Solar modules are supposed to make kWh. So why do we care so much about kWp?
By James Plastow
Solar module efficiency is a funny thing. It is overwhelmingly equated with performance at Standard Test Conditions (STC), which specify a temperature of 25¡ÆC and an irradiance of 1000 W/m©÷ with an air mass 1.5 (AM1.5) spectrum. In the real world, you would come closest to these conditions on a clear day at noon, with your panel facing the sun directly, preferably around the time of an equinox. These are the so-called ¡®peak¡¯ conditions that give us PV¡¯s quirky unit ¡®Watts peak¡¯ or ¡®Wp¡¯. The funny thing is that such conditions almost never happen in real life. Sunshine is rarely exactly 1000W/m©÷. During an average sunny day, the moment that the amount of sunshine hitting the ground is 1000W/m©÷, if it even reaches this so-called peak, would never be more than a small fraction of the total time of sunshine. The rest of the day, it¡¯s going to be something quite different. Moreover, when the irradiance actually is 1000 W (strong sunshine), the chance that the module temperature is exactly 25¡ÆC is virtually nil. It gets funnier when you consider that module efficiency changes considerably with variations in temperature and light level. So, STC, after all, tells us a lot about efficiency under very specific lab conditions¦¡Watt peak (Wp)¦¡but very little about how efficient a module is going to be under real conditions.
Let¡¯s not forget what a solar module is supposed to do: convert one kind of kilowatt hours (kWh), the energy in sunshine, into another kind of kilowatt hours, electrical energy. So why not use that as a metric for determining efficiency: kWh of electricity output divided by kWh of sunshine energy input? The reason, of course, is that this measure would mean a module¡¯s efficiency would be dependent on where it was mounted. So the compromise is STC, and it should be understood as such. Calling a module¡¯s efficiency x% based on STC/ kWp is an indication, but hardly a definitive one, and should be considered with other equally or even more important indicators. The more important indicators are arguably those that use real-world data.
As the solar industry grows, such indicators are increasing in accuracy and sophistication through data intensive software applications such as PVSYST and PVSOL. These programs use real-world solar and climate data and calculate performance based on panel characteristics beyond the basic STC rating. Data-intensive work like this can show, for example, the relative merits of different technologies by climate, as crystalline silicon, amorphous silicon, CdTe and CIS all perform differently under different temperatures and light levels. Generally, thin films do better in high temperature and cloudy locations while crystalline modules prefer cold, clear conditions.
Using PVSYST, Table 1 shows some calculations of ¡®kWh efficiency¡¯¦¡DC output in kWh/m©÷ divided by irradiation in kWh/m2 for different locations around the world. The results, in this case for 155 Wp CIS modules, should be startling to anyone in the habit of judging a panel first by its STC-based Wp efficiency rating. We can see from the results that despite a significant gap in kWp efficiency between crystalline and thin film, the actual performance gap in terms of kWh is clearly smaller.
While the difference in efficiency between CIS and Poly-Si is 17% from an STC perspective, the gap in terms of actual kWh produced per m2 is as low as 5% in Calcutta. The results of Table 1 can be viewed in two very useful ways. 1) In terms of the space required for an annual energy output goal, the difference between CIS and Poly-Si will be in the neighborhood of 5-7%. 2) In terms of the energy you can expect for a predetermined area available for installation, again the difference is much less than if you simply compared Wp. Case 1) is illustrated below with the additional amount of space, indicated in red, needed for CIS to achieve the same kWh as Poly-Si, whether rooftop or large scale.
If you were only considering Wp, then the red strip would have to be more than double in area. Moreover, while poly-Si only delivers 5% more kWh per m©÷, you must pay 17% more per m©÷ for the panels, since the price is based on the Wp rating. (In fact the price premium is even larger since thin-film modules are generally priced below crystalline modules).
Whether through real estate costs or system costs, or combined, the key is to understand your returns in terms of your kWh outlook, and the premium you pay or discount you receive should be considered in terms of a much narrower difference in outlook than STC indicates between the two technologies. If you are paying for Wp, as is the most common case today, then it is clear that CIS gives you more ¡®bang for your buck¡¯.
Despite this real-world evidence, there is still a prevailing assumption that Wh output rises in proportion to STC-measured Wp. Even on large projects, goals are set to achieve a certain MWp without fully realizing what is needed to achieve the more important goal of a certain number of MWh annually, and for the lifetime of the project. Where space is not a consideration, lower STC-rated modules are assumed to require higher costs in racking, cabling, and land (known as the ¡®BOS penalty¡¯) to achieve a certain kWp. Where space is limited, projects are often over-focused on fitting as much kWp or MWp as possible. Yet neither revenue nor energy cost reductions come from Wp capacity. The income or energy savings for a solar project will always come from Wh generated.
So the real question is¦¡how many Wh can I squeeze out of my limited area? Or how much racking, cabling and land do I need to generate a given Wh output? It is on this basis that any component of a solar solution should be considered, and in the case of the panels, your price per Wp is a poor indicator of what your cost per Wh will be.
STC efficiency no doubt has its place in the laboratory. But for real PV projects, it is of limited practical value. When considering which type of solar module to choose, end users and investors should take into account the different yields of different PV technologies and consider foremost how many Wh a given ¡®Watt peak¡¯ will actually produce under real-world conditions.
James Plastow has worked in the PV industry based in Tokyo since 2002, first at MSK, then Suntech before joining Solar Frontier (www.solar-frontier.com). His first degree is in Electrical Engineering from Imperial College, London and he holds a Master¡¯s degree in Renewable Energy Systems Technology. Plastow has presented at numerous solar exhibitions around the world and has published a variety of articles on solar.
For more information, please send your e-mails to email@example.com.
¨Ï2011 www.interpv.net All rights reserved.