OUR POWER SYSTEM IS CHANGING. Historically, we’ve operated the supply side of the power system deterministically. We estimated the load and then dialed in the required amount of generation. The U.S. Department of Energy said that 32 percent of added generation in 2011 came from wind energy. The supply side now includes large quantities of renewable resources that are intermittent in nature. They are somewhat but not completely predictable. The supply side has become stochastic.
The load side has changed, too. It is still stochastic, but it is also becoming intelligent. The number of communicating and sometimes intelligent devices capable of changing their energy consumption patterns is growing. New loads in the form of electric vehicles have consumption patterns that must be managed to avoid peak load growth. While still stochastic, the load side is now capable of responding to information and adjusting consumption behavior in useful ways.
Finally, the grid itself is changing with the introduction of distributed energy resources. These resources require bidirectional power flows and in the long run will be intelligent and communicating.
Transformation is clearly upon us, calling for profoundly different approaches. One such approach, referred to as transactive control, will provide the opportunity to leverage smart devices to operate the grid with greater precision and efficiency while optimizing generation and end use to unprecedented levels.
Through its role leading the largest regional smart grid demonstration in the nation today, the Battelle Pacific Northwest Division and its partners are developing and deploying a transactive control and coordination technology as a key component of the Pacific Northwest Smart Grid Demonstration Project funded by the American Recovery and Reinvestment Act. The team is building on previous, pioneering work by the Department of Energy’s Pacific Northwest National Laboratory, which first tested the transactive control concept in the GridWise Olympic Peninsula Demonstration – the “Oly-Pen” project. In Oly-Pen, PNNL for the first time successfully demonstrated the use of real-time pricing to retail customers as a tool for managing grid constraints. Working with the Bonneville Power Administration, the follow-on demonstration project was conceived to scale up the technology and demonstrate regional interoperability, with the notion of addressing several of the changes outlined above.
This project has a total budget of $178 million with $89 million of federal funds and a match of $89 million from the project participants, including BPA and participants from 11 utilities located in the states of Washington, Oregon, Idaho, Montana and Wyoming. The technology development effort includes five providers, including IBM Research, IBM-Netezza, Alstom Grid, 3TIER and QualityLogic.
The transactive control and coordination technology being developed and deployed on the demonstration project directly responds to today’s new grid boundary conditions. Transactive control is a distributed system using a transactive incentive signal and a transactive feedback signal that provides global information to local decision making at any point in the power system. The system essentially opens up a two-way line of communication between generation and load that previously didn’t exist. This communication follows the flow of electricity, enabling local decision-making at any point where the flow of electricity may be modified. Through this approach, grid conditions throughout the system may be taken into account to coordinate the behavior of generation resources, loads and distributed energy resources.
If we could visualize the conversation that load and generation were having using this system, we would see the supply side originating an incentive signal, communicating the expected future cost of delivering power to specific locations and taking into account system conditions along the path from supply to load. Once received, the demand side reacts to the price incentive and replies by sending a feedback signal, communicating the energy consumption plans in the future. In a simplified way, a wind turbine could say to the system, “At 3:00 a.m., I have a lot of energy. It will be a bargain.” Your car, seeing lower prices, would reply, “I will plan to charge then, because power will be cheap.”
Through the use of a basic five-minute forecast interval, the transactive control approach also allows us to reflect the intermittent fluctuations of wind energy due to natural variations in wind.
As the example hints, transactive control can help support the integration of renewable energy and the engagement of new loads without straining the grid. Let’s look at these two characteristics in detail.
Over the next couple of years, the amount of wind generation in the Pacific Northwest is projected to double. Future regional wind capacity of about 12,000 megawatts will equal the amount of hydropower generated by the federal dams along the Columbia and Snake rivers. In a way similar to the mock dialogue above, transactive control engaging the responsive assets helps to balance the intermittent nature of wind energy and allows optimal operation of generation resources such as the hydro system. Transactive control is also capable of incentivizing the consumption of renewable energy when it is most abundant. This feature will help reduce the amount of “spilled” wind power – excess wind-generated electricity that has to be curtailed.
Our system also allows responsive assets, such as electric vehicles, to engage without straining the grid. As an example, we could think of a pole-top transformer serving a number of electric vehicles. If the transformer is overloaded, its service life will be reduced. To avert that, the transformer increases the forecast value of the incentive signal based on its current state, weather conditions and feedback from the vehicle chargers about future charging plans. The forecast increase in incentive signal value will cause vehicles to modify their plans, thus decreasing the strain on the transformer. When the transformer changes the future value of the incentive signal based on the charging plans, the plans may change as a result. Because the system is a form of closed-loop control, convergence of the give and take between incentive signal changes and charging plans may be ensured through careful algorithm design.
PNWSGD is testing the technology on a limited scale with about 12,000 smart grid-responsive assets, which include water-heater load controllers, solar panels, battery storage units, and backup generators. This fall, our transactive control system will go live and allow the installed assets to start responding to electric power system conditions. During the project’s two-year data collection and analysis period, we will be working to confirm that the transactive control technique offers the operational benefits described here and to estimate the benefits of regional scale-up. In addition, the data we gather over the next two years will help us improve our knowledge of how the nontransactive smart grid technology deployed by the 11 utilities performs, identify what types of technologies will best suit our region and evaluate how the expected benefits could be spread through deployment beyond our current footprint. Starting this winter, status updates will be available on our website, www.pnwsmartgrid.org.
As the boundary conditions have changed, power system operators are challenged to ensure that we continue to receive reliable electricity regardless of whether it is green, used by new loads, or flowing from distributed energy resources at points throughout the system. Transactive control is our answer to this challenge. Adapting to the changing boundaries, we are taking a significant step that will move the nation closer to a more efficient, sustainable and resilient power system.