Solar PV: The cause of, and solution to, instability at the grid edge

Utilities and ISOs across the country are in the hunt for flexible assets that can compensate for intermittent resources like solar photovoltaics (PV). Not only is the generation capacity of these PV resources variable, when interconnected at the grid edge they can give rise to distribution challenges such as over-voltage and reverse power flow and even drive negative prices in wholesale markets.

Surprisingly, it turns out that the best-performing asset to address these challenges is… solar PV itself.

Recent studies conducted by NREL and their research partners have explored how well smart inverters can supply grid support services typically provided by conventional generators and dedicated distribution equipment. These services directly address the distribution and market challenges above. A 2016 collaboration with Hawaiian Electric Company tested the ability of residential smart inverters to mitigate voltage rise through fixed, absorbing power factor and volt/watt control. Not only was the testing successful, it was concluded that “a ‘critical mass’ of grid supportive inverters is needed to effectively mitigate high voltages,” implying the approach not only works but also improves at scale.

In a subsequent 2017 study jointly published with CAISO, NREL proved the capability of a 300 MW FirstSolar PV plant to perform multiple grid support services, including primary and secondary frequency response and voltage regulation. Primary response is important because it helps a balancing authority meet its Frequency Response Obligation imposed by NERC BAL-003, which requires it to respond to and stabilize frequency excursions within twenty seconds of their start. Secondary frequency response, a lucrative service known as frequency regulation (FR) in ancillary service markets, was evaluated with respect to an actual, historical automatic generation control (AGC) signal dispatched by CAISO. The plant was able to provide 30 MW of FR capacity by dynamically measuring its real power capacity and setting its operating point to be 30 MW below that, plus the AGC signal. The accuracy of the plant in matching the 4-second signal, illustrated in Fig 1 below, was more than 24 percentage points better than the typical performance of a fast gas turbine generator in CAISO’s regulation-up market. Early morning performance was even stronger.

To demonstrate its ability to provide voltage regulation, the plant was tested to CAISO’s aggressive expectations for asynchronous generators to absorb or deliver a third as much reactive power as real power, at all real output levels. Not only was this standard achieved, the plant showed it could operate like a synchronous condenser, using its full MVA capacity for reactive power — a capability requiring a costly retrofit to a synchronous generator.

While the FirstSolar plant happened to be tied to CAISO’s bulk power system, a distribution-tied plant would perform identically, providing the same grid services at the same level of performance.

The same may even be achievable for aggregated residential solar as well, given advances in smart inverter telemetry and availability of local, revenue-grade production metering.

Taken together, these studies show that, contrary to conventional wisdom, grid-edge PV is well-positioned to provide the grid services necessary to maintain stability in a high-DER penetration world. In fact, their siting at the grid edge makes DERs even better positioned to provide such services than conventional resources in the bulk power system because they are able to react to power and voltage incidents closer to where they occur on the distribution system. This is particularly true for voltage support and power factor correction because reactive power does not travel far on the transmission system, so proximity is key to effectively providing these services.

It is worthwhile asking whether curtailing PV in order to provide regulation capacity is actually helpful. Can we take credit for reintroducing capacity that we just took off the table? The answer, significantly, is yes. The flexible capacity demonstrated by PV in these studies is more valuable than the inflexible capacity provided by slow-ramping and non-dispatchable generators because it can respond to system needs in real-time, rather than requiring hour or even day-ahead commitment. Every megawatt that a DER can convert to regulation capacity creates optionality for the distribution and/or transmission operator, providing value and increasing the responsiveness of the generation fleet. Flexible capacity not only contributes to voltage and frequency stability, it can address the negative prices seen in CAISO’s real-time market, which may be on the horizon for other markets with increasing PV penetration.

The real question is not whether flexible PV curtailment is valuable, but how we can align technical standards, wholesale and retail market reform, and financing of DERs to make this vision of grid-stabilizing DERs a reality. Today’s markets do not incentivize PV investors to flexibly curtail: Retail customers cannot participate in wholesale markets, project developers cannot lock in a predictable return on capital as they can through capacity markets and power purchase agreements, and both stand to lose energy revenue by curtailing. Without holistic reform, investors will continue to use PV simply for energy, failing to leverage its potential to provide higher value grid services.

While such industry-wide alignment is years off, a way forward can be seen from the widely successful Bring Your Own Device demand response program model used by utilities today. By enrolling flexible resources in Bring Your Own Device programs, utilities are effectively purchasing flexible load capacity out of market with an annual incentive. The same could be achieved on the generation side with the right hardware integrations and customer incentives.

As the leading provider of Bring Your Own Device services to utilities, EnergyHub has led the way in working with utilities to more effectively utilize thermostats and other DERs to provide demand response. Now EnergyHub is working with smart inverter partners to broaden these programs to include behind-the-meter smart inverters in solar and energy storage systems, leveraging new communication capabilities to bring inverter fleet control into the control room. Combining flexible PV generation capacity with flexible load capacity makes advanced grid services possible, including grid monitoring, voltage regulation, frequency response, and targeted congestion relief. With easy-to-understand incentives and device control that does not impact customer comfort, these programs are a win-win for the customer and utility.