By Mark Triplett, Chief Operating Officer, Stem
The solar industry has long viewed operations and maintenance (O&M) as basic tasks: checking solar panel output and inverter status, dispatching trucks for repairs and conducting annual maintenance, periodic cleaning and vegetation management. The biggest concern was mainly how much electricity can be generated based on the amount of sunlight hitting the panels.
But with the addition of energy storage systems (ESSs) to solar projects, an array of new use cases has arisen, fueled by incentive programs and market opportunities. As a result, the industry is focusing more on how to optimize storage to save costs, generate revenue and improve resilience and sustainability.
Optimizing ESSs is a critical task for solar project operators. It requires new strategies, the ability to analyze vast quantities of data in real-time and a deep understanding of compliance requirements and market opportunities. The associated complexity is more than human operators can manage, opening the door to the integration of artificial intelligence (AI) to deliver effective energy storage O&M.
Energy storage operation is a proactive job since real-time decisions must be made about the best time to charge and discharge the ESS. Optimization helps the system maximize its financial potential, while operating under federal and state incentive program parameters.
Solar operators first need to develop a strategic plan for using their ESS to save on customersâ€™ energy costs and determine if they can generate revenue from utility and energy market participation. Then they need to figure out how to execute on that strategy daily. This is where AI is critical.
For behind-the-meter (BTM) storage systems, AI can be used to forecast customer site load, solar production and other co-located generation. AI also can help determine when to use the ESS to avert costly energy spikes and how to maximize the charging and discharging value while operating under a variety of program constraints, such as compliance and interconnection rules, utility tariffs and equipment warranty operating parameters. AI is critical to market participation too. It can assist in management and optimization of concurrent revenue streams, including participation in utility demand response programs and local capacity programs.
Front-of-the-meter (FTM) systems require AI to forecast co-located solar production and energy market prices to financially optimize usage of the battery within the energy markets. There are multiple value streams to consider: forward capacity, day-ahead and real-time energy, and ancillary services, such as operating reserves and frequency regulation that each have a different price value and curve, risk and wear and tear on the ESS that must be optimized.
Much like solar projects themselves, energy storage in the United States is partially funded by the federal solar investment tax credit (ITC) program and local renewable energy incentives, such as the Self-Generation Incentive Program (SGIP) in California, Solar Massachusetts Renewable Target (SMART) and Value of Distributed Energy Resources (VDER) in New York. AI can aid in compliance, ensuring that the ESS operates under the various program rules and constraints, which becomes part of the AI co-optimization algorithm.
However, monitoring ESSs is a complex process. Network operations centers (NOCs) work 24/7 and track the ESS, its components and related ancillary equipment, such as non-export relays, load meters and data feeds from solar and other onsite generation. The NOCs also collect information about site load, weather conditions and market pricing.
With such a high volume of data continually coming in, itâ€™s impossible for human operators to process everything. Thatâ€™s where AI comes in to:
AI also supports asset optimization, which looks at the resting state of charge, cycling, depth of discharge and how these factors degrade the battery over time. By understanding operating conditions, AI can predict wear and tear on the ESS and help maximize the life of the battery system.
AI paired with solar + storage is delivering real results for solar developers. For example, consider a 2-MW/4-MWh front-of-meter project paired with a SMART-eligible PV facility â€” the combination of ITCs, SMART incentives and accelerated depreciation exceeds the total project costs (see graphic). But with ISO New England (ISO-NE) market participation, the developer can add new revenue streams, including those for forward capacity, forward reserves, real-time reserves and day-ahead energy. Those participating in ISO-NE wholesale markets can expect 50 to 75% returns on their ESS projects, with the addition of intelligent storage.
The role AI plays in the energy sector is only going to grow in importance with the transformation from an environment where utilities have centralized control to a more vibrant ecosystem with intelligent distributed resources. AI is a critical tool for optimizing O&M of energy storage and will become an even more important technology to support the evolving, and increasingly complex, interactive energy infrastructure.
Mark Triplett serves as Stemâ€™s chief operating officer, responsible for energy storage system deployments, supply chain, network operations, asset management, market participation and utility programs. He also ensures effective deployment and usage of Stemâ€™s energy storage systems and cutting-edge AI, Athena. Mark brings more than 20 years of experience developing, selling and delivering complex technology solutions in energy storage, smart grid, renewable energy, DERS, utility operations, telecommunications and enterprise software. Prior to working at Stem, Mark was the COO of Green Charge Networks, and earlier served as president of software and COO at UISOL, an Alstom-owned DRMS enterprise software application product exclusively for the utility industry.