At a recent conference, I argued that current utility-scale PV operations and maintenance prices in the U.S. were dangerously low, posing a threat to quality of service and vendor survival. I added that O&M prices might bounce back after bottoming out, as they did in the wind industry a few years ago.
A fellow presenter and esteemed industry colleague rejected this idea: He thought that prices never would go back up, citing the computer and cellphone markets as examples. But this comparison ignores that O&M is a service, not a piece of hardware that gets cheaper with manufacturing volumes and experience.
The primary sources of O&M costs are people, vehicles and tools, none of which are getting cheaper over time. And, in fact, they are becoming more expensive.
A seemingly impossible equation
How can O&M providers resolve the head-scratching equation of continuously reducing costs when the main resources they tap (people, vehicles and tools) are becoming more expensive?
Scale matters, so O&M vendors have to race to increase their service-portfolio density to lower the cost of maintaining each plant (a battle mostly fought at the local level). O&M providers must also continuously improve the efficiency of operational processes and use modern technology, such as automatic detection of system issues and computerized maintenance management systems, to make their service staff more efficient. Furthermore, PV plants must be designed with serviceability in mind to minimize ongoing O&M costs.
All these things are already happening — but at current and future price points, they won’t be sufficient for vendors to offer quality service and remain profitable. The only path forward is to use significantly fewer labor hours, vehicles and tools per PV capacity unit. That means spending less time in the field, relying on technology rather than people to detect, troubleshoot and resolve more issues, and automating low-value field tasks like cleaning and mowing.
Here are some of the technology innovations tracked in the new GTM Research report Global Solar PV O&M 2017-2022: Markets, Services and Competitors — all of which are part of the industrywide trend termed “digitalization.”
Power electronics innovation
String inverters, microinverters and power optimizers (together with the “smart modules” that integrate them) increase the granularity of monitoring and analytics by providing data at the string and module level. This enables finer detection of issues inside the DC array (like faulty or underperforming strings or modules) without the addition of additional hardware like smart string combiners. These technologies are gradually gaining market share, which is a clear sign that the industry is letting go of its tunnel vision on dollar-per-watt metrics, to embrace dollar-per-megawatt-hour metrics as more important data points to consider when making design and purchase decisions.
Going one step further in innovation, Huawei is now offering three-phase string inverters with built-in IV curve tracing functions. Traditionally, IV curve tracing requires costly on-site work that exposes technicians to safety risks. Overall, the O&M vendors interviewed for the new GTM Research report are excited about the prospect of this capability being built into the inverter and accessible remotely via the monitoring software. They also remain skeptical because none of them had operational experience with the technology, which is still very new.
While microinverters and some string inverters require no preventive maintenance at all, new central inverters also require less preventive maintenance (once every two to three years vs. once or twice a year a few years ago). Schneider Electric differentiates itself in this area by calculating plant-specific maintenance intervals based on design and environmental factors instead of mandating one-size-fits-all maintenance schedules.
To reduce maintenance costs and increase plant performance, O&M practices must evolve toward more detection and diagnosis of issues by remote supervision and operations centers. Both O&M providers and independent software vendors are investing in advanced data analytics. Buzzwords like “big data” and “deep machine learning” were mentioned by most O&M professionals interviewed for the new GTM report as some of the key innovations that will help them become more efficient and increase plant performance.
Advanced analytics features include, for example, benchmarking of performance against similar plants (normalized for capacity and weather), and ongoing comparison of a given plant’s performance against a “digital twin” that models the expected behavior of the plant based on historical data and observed weather conditions. However, use of these technologies so far remains limited. Monitoring software provider QOS Energy reports that only 10 to 20 percent of its clients currently use the benchmarking and digital twin functions built into the platform.
Another application of advanced analytics is to predict a future problem before it translates into a failure. The technology, known as condition monitoring in the wind industry, is now emerging as a differentiated software feature in the PV world. The gain is twofold: increased performance if the problem is resolved before production losses happen, and reduced maintenance costs if the O&M provider can replace or repair the component during a scheduled maintenance visit instead of reacting when it breaks.
