A lot has changed in the production of solar energy since Ron Corio first entered the game 37 years ago. Known to some as the "Godfather of Solar" for a combination of Italian heritage, New Jersey upbringing and enduring presence in the solar industry, Corio has seen demand grow, costs plummet and new technologies to draw maximum power from the sun emerge. We sat down with the CIO (Chief Innovation Officer) of Array Technologies in Melbourne to discuss the latest innovations in the realm of solar tracking, and what the future may hold for this increasingly vital form of renewable energy.
Corio was in town showing off a shiny new tracking technology built to control the orientation of large solar arrays modules on the fly. Using machine learning algorithms, this so-called SmarTrack system familiarizes itself with the conditions of the site and makes adjustments to extract as much solar power as it can, getting better at that task as time goes on.
We got hold of Corio at the All Energy conference, and picked his brains on the finer details of this new system, the role of industry players like Tesla and the path forward for renewable energy. What follows is a transcript of our chat, edited for length and clarity.
What is solar tracking and why is it important?
Solar tracking is basically following the sun throughout the day, keeping the modules perpendicular to the sun so that they produce the most amount of energy they can. We produce what are called horizontal single axis trackers for utility scale, but there are also dual-axis trackers that have been used in previous years. Today it's really about minimizing the cost of the energy produced, and single axis trackers provide high density while increasing the yield of a power plant by 20 to 25 percent.
What are some of the less obvious ways weather and site conditions can affect solar yields, that people mightn't be aware of?
I mean, you are building a piece of equipment that has to last 30, sometimes 40 years and you have to design for all the weather that happens in that period, all the corrosion and the durability of the product itself. We've come up with many patented and unique ways to do this, we've been in business for 29 years, building solar trackers for remote homes in the early days, to 10 GW trackers.
One of the things we do to combat the weather effects is a patented wind mitigation system for our tracker. For many trackers, when the wind blows, they have an anemometer that measures the wind and sends radio signals out to all the individual trackers and tells them to go into a safe position to withstand the wind load.
With our system, as the wind load blows onto the system it creates this hinge moment force (the force required to shift a control surface). Internal to each row we have basically a clutch type mechanism, where if the hinge moment gets too large or if there's dynamics vibration interaction where the load gets high, its releases. It's like a safety valve, to the wind, built into every row.
The outer edges of a field get more wind load than the internals. So what might happen, it usually happens at about 100 km/h in a controlled manner, the outer rows would actually rotate in a more vertical position, creating essentially a wind fence for the rest of the field.
The beauty about our system is it only moves in defense of the wind when the wind acts upon it. It's not that an anemometer says "stow the entire field," which can impact your production and the severity of the results can be very high. So we have, I believe a very holistic, passive and reliable way of minimizing risk and optimizing performance in power plants.
How much can this SmarTrack technology boost solar yields and why does it represent the "next generation" of solar tracking?
There's really two main components to SmarTrack. One is for backtracking optimization, and what that means is you follow the sun, and keep the panels perpendicular to the sun in the east-west motion. When they reach their limits, the sun keeps moving obviously and to keep one row from shading the other, you start backtracking, you start taking the modules off the sun so that the shade from one row doesn't shade the next.
That works perfect on a flat site, you know your algorithms to make that calculation are fine as long as the site is flat. But if you get into a undulating site, you can optimize that backtracking better. So what we've developed with SmarTrack is a system that actually monitors the output of the inverter that these arrays are attached to.
During the backtracking we know where the sun angle is and we jog the system around its programmed point to find the maximum power output of the inverter for the as-built conditions of the site. We then create a table that says when the sun is at this angle the tracker doing backtracking should be at another angle. And once we fill that table for each tracker motor, then that's forever, we don't have to sense anything anymore, we don't have to have any active components and its just a geometric relationship to the sun angle and the tracking position.
And how many different modules or panels are we talking about here, across an entire field?
So one of these motors drives about two soccer fields of modules. Each row is 90 modules, or 90 meters (295 ft) long, and then we link 32 rows together, so it's a significant area with one small motor. We optimize that block for the as-built conditions and we do that during commissioning. So that's SmarTrack for back-tracking and obviously the gain you get from that is really about the site specific conditions, how undulating it is on so on and so forth.
The other area of SmarTrack is about diffuse light capture. We grew up in this industry building optical tracking systems and we used to watch them sky upward when clouds passed by, because they were looking for the brightest part of the sky.
So what's an optical tracker?
Today we use astronomical algorithms to say, it's this date, it's this time, we're at this part of the Earth and the sun is over here. We actually just open loop calculate where the sun is and then we move the array into the position, and we're not looking at anything in the sky. Optical tracking systems, which is something we've built for many years in the remote home industry and off grid industry, are systems that actually look for the brightest spot in the sky and then point the array towards that.
