I received a copy of this PV testing report from very helpful guys at RETScreen/NRCAN. The paper is put together by TUV Rheinland – flipping through it, it looks pretty interesting and contains a few notable pieces of information. I thought I would share it.
Information that was of interest to me:
- Global PV market growth was projected to be around 25% annual compounded increase between 2008 and 2012
- (this was before the credit crisis, but to counter that, it probably did not factor new government subsidies in US and Canada which were announced after 2008)
- US Department of Energy estimates grid parity between solar and utility systems in cost per Watt produced by 2015
- I have to wonder here, if Department of Energy calculated True Cost of coal-powered electricity, factoring not only environmental damage resulting from mining coal, but also medical costs associated with breathing polluted air, drinking polluted water, etc
More to the point:
- Power output of solar panel is influenced by light spectrum, temperature, and irradiance
- Light spectrum was measured at the test site in Arizona and was found to be about the same throughout the year
- Solar panel temperature can reach levels more than double the air temperature, raising from 40 C (Arizona or Colorado test sites) to as much as 90 C. At those levels performance may decrease fairly noticeably
- Larger gap between roof and panels allows for better air circulation and results in better natural cooling, taking the worst case scenario from around 89 C to about 77 C (this drop in temperature represents a 4″ air gap between roof and solar panel)
- Long-term field testing gives different baseline data for solar panel performance degradation:
- BP Solar panels were seen to lose 0.25% per year in output (8-year test run)
- Another 10-year test of BP solar panels revealed a 0.1% annual drop in output
- Third test reveals 0.4% annual degradation over 8-year test run
- Assuming worst case scenario of presented test data, solar panel output may drop by 8% of its initial installed output by the end of the 20-year OPA microFIT contract
- Several solar panel performance modeling tools are compared
- Measured data from test runs = 100%
- First and second best modeled results were given by MAUI and RETScreen, 2% accuracy
- SAM scored much worse than RETScreen, 6-7% accuracy
- RETScreen is a clear analysis leader, versus SAM, relative to modeling future output
- For c-Si solar panel modules,
- 12% failed after 200 thermal cycles
- 13% failed after 10 humidity freeze cycles
- almost 30% failed after 1000 hours in damp heat
- 12% failed at termination point
- diodes failed in more than 30% of panels
- THIS DATA MAKES ME NERVOUS, if I am reading it correctly. Specifically, two data points are of great interest: diode failures and damp heat failures, which are recorded at 30% failure rate. This could be a problem for poorly sealed panels in Toronto, as we are known for very hot and humid summers. Furthermore, diode failures could have a large impact on panels that get some shade. Diodes ensure that electricity can flow only in one direction, and they subdivide each panel into 2, 3, 4 smaller groups of cells. Shading on one group of cells does not affect the output of the other non-shaded groups cells, because of diodes.
- Thin film panels recorded much lower failure rates throughout the chart, with one key exception: damp heat failure rate is a shocking 70%, which makes me wonder if this makes thin film technology financially unsustainable at current panel price levels in hot/humid Southern Ontario geography.
- According to BP, 45% of failures they recorded were due to corrosion, while another 40% were due to cell or interconnect break. This is another proof that humidity IS a factor in solar power collection, and that panels must absolutely be sealed and weatherproof, in order to stand the test of time in Toronto.
- Common reliability problems of solar panels, according to US Department of Energy:
- Glass breakage (presumably in tempered glass-covered poly c-Si panels)
- Delamination
- Broken interconnects
- Corrosion of soldering joints
- Note that some of these issues may expose high-voltage current or cause fire, and present serious safety problems. Buying modules that are certified reduces this risk.
- More than 20 years of reliable operation is required for most solar systems, if grid parity is to be achieved
- In 2010, cost of solar system should be around $0.15 per kWh. That works out as follows:
- ~13,000 kWh/year from a 10 kW solar power plant yield
- 13,000 * 0.15 = $1,950 per year
- $1,950/year * 20 year microFIT lifespan = $39,000 cost of entry in 2010
- As it becomes obvious, cost of entry in Ontario today is nearly 100% higher than this projection made by a US agency for the US market
- Granted, 2010 is not over yet, and we may see potentially rapid price drop into the end of the year, but it is becoming very obvious that as many parts for OPA microFIT installation as possible should be imported from USA or elsewhere. Buying from Canadian resellers/manufacturers does not make sense at this time.
- In 2010, cost proportions of a solar system installation should be something like:
- 30% – solar panels
- 10% – inverters
- <10% – installation
- Solar panels are tested for qualification, and reliability
- Qualification test = market-ready certification, ensures low quantity of defective product shipped, achieves higher-quality product assembly (test-to-pass)
- Reliability test = lifetime prediction, ensures low constant failure rate during life cycle and means to achieve higher-quality product design/technology (test-to-fail)
- Photovoltaic Standards
- Power rating: IEC 60904 and ASTM 1036
- Energy rating: IEC 61853
- Confidence: c-Si – IEC 61215; a-Si – IEC 61646; concentrated PV – IEC 62108
- Safety: USA – ANSI/UL 1703; Europe/Asia – IEC 61730
- Most recent solar panel certification standards:
- 60904-1 to -10 (performance)
- 61701 :2000 (salt mist corrosion)
- 62446 :2007 (grid connected system)
- 50380 :2003 (module datasheet specs)
- UL 1703 :2003 (USA safety)
- 50521 :2008 or DIN 0126-3 2006 (connectors)
- DIN 0126-5 2008 (junction box)
- TUV 2Pfg1169 2007 (cables)
- Also UL 746, UL 484, UL 514, UL 44 / UL 83
- Read the PDF to make more sense of these numbers AND KNOW WHEN TO REFERENCE THEM before pulling the trigger on component purchase!
- Rating from least reliable, to most reliable, your preference in purchasing from a vendor who makes quality claims should be (in order of increasing reliability):
- IEC 60904-1 (out of box performance at static irradiance level)
- IEC 61853-1 (out of box performance at varying irradiance levels)
- IEC 61215 or IEC 61646 PLUS Sandia or IEC 61853-1 (qualification test)
- NREL protocol PLUS qualification test PLUS Sandia or IEC 61853-1 (test to failure)
At some point going down this standards list, you may well find that the price of panel grows out of your budget. This is normal; higher quality product that has longer duration and more expensive/extensive testing to prove it, should cost a bit more than run-of-the-mill Chinese operation, but… We know that solar panels cost a lot more than they should, and we have a reason to believe that due to large number of manufacturing entries and ongoing technology development, these things can only become cheaper, and fast. So, do you really care about getting a very high-end product, for MUCH more money, to only produce marginally better power outputs, but which would add little to your OPA income account?
It is an individual question that each of us will need to answer personally. My gut tells me that I would rather gamble on commodity-priced solar panels, and replace failing panels at some lower future price, when and IF they fail, as opposed to spending a ton of money on something that may fail anyway. While these things are becoming rapidly cheaper, it makes no sense in buying at the top of the price range – when price is stable, then perhaps it makes better sense.
If you are an IT guy, think RAID in storage world, and you’ll know what I mean.
But if there is one conclusion that we can draw safely from this PDF is: know how much a PV system should cost, and know how to read manufacturer’s quality claims against standardized testing. If you chose to pay more for a higher quality product, make sure that it is certified appropriately, and don’t touch it if vendor claims are unsubstantiated.
