Manufacturer Perspectives on Array Tracking Markets, Equipment and Innovation: Page 4 of 7
Inside this Article
IHS Technology forecasts 9 GW of single-axis tracker installations worldwide in 2019, with the US...
A wealth of examples of innovative equipment, systems and services developed with LCOE in mind were...
In an era where 1%-2% efficiency gains define an inverters competitiveness, utilizing a solar...
JS: NEXTracker’s products seem to go through frequent design iterations and refinements.
MM: Yes, that’s accurate. The major innovation was the self-power concept. NEXTracker was the first company to commercially scale a self-powered horizontal single-axis tracker. We’ve continued to refine the design, making it more efficient and more reliable. What we’re focused on now is pairing software with hardware to create more flexibility and future-proof the platform to an extent.
We’re trying to get away from this mentality of thinking about balance of system components that you install and have to live with for 30 years and to start approaching trackers as equipment that’s dynamic and upgradable over the life of a system. It’s analogous in a lot of ways to how Tesla approaches its vehicles. Tesla treats the vehicle as a platform. When you purchase the vehicle, you get a certain set of features and capabilities; and as Tesla improves its technology, you can upgrade it with the new functionality.
The other high-level shift in the product has been a move toward more-integrative solutions. We’ve stopped thinking of the tracker as a single product and started to think of it as part of a value-added system. The first iteration of that is our NX Fusion Plus, our tracker plus inverter plus battery storage solution. We launched it in December last year.
JS: I’m curious about the decision to integrate with flow rather than lithium-ion technology, at least initially.
MM: Both the applications we’re looking at and the cost structure drove that decision. In pairing NEXTracker with storage, our focus has been on storage duration in the range of 2 to 4 hours. Some initial market opportunities we have access to are demand charge mitigation and load shifting. For these applications in particular, a flow battery is advantageous on a cost basis. More broadly speaking, flow has a lot of intrinsic cost advantages, particularly in terms of degradation. With flow batteries, there’s essentially no degradation for the life of a PV plant, so you’re able to maintain the capacity over time. For our current applications, it’s a more elegant, self-contained, modular solution than containerized lithium solutions, which can be custom projects in and of themselves.
JS: What do you see on the horizon for tracking systems in terms of innovation or markets?
MM: The competition in the industry has definitely increased, and we consider that a good thing. We want other good tracker companies out there providing a positive image for trackers as a general technology to our customers. At the same time, competition drives the need to maintain differentiation and to continue innovating. One area we’re really focused on is improving the performance of tracked array systems. If you can get several percentage points more energy production from your tracker, that value accrues relative to the total installation cost of the system.
In other words, if I’m installing a system for $1 a watt and I’m getting 2% more energy, you can think about the increase as 2 cents a watt of lifetime value, which is a huge number relative to the cost of the tracker. This focus on performance led us to the launch of our TrueCapture feature at Intersolar North America, which our customers received very well. It’s a completely new way to think about the control system of the tracker and builds on what Dan originally worked on back in the ’90s with backtracking—a technique to improve the performance of a tracking system by avoiding shading.
With TrueCapture, we’re dramatically re-envisioning tracker control with advanced machine-learning software technology that enables each individual tracker row to essentially learn and execute a completely unique tracking algorithm relative to every other row. The software takes into account real-world conditions, such as site topography, construction variances and weather conditions at any given time. It integrates all that data and applies machine-learning software techniques to compute unique tracking commands for each row that it then dispatches to the plant in real time. It allows us to optimize the performance at a much more granular level than has ever been done before. Just like we broke the physical link of having tracker rows linked together, now we’re breaking the virtual link by allowing each row to optimize its performance. Based on analysis we’ve done at a real-world site, TrueCapture can drive anywhere from 2% to 6% more energy. That’s huge, especially when you apply that kind of increase across multiple gigawatts.