Using automated system controls will make your plant perform better overall.
By: Adam Baker
Over my career, I’ve been involved in more than 1,000 MW of solar tracker installations at various sites around the US. I’ve seen a lot of systems that worked pretty well…but there are always integration opportunities that allow the overall site to operate even better.
Like many solar site components, trackers are designed to operate independent of the plant control system. Similarly, inverters aren’t designed to have a plant control system above them. Also, met stations are simply designed to be data collection systems that feed information to SCADA.
This lack of system communication/integration means trackers don’t know if the rest of the site is running well or not...and vice versa.
All these separate pieces of equipment in their own islands of automation can derive increased value by working together. I want to review seven small improvements to trackers using controls that will increase plant-wide optimization.
1. Actuator movement
The first of my seven ways revolves around a common misconception that trackers should continuously move to follow the sun as closely as possible to optimize module output.
The downside that comes with this perceived need is the nature of mechanical engineering around the bearings that support the tracker system. Every time the torque tube is moved, bearings are forced to overcome a force of stiction. Every time you start and stop the system, the more wear and tear occurs from stiction. Fewer tracker moves mean fewer stiction events, which equates to a potentially longer tracker system lifecycle.
Most trackers move every 5-10 minutes. If you step back and do the math, it’s unnecessary for trackers to move that often.
As the earth rotates, the sun moves a quarter of a degree every minute (15° per hour). My back-of-the-envelope calculation says that if we did a 5° tracker move (that’s every 20 minutes) with 2.5° behind the sun and 2.5° ahead of the sun, a cosine of 2.5° keeps potential module output at 99.90% of rated output. That’s less than 1/10th of one percent loss due to misalignment with the sun.
To be fair, let’s calculate for overall system losses.
If you have 6 MW of DC installed, there’s likely a 1% loss due to module mismatch, 1% loss from conductors carrying current, and 1%-3% loss due to accumulation of dirt. This 5% overall loss in DC power is pretty typical.
A typical tracker site design has a DC/AC ratio of somewhere between 1.1 and 1.5. That translates to an extra 150kW DC power in every 1MW inverter. When you remove 5% to account for the losses we talked about earlier, there’s still 5%-10% excess DC capacity.
If you set the module to 18° behind the sun, that would give you 95% of rated output which would still be within the site’s DC/AC ratio losses. A 36° move (18° behind the sun to 18° ahead of the sun) would still get you full output from your inverter, and would only require a move once every 2.5 hours!
In summary, 5, 10, or 15-minute moves to keep trackers pointed tightly at the sun are completely unnecessary. Moving panels as infrequently as once every hour will still get you better than 95% of DC energy…which is in excess of what the inverter can convert anyway.
2. DC overvoltage
Early in my career, we put in electrical mechanical switching to drop DC when it was getting too high and the inverter was at risk of tripping. Today’s inverters are a little more tolerant of wide DC voltage ranges, but there’s still an upper limit to what an inverter will accept and operate in MPPT.
Those most likely to run into trouble are sites with a high DC/AC ratio or sites using less-expensive inverters. But controls can definitely help. DC voltage going into the inverter is probably already being monitored in SCADA.
If integrated, the plant control system can command trackers to move ahead of the sun when DC voltage gets too high, which brings voltage back into normal range. Inverter output remains the same, you’re simply using trackers to manage the DC voltage going into the inverters.
3. Rain as cleaner
This is the lowest hanging fruit of tracker optimization. I’ve already written about the soiling problem at solar sites, but in brief, when various forms of soiling like pollen, dust, and leaves get significant enough, there’s a difficult decision to be made.
Do you roll out a truck to rinse dust and dirt off the modules or send people out with a cleaning system? Both are pretty expensive options.
Luckily, most of the country has periodic rain events. Because met stations know if it’s raining or snowing, it’s a fairly simple exercise to integrate that weather data in order to affect the tracker system.
You may lose a little energy by moving trackers to a somewhat flat orientation for rinsing, but the energy loss will be easily recovered in a single day if you can reduce the amount of soiling by 5%-6%. Besides, if it’s raining early or late in the day (or if it’s overcast), you’ve got low generation anyway.
