tracking applications present a series of tough challenges for automation and
control. Even though these systems don't require high-speed operation, system
integration is a major issue with the need to network thousands of devices
together, and ensure precise sun tracking and data logging. Mirroring the rise
of large systems and concentrated
photovoltaics (CPV), control solutions are adapting along with the major
opportunity that the solar power industry represents.
"Our goal is to address the technology needs for solar power generation using automation solutions that enhance efficiency," says Paul Ruland, product marketing - automation systems for Siemens Industry Inc. "One way we are doing that is through a new algorithm from the National Renewable Energy Labs (NREL) that we put into a PLC function block.
"We took that very complex calculation, a euro Ssystemized it into our code and made a usable function block that customers can parameterize themselves and use it with their particular solar technology to track the sun in the most efficient manner."
Quite often, higher-end algorithms for solar positioning (SPAs) must run on an expensive embedded controller or a higher-end industrial PC. But the Siemens solution is able to use the NREL algorithm on its newest and smallest S7-1200 controller that sells for around $300.
"This has been very attractive to larger solar installations that have many nodes and not necessarily a lot of I/O, but need this powerful computation capability and advanced algorithm," Ruland says. "It accepts computations that very few traditional PLCs can handle."
Siemens' latest S7-1200 compact controller, which would be considered by most to be in the micro PLC class, supports both 64-bit floating point math and the long real data type. For two-axis solar tracking applications, this capability enables use of complex math calculations such as tracking the sun's azimuth (vertical) and zenith (horizon) angles.
The advantage of two-axis tracking is that, as the months and seasons pass, the azimuth and zenith angles change with the rotation of the earth and the angle of rotation of the earth around the sun. These two outputs, coupled with additional environment-related inputs into the function block, allow the system to track the actual position of the node (latitude and longitude), the average barometric pressure, temperature, elevation, atmospheric refraction and other parameters that the NREL algorithm (endorsed by the U.S. Dept. of Energy) supports.
All of these input parameters, as well as connecting the PLC to NTP protocol (National Time Protocol), assures that the system maintains the exact time of day at a particular location. a euro ...The combination of input parameters provides the intelligence to move the solar collection technology exactly at the angle of the sun to get maximum efficiency.
"Compared to traditional technologies, it provides an increase in accuracy of 30 percent or more," says Ruland.
The NREL algorithm also handles calculations up to the year 6000, where the previous version of the algorithm was limited to 15 years and created an uncertainty in the calculation as well as a long-term sustainability question. Use of the PLC function block means customers don't have to implement the algorithm as C source code and develop a runtime engine.
Another advantage of the system is that an automation controller at each solar tracking node provides multiple diagnostics, data logging and functions such as a local HMI for service. Technicians can open the control cabinet and access predictive or preventive maintenance features normally found in a PLC used in the manufacturing industries. The S7-1200 also provides a built-in Ethernet port, internal data logging and the ability to connect to an NTP server to keep its real-time clock accurate.
System Integration Challenges
"The biggest challenge that solar tracking systems face
is deployment of a huge number of solar devices, such as concentrators or
photovoltaic technology, so communication to the different devices is a
problem," says Ed Schultz, renewable energy business development manager for Beckhoff Automation.
Click here for larger image
"The market is also highly cost driven, so there is a struggle to understand what technology works best in the field. Many companies don't have years of experience to draw on, so they go with what they know best," he says. "A recent customer reported use of Ethernet to connect multiplexing (MUX) boxes, which then communicate via serial to the 17,000 heliostat mirrors in the system. We see interest in networking technology such as EtherCAT that offers alternative solutions."
Schultz says EtherCAT technology lends itself to solar applications because the technology can use one master to communicate to as many as 65,000 different nodes. But there also needs to be a business case and risk management decision on how many nodes to address using a single master. Some customers decide to use one master to communicate to 400-500 slaves, and leverage built-in features such as media redundancy, diagnostics and the ability to use Ethernet cabling.
"The combination of EtherCAT hardware architecture and advanced software features provides networking technology that is well-suited for solar tracking applications," says Schultz. "The networks are not high speed, so performance is an advantage but not a requirement, but topology is very important."
