How photovoltaics are tested and rated. Richard Perez Have you ever wondered how PV modules are rated for power output? How do those magic wattage numbers appear on the back of every module? Well, virtually every module is tested by their manufacturers. This article discusses how PV makers test and rate their modules. And how these power ratings may be different from actual module performance out in the sunshine. A long and winding road... This series of articles grew from PV testing we have been doing over the last three years. We found differences between the performance ratings printed on modules and their actual performance in the sun. We set out to find out why. This turned out to be a very long journey indeed. We secured information from the modules' makers, we talked to the govt-sponsored Solar Energy Research Institute (SERI), and we set up module "test beds" for evaluating modules ourselves. During the next few issues of Home Power, we will be printing the actual performance data of virtually every module, new and used, now available. This article defines the terms, standards and procedures used by PV makers and by us during our "in the sun" PV testing. The Standards All measurement depends on standards. Without using clearly defined standards, measurement is meaningless. Rating the power output of a photovoltaic module is done in a highly structured and standardized fashion. We, for example, only take module data between the hours of 10 AM and 2 PM sun time. We insure that the module is perpendicular to the incoming sun. Here are the various measurement parameters and a schematic of our test jig. INSERT TEST JIG SCHEMATIC Voltage Modules are rated at two voltage levels. The first is called "Open Circuit Voltage (Voc)" and is just that. The voltage output of the module is measured with the module disconnected from any load (hence the name- open circuit.). The second voltage rating point is called "Voltage a maximum power point (Vmp)" and is the voltage at which the module puts out the most power. All voltage measurements are made at the module's electrical terminals mounted on the module's back. These measurements are made with a highly accurate voltmeter. We use the Fluke 87s with 0.1% accuracy. Current Current is also rated at two important levels. The first is called "Short Circuit Current (Isc)" and is the amount of current that the module supplies into a dead short. The second current rating is called "Current at maximum power point (Imp)" and is the number of Amperes delivered by the module at its maximum power point. Current is measured with a shunt in series with one of the PVs lead. The voltage drop across the shunt provides accurate current measurements. We use 10 Ampere, 100 mV. shunts made by Deltech that have an accuracy of 0.1%. We use a Fluke 87 in 4 1/2 digit mode to take the measurements. Maximum Power and Maximum Power Point Power is equal to Amperes times Volts (P=IE, or Watts=Amperes X Volts). Every module has a specific point on its power curve where the product of Amps times Volts yields the greatest Wattage. This is the Maximum Power Point, and the module's wattage output is rated at this point's voltage and current. So to find the module's maximum power point we take data over the entire range of voltage and current. Because we have taken the modules voltage and current data, we can compute the wattage for each Current and Voltage data point. By doing this we can easily find the Maximum Power Point in the sea of Current versus Voltage data. The charts and table detail a single test run on a 10.8 Watt multicrystal PV module. All the data appears on the table. The graphs detail the data as Volts versus Amps curves and Power versus Voltage curves. We took the data with a module temperature of 41.5øC. (104øF.). The curves of performance at 25øC. and 60øC. where derived from the 41.øC. data. Here's what the finished product looks like. INSERT PV DATA Effect of Temperature on PV Module Performance As the temperature of a module increases two things happen. One, the voltage output of each cell decreases, and two, the current output of each cell increases very slightly. The graphs show the effect of temperature on module performance. If the module is at its rated temperature of 25øC., then the module will supply its rated power output. If the module's temperature is increased to 40øC., then its output drops to 94% of rated. If the module's temperature is increased to 60øC., then its output drops to 87% of rated. This is why we don't see rated output from modules on hot days. The use of 25øC. as a temperature standard at which all other data is taken, leads to less than rated performance in the sun. When modules are actually doing their work, they are temperatures greater than 25øC. We have measured module temperatures as high as 76øC. (169øF.) on very sunny, hot (air temp 38øC. [100øF.]), and windless days. The point here is that, with the exception of cold winter days, the modules are always running at 40øC. or greater. We measure the temperature on the back of the module with a Fluke 80T-150U temperature probe. Air temperature and wind play a big part in the module's operating temperature. Solar Insolation Solar insolation is a fancy term for how much sunshine is an object receiving. All modules are rated using a standard solar insolation of 1000 Watts per square meter or also as 100 milliWatts per square centimeter. This standard insolation is rarely seen anywhere on the face of the earth, other than in laboratories. This is because solar radiation is never uniform and stolidly refuses to be consistent. Too many factors affect the amount solar radiation a body receives. Small items like weather, altitude, and reflection all make realistic standardization of sunshine impossible. So we do the best we can and measure the amount of sunshine hitting an object. There are two ways to measure sunshine. One is with a PV module that has been calibrated against a standard radiation of 1000 Watts per square meter. The second instrument is called a pyranometer. We are sending two PV modules to SERI for calibration and future use. Right now we are measuring solar insolation