Stop paying extra for a bunch of useless data you don’t need.

By: Adam Baker

From an integration standpoint, controlling a tracker is not rocket science. In fact, it’s extremely predictable. We can predict where the sun is going to be tomorrow at 10:06 a.m., or at 4:30 in the afternoon on June 30 of 2017.

The problem is, trackers can generate a lot of data…and data is expensive. Solar farm owners have no clue what information they actually need from their trackers…and they’re paying for it.

Let’s talk about what information you should and shouldn’t specify when installing solar trackers. But first we need to discuss why someone would install a tracker in the first place.

 

Why Install PV Trackers?

There’s a fine line between how much extra energy you can squeeze out of a solar plant, and the cost it takes to get that energy.

If you read my last post about clipping, you may remember how with fixed racking, you can increase shoulder energy by increasing the amount of DC behind the inverter, but extra modules (and BOS) provide diminishing returns when the DC/AC ratio gets larger than ~ 1.4. This is where trackers become important.

  • On a normal, fixed rack site, irradiance on a clear sky day looks like a bell-shaped curve.The higher a solar site’s DC/AC ratio, the wider the shoulders of energy on a normal irradiance bell curve. The problem is, while wide shoulder energy becomes AC energy, the clipped potential energy remains unrealized, and gets exponentially more expensive. Purchasing more modules to up the ratio increases costs without growing realized energy by much. Financially, anything above a 1.5 to 1.6 ratio doesn’t pencil out.

 

  • On a site that uses trackers, output on a clear sky day goes from 0% to100% very fast. Trackers point to east in the morning and follow the sun during the day. Because they don’t need to wait for the sun to reach its zenith to reach peak energy output, tracker output extends the typical bell curve into more of a long trapezoid. Steep rise in the morning, steep fall as the sun goes down, and a mid-day dip. Trackers mean faster realized energy while you can keep a relatively low DC/AC ratio. The DC/AC ratio can be 1.1 or 1.2, which offsets some of the price penalty for tracker hardware.

Trackers make more financial sense as their costs decrease. A few years ago, trackers cost $0.30/watt. Now they’re down to $0.08 to $0.10/watt. Due to their ability to gather highly profitable afternoon energy, we have customers who have decided never to install a fixed rack site again.

 

Downsides to Solar Trackers

Although trackers increase energy, they also add cost and complexity.

  • With many moving parts, more intense monitoring and regular maintenance is required to keep actuators running without incident.
  • Integration costs increase as complexity, data points, and additional considerations increase. For example, fixed racking is engineered for the maximum expected wind load at the worst possible incident angle. Trackers have a variable incident angle, so typically anemometers are employed to ‘flatten’ modules as wind speeds rise.
  • Land cost may disqualify sites for trackers. Tracker rows are placed north/south instead of east/west like typical fixed racked systems. Some trackers can tilt +/- 60 degrees, but the steeper their tilt, the longer shadow they cast. The longer the shadow, the more spacing is required between rows. The more spacing between rows means more land is required. Backtracking can allow for tighter row spacing, but now you are moving the panels away from the sun during the earliest and latest parts of the day.

How a Control System Integrator Views Trackers

As I stated before, the biggest problem I see with trackers is their abundant data, much of which is unnecessary for day to day monitoring. In one of my earliest tracker experiences, I saw a SCADA system collecting 28,000 tracker data points per MW. 28,000! Of which, only about 250 were actually useful to the owner. Hard specifications sometimes state that all data from instrumentation must be integrated in SCADA, and all points in SCADA must be historized.

When customers specify the inclusion of all (or most) tracker data points in their SCADA system and historian, it’s usually because they don’t know what they need…and it’s a costly mistake. At $10 per point over the life of a solar plant, it doesn’t take much to rack up the bill.

A 5MW tracker site with a million-dollar historian? I’ve been dealt one, and talked the customer off the cliff.

Pet Peeve: Dual Axis Trackers are a Waste of Time

For some reason, people are still developing solar sites with dual axis trackers. Every calculation I’ve run screams that dual axis trackers are never financially worth it.

  • You’re wasting land. Not only do you have to worry about row shading from north/south, duel axis trackers must also worry about east/west spacing. Instead of utilizing 40% of the land with a single axis tracker system, you’re only using 1/9th of land.
  • You don’t need the energy. Sites using single axis trackers reach clipping shortly after sunrise. As I elaborated on in a previous post on solar clipping, anything above the clipping line is unrealized energy. The cost of duel axis isn’t worth getting to clipping one minute faster.
  • You’re spending more. Due to their complex mechanics, dual axis trackers are extremely expensive compared to fixed or single axis.

 

Psst! Here’s the Data You Actually Need to Monitor

It’s best practice to involve a system integrator to help you identify which tracker information is important enough to pay for and integrate. For example, I can tell you that some of the useless tracker information owners pay for includes:

  • How much time it takes each actuator to execute its last move
  • The average actuator motor current over the day
  • Each actuator’s individual ideal position
  • The actual position of each actuator

Here’s the data you should be tracking that can actually help you monitor and make future decisions about plant performance.

  • The max current for each tracker all day long, and a report of which were in the top 5
  • Commanded position, and any that aren’t within a few degrees of that position
  • Alarms associated with trackers
  • Maximum actuator current per day

With just the basics, you can identify if things are working as expected or if something needs to be examined. If the same actuator draws the highest amount of current every day, it’s probably in need of some maintenance. If one tracker is pointing 20 degrees away from target angle, there might be an obstruction. Other than that, let it do what it does. Even 10 degrees of error from the ideal means the module will see 98.5% of the potential (COS 10 degrees).

Also, remember that a tracker’s failure mode only makes the shoulders a little narrower on inverter output. The site doesn’t shut down, and nothing catches on fire. Identify and correct the issue, but failure is generally not catastrophic.

 

Don’t Overdo Tracker Controls or Monitoring

Just because data exists, doesn’t mean you need to monitor it. Screen minimalism brings attention to problems quicker than a screen full of data an operator has to hunt through to find the anomaly.

If you decide to increase the width of clipping shoulders without increasing peak DC capacity by installing a single axis tracker, make sure you talk to a system integrator before moving forward. By understanding exactly what data you need to monitor, you’ll avoid paying extra for information you don’t actually need.