German solar photovoltaic performance – informative graphic site
A friend just posted a link to a site that everyone who wants effective energy decision making should bookmark. It is titled Performance of Photovoltaics (PV) in Germany.
The site includes animation features that illustrate the modeled output of solar installations over the course of each day. The model used should provide a reasonably good match with actual output; it is based on reported production from a sample of installations spread around the country that represent 13% of all installed systems. (You can find out the details of the model by clicking on the link titled “Our Data Calculation Model” on the solar performance website.)
I did a quick scan through the month of January 2013. So far, the absolute peak performance was 5.7 GWe on January 12. Throughout the month, most days peaked at less than 2.5 GWe. This peak is coming from an installed base of 32 GWe spread throughout the country.
Here is a representative sample of the informative pictures you can discover. It shows the absolute maximum output from January 19, 2013 as 0.9 GWe at noon.
Hat tip to Jeremy Gordon (@JeremyGordonWNN)
I’ve linked to the SMA website various times, it’s such a telling graphical depiction.
Even though they are being dishonest in their actual presentation – for example, nighttime hours are not shown, the graph is basically self-zooming, making it look like PV performs much better than it actually does.
About a year ago, I did a fun excercise with this website. I noted down the output at each hour, added it all up and divided by the nameplate capacity x 24 hours. This gave me the capacity factor.
In winter, which is right now, it turns out that capacity factors between 1% and 2% are typical. There are also many days when there’s snow on the panels (I can verify this because I live near Germany). At thos times the capacity factor actually was 0 to 0.1%.
This is abominable. The energy source is not there 99% of the time when it is needed most, in cold dark winter Germany.
Of further note is that SMA is actually a solar equipment manufacturer/seller/installer (inverters and stuff). So they are not likely to have any incentive to make solar look bad. Quite the opposite.
So get that – this is the optimistic version.
The figures and chart provided come as no surprise, given the nature of solar power generation. The current German government finds itself in a serious jam, because on one hand they are banking on solar (its the second pillar of the so-called energy transformation). On the other hand it cannot be depended on since long periods without sun are common in Germany, even between May and September.
Rod,
You provided us the link to that site many months ago. Thanks again !
@Daniel
Your memory is better than mine. I have an excuse; I’m a granddad.
Oh well, some things are worth repeating.
Here’s a link to an article in Journal of Power Sources.
Cost-minimized combinations of wind power, solar power and electrochemical storage, powering the grid up to 99.9% of the time
On the surface it seems like a good study, but I think the devil is definitely in the details. Also it references a study by Jacobson (the guy who attributed nuclear war CO2 to nuclear power).
Hi Rod,
Here is a another good link. It shows the wind power output over a week at the Bonneville Power Authority in the Northwest. The data is collected every 5 minutes, so you can see all the variations. This week it’s practically flat. More interesting is when it is jumping around like crazy, which is typical. It doesn’t take a lot of technical knowledge to look at that chart and imagine the difficulty of using such a source in a power grid.
http://transmission.bpa.gov/Business/Operations/Wind/baltwg.aspx
Instead of ‘flat’ I should have said ‘zero’, which is the only time it is flat.
Wow, there’s been zero wind output since saturday night! Clearly wind sucks as bad as solar or worse.
Hydro’s excellent load following performance shows its usefulness. Unfortunately, greens don’t like clean energy that works, including hydroelectric dams.
What’s the wind capacity (peak) on the BPA grid?
Intalled peak capacity in BPA system is 4711 MWe as of May 2012.
http://transmission.bpa.gov/Business/Operations/Wind/WIND_InstalledCapacity_Plot.pdf
So putting two-and-two together, the max wind generation peak of the last week or so was under 300 MWe out of 4711 MWe installed capacity.
That’s horrible! Even worse than German solar in winter!
God thats embarrassing. Tragic even.
Rod,
It’s winter and it’s snowing in Germany right now.
To be fair to them, they have better days, like 13.1 GW peak on July 16 2012 and, eye balling it, about 4.5 GW average over 24 hours. But then, it was mid summer on one of the longest day of the year.
