Researchers at the MIT Laboratory for Information and Decision Systems (LIDS) have shown that using decision-making software and dynamic monitoring of weather and energy usage can greatly improve resilience to weather-related outages and efficiently integrate renewable energy sources into the grid.
The researchers note that their proposed system could have prevented or at least mitigated the widespread blackouts that occurred in Puerto Rico last week by providing analysis for rerouting power between different transmission lines and limiting the outage’s spread.
The computing platform, which the researchers call “DyMonDS (Dynamic Monitoring and Decision System),” can be used to augment existing operational and planning methods used in the power industry. The platform supports interactive information exchange and decision-making between grid operators and grid edge users (all distributed generation, storage, and software contributing to the grid). It also supports optimization of available resources and controllable grid equipment in response to changing system conditions. It also enables collaborative decision-making by a variety of utility-owned and non-utility-owned grid users, including mixed resource, user, and storage portfolios. Operational and interactive planning of high-voltage transmission and local distribution networks and microgrids is another major potential application of the platform.
This general approach is illustrated with a public dataset that contains both meteorological information and detailed information on electricity generation and distribution in Puerto Rico. The enhanced AC optimal power flow software by SmartGridz Inc. Development is used to optimize controllable devices at the system level. The software provides real-time guidance in determining how much power to send and on which transmission lines by adjusting plant dispatch and voltage-related setpoints. It also provides guidance in determining when, in extreme cases, curtailments or blackouts are required to maintain services that can be realistically deployed to as many customers as possible. The research team found that using such a system could ensure that the maximum number of essential services remain powered during a storm, as well as significantly reduce the need for new power plants by using existing resources more efficiently.
The findings are described in a paper written by MIT LIDS researchers Marija Ilić and Laurentiu Anton, along with former student Ramapati Jadivada, in the journal Foundations and Trends in Electric Energy Systems.
“We’ve demonstrated that by using this software, we can predict equipment failures during severe weather and then use that information exchange to identify the impact of the equipment failure and continue to serve a large number of customers, 50 percent of our customers, who would otherwise lose all power,” Ilich said.
“The way many power grids are currently operated is not optimal,” Anton said. “We therefore demonstrated that with this software, they can perform much better under normal conditions without errors.” They say that savings from this optimization under everyday conditions can reach tens of percent.
Regarding how utilities currently plan, Ilitch said, “The standard is often to have enough capacity and operate in real time to ensure that if a major piece of equipment, like a large generator or transmission line, fails, service to customers is not interrupted. This is called N minus 1.” This policy ensures that service can be maintained for at least 30 minutes even if a major component of the system fails. The system allows utilities to plan how much backup generating capacity they need on hand. Ilitch noted that this reserve capacity is costly because it must be maintained at all times, even under normal operating conditions when it isn’t needed.
Moreover, “there is no standard right now for what I call N minus K,” she said. If severe weather causes five pieces of equipment to fail at the same time, “there is no software to help utilities make schedule decisions in terms of keeping power up to the most customers and the most important services, like hospitals and emergency services.” They showed that they could maintain power to the majority of customers even if 50 percent of the infrastructure was down.
Their work analyzing Puerto Rico’s power situation began after the island was devastated by Hurricanes Irma and Maria. While most of the generating capacity is in the south, the heaviest loads are in San Juan in the north and Mayagüez in the west. When transmission lines are destroyed, large amounts of power must be rerouted quickly.
With the new system, “the software will find the optimum adjustments to the setpoints,” Anton said, adding that it could change the voltage, for example, to direct power to less congested lines or increase the voltage to reduce power losses.
The software can also help with long-term planning for the power grid. With many fossil fuel power plants due to be retired soon in Puerto Rico and other countries, developing a plan to replace that power without relying on greenhouse gas-emitting sources is key to meeting carbon reduction goals. And by analyzing usage patterns, the software can guide the placement of new renewable energy sources that can most efficiently deliver power where and when it’s needed.
“We want to ensure that we have the ability to deliver power as loads change if a power plant goes down or if weather impacts a component,” Anton said, “but we also want to ensure stability if we lose a critical component and ensure sustainability at every stage of the outage schedule.”
One thing they found was that “if you look at the available generating capacity, even after you shut down some fossil fuel plants, you have more than your peak load,” Ilitch said. “But it’s hard to deliver.” Strategic planning for new distribution lines can make a big difference.
“We evaluated different architectures in Puerto Rico and demonstrated that this software has the ability to guarantee an uninterrupted power supply,” said Jadivada, SmartGridz’s innovation director. This is the most important challenge facing utility companies today. They have to go through a tedious calculation process to ensure that the power grid works during any outage that may occur in the system. And it can be done much more efficiently through the software that the company developed.”
The project is a collaboration between researchers at MIT LIDS, MIT Lincoln Laboratory, and Pacific Northwest National Laboratory, and is fully supported by SmartGridz software.