Another advantage with EtherCAT for the burgeoning solar industry is that the communication technology doesn't require switches, routers or hubs, and the user doesn't need to assign IP addresses. When an application incorporates 17,000 heliostats and 3,000-4,000 stations for the troughs, there is a clear benefit in eliminating networking switchgear which is both cumbersome and expensive.
With a system architecture like that featured by EtherCAT, all of the controls are running through one main controller - a PC back at the main facility - rather than through distributed controls in the field and having them communicate back to the host. Schultz says that many users often have multiple embedded PCs deployed in the field to control a couple hundred panels.
In these distributed applications, "the more you can eliminate hardware makes an impact on reliability and uptime," he says.
One trend Schultz has seen play out in the development of solar plants is the move toward larger systems.
"There has been a trend toward huge 850 MW to 1 GW solar power plants. But with 7,000 acres of solar dishes or troughs, there is also an environmental concern," he says.
In recognition of the environmental issues with such large deployments, Schultz says Beckhoff is seeing increased interest in developing smaller 50-100 MW plants that can be done in a 700 to 1,000 acre field where combining the system together makes more sense rather than attempting to overcome the problems with permitting for a larger system and the need for an environmental impact statement.
The increased use of concentrated photovoltaic technology
(CPV) is considered a significant trend that shows the solar industry is
starting to reach commercial maturity. Some companies are predicting that it
could become one of the lowest cost forms of solar power.
"CPV systems require very precise tracking because they use a high level of solar concentration," says Brian MacCleery, principal product manager for Clean Energy at National Instruments. "Typically a Fresnel lens in front of the solar cell, similar to the effect of looking through binoculars at the sun, creates a tiny beam of light that is high intensity and can be 300 times more intense than ambient sunlight."
He says that the sun moving even a small amount in the sky can cause the highly concentrated beam, even though it is only travelling a few millimeters, to move off the tiny solar cell if it isn't tracked
properly. To account for this, CPV systems must track in two dimensions to provide sub-degree accuracies and maintain maximum power output. Encoders are key to helping the CPV systems maintain arc-seconds of precision.
The main benefit of CPV is that it reduces costs by minimizing the amount of PV material required, using glass or plastic for the lenses and a metal backing similar to a headlamp.
National Instruments has worked with a company developing standards for characterizing the performance of concentrated photovoltaics. Testing standards for flat-panel modules have become very well-defined. But since CPV is a new area, the standards body has been working to define the standards for its characterization, as well.
"It's an interesting challenge because to characterize the performance, you need a control system that tracks the sun. Unlike a flat panel system, the sun can move enough in a few minutes to affect the power output of a CPV system," says MacCleery. "The testing system uses advanced motion control algorithms that automatically scan to determine the performance of the cells."
A key performance measurement is the sensitivity of both the cell and lens design to sun tracking error. If the user wants to get 90 percent of the maximum theoretical performance output, the goal might be to maintain tracking accuracy to within 0.5 degrees, depending on the system design.
The system is an interesting example of using advanced analysis of measurement data, and feeding that information directly into the control system. It automatically moves in a motion pattern to characterize the "acceptance angle" performance of the cells, based on tracking error, and uses actual performance data to develop an I-V characterization of the cell.
The theory of I-V characterization is that PV cells can be modeled as a current source in parallel with a diode. When there is no light present to generate any current, the PV cell behaves like a diode. As the intensity of incident light increases, current is generated by the PV cell. In an ideal cell, the total current I is equal to the current I? generated by the photoelectric effect minus the diode current ID.
To do the analysis, the system captures voltage and current information as it sweeps through the cell's performance range. Based on that information, the application calculates the resistance of the cell, its efficiency in converting sunlight, as well as a full analysis of cell performance. Historically, MacCleery says this was done only in the testing lab as the panels were being manufactured but now, with the availability of advanced instrumentation capabilities, more analysis and characterization is moving into the field and fusing with the control systems. The reason for this is because instrumentation provides valuable data for optimizing the control and achieving the goal of getting peak power out of the solar farm.
"This combination of instrumentation and analytics, combined with the control system, is a general trend we see in many applications," says MacCleery. "Higher quality measurement and advanced analysis are creating more sophisticated control systems, enabling automation systems to be optimized. This isn't just something happening in factories like we typically hear about; it's also happening in field applications such as sun tracking."