Oh well…
The output today is zero.
Solar panel, meet snow.
To gullible solar enthusiasts: snow is the little white powder that makes the power output zero.
To lazy solar enthusiasts: you will, actually, have to go and get rid of the snow. Yes. So much for zero maintenance. Be careful; working on roofs is far more dangerous than working in a nuclear power plant.
Be careful; working on roofs is far more dangerous than working in a nuclear power plant.
That is especially true when the steeply slanted roofs that are typical for areas that have winter weather are covered with white, slippery stuff that makes footing rather treacherous. Since solar panels are most beneficially installed on southern facing slopes, I would also expect that there is some amount of daytime melting and nighttime refreezing that occurs, making the clearing process doubly difficult and hazardous.
The experiences of people attempting to use solar PV in winter in Germany should apply equally well to other locales that are rather illogically devoting public resources to solar energy installation programs – like Vermont and Massachusetts.
Its even more tragic when you realize even what is produced isn’t necessarily used:
“Operators may just decide to accept that they are giving away power and money during the Christmas week for a few hours rather than putting up with even higher costs for switching their units off,” Juergen Rogalla, head of power plant operations at Stadtwerke Bielefeld GmbH, a regional utility supplying 280,000 households with power and gas, said Dec. 21. ( http://www.bloomberg.com/news/2012-12-24/german-utilities-pay-power-users-as-warm-windy-christmas-looms.html )
Ive yet to see anyone acknowledge this situation exists, much less discus it.
German law states that electricity from wind and solar MUST always be bought up. Even when there’s no demand.
Such a law is perfectly fitting to the naïve energy policy of paying more the less reliable it gets (solar is the least reliable renewable energy but gets the most subsidy).
It’s a good lesson in “how not to devise a good energy policy”.
To be fair, the capacity factor of PV on the very best days in summer is about 25%.
That means, to be fair, that the energy isn’t there 75% of the time on the very best days.
Do you really want to hear about the not-so-good days with such performance?
So sometimes they “trow away” the fossil fuel energy because its cheaper than switching it off? jheezzzz. Cooked books by another name.
I’m not sure if electricity is thrown away. Money is being thrown away, for sure. But it’s likely that the excess capacity is dumped for low prices onto the EU grid market. Germany is quite well interconnected with neighboring countries and exports considerably.
This is a sandbox economy; PV electricity bought and paid for 30-60 eurocents per kWh is being dumped for typical EU market prices of 4-6 cents/kWh. Actually that’s not what’s happening now, because there ain’t any PV electricity. It’s what happens mostly in summer, when German electric demand is at a low.
Of course one could argue the reverse: that the PV electricity is used by Germany and dirty fossil fuel electricity is being dumped onto other countries. That argument, asides from not reducing fossil fuel usage, will soon be moot, as Germany’s PV and wind fleet is starting to outgrow even peak demand at times. PV and wind dumping will be increasingly used in the future.
The reality is that PV is very low marginal cost and can’t be controlled. It’s a fact of life. Once installed, the money is essentially wasted already, and the uncontrollable nature of PV (utilities, who control the grid, can’t control PV! imagine driving a car and you have no control over the accellerator pedal!) will make PV the most attractive energy to dump onto other countries at low price.
I wouldn’t be suprised if neighboring countries would buy up excess PV and wind for dump prices of a few cents per kWh, and then apply a double counting trick to make it look like they’ve achieved the “renewable energy” builds. The renewables industry is certainly dishonest enough for that, and with a technology that can’t perform, politicians pushing renewables are asking for it.
We have a saying for that here: if you ask for a monkey, you will get a monkey. Asking for an energy source that isn’t there 90% of the time to power a country is asking for fraud.
Where is it going ? If they leave the fossil generators on. No one elsewhere has the capability or insight to respond to those types of fluctuations via imports on that scale. Especially from unanticipated imports.
I think every one just assumed it would be used C.
I did see where the main driver of German inflation was power prices.
