Hi Readers,
back to the topic of testing PV.
I have dabbled with PV for many years. Since early mono days using ARCO panels made in WA, Solarex poly cells and panels in the 1980's and even Green cells from the one and only. Have not had the funds to go the GaAs, or hybrid cell route yet.
Solar panel performance data is only as good as the incident solar radiation measuring equipment used to compare panel output with.
I made a solar radiation [pyranometer] sensor from a single silicon solar cell. I was given some space grade cells [75mm round], but any silicon solar cell will do, with suitable current output. I connected an Alegro Hall effect current sensor, directly across the cell, with the shortest leads. This measures cell current, that happens to be directly proportional to incident solar radiation. Temperature effects are minimal, as this effects the cell voltage, and in this instance this is close to "0V". I used a current sensor with a 0-5A range, and this worked well with my solar cell that produced 3A at 1050W/m^2 solar radiation. This was calibrated against my Davis weather station incident solar radiation sensor, and gives better than 1% correlation.
Visual human interpretation of how much solar is available is a guess at best. Probably have more luck winning Lotto.
It is surprising the effects of high cloud, [glare] and haze have on solar energy harvest, compared to cloud free conditions.
I have incident solar records going back years, logged at 10s intervals at my location. I have test records for many comparison solar systems including Amorphous panels, Poly-Crystalline, Mono-Crystalline [Std] and Sanyo [HIP] and mono-Crystalline [Sunpower buried conductor] types. I have found that most panels produce their rated power. Deficiencies occur usually with MPPT equipment connected to them. In certain weather conditions I have measured up to 20% difference in energy harvest from the same panels in the same orientation, but with different inverters.
It is very difficult to measure objectively and conclusively say one particular system is better than another, or a system is not performing properly, based on spot measurements.
I have spent a lot of time identifying the interactions of weather and MPPT equipment on solar panel energy yields. I am no expert yet.
I have to conclude that even from an informed position, removing the confusion is difficult.
The one thing that has happened over the years is that solar panels are far cheaper now than even 6 years ago. The low price makes the confusion dated. When I first looked at home solar, I had to really work the numbers to justify a system. The honeymoon period has passed with free systems, but systems are still about 1/10 the price I paid in 2006, for a 1.5kW installed. I will not go off grid, but continue with a hybrid system. I have made the grid my friend, and look forward to the checks they pay me.
At the end of the day, as long as the panel voltage is reasonably matched to the batteries [inverter], and the regulator can handle the maximum power, and the load demands are met, then you are most of the way there. The OFF grid systems I have seen that have the most problems are related to insufficient, battery capacity and recharging sources, to meet the demands of the loading. Typically battery loss of capacity, due to prolonged periods of low SOC. It really takes a typical lead acid battery many more hours than are left in a solar day to fully replace the energy that is drawn out. This is especially noticeable if the battery has been drawn down to the 50%SOC level. The best system I have monitored has 1300Ah @ 48Vnom battery, with 7kW of solar RE, and a 2.5kW windmill, with an auto start Genset, supplying a house with 10-14kWh daily loading. There are times during the year when the genset is called due to low battery SOC [<50%]. When there is plenty of solar, the battery is effectively charged by 10AM, and the bulk of the day is holding the battery at float voltage. You really can't beat a surplus of RE, in reducing the worry of maintaining an OFF-grid system.
Goood luck deConfusing.
Gordon.