When running a power system to supply electrical energy to homes and businesses, engineers strive to satisfy three objectives:
Providing a reliable service
Minimizing the cost of providing this energy
Reducing the environmental impact of the system, in particular facilitating the use of renewable energy sources
Unfortunately, these objectives conflict: improving reliability increases costs and renewable energy sources affect reliability unless costly security measures are put in place. The main objective of our research is to develop techniques that achieve the optimal balance between these three essential goals. In particular, using advanced optimization techniques, we explore how flexibility resources (such as demand-side participation, energy storage and agile generating units) should be deployed and operated. This is the essence of what has recently been called the "smart grid." In this respect, we are also working with colleagues who are specialists in communications, sensors and control to explore how these rapidly evolving technologies can be harnessed to help meet the objectives mentioned above.
While integrating these objectives is an important part of our work, we also do research on each of these topics taken separately. For example, we have developed probabilistic techniques to assess the operational security of power systems. As part of this strand of work, we are exploring the complex mechanisms that lead to blackouts in large power systems. We also study electricity markets. The next challenge in this area is to develop mechanisms which will ensure that competitive electricity markets deliver sufficient investments to meet the evolving needs of the electricity supply system.
Current Research Projects
Power System Flexibility
Sponsor: University of Washington
Abstract: Recent years have witnessed the integration of large amounts of stochastic renewable energy sources, such as wind and solar photovoltaic. This is likely to continue and will probably be accompanied by the deployment of a significant amount of demand response. While these developments are desirable, they are also likely to increase the uncertainty on the load/generation balance. The standard answer to this problem is to say that the system needs more "flexibility" to handle this uncertainty. However, installing and deploying flexibility costs money. On the other hand, if the system is not sufficiently flexible, operators may have to resort to load shedding or the curtailment of renewable generation to maintain the stability of the system. We are therefore investigating the following questions: How do we quantify flexibility on various timescales? How much flexibility do we actually need? How much physical flexibility (i.e. from generation, storage, and demand response) is needed and how much can be accomplished using virtual flexibility (i.e. improved operating procedures and market design)? What is the optimal portfolio of flexibility resources?
Certifiable, Scalable, and Attack-resilient Submodular Control Framework for Smart Grid Stability
Sponsor: National Science Foundation
- Prof. Linda Bushnell
- Prof. Radha Poovendran
- Prof. Daniel Kirschen
- Prof. Andrew Clark (WPI)
- PhD student Philip Lee
Abstract: The smart grid is a large-scale, societal-level hybrid cyber-physical system with tight coupling between
cyber (sensing, communication, and computation) and physical (actuation) components. Ensuring
availability and reliability of power requires maintaining stability of the power grid even as increasing
demand and uncertain renewable power sources push the power system close to its operation limit. In
addition, the cyber-enabled grid has multiple entry points, leaving it highly susceptible to cyber attacks
by malicious adversaries. Currently, however, developing scalable, certifiable, bound-achieving, and
attack-resilient control methodologies for power system stability in the context of Science, Technology,
and Engineering of CPS is an open and vital research area.
We propose to research and develop a submodular framework that will provide control methodologies
that are scalable, certifiable, and attack-resilient for the following power system problems:
Submodularity is a diminishing returns property that enables development of efficient algorithms with provable optimality guarantees. Our
main insight is that the grid stability problems have inherent physical invariants that exhibit
submodularity in terms of control variables. When submodular structures are exhibited by the physical
dynamics, scalable algorithms can be developed to select control actions with certifiable stability
guarantees, thus eliminating the computationally inefficient current practice of computing control
actions and verifying stability through simulation. Since the physical dynamics are invariant from attacks
on cyber components, submodular structures remain intact even under cyber attacks. Hence, our
proposed approach is a fundamental contribution towards attack-resilient control