The switches and converters needed to plug new clean energy technologies into the grid are going to be needed soon. The cheaper and smaller they are, the easier it will be to cope with the wide range of solutions coming down the pipe. NREL and its partners are building a megawatt-scale prototype converter that could fit the bill. They will be 1/5th the size and 1/10th the weight of existing alternatives: small enough to place almost anywhere. They should improve grid stability, frequency support, power re-start, fault catching and protection. The Silicon Carbide switches enable “back-to-back” conversion — AC to AC power — at distribution voltages, avoiding the need for expensive and bulky transformers. Standardisation (starting up procedures, syncing to the grid, etc.) is being addressed, too, to make sure this solution is ready to be deployed in the near term. Already, the Distributed Energy Resources (DERs) like rooftop solar, battery storage, EVs, local waste heat and biomass, are queueing up to join the grid.
As we trend toward more renewables and distributed energy resources (DERs), the design of the electric distribution system itself imposes physical limitations. These system constraints could lead to issues like overloaded power lines and faults that propagate freely.
But what if we could restructure the underlying system to support greater renewable integration and system resilience? To that end, a National Renewable Energy Laboratory (NREL)-led project is working on a new type of grid device enabled by silicon carbide (SiC) switches and other medium voltage (MV) power electronics that could segment sections of the grid, providing advanced control for flexibility and resilience for our power systems.
Silicon Carbide switches that segment the grid
The project team is first designing a megawatt-scale prototype converter that provides native “back-to-back” conversion—AC to AC power—at distribution voltages (i.e., not requiring transformers to step down voltage to levels typically used in electronic power conversion). By using MV SiC-based power modules, the converters could be 1/5th the size and 1/10th the weight of alternate equivalent systems, which are trailer-sized and include heavy transformers. Then the team will connect the power converter into NREL’s MV testbed to validate new grid control approaches that the prototype enables.
The project is named “Grid Application Development, Testbed, and Analysis for MV SiC (GADTAMS)” and is funded by the Department of Energy’s Advanced Manufacturing Office.
“With back-to-back converters between feeders, we can go one step higher in providing resilience across the distribution system,” said Akanksha Singh, a project lead at NREL.
Injecting DER power into the grid, easily
“This technology wasn’t necessary before because we didn’t have so many distributed energy resources on the system, but now we have feeders that are becoming saturated with PV; apart from storage, these feeders don’t have anywhere to inject that excess power,” Singh said. “A new approach to grid interconnection could enable advanced forms of power sharing and provide much-enhanced grid resilience.”
A future grid that features such converters would have the capability to control the flow of power between sections of the grid, shunting excess load or DER-based generation to feeder sections or adjacent circuits as needed, adding new versatility to power distribution. Networked microgrids could protect against the propagation of faults from one microgrid to the next while still allowing controlled power dispatch between the two systems and the macrogrid as well.
During outage recovery, microgrids could be formed that then stabilise neighbouring microgrid systems, as envisioned in NREL’s autonomous energy systems research. In general, the two sides of the converter do not need to be synchronised in frequency or even exact voltage level at all—a major shift from the modern power system. But prior to proving any of these applications, NREL and others will first need to build the necessary controls.
“We are developing very novel controls for upcoming grid architectures,” Singh said. “We have local controls on inverters, and we have hierarchical controls that coordinate between grid partitions. With regard to grid support, these controls can do it all: dynamic stability, frequency support, black start, fault ride-through and protection.”
Unlike anything currently available, the NREL testbed provides an environment to validate medium-voltage grid solutions with real power hardware-in-the-loop and real-time grid simulation. For this project, NREL and partners are interested in the full range of use cases for back-to-back SiC converters and have teamed with utility Southern California Edison to inform on utility applications, as well as industry partners General Atomics and Eaton to seek out a commercial path for the technology.
The SiC converter is being built in two halves by project partners Ohio State University and Florida State University. The three-phase converter prototype will be rated for 330 kW and will implement a full thermal and electrical design appropriate for utility use. Traditionally, the same AC-to-AC conversion process requires stepping-down the voltage to low-voltage levels where conventional power electronics can be used, which results in heavy and expensive transformer equipment. The MV SiC option takes advantage of the superior voltage ratings of devices to minimise weight, cost, and size, which makes the technology far more practical and economical for system-wide deployment.
Still, the converter technology is only one aspect of fulfilling flexible interconnections. This framework currently lacks the standardisation that exists for so many other recent grid innovations. At NREL, the project team hopes to collect baseline operational data to jumpstart the conversation around how to integrate MV converters in future grids.
“This is a new application that doesn’t exist anywhere yet. We need standards that apply to how the converters can integrate with regular system operation, like starting up, syncing to the grid, etc.,” Singh said. “We are using IEEE Standards 1547 and 2030.8 as a base, interpreting their rules to implement new controls on MV systems. We are trying to merge the two to understand what will apply to this new approach.”
An entirely new grid architecture and operational flexibility could seem far-out for now, but NREL and partners are showing that these options are viable in the near-term and that NREL has the capability to prepare these solutions for real systems.