For example, NEXTracker offers predictive analytics designed to reduce maintenance costs and increase availability of their tracking systems. But another recent GTM Research report, Global PV Monitoring 2017-2022: Markets, Trends and Leading Players, finds that only 13 percent of monitoring solutions by independent software vendors declare full support for inverter condition monitoring (and 28 percent partial support). Conversely, only the largest and most advanced O&M players such as First Solar and SMA have been performing predictive maintenance to date. We can expect the practice to spread as predictive analytics become more broadly available in monitoring solutions in the next five years.
With recent advances in digital infrared cameras, image processing and unmanned aerial vehicles (drones), aerial thermography is gaining adoption in most markets, either as a substitute for IV curve tracing and VOC testing, or as an additional value-add service.
At a basic level, aerial thermography allows clients to pinpoint dead strings and modules that may elude detection by the monitoring system in plants with central inverters that are not equipped with smart combiners for string monitoring. Images with higher resolution, coupled with more advanced processing software, also enable detection of subtler issues like underperforming strings and modules, hotspots, bypass diode failures, cracks and abnormal soiling.
Aerial thermography has become popular for the commissioning of new plants, for inspections prior to warranty expiration, and in preparation for an asset/portfolio sale. It is also rapidly gaining adoption as part of standard preventive maintenance plans.
IEC standards now provide technical requirements for aerial thermography, and most O&M firms interviewed in the new GTM Research report are already using the technology, either with their own drone pilots and equipment, or via service providers (using drones or aircrafts).
Robotic cleaning and mowing
Robotic cleaning solutions are usually built into the power plant, although some solutions are designed to work with existing plant designs. They remain a marginal phenomenon limited to markets with large amounts of dust, high labor costs, or water-access and cost issues. However, as water scarcity increases in many parts of the world, especially the drier locations where most utility-scale PV plants are located, we can expect the technology to gain traction.
In the meantime, semi-automated cleaning methods are gradually replacing manual ones with improved efficiency and costs.
In the rooftop segment, some installations (particularly in France) are equipped with innovative technology that reclaims rainwater and uses it to clean the modules and keep them cool on hot days (thus increasing performance). The solution is mostly deployed in farming environments with large amounts of dust.
Cutting grass remains a challenge in many markets, and O&M vendors are actively investigating and testing new solutions such as semi-automated mowing. In the future, we can expect fully autonomous mowing equipment to become an option as well — just like your Roomba vacuum cleaner robot, except smarter (since bumping into the array is not desirable behavior). Autonomous cars are getting ready for prime time, which means autonomous mowers for PV plants can’t be too far away.
The ultimate paradox
As PPA prices keep reaching new lows, ensuring that plants perform at or beyond expectations is more critical than ever, but unfortunately, the value of each kilowatt-hour produced has become so low that the ROI for upfront investments that increase production keeps getting longer. While some of the technologies mentioned in this article deliver strong returns when solar energy is priced at $100 per megawatt-hour, the economics become challenging when PPAs go below $50 per megawatt-hour.
As a result, the balance between O&M expenditure reduction and plant performance improvement is likely to tilt toward a cost focus — at least for new projects.
For more details about O&M markets and trends and an in-depth competitive landscape analysis by country and segment (including 128 vendors), as well as market forecasts to 2022 and more than 70 detailed vendor profiles, refer to the GTM Research and SOLICHAMBA report Global Solar PV O&M 2017-2022: Markets, Services and Competitors.
For more details about PV monitoring markets and trends, globally and country-by-country, with market size and forecasts to 2022 and competitive landscape analysis by segment (residential, commercial, industrial, utility) and key country, please refer to the GTM Research and SOLICHAMBA report Global PV Monitoring 2017-2022: Markets, Trends and Leading Players.