with a Li-Cor 200SB Pyranometer. This pyranometer produces 1 mV. DC per 10 milliWatts per square centimeter with an accuracy ñ5%. We measure the pyranometer's output with a Fluke 87 DMM in 4 1/2 digit mode. Flash Testing Modules The folks who make the PVs test them under artificial light inside a building. These folks need reproducible lab standards that are not at the mercy of solar insolation and weather. Most manufacturers use what is called "flash testing". This means that the module is exposed to a short (1ms. to 30 ms.), bright (100 mW. per sq. cm.) flash of light from a xenon filled arc lamp. The output spectrum of this lamp is as close to the spectrum of the sun as possible. A computer watches the modules output and gathers the same data as we did above- voltage and current. This data is compared to a reference module located in the flash chamber with the module under test. The reference module has its power output calibrated to solar insolation by SERI or by Sandia National Labs. Flash testing is done at temperatures between 25øC. and 28øC., depending on the particular PV manufacturer. The results of flash testing determine the numbers you see printed on the module's back. Every maker we talked to, flash tests each and every module. Testing Modules in the Sun Testing modules in the sun produces different results than rating them with a flash tester. The main difference is caused by temperature. Manufacturers of PV modules must test modules in artificial conditions because they mass produce these modules. The rated performance that flash testing produces is not what a user will actually see in the sun. This difference is why we are going to test virtually every module available and report on the results in Home Power. I think that the manufacturers of PVs could better serve their customers by rating modules at between 40øC and 50øC. Just making this one change in standards would do much to bring manufacturers' rating into line with actual module performance in the sun. While gathering information for this article, I talked to many PV industry folks. Many of them expressed the same desire- to use standards that more closely reflect actual operating conditions. For example, here is an excerpt from a letter regarding ratings from Mike Elliston of Carrizo Solar. "Carrizo Solar Corp. purchased the Carrizo Plains solar power plant in January 1990. In June of 1990, we begin taking down the ARCO M52, 4 volt laminates from that field. We devised a laminate rating procedure using the industry standard test conditions which are cell temperature of 25ø C. and 1000 watts/sq. m of solar insolation. We have relied on a comparison to a "reference cell". This is a laminate that has been "flashed", which means rated under standard test conditions by Siemens Solar at Camarillo, CA. We compare the output of this reference cell to the output of a laminate under test. This is done in the sun using a PVI brand automatic IV curve tracer. We sort the laminates into these groups: 32+ watts, 27-32 watts, and below 27 watts and price accordingly. This method gives us an output rating which is comparable to that of the other manufacturers. How useful is this standard rating? The standard rating is more optimistic than useful. 25ø C. is not a typical cell temperature. If it is 25ø C. and sunny, look for cell temperatures of 40ø C. to 65ø C. If it is 35ø C. (95ø), cell temperatures could reach 75ø C. with no wind. The voltage and power drop 0.4% per degree C. A 40 watt (25ø C.) module is only producing 33.6 watts at 65ø C. and 15 volts sinks to 12.6 volts. Under these conditions this 40 watt, 15 volt rated module would no be able to charge a battery (where 14 volts are required). What the module buyer needs is more than one 25ø C. power curve. He needs 2 or 3 power vs. temperature curves to try and match his location to the appropriate curve. Only with accurate information on his charging system and the power curve for his location can an informed decision be made about modules. Michael Elliston, Carrizo Solar" Home Power's PV Testing Program So we are setting up a large test bed out in the sun. We will be testing just about every maker's new modules and also the used modules now available. We will run all the modules side-by-side, under the same solar insolation and at the same temperature. We will report extensively on our results in the next issue of Home Power. Meanwhile, if you would like to set up your own test bed and take data from your modules, please do. Please send us a copy of your data and we'll include it in our PV survey. The more data we collect about module performance, out in the sun and heat, the better we will be able to design, purchase, and/or use our systems. Access Author: Richard Perez, C/O Home Power, POB 130, Hornbrook, CA 96044 ù 916-475-3179. Information about PV Testing supplied by the following organizations: Keith Emery, Solar Energy Research Institute (SERI), 1617 Cole Blvd., Golden, CO 80401 ù 303-231-1032. Michael Elliston, Carrizo Solar, 1011-C Sawmill Rd. N.W., Albuquerque, NM 87184 ù 505-764-0345. Al Panton, Kyocera America, 8611 Balboa Ave., San Diego, CA 92123 ù 619-576-2647. Ramon Dominguez, Solarex, 1335 Piccard Dr., Rockville, MD 20850 ù 301- 698-4468. John Loveless, Siemens Solar, 4650 Adohr Lane, Camarillo, CA 93012 ù 805-388-6254. Joel Davidson, Hoxan America, POB 5089, Culver City, CA 90231 ù 213- 202-7882. Where to find the instruments necessary to test PV modules. Pyranometers: LI-COR, Inc., Box 4425, Lincoln, NE 68504 ù 402-467- 3576. The LI-COR model LI-200SB pyranometer sensor costs $200. plus shipping. Shunts: Deltech, 13065-H Tom White Way, Norwalk, CA 90650 ù 213- 926-2304. They make a 10 Ampere, 100 mV., 0.1% accuracy shunt (model MKA-10-100) that is ideal for measuring the current output of a single module. Cost is $12.20 plus shipping. Digital Multimeters and Temperature probes: Fluke instruments are available everywhere, check your local phone book or with HP advertisers. Rheostats and high wattage resistors: Fair Radio Sales, POB 1105, Lima, OH 45802 ù 419-223-2196. Fair Radio sells a 1.6ê, 220 Watt resistor for $2.50 and 3ê, 300 Watt rheostat for $6.50.