass solar power due to lower costs.

Total new solar installations for 2010 will be around 11 gigawatts. By contrast, 7 gigawatts of solar was installed in 2009. This increase is partly driven by government policies around the world. But the rapid decline in solar photovoltaic (PV) module costs by about half from late 2008 till today also lowers the threshold for profitable solar projects.

Suntech Power VP Andrew Beebe says SunTech has a single manufacturing building in China which will have the capacity to produce 1 gigawatt of PV per year.

We have a building (that's one building!) in China that this year should be capable of cranking out one gigawatt of product per year. I think that's larger than the entire industry's capacity ten years ago.

Solar's got one big advantage over wind: electric power demand is strongest (and wholesale electric power spot market prices are highest) when the sun shines. Solar's output profile peaks earlier in the day than overall electric power demand. But solar's power output peak is much closer to peak demand than wind's night time output peak. Therefore solar doesn't have to be as cheap per kilowatt-hour to compete against wind.

Share |      Randall Parker, 2010 June 14 11:28 PM  Energy Solar


Comments
roy said at June 15, 2010 7:51 AM:

Other than peak summer hours, most of the US currently has plenty of electricity generation. While solar may help with this, wind just doesn't help much. At peak times, wind is only worth 13% to 20% of its total installed capacity.
[ http://www.pjm.com/~/media/committees-groups/working-groups/irwg/20091217/20091217-item-06-wind-interconnection-study-assumptions.ashx ]

In other words, to get the equivalent of a single 100 MW gas turbine, you'd have to install at least 500 MWs wind generation. And this doesn't even include the massive amount of investment in new trasmission infrastructure needed to support significant wind generation.

LL said at June 15, 2010 8:26 AM:

Roy,

Your comment is very valid. Wind's value will come only when the power generated at night can be stored, then that power can be released during peak time. If we go down the rabbit hole of wind generation and it look like we certainly will, storage technologies, transmission infrastructure and policies to flatten the demand curve will be in our future.

In terms of flattening the demand curve, Tiered pricing to get consumers to use more of their electricity at night will be the norm. Plug in hybrids and EVs will also add value since they can be charged at night adding value to the wind farm and help to displace dirty fuels.

Of course more solar is always a good thing especially when it keeps getting cheaper. 11GW is a great thing....I can see a tipping point with a 15-20% cost reduction.

Nick G said at June 15, 2010 11:13 AM:

Randall,

The 200GW figure for wind is for the cumulative installed base. The new installations for 2010 would appear to be in the 40-50GW range.

So, the ratio of wind:solar appears to be about 4.5:1. You'd expect a higher ratio based on the simple cost per average output, which is probably about $6/W for wind and at least $15/W for PV.

There's an additional factor helping PV: it's often on the customer side, so it competes with utility retail pricing, where wind competes with wholesale pricing. So, in California a large industrial/commercial customer may pay $.20/kWh during the afternoon peak - PV can compete with that. And, a large residential customer may pay $.38/kWh, which allows PV to compete there too, despite the higher installation costs.

Tim Dowty said at June 15, 2010 12:25 PM:

Nick:

Can you break down your claim that PV kWh can cost less than $.20/kWh? Including all burdened costs and free of subsidies?

Nick G said at June 15, 2010 12:48 PM:

Tim,

Large PV installations can be done for less than $4/W, with no subsidies. If we assume 20% capacity factor, 7% ROI, and a 25 year life, we get $.196 per kWh.

shannonlove said at June 15, 2010 12:51 PM:

This is a nice milestone on alternative powers road to irrelevance. Alternative power is about to hit its technological installed ceiling.

Without some kind of technological breakthrough, it will be impossible for alternative power to provide more than 15% or so of total baseline power without bringing the entire grid down. Generating and distributing electricity isn't like pumping water. Every device that creates or consumes electricity has to be balanced in a circuit with every other device. If the grid becomes unbalanced the entire grid goes down not just the parts that caused the imbalance. Since alternative power simply disappears unpredictably with no relationship to demand other power sources to have to quickly take up the slack to prevent the collapse of the grid. There is only so much redundancy in the total grid and only so much power that can be replaced on short notice. When the amount of alternative power supplied exceeds the fast spin up non-alternative backup, the grid will come down.

The Texas power grid was almost brought down by wind power in Feb 2008 when wind nominally provided only 6% of power. Backup generators prevented a statewide blackout but the system simply doesn't have more than 10% backup. If the same conditions would to occur today, the grid probably would collapse. Once alternative power supplies more than 10%-15% of baseline power it will inevitably fail at some point triggering a cascade blackout.