We have a lot of that information and experience through the years and what we do with these systems to harness the maximum power during a cloudy day is have sensors that look up at the sky. We threw in machine learning algorithms that understand when we want to move to sky, or flatten the array so it gets greater irradiance and greater power output. We also have to know when to come off of that, and so that's all part of smart capture.
Can you go into a bit more detail on the "self-learning" element of these algorithms? How do they adapt and improve over time?
We can try and map the conditions of a specific site. We know how much cloud cover there is, we can correlate that with seasonality and so forth and we can hone when we go flat. The trick of this is not to go flat every time there's a cloud, because there is a lot of movement and you might have one cloud pass by and the thing flattens out and then we have to move again.
The idea is to aggregate this, at least in the beginning, and learn about the optimum conditions on when to go in a more horizontal position, and also when to leave that horizontal position. And over time that you can learn from a single site or single geography and correlate that with seasonality, and basically hone those times at which you decide "I'm going to move flat" or "I'm not going to move flat," and refine that algorithm.
So rather than having someone pull the strings on that manually, you setup these algorithms …
… and they reiterate over time, and they're date and season specific. It's machine learning about how to make better choices about when to do certain things.
Is this a new thing in the solar industry?
I wouldn't say it's completely new, I'd say there's some of this going on in all industries today.
What are the main technical obstacles you see preventing the further adoption of solar energy?
Storage. So I've been doing this for almost three decades and I've watched solar prices go from US$12 a watt to 30 cents a watt today, or less. So the one thing I've learned is as the price drops, the market is elastic and just grows. In many parts of the world if not most parts of the world, solar on a kilowatt basis is cost-competitive with fossil fuels today.
That's where we are now. The issue is, we can produce the energy when the sun's shining, but we need the energies 24/7. We have to advance the solar power plant to do things that traditional power plants do, with things like frequency regulations of the grid, and that's happening today with solar power. But the real goal is to provide base load power to the grid.
In order to do that, we have to store energy. There are many ways to store energy, there's water, there are mechanical means, there's lithium ion batteries, there's flow batteries, there are many different technologies out there vying for their position, there's even creating hydrogen. You name it, there are lot of ways to take electricity and put it in to something you can store and use later.
Obviously lithium-ion is growing, primarily because of the transportation industry and the use of electric vehicles. We'll see which one wins out, but like I said, solar was at $12 a watt, now it's at 30 cents a watt.
I feel like give it time, let the technology develop, let it be produced at scale to reduce costs, innovation will take place. Technology driven energy will eclipse fossil fuels, because fossil fuels you have to dig out of the ground, you have to transport them, you have to burn them, let's not even talk about pollution, and pollution is a big thing. Technology-based energy sources from renewables is constantly on a path of cost-reduction and efficiency and improvement.
You might be aware of the Tesla battery that they installed in South Australia, the world's largest battery. How do you see the role of that and the example that it sets as a proof of concept for energy storage?
So these all start as proofs of concept and then they get larger and larger, and that's a new proof of concept at scale … I'm not a storage expert, I can't call out whether lithium ions are going to win over other technologies, but it definitely has some momentum behind it because it's being used at scale.
I guess it breaks down some kind of barrier?
That's right. And that's what happened with solar by the way, when it was $12 a watt, it was because all the silicon was made for calculators and computers and very small quantities. And when there were breakthroughs in the mining of silicon and the production of high-grade silicon, that's when the price of solar modules plummeted.
And I think the same is true with storage. The interesting thing about storage is there are multiple ways to go about it. Flow batteries are interesting because you have a unit and you can store lots of energy in a tank and flow it through the battery. But all these technologies have different applications, right? Some are better suited for transportation, that may be the same technology that works for utility storage, but it may not. It remains to be seen.
Where do you see the solar industry in 10 years, in terms of its contribution to our overall energy needs?
We are now seeing the world sort of wake up to solar, because like I said, its on grid parity with other sources, its easily dispatch-able and it comes with zero fuel cost risk. Once you build it, the sun is free so as long as you know what your operation and maintenance costs are, you know what your overall costs are.
It's very low-risk and easily deployable. Small systems, large systems or residential rooftops, you can deploy it in many places. I think solar is going to continue to decline in price, it's going to continue to get more efficient, and I think the market has demonstrated that it's incredibly elastic and will continue to grow. And I see renewables, solar, wind and maybe hydro, continuing to take over more and more of the global energy market until it takes over nearly all of it.
It's the way of the future, I think it's the mega trend in energy for the next 50 years and it's just stepping through this, "we only produce energy when the sun shines" to "we provide more value" to then "we provide base load value." And that's the path for renewables. The interesting thing is in most places, solar today can makeup 30 percent of the grid, without too much impact on the grid. So there's lots of market for today's solar right now and that will give us a runway to Solar 2.0 to Solar 3.0.