The average inflation for Germany in 2012 was reported at 2.0%, largely driven by a 5.7% annual increase in the price of energy products. Food and non-alcoholic beverages saw a 3.2% average increase in prices over 2012. ( http://www.dailyfx.com/forex/market_alert/2013/01/15/german_Inflation_Reported_at_2_for_2012.html )
Wholesale electricity prices and use dropped in the recession since 2009 :
Electricity for 2013 fell to a record 44 euros ($58) a megawatt-hour on Dec. 27, the lowest for a year-ahead contract since 2009, and last traded at 44.05 euros Dec. 28, according to broker data compiled by Bloomberg. ( http://www.bloomberg.com/news/2013-01-02/record-low-power-prices-stung-by-plunging-use-in-europe.html )
Of course all the extra power didnt stop them form cutting the power to Greece.
Jun 22, 2012 – Trading sources told Reuters that at least four trading companies in Switzerland, Italy, Bulgaria and Germany have either lowered or cut off sales to Greece due to high credit risk and delays in payments over the past 2-3 months for power they sold. ( http://in.reuters.com/article/2012/06/22/greece-power-blackouts-idINL5E8HM8L920120622 )
Today :
In Athens protesters gathered outside the state controlled power company PPC. They’re angry at power cuts in homes where residents have failed to pay their electricity bills.
The power company raised prices by 15 percent this year. ( http://www.euronews.com/2013/01/22/athens-protest-over-power-cuts/ )
It all rolls downhill and lands on the poor. Even in times of excess.
Isn’t it a well known fact that you don’t really need electricity when it’s snowing? I mean, you just stay home, nicely tucked by the fireplace, reading a book by candle light.
Or at least, that’s what Hallmark Christmas cards say 🙂
The Germans are very enterprising and soon they will install heating coils, like you can install on the eaves of your roof to prevent snow/ice buildup, and recover this lost electricity. They will do like all good power plants, sell the solar generated power to the grid at 30 – 60 cents per kWh and buy electricity from the grid for melting the snow/ice at 10-20 cents per kWh, thus still making money (from you.) Every power plant I have ever worked at has separate meters for power sold and power used.
Here’s another site for German wind and solar power. You can get the numerical data at 15 minute intervals.
http://www.transparency.eex.com/en/Statutory%20Publication%20Requirements%20of%20the%20Transmission%20System%20Operators/Power%20generation/Actual%20solar%20power%20generation
For the solar capacity,
http://www.germanenergyblog.de/?p=9771
http://www.germanenergyblog.de/?p=11832
show the increase each month for the last few years.
To be even fairer, they’ve done substantially better than that. On 25 May 2012, German solar was 22 GW at peak and averaged almost 8 GW.
The math does not work for me. Are you averaging over a 24 hour period or just during daylight hours?
It says here,
http://cleantechnica.com/2012/05/31/in-depth-germanys-22-gw-solar-energy-record/
that German solar produced 189.24 GW-hours (= 7.9 GW-days) that day.
The peak must have been greater than 22 GWe to have produced 189.24 GW-hours.
If it is worth one child’s life to ban all guns, how much is an eagle worth? You would be surprised what you can learn if you frequent some “skeptic” sites like this one – http://toryaardvark.com/2013/01/08/bald-eagles-nest-and-habitat-destroyed-for-wind-farm/
Even with a 100% efficient “storage system” (impossible) and solar/wind generators 5 times as efficient as they are now (maybe when they perfect superconductivity) the math just does not support “unreliable” energy.
If you run the program on the 12.18.2011 the 21 GW PV power installed in Germany (cost 85 billion euros) at noon you have a peak output of only 1.6 GW. This means that yesterday a 350 MW nuclear power plant (cost less than one billion euros) would produce energy as all the 21000 MW of installed PV power in Germany (24 hours).
I simplified by looking at peak output (12.18.2011) which was 1,6 GW.
Area under Gaussian peak = (peak * full width at half maximum, sice 10,30 to 14,30 = 4h)/(2,35*0.4) = (1.6* 4)/(2,35*0,4) = 6,8 GWh then 6,8 GWh/(21GW*24)= 0,013 = 1,3% “capacity factor”.
