Whether it deserves it or not, the solar energy industry can’t count on continued government largess, thanks in part to the Solyndra mess. But in the end, Solyndra’s demise might be exactly what...
PV System Performance
DC monitoring raises the bar for solar power plants
DC-DC converters, and depending on the particular product, generally has some capability to optimize DC PV array performance and reduce the amount of balance of system (BOS) equipment required. String monitoring boxes add DC string monitoring functionality to combiner boxes in central inverter systems, but these products don’t offer any performance-optimizing capabilities and, like all DC monitoring systems, they pose increased labor and equipment costs. It should be noted that string inverters aren’t typically employed in large scale applications in the United States, except for systems that have a heterogeneous mix of module types and orientations.
Feeder-level DC monitoring, implemented at the recombiner box or inverter level, might be simpler, easier and less costly to implement than at a string or module level. However, the data provided from such systems are less sensitive to trends that occur from a change in system status and accordingly, underperforming modules or strings might be harder to identify and correct. Trend analysis requires voluminous amounts of detailed historical data and is much more difficult to conduct precisely without a DC monitoring system, regardless of whether it’s implemented at the feeder, string or module level.
To characterize the DC performance of a PV array, feeder, string and module-level monitoring systems measure the electrical current flow. Module-level monitoring systems also measure the voltage of each module and calculate its power output. By contrast, feeder and string-level monitoring systems typically make a single measurement of system DC bus voltage and—ignoring voltage drop due to copper loss—apply that value to all strings to calculate string power values. Although quantization strategies vary, values for voltage, current and power aren’t typically instantaneous: they’re averaged and reported over short periods of time, usually 15 minutes or less.
As the location of the DC monitoring system is moved further away from PV modules, the data set characterizing performance of a PV system becomes smaller and less detailed, but retains great value. Comparing the sum total number of module-monitored datapoints to the sum total number of string-monitored datapoints produces a ratio whose maximum values asymptotically approach a number approximately equal to the number of series-connected modules in a string. In other words, at string sizes in use today, a module-level monitoring system can produce up to 20 times as much data as a string-level monitoring system that monitors the same crystalline module PV array and up to 14 times as much data for thin-film PV array systems. Analogously, depending on the number of strings in a combiner box, a string-level monitoring system might produce anywhere from six to 30 times the amount of information produced by the corresponding feeder-level monitoring system.
Many DC monitoring products come with technologies baked into the product, or available as an option, that reduce inefficiencies in an array and allow PV modules to operate more closely to each module’s individual maximum power operating point. The proprietary nature of some of these technologies makes their additional yield hard to quantify in performance estimates, but their inclusion in a PV system enables the quantification of DC mismatch losses and inefficiencies