Snow is actually the best way of cleaning soiling off a panel. Instead of parking modules at a fairly high angle at night during predicted snowfall, I recommend allowing 1-2 inches of snow to accumulate on a flat surface. When you move the modules to a higher angle in the morning and the snow slides off, it takes dirt with it. It’s the same as paying someone to squeegee off each module.
4. Mowing under panels
Here in the southeast, vegetation is a persistent problem. Mowing at a solar site is pretty difficult because they must navigate around modules 18 inches off the ground. It’s pretty hard to get a mower under that.
Parking trackers in a flat position makes it a whole lot easier for vegetation management people to do their job. The great thing about trackers is, even flat modules will get you some generation. It might not be pointed at the exact right angle, but as we discovered in point #1, even if you’re off by 10 degrees, you’re not losing much on the energy side.
5. Predicting wind speed
Almost all trackers have an associated wind sensor that detects wind conditions and commands modules to a near-flat position to reduce mechanical stresses from wind loads.
The problem with that system is, it’s not predictive. Wind sensors collect wind speed as it exists in that moment and can’t predict when a wind event is coming. However, a SCADA system getting its weather information from a met station or RSS feed can. It can anticipate if wind is coming in an hour, and you can start to move the modules to a safer stow position.
After all, it might take 15-20 minutes to get all modules pointed in a wind stow position, especially if not all actuators move at the same time. As the wind starts to build, you’ll already be stowed ahead of time.
6. Actuator binding
Predicting when an actuator will stop working is not as difficult as you might think. All you need is each actuator’s maximum current over the course of the day. In normal operations, the actuator demanding the highest current should change from one to the next over time. But if you see the same tracker with high current draw over and over again, that’s indicative of binding.
Binding comes from a number of sources. In the southeast I see quite a bit of vines and grass that grow up into the mechanical structure of the tracker. It could also be a piece of equipment left in the field that’s impeding the tracker from smooth operation. It might just be an old bearing.
Whatever the reason, a high current over multiple days is a trigger to roll someone from O&M to investigate that specific actuator.
7. Minimize data points
For some reason, people want to collect and historize all tracker data possible. It’s important to filter out which data you can actually turn into actionable information that keeps the system efficient and running. I once counted a site with 28,000 tracker data points per MW. Of all those data points, only several hundred were useful.
Not every point is necessary to determine if the system is working well or not. The time it takes for an actuator to execute a move, each actuator’s individual position, or average motor current per day, are prime examples.
How about actuators with the highest current draw over the course of a day? If the same actuators show up over and over again, you’ve got a potential O&M issue and time to plan on what to do about it (see #6).
Instead of telling me all tracker positions, what about only the trackers that aren’t within a few degrees of where they’re supposed to be pointing?
It’s not always obvious how to figure out what tracker system data is useful vs. useless, but Affinity Energy would certainly love to help you work that out.
There’s a small opportunity cost to integrate trackers together with a plant control system up front, but I would expect the return to be many times that over the life of the plant. You get more energy output, better control of maintenance activities, and a system that performs better overall.
Want to talk about solar trackers with our solar monitoring specialist? Contact us!
Adam Baker is Senior Sales Executive at Affinity Energy with responsibility for providing subject matter expertise in utility-scale solar plant controls, instrumentation, and data acquisition. With 23 years of experience in automation and control, Adam’s previous companies include Rockwell Automation (Allen-Bradley), First Solar, DEPCOM Power, and GE Fanuc Automation.
Adam was instrumental in the development and deployment of three of the largest PV solar power plants in the United States, including 550 MW Topaz Solar in California, 290 MW Agua Caliente Solar in Arizona, and 550 MW Desert Sunlight in the Mojave Desert.
After a 6-year stint in controls design and architecture for the PV solar market, Adam joined Affinity Energy in 2016 and returned to sales leadership, where he has spent most of his career. Adam has a B.S. in Electrical Engineering from the University of Massachusetts, and has been active in environmental and good food movements for several years.