If you run the program on the 25. 05. 2011, you register a peak output of 12 GW (75% of 17 GW). The area under the Gaussian was then calculated in the following way.
Area = (* peak width at half maximum of the curve) / (2.35 * 0.4) = (12 * 6) / (2.35 * 0.4) = 82 GWh. So 82 GWh / (17Gw * 24) = 0.2 = 20%, which we call capacity factor.
Then we can estimate roughly an average (12 months) capacity the all GW PV power installed in Germany about 10%.
The energy output throughout 2011 would have been the same of a couple of 1000 MW nuclear power plants (cost 10-12 billion euros)
No, the numbers match those I get from the EEX website I linked to above:
Peak power: 22,274.5 MW at 1:45 pm
Total electricity: 189,242.825 MW-h
The numbers on the sma.de site are slightly different: 21.1 GW and 179 GW-h.
22 GW * 24 h = 528 GW-h ; 189/528 = 36%.
At the time Germany had about 29 GW of PV online, so this is a capacity factor of 27%, which is indeed very high for PV. But southern Germany is about as far north as Newfoundland, and this is on a day near midsummer. On its best day last week, German PV produced 15 GW-h, with a peak power of 3.4 GW. Between seasonal variation and weather, Germany’s annual capacity factor is about 10%.
They don’t match with SMA data.
SMA data gives “profit yield days”. The best being 179 GWh (7.45 GWe average) on may 5, 2012. There was 27.16 GWe peak at the time.
Therefore the capacity factor was about 27%.
The Cleantechnica website clearly uses different data than the SMA model produces. The SMA model gives 179 GWh (7.45 GWe average). From a peak installed capacity of 27.16 GWe peak.
This is a capacity factor of 27%.
They installed heating coils on a solar project here in the Netherlands. They also had a heat tracing system, and additional cooling systems (for summer), on the inverter and other equipment.
It reduced the output by over 30%, to an abominable 6% capacity factor.
This is a good idea to make a bad idea even worse.
Ok, I added up the output of each hour yesterday. It adds up (rounded up for optimism) to 2.7 GWh.
Out of a theoretical yield of 770 GWh. That’s 0.4% capacity factor (0.004), again rounding up…
This is embarrassing. Energy that goes on vacation in autumn and doesn’t come back till the daisies have popped up.
That’s an interesting way to calculate it. As a sanity check, I added up all the GWe at every hour, this gave 118 GWh. Much higher than your figure. Perhaps the division isn’t entirely Gaussian. This will be even more so with changing weather within a day, for example a sunny noon and cloudy afternoon would overestimate the real output.
118 GWh was about 27% capacity factor.
This 27% appears to be the ceiling of what’s possible on the best day.
Actually the capacity of a typical photovoltaic solar power plant in Massachusetts is 12-15%. In Arizona is 17-19%
http://en.wikipedia.org/wiki/Capacity_factor
The German insulation is 1100 KWh/mq, close to the Canada insulation.
In my opinion the German PV capacity factor (world leader as PV power installed – 32GW ) is much more close to the 10% than 20%.
http://en.wikipedia.org/wiki/Insolation
Well the very first thing to say is that if you have electricity 99.9% of the time, your economy is *dead*. This means you have around 10 hours of complete blackout every year, the economic cost would be atrocious.
Allianz made an estimate of such costs, it’s telling : http://www.agcs.allianz.com/assets/PDFs/Special%20and%20stand-alone%20articles/Power_Blackout_Risks.pdf
Usually the legal requirement is 5 or 6 nine. Here in France it’s n-K, which means that there’s constantly enough stand-by power to compensate either the one or the two biggest units going off the grid, the choice being made depending on how likely it is that the demand would be higher than initially predicted.
Two weeks ago, there was one small city near Paris that had a day long blackout because of a local distribution center that burnt, it has been already very costly for the local shops and business.
BTW I wonder if it’s been reported in the US that gas price reached 14$ at some places in New England earlier this week because of the cold and lack of supply ? See http://gastopowerjournal.com/markets/item/1294-new-englands-high-gas-prices-rival-northwestern-european-prices-eia
Yes, I know, but we were talking about the best days. In the desert you should be able to push 40% on the best days.