Every new windmill and every new solar install inches uw that much closer to a massive failure. Very soon alternative power will not be viewed as the future but as an active threat to keeping the lights on. The faster we install alternative power the sooner it will fail and we can dispense with it and get to the grownup energy policy discussion.

Tim Dowty said at June 15, 2010 12:51 PM:

Forgot to include this analysis which claims that PV breakeven is ~$1 per peak watt. Do you disagree with Don Lancaster's numbers? His conclusions say that PV panel manufacture is a net energy-destroying activity.

Tim Dowty said at June 15, 2010 12:58 PM:

Link fail!!

Corrected here:

Thanks for your response Nick.

Forgot to include this analysis which claims that PV breakeven is ~$1 per peak watt. Do you disagree with Don Lancaster's numbers? His conclusions say that PV panel manufacture is a net energy-destroying activity.

Nicholas Guererro said at June 15, 2010 12:59 PM:

Solar is more predictable than wind in desert areas, which for a grid manager makes a big difference. A big advantage over wind.
But solar is not a good baseload power source and it is actually not a good load following power source.
Put solar panels in space and you may just have something.

Can you build it without massive government subsidies? No you can't.

Backers of big wind and big solar invariably are true believers in carbon doom. One trick ponies, unless you consider the capacity to also believe in peak oil doom as another trick.

Their thinking is already impaired, they are under compulsion.

crosspatch said at June 15, 2010 1:01 PM:

Are those power ratings actual generation numbers or rated capacity numbers?

The problem I have with most "renewable" installations is that they advertise the potential capacity as if it is actually the power produced. Recent wind installations are generating from 5 to 10 percent of their rated capacity. I would be more interested in an annual average output number. Is such a thing available for solar installations?

matt said at June 15, 2010 1:58 PM:

I'll call BS on that. The contracts are out there but are not yet funded. No tickee, no 10Gw. I was at Suntech 4 weeks ago and yes, they have the capacity. They just need the orders. PV is up 35% in China, but down in much of the ROW.

and cross, the problem is that yes, it is rated capacity, which tends to be a bit starry eyed.

wGraves said at June 15, 2010 1:59 PM:

Before we all go nuts celebrating, US generation is about 300 TW. So this is about 0.003% of total US generation.

Tim Dowty said at June 15, 2010 2:03 PM:

Nick:

If I understand your numbers:

- For an installation that produces 10000 peak watts, if we assume $4/W, our cost is $40000.

- Assuming an capacity of 20%, the installation would produce about 50kWh each day. (24hr * .2 = 4.8 hrs)

- If we assume a value of $.20/kWh, then we are producing $10.00/day worth of electricity (assuming 100% conversion efficiency), or $300.00/mo

- Assuming 7% financing over 25 years, the installation cost is ~$283/mo.

This confirms your $.20/kWh claim, but to get there, we must
* assume 100% conversion efficiency (Lancaster claims it is 10% at best)
* assume that the useful life of the installation is 25 years
* assume that there will be no maintainance or repair required over the installation's life

For comparision, the cost over ten years assuming perfect conversion efficiency would be $464/mo, and the value of the electricity produces at $.20/kWh @ 10% efficiency would be only $30/mo.

Am I off base here?

-Tim

Realist said at June 15, 2010 2:14 PM:

which will cover how many acres of land somewhere?

You know, a nuke plant would use a lot less space, match the demand curve in any weather and produce cheaper electricity.

Jon said at June 15, 2010 2:43 PM:

wGraves - I believe US generation has been running ca. 300-400 TW-hours/month (ref: http://www.eia.doe.gov/cneaf/electricity/epm/table1_1.html ). It looks like 10 GW peak solar capacity would deliver well over 1 TW-hour/month avg. over most of the US (ref: http://www.solar4power.com/solar-power-insolation-window.html), or say .3%, rather than .003%.

Greg F said at June 15, 2010 3:03 PM:

If you want to know how much energy you will really get from solar panels browse real data at Fat Spaniel . Pay no attention to the Environmental Benefits" as they all start at zero ignoring the energy used to make, deliver, and install the panels. You will also notice that there is no easy way to extract the data for analysis.

Nick G said at June 15, 2010 3:42 PM:

Tim,

The power rating of the installation already takes the conversion efficiency into account. If you buy a panel rated at 1kW, then it will produce a kilowatt of power on a sunny day at noon.

PV lifetime is well above 25 years, but output tends to start declining very gradually well before that. Cumulative lifetime output will be well above 25 x first year output.

PV tends to be very low maintenance. It helps to get out and spray water on it occasionally to remove dust - that's about it.

