In previous articles (here and here), we looked at solar monitoring first as more than just an energy production reporter and then at what really critical data needs to be gathered. In this third session, we’ll look at the basic hardware required to accomplish these tasks and how the many different components communicate with each other and the outside world.
Thinking back to what we want to monitor, the system needs to be able to collect inverter performance data, site weather conditions and the energy production meter information, and then assimilate all of these bits of data into a format that can be exported to the monitoring portal provider. Sounds like a daunting task, but in reality it’s not too difficult since the data collection industry is very mature and robust. When we know what we want to collect, the correct data logger can be chosen that will meet our needs and collect data reliably.
As the core of all energy monitoring systems large or small, the data logger is critical to acquiring the necessary data using a suite of communications protocols that the connected sensors may employ. Think of it first as a translator that can listen to Modbus RTU, DNP-3, ASCII, BacNet, Ethernet TCP and other “languages” and convert each data stream into a common format that can then be readily understood and used. Secondly, think of the data logger as a storage device holding many files with the raw information generated by the translator section. Thirdly, think of it as a publisher that takes the many disparate data pieces and assembles them into a common file, typically in CSV or XML format which then can be used and read by a wide range of external devices. It’s one smart box that is for sure.
As you might guess there are different degrees of data logger capability depending upon the needs of the system. One manufacturer may only provide the ability to sample data once per minute while another device is capable of taking once per second snapshots of how the system is performing. The smaller device may only provide 5 MB of storage while an industrial machine might provide 1 TB of data collection capacity. So it’s really up to the monitoring portal provider to determine what data is relevant, how often it is collected, what reports are generated and in what format. Those decisions along with installed cost will drive the final choice of data loggers.
Data loggers need data, so where does that come from? The short answer is from intelligent devices located on the project site and those devices can take on many forms. Let’s look at several different categories briefly.
The heart of any PV system is the inverter or multiple inverters depending upon the size of the project. By their very nature inverters are extremely “smart” devices and have the capability to gather their own internal operating data and either export it immediately via one of several different communications protocols to the data logger, or in some cases store the data for future export when asked to do so by the data logger. Either way, there is a host of data that is generated by the inverter(s).
What kind of data is being generated? LOTS! Information that we all think about like energy production, line voltages and currents, DC input power and ground fault currents are all readily available just as if you were scrolling through the HMI screen on the machine itself. The inverter data system also collects other data that is critical as far as ensuring machine longevity too. Information on internal enclosure temperatures, cooling fan speeds, fan run time, IGBT status, master/slave section relationships, power factors and phase angles are all collected and available as well. Small string inverters may have 30 or so registers that can be read while the huge megawatt-plus machines can have hundreds of data points! It’s all there if you want and need it.
Weather information is another critical part of the data puzzle. It is quite important to be able to calculate a system’s potential performance based on key weather parameters such as irradiance, temperature, wind speed and barometric pressure. To be able to do so, many different devices are available either as separate instruments or an assembly that will measure those variables and convert analog values into a digital format which then can be gathered, stored and delivered by the data logger.
Irradiance is typically measured by a very important device called a pyranometer and typically there are two located on each site. There are more on larger sites, but their function is the same. One device is mounted level to provide the Global Horizontal Irradiance values in W/M2, and the second is mounted at the same angle as the modules to provide the Plane of Array Irradiance. Those two values are the most critical to monitor and use to compare actual performance to calculated performance.
Another key data point is the back-of-module temperature. As we know, module performance is inverse to the crystalline temperature, so knowing this variable is very important. Typically this is a small thermosistor where its resistance changes based on the temperature of the backplane of the module and the data logger can read those minute changes and calculate the actual module temperature.
Optionally, you can add ambient temperature sensors, wind speed and direction devices and barometric pressure to make a complete package of weather data. Yet more data to be logged and delivered.
The last monitored device is typically the energy production revenue grade meter. This can be one or several, depending upon the site design, and it provides the owner the knowledge that the system is indeed performing at the level the utility meter says. Accuracy is very important as well as ease of set up and use so the installer can configure the device correctly to match the systems operational parameters. Setting things like CT and PT ratios are critical to gathering the correct performance data. Capturing wave forms, power factors and VAR’s are all meter data registers, so selecting the right meter and being able to communicate with it allows the system to provide a very high degree of performance accuracy.
One last comment about communications protocol. Most of today’s hardware is capable of communicating via Modbus RTU which is very robust and simple to configure “language.” It is capable of being transmitted over a much longer distance than TCP/IP over copper and is economical too, so often times it is the perfect choice.
But as more and more of the solar hardware comes factory ready with a small web server embedded, Ethernet makes more sense than Modbus. You can still get the same data, but also you get real time connectivity to the web enabled device, which can be great for troubleshooting and checking performance levels. But there is that 100-M distance limitation over copper, so many times fiber becomes the communication medium of necessity. Or Ethernet over radio which has become quite competitive in the last year or so. More and more we specify Ethernet over Modbus on projects.
So hopefully now you know more about basic monitoring hardware and data sources. In our next article we’ll look at some more advanced applications and how they are handled out on the site. Remember, anything with a digital processor inside is usually open to be monitored, so we’ll look at some of the other valuable information that is available and how to acquire and log-it. We’ll see you then.
John Ruben de la Fuente says
I love my smart box !