When contemplating the use of a solar electric photovoltaic system, it is important to assess how much energy in theory the system can produce according to location, orientation and plant conversion efficiency. Employing a performance monitoring system is a must for being able to account for the amount of energy produced by a system in real time, and to ensure the forecast conversion efficiency will remain intact over its service life.
We are all familiar with our residential electric meter used by the utility company to record and bill us monthly the kilowatt-hours consumed. Over the course of a year, these bills can be compared to determine monthly consumption. While this scenario illustrates usage consumption, it is different for monitoring production with PV systems. A meter is also used to measure the energy produced, but instead of a monthly basis, we are interested in the amount of energy produced during short time intervals – perhaps every hour or every five minutes.
The recording frequency requires more sophisticated data loggers than the meters installed for residential systems. Data loggers feed data into a database that can be archived for use at a later time. They also have communication interfaces through Ethernet or serial ports that allow computers to access the data.
Most electric utilities in the U.S. have adopted standard criteria and guidelines for interconnection of distributed generation (DG) into their electric distribution systems. Solar PV installations effectively reduce the customer load and, during minimum loading conditions, may export energy back to the utility in a net energy metering (NEM)transaction. The Institute of Electrical and Electronic Engineers (IEEE) recommended a set of guidelines (IEEE P1547.6) to PV system integrators to help them design systems that are operable with utility systems.
At Staer Sistemi, we tackled the design of an automated PV management system in late 2009 and have since revised it. The initial design incorporated a simple data acquisition system, but we quickly saw the need to add more sophistication. The volatility of solar radiation at ground level – mainly due to atmospheric turbulence – required a fast sampling rate of five seconds or less.
Due to this requirement, we decided to develop an industrial supervisory control and data acquisition (SCADA) environment. This way, PV system designers could manage data streams in the range of several thousand measures per second. After conducting numerous system tests as a proof of concept, we configured ARC Informatique's PcVue software to meet PV application requirements. This SCADA provided flexibility in monitoring and controlling the various plant component and operations, including trackers, inverters, substations and meters. The system has proven capable of managing PV plants exceeding 5 MW individually, as well as some larger multi-tenant, multi-site systems.
The system monitors PV plant performance by means of a mathematical model initialized at installation with plant design data: PV panels' peak power, inverter specifications, manufacturer-provided electric parameters, number of strings, strings length, etc. The model is continuously fed with local weather data, and it calculates in real time the correct energy production at 100% plant capacity. The automatic comparison between the calculated and the real production figures – supplied by the data logger – gives a precise measure of plant performance and plant health every minute or less.
Under the hood
From a technical point of view, it is interesting to examine how the overall data acquisition is performed, starting at the DC level. Here, string combiner boxes designed for PV installations have built-in string probe units that measure the values of DC current voltage and power made available through a serial RS485 port – although different methods or wireless may be used – for communication to the SCADA via ModBus. Some remote terminal units (RTUs) are installed at the field location that connects to multiple-string junction boxes on the RS485 multi-drop loops.
At the AC level, inverters expose RS485 ports to allow an easy connection. The native communication drivers from PcVue collect data from control boxes and RTUs with time stamps for real-time processing, storage, alarming, reporting and display. Both DC and AC side parameters, status and diagnostics are continuously acquired. The SCADA capabilities are further used in monitoring of grid protection relays, energy meters, weather monitoring station/sensors, low-tension and high-tension control panels, DC switches, transformers, and any devices capable of affecting plant production.
Additionally, the SCADA application enables dynamic configuration, stand-alone and client-server configurations, redundancy for data protection, and historical and real-time trends analysis, as well as advanced alarm management. Looking further at compliance, the support of such protocols as IEC 61850 and DNP3 are considered an asset if you have to communicate with various electric substation devices, for example.
In order to access all data points, the system has a graphical interface with 2D and 3D displays, report generator, scheduler, and an event-driven engine all to make the process much smoother. Finally, Web-access capabilities provide all kinds of mobility and access to remote devices the application may need.
Lino Picheo is president of Staer Sistemi in Rome. Emmanuel Ecochard is the general manager for North America at PcVue in Woburn, Mass.