Grid flexibility is a critical piece of today’s energy landscape. It is becoming clear that in markets with a high percentage of renewable power, making sure that clean energy is available when and where it is needed, often referred to as ‘resource adequacy’ (RA), is becoming a central issue in evolving utility markets. Nowhere is this more evident than in California, with the recent decision to give the state’s two largest utilities an expanded role in the procurement plan for grid reliability and it’s affect on how renewable resources are able to participate in the state’s RA requirements. Energy storage has the opportunity to play a significant role in meeting the RA needs of grid operators, whether they have mandated RA requirements like California or there is a market-based approach to capacity like ERCOT.
The idea behind RA is simple enough, grid operators want to makesure that there is enough power capacity to meet demand despite potential changes to that demand related to weather, power plant maintenance or grid emergencies. RA is a form of insurance that can de-risk the operation of the grid by essentially paying or otherwise incentivizing power producers to have additional generation capacity on standby. California’s RA program was developed in response to the crisis of 2000-01 where rolling blackouts and pricing spikes led to customer dissatisfaction and chaotic markets. The state does not run a capacity market but instead relies on retailers to provide RA. In other markets, like PJM or ISO-NE, reserve margin procurement is driven more by modelling demand curves and generators bid into a capacity market auction. ERCOT has a market-based approach to RA whereby pricing is expected to drive power producers to make capacity available to the grid
when needed. All of these markets have one thing in common and that is the impact that clean energy goals are having on how reserve capacity is viewed and what role energy storage will play in the wholesale and retail markets going forward.
Energy storage has traditionally meant pumped hydro storage, which currently accounts for over 90% of grid energy storage in the United States, but the increased use of battery energy storage systems (BESS) due to reduced cost and increased flexibility in how the BESS can participate in the market has led to a rethinking of the role of battery storage in capacity and ancillary services markets. BESS participation in these markets can vary across ISO’s and RTO’s, but with FERC Order 841, the rules now necessitate fair compensation for BESS participation in wholesale markets and may open up opportunities in grid markets like ERCOT’s fast-responding
regulation service or PJM’s Reg-D product (both are dispatched frequency regulation services). With this increased diversity of participation, meant to take advantage of the flexibility allowed by BESS installations, has come a need to monitor, model and predict how energy storage resources are performing so that they can optimally provide both RA and ancillary services like frequency regulation in increasingly diverse markets. Owner/operators of energy storage will compete in markets where the volume of storage participation is growing and efficient operation of their assets will determine how competitive they are able to be. Large utilities and operators will likely have in-house resources to work with the data streams available from their installed storage, but the smaller utilities and owner/ operators will benefit from companies like NexESS Analytics and the services they provide as an arbiter of operational data on performance and reliability. “NexESS was founded to empower BESS owners and operatorsby converting the massive amount of data generated by these assets into useful business intelligence. This empowerment allows operators to maximize the utility of their investment while ensuring warranty and off-take compliance through the use of machine learning and advanced analytics” according to NexESS co-founder and CEO, Cory Schaeffer.
Grid-side data (load, power quality, frequency, etc.) combined with battery operational characteristics (SOC, round trip efficiency, etc.) contribute to anorchestrated set of analytics designed to optimize asset participation in the varied markets. Storage owner/ operators can use these insights to help manage decisions on how hard to push their assets or whether to participate in a particular service. For example, frequency regulation may require a fast ramp rate that could put an increased stress on the batteries due to the thermal characteristics of different chemistries. By modelling the actual data on conditions encountered by the BESS, the resulting battery lifetime may indicate that costs above and beyond the warranty may be necessary in order to keep up with the service requirements. Changing BESS operational limits may be able to mitigate some of these costs and implementing an operational
strategy that is monitored in real time for compliance can allow participation in varied markets with a confidence based on data.
Operation and maintenance (O&M) costs can also be reduced with curated data streams feeding machine learning (ML) algorithms predicting maintenance events. Downtime is costly for any asset and predicting when component failure will occur can allow for scheduling repairs at convenient and less expensive times. Statistical analysis and ML work better with high quality data streams (as the saying goes, garbage in/garbage out) but real-time operational data can be noisy and filled with errors. The ability to clean or curate this data stream for these orchestrated ML routines is critical to successful O&M strategies. Performance and reliability gains increase the marketability and ensure that investment in storage assets goes according to plan.
Energy storage will increasingly become a key component in the operation of the grid as the drive for cleaner energy will see more renewable generation deployed as decarbonization goals are met. Grids with high renewable penetration will benefit from the flexibility that storage can bring to the market and the participation of assets like BESS and other storage solutions will require an architected approach to using the data generated from these assets to provide the best performance and bankable reliability in the capacity and ancillary markets.