In terms of realistic averages over the year, Germany gets 10%. Average USA would be 15%. High insolation deserts are 20%.
In my opinion it is pretty terrible even in the desert, as 20% capacity factor means the energy isn’t there 80% of the time. Solar also fails common-mode, with the sun basically. So there’s no advantage that independent power units get. With wind it’s slightly better but still pretty bad because the long term trend follows the climate, so even large countries can go without wind for days at a time. With nuclear, the units are all independent, and in fact, most of the “failure” can be scheduled (refuelling) to be in a low demand season. That is common practise in the USA right now, not some academic study that talks about what theoretically could be done with wind and solar.
4-6 cents/kWh is the typical EU market prices value. *When* German has too much solar and must export it, it basically never reaches 6, and was down to only 1 in May last year.
Now May this year might be fun, first because the capacity is about 7 GW higher than last year, but second because it seems that more and more people have decided that even unsubsidized solar is worth it when the individual price reaches around 30 cent/kWh. So it may be that when the subsidized 32 GW of solar will produce most, the demand will also go down because of additional hidden capacities.
Prices have been negative for several days, with peaks at -500€ as can be seen here :
http://www.epexspot.com/en/market-data/intraday/chart/intraday-chart/2012-12-26/DE
http://www.epexspot.com/en/market-data/intraday/chart/intraday-chart/2012-12-28/DE
http://www.epexspot.com/en/market-data/intraday/chart/intraday-chart/2012-12-30/DE
and also
http://www.epexspot.com/en/market-data/auction/chart/auction-chart/2012-12-26/DE
http://www.epexspot.com/en/market-data/auction/chart/auction-chart/2012-12-25/DE
It should be said that major producers sell most of their production (actually almost everything) in advance contract. And they also are both producer and distributors, which means most of the consumption is paid at the usual fixed-contract price (see here http://www.rwe.com/web/cms/mediablob/en/1702648/data/280030/6/rwe/investor-relations/reports/2012/RWE-Interim-Report-first-three-quarters-of-2012.pdf
“We sell forward nearly all of the output of our German power plants and secure the prices of the required fuel and emission allowances in order to reduce short-term volume and price risks. Therefore, the market’s most recent development only had a minor impact on our income in the period under review”)
But still this kind of thing is an economic disaster in the making. Since the spot price goes that low, the advance contract price are also going lower and lower. The standard is that they are the average of the spot price, which was around 10€/MWh for several days. At this price, only nuclear and lignite barely cover their marginal cost, and no capital one. As RWE also reports a little later : “The average price we realised in such transactions for the electricity we generated in the period being reviewed was lower than the comparable figure for 2011. This had a negative effect on our generation spreads”. Profitability margins on nuclear, coal, gas were lower, only lignite staied the same because the CO2 certificate price went down.
Actually it’s not good news at all because it means there’s way too much variability between the very good days in May, and the extremely poor one in December.
At this point solar still brings almost nothing in December, but is already very hard and costly to integrate in early summer.
Don’t forget it’s hotter in the desert and the performance of Solar goes down with temperature. The panels get much hotter than the surrounding air (just like your car when you leave it under the sun).
I believe that’s the most likely explanation of why Solar in Germany last year generated more in May than in June, July and August (even if May in Germany is one of the 2 month with the most luminosity).
The effect is there but it’s not very large. Like a few percent. The higher insolation in a desert more than makes up for it.
In Germany it’s to do with the climate, I live in a similar climate and May is often one of the most pleasant months, usually lots of sunshine and not the humidity and rain that we usually get in july-august. Of course May is among the lowest electricity demand months (very little heating, less lighting, almost no aircon demand, because of the pleasant weather).
Using the EEX website, I get
25 May 2010: 3912.5 MW(peak); 31.7 GW-h; 11.7 GW(capacity)
18 Dec 2010: 1354.2 MW; 4.5 GW-h; 17.3 GW
25 May 2011: 12,840.8 MW; 108 GW-h; 17.4 GW
18 Dec 2011: 2,304.1 MW; 3.4 GW-h; 24.8 GW
25 May 2012: 22,274.5 MW; 189 GW-h; 27.4 GW
18 Dec 2012: 2602.5 MW; 11.4 GW-h; 32+ GW
It will be obvious I’m a data sourcing maniac.
The 19% in Arizona above is based on the numbers from this study about the Springerville array : http://wpweb2.tepper.cmu.edu/ceic/papers/ceic-08-04.asp
Usually to enhance the performance under heat, CadTe pannel are used instead of silicium one, but here only a few of use are used there, and none in the recent 4.6 to 6.4 MW extension.
https://www.tep.com/tracker/systems/springerville/
https://www.tep.com/doc/renewable/springerville-original.pdf
Another study about the five first years of operation did find the same 19% http://www1.eere.energy.gov/solar/review_meeting/pdfs/p_20_moore_snl.pdf
It does rise interesting questions about the effectiveness in actually reducing carbon : “PV generating intermittencies associated with short timescale events, such as cloud passage and storms, are in fact swinging the controls of a 420 MW coal fired unit at the generating station. These impacts bring into question the capacity value of solar in the utility plant operations”
I don’t know how this is handled at a larger scale, it seems the grid operators invest the money required to reinforcement their equipment and compensate, but this means in countries with high renewables penetration the grid costs rise very strongly in addition to the direct cost of renewables.
Another thing is the above site gives a production of only 6.58 MWh production for 2011 which is only 11.6% of the 6.48 MW DC rating, and 14.1% of the 5.32 MW AC rating. The monthly production numbers show that production does not vary as little as one could think for a desert. Also July production was very low compared to the peak power of the same month, which may mean that there has been a production outage.
Even though the production result for 2011 seems abnormally low, the web site does only announce a 1504 kWh/kW DC Energy Yield (17.2%) whilst the earlier study says 1707 (19.5%). The newer Gloria Solar modules do seem to have a slightly higher efficiency than the older ASE-300-DG/50.
I’ve seen studies that claim panels produced after 2000 have a very low ageing coefficient. I don’t know for sure what to attribute the above effect to.
And to come to our initial subject, those earlier panel had a Ppeaf of -0.47%/°C (http://www.asepv.com/asandas.htm , compensated by the better performing CadTe), and the Gloria Solar around -0.41%/ oC ( http://www.gloriasolar.com/adobe_file/GSM6_6x10.pdf or http://www.gloriasolar.com/adobe_file/GSM6_6x8.pdf ). At 45°C, this means you loose around 8 % of power. But when the air is at 45°, the panel is much hotter than 45°. And some have found that the loss is higher than temperature would theoretically give : https://groups.google.com/forum/?fromgroups=#!topic/sam-user-group/E629_spgPhA
The high use of renewable energy in eastern Germany driven by government green energy policies is causing instability to its own electric grid as well as to neighboring countries, resulting in industrial companies having to purchase generators and emergency back-up systems rather than face replacing equipment damaged during disruptions of service. Electricity bills are also expected to go up by 10 percent this year.
Poland and the Czech Republic, are taking action on Germany’s use of their power grid that Germany undertook without asking permission and without paying for its use. These countries are building a huge switch-off at their borders to block the import of green energy that is destabilizing their grids and causing potential blackouts in their countries. This action by German’s neighbors fragments the European electrical grid, turning Germany into an electrical island.
Is it just me but I can’t wait for an European wide cold snap. See how this newly added German capacity is going to perform.
Last year, Germany used less electricity than in 2008. But they’ve added a new block of 675 MW lignite unit in Boxberg (apparently the new very large BoA plant in Neurath, of 2,2GW only replaces older capacity). So not having enough electricity is not about to be a problem, the more given that almost none of their electricity is used for heating (it’s just way too expensive).
The trouble is more when there’s too much electricity, and that’s what the Czech are protecting against. We’ll see what happens in May, even if the new Neurath plant can follow load faster than before …