Don Lancaster makes a number of unrealistic assumptions: that peak power is worth only $.10; a financial amortization of only 10 years; and an interest rate of 10%. Other unrealistic assumptions, not part of this particular cost analysis: that net energy can be estimated using purchase price; that external costs for fossil fuels are unimportant; and 10% conversion efficiency - this is more of a floor than a ceiling these days (though cost per peak watt is far more important than efficiency).

That said, there's no question that solar power is still more expensive than would be desirable or affordable for the majority of our power - either wind power or nuclear would be substantially cheaper. To a large extent, our current subsidies are an investment in bringing down the cost of PV.

jpintx said at June 15, 2010 4:01 PM:

I fear shannonlove is on target. As usual with such announcements, the number sounds impressive, but as I have said before, many folks hear such a seemingly large number and rejoice in triumph, failing to compare it to reality. So here are the numbers:
Total electricity generated in the USA in the rolling 12 month period ended February 2010..............3,975,732, 000MW or 397,573.2Gigawatts, so the addition world wide equals .000025 of the US total demand.
Is this even keeping up with increased demand from Nissan Leafs? LINK http://www.eia.doe.gov/cneaf/electricity/epm/table1_1.html

Ari Tai said at June 15, 2010 4:02 PM:

I must be missing something fundamental.

A large panel is around 2KW (when well lit). Back-of-the-envelope: 10 GW is 10,000 MW is 10,000,000 KW, or 5 million (full-length mirror man-sized) 2KW panels. They produce at maybe 20% of the time over the period of a year. Which means 25 million panels (w/ 2,000 silicon chips per panel or 50 billion chips).

Would take the entire silicon output of all existing factories to do a fraction of this(?).

Perhaps the article (or how it's been reported) has confused GW with GWH (unit of force v. unit of energy - "work"). Call it 10K hours per year, and the numbers look a bit more feasible. Still a small fraction of one ng or coal plant. An aside. It's just amazing how the advanced combined cycle plants efficiency numbers continue to rise.

Tim Dowty said at June 15, 2010 4:13 PM:

Thanks Nick,

So, if I buy a PV panel rated at 10kW that means that--at its peak output--I will see 10kW put on the grid after the inverter? I assumed that the panels were rated at their raw output.

Stepping back a bit, I guess the important question is whether you foresee a time when PV will ever generate more value in power than it takes to pay for it (even ignoring the billions that have been poured into it to this point and must be recouped). The external costs of fossil fuels are absolutely relevant when you incur a net loss of energy (exergy, actually) whenever you build, transport, and install a PV cell. Subsidizing the destruction of usable energy makes no sense unless we are *very* confident that it will lead to net positive energy within a reasonable timeframe.

I have no practical experience with PV; I am just an interested citizen, but I am very skeptical.

-Tim

Daniel said at June 15, 2010 5:46 PM:

They are building a lot of solar, especially PV, but let's not forget how expensive it is. Solar PV is about $396 per mWh, compared to $83 per mWh for combined cycle natural gas or 100 per mWh for coal (this includes a carbon fee) according to EIA data: http://www.instituteforenergyresearch.org/2009/05/12/levelized-cost-of-new-generating-technologies/

Greg F said at June 15, 2010 7:15 PM:
The power rating of the installation already takes the conversion efficiency into account.

Unless your in California or a state that requires it, no it doesn't. Where I work we have one hundred and ninty 208 Watt panels. That is 39.52 kW and that is how it is rated. The converters are not even considered.

If you buy a panel rated at 1kW, then it will produce a kilowatt of power on a sunny day at noon.

No it will not. All manufacturers rate their panels are under "Standard Test Conditions" (1000 W/M^2, cell temperature 25 degrees C). The tests are pulse tests that do not represent real life conditions. There is no way your going to have 1000 W/M^2 and maintain the cell temperature at 25 degrees C. Silicon has a negative temperature coefficient which results in a drop in power with increasing temperature. An example of this is shown in the data sheet for a Sanyo solar panel on the second page (Dependence on Temperature).

Engineer-Poet said at June 15, 2010 7:51 PM:

wGraves doesn't know what he's talking about; US electric generation was about 4110 billion kWh/year in 2008, or about 470 GW average.  That's gigawatts, not terawatts.  jpintx can't even get the units right; the number he's quoting is thousands of megawatt-hours (per year), not gigawatts (it averages to about 454 GW from Mar09-Feb10).  And Ari Tai, the biggest panels you're likely to see are rated at 200 W, not 2 kW.

I'd like to know why people think they can come here and post things which show that they're totally clueless about the basics of the thing they're talking about.  Are they just trying to bury worthwhile points under crap?

Getting back to Ari Tai, PV panels are rated at 1000 W/m² incoming flux and 25°C temperature.  Different cell types change characteristics differently with temperature.  If you don't know this, you have no business expounding on the issue.

Silicon is scarce or plentiful, depending on the grade.  Metallurgical-grade silicon is like dirt ($1.45/lb in 9/2008), semiconductor grade is much higher and depends on wafer diameter.  PV grade is in between, but the PV market wasn't big enough to justify its own production capacity until recently.  That's no longer the case.  There are also PV cells which don't use silicon (gallium arsenide, cadmium selenide).

Tim:  You're not going to see 10 kW on the grid from 10 kW of panels unless they're getting full sun on a cold winter day.  PV must be de-rated for temperature above 25°C and the inverter imposes losses also.  If you use DC, you can avoid part of that.

PV is good for off-grid applications in dry southern areas.  The efficiency of a panel doesn't depend on its size, so it's good for small installations.  It's often cheaper to buy a PV panel and battery than to run grid power to e.g. a traffic warning sign.  But as providers of bulk power, PV is about 20 years behind wind in running down the cost curve.  That's why wind added 38.3 GW of capacity last year, and PV may just pass 10 GW.

C. Delmain said at June 15, 2010 8:16 PM:

As a past consulting engineer to a large utility I am sick and tired of people who know nothing about the electrical energy generation sector saying how wonderful Solar power will be. It will NEVER replace a coal or any other type of generation. For several reasons 1. When the sun does not shine there is no power this means to make up for the loss power has to come from another source this means no coal, gas or any other source can be taken off line at anytime (we call it spinning reserve) the same goes when there are clouds or at any time of low levels of sunlight the cells cannot supply rated levels. 2. There is NO current battery technology that will store enough energy to compensate for loss (especially at peak periods) of any sunlight unless the batteries are the size of a house. 3. It costs upwards of a million dollars (and more) a mile to run transmission lines. On top of that the NIMBY"s and especially the environmentalists hate transmission lines. Solar power and wind power will NEVER NEVER be more then 3 to 5% of the supply. If you don't believe me I suggest you research Spain's problems with it's green power experiment.

Wolf-Dog said at June 16, 2010 12:57 AM:

Engineer-Poet: Please clarify the physical units to make sure that there is no misunderstanding. Despite the economic crisis in Greece, credit must be given where it is due, and the ancient Greek words have enriched the English language (and made it more confusing). 1 Megawatt = 1 million Watts, and 1 Gigawatt = 1,000 Megawatt.

http://wiki.answers.com/Q/What_is_the_average_output_MW_of_a_nuclear_power_plant
EXCERPT:"As of 2007, worldwide there were 439 operational nuclear reactors with total capacity of 372,002 MW; making the average output 846 MW. However, the average new reactor can be expected to be larger. The reactors so far built in the US range up to about 1100 MWe electrical output, but new ones up to about 1500 MWe are planned "

But in the United States there are 104 reactors, and approximately 20 % of the electricity in the US is nuclear.

On the other hand, by 2020 China is planning to build 28 reactors of size 1 Gigawatts, almost 2.8 reactors per year until the end of this decade:
http://www.businessweek.com/news/2010-03-23/china-to-build-28-more-nuclear-power-reactors-by-2020-update1-.html

Thus for the sake of comparison, within 10 years, 10 Gigawatts of new solar power installed per year would equal all the nuclear reactors of the US.
So maybe the 10 Gigawatt figure for 2010 is inflated, otherwise it would be impressive.


Wolf-Dog said at June 16, 2010 1:02 AM:

Engineer-Poet: Please clarify the physical units to make sure that there is no misunderstanding. Despite the economic crisis in Greece, credit must be given where it is due, and the ancient Greek words have enriched the English language (and made it more confusing). 1 Megawatt = 1 million Watts, and 1 Gigawatt = 1,000 Megawatt.


As of 2007, worldwide there were 439 operational nuclear reactors with total capacity of 372,002 MW; making the average output 846 MW. However, the average new reactor will be at least 1,000 Megawatt.

But in the United States there are 104 reactors, and approximately 20 % of the electricity in the US is nuclear.

On the other hand, by 2020 China is planning to build 28 reactors of size 1 Gigawatts, almost 2.8 reactors per year until the end of this decade:

Thus for the sake of comparison, within 10 years, 10 Gigawatts of new solar power installed per year would equal all the nuclear reactors of the US.

So unless I got confused by the Gigawatts and Megawatts, maybe the 10 Gigawatt figure for 2010 is inflated, otherwise it would be impressive.

BigRed1 said at June 16, 2010 7:36 AM:

The dream of Solar Power, at this point in history, is essentially a sham. It can't function without massive subsidies. Even with subsidies on home installations the ROI is very questionable at best. Solar installations are 10-15% efficient. The peak production point during the day (mid day) for solar does not match up well with the peak of usage (evening). Until some method to store excess electricity produced is commercially viable, much of the power produced from solar may be lost (grid power is lost if not used immediately). ANY solar power production that the grid actually relies on must have redundant production capacity for off peak hours (most hours in the day are off peak). Commercial Solar power requires huge sums of water for cooling. There are also huge costs associated with getting the power generated from the source (middle of the desert for instance) to the grid. We need to re-think the promise of Solar power. Likely viable for home use some time in the future. Commercial viability is a long time off without a technology breakthrough (changing the slope of efficiency improvement).

Rob said at June 16, 2010 9:38 AM:

Solar power has no current future as base load generation. It may work as a peaking resource, a little bit, but as others have noted, it's peak is around noon and peak demand is more like 5-6PM. Solar has many cool uses for off-grid sites or small things that can store the power and work around the clock, but as a big fraction of our daily power I just don't see it.

The full costs of solar and wind isn't talked about. Imagine that you had a city in Texas or some other sunny place and it's getting 25% of its power from solar panels. Now, a big cold front rumbles through, blocking 90% of the sunlight for an hour or so (and with wind gusts high enough to knock out the wind generation too). Where are you going to suddenly pick up all of those additional megawatts? Well, you HAVE to fire up a conventional power plant. If you have to have the conventional plant standing by, ready to pick up the load, then the COST of that power plant idling has to be considered part of the cost of the solar power. This is almost never done, so the cost figures you see are laughably inappropriate.

If you think that's somehow unlikely, you're wrong, a similar situation has already happened. In California a couple of years ago, when they had the big fires, the smoke actually took out the solar generation at the same time the high Santa Anna winds took out the wind farms. This lasted for almost a week; no known storage technology could have helped.

Also, imagine how difficult it will become for your local utility to predict the amount of electricity they have to produce, when they have a lot of private citizens with solar panels all over the place coming on and off the grid at whim. The grid was designed to move electricity from producer to consumer, but when all of the consumers become producers the grid you want to have looks a lot different from the grid we've built. The answer isn't what's touted as "smart grid technology" today, that describes "smart meters" and Demand Response stuff. The answer is billions and billions of dollars of infrastructure improvements - new transformers and transmission line upgrades, smarter interconnects and so on. Will the COST of all this be counted as part of the cost of "free" energy? Hell no.

Nick G said at June 16, 2010 11:39 AM:

Tim,

In a previous post, you referred to conversion efficiency of about 10%, i.e., conversion of light to electricity (* assume 100% conversion efficiency (Lancaster claims it is 10% at best)". That's what I was referring to. After the DC electricity leaves the panel, there will be inverter losses, and there will be other losses due to temperature, dust, etc. Those losses, of course, are much, much smaller than the difference between 10% and 100% that started the discussion.

Wolf-dog,

10GW of PV will give average output of perhaps 1.5GW, the equivalent of roughly 1.5 large nuclear plants. If you're willing to accept the various risks of nuclear, and able to ramp it up to the point of gaining economies of scale, nuclear will work quite nicely. Wind will too. Solar hasn't reached the point of cost effectiveness yet for most locations and customers, though there's no reason to think it won't.

anonyq said at June 16, 2010 3:20 PM:

Rob,

Solar power doesn't peak at noon but at 1 (at least during summertime) and it doesn't need to be base because for that there are other solutions.

Realist,

Nuclear has a very hard time with following the demand curb and with peak power, that is why no country uses nuclear for more than 65% of its demand and this includes France

Daniel,

Solar is more peak, no 24:00 - 6:00, power supplier so it can be more expensive and still be profitable.

Engineer-Poet said at June 16, 2010 6:25 PM:

Lancaster shouldn't be trusted on this issue.  Among other errors he makes, PV efficiency is considerably higher than 10%.  Even cheap polycrystalline cells beat 11%, and Evergreen Solar has hit 17.8%.  Those are consumer-grade cells, nothing special.

As for the match of demand curve to supply, the afternoon peak load is for air conditioning.  Instead of generating power in real time, running the A/C in the morning to make ice can supply cooling demand hours later (this might even increase efficiency by running the system before outdoor temperatures peak).  As electric vehicles become more common, early-peaking generation can be used to feed them too.  Last, siting generation at the consumer's site removes load from the grid's lines and transformers.  This all improves the resilience of the system.

Post a comment
Comments:
Name (not anon or anonymous):
Email Address:
URL:
Remember info?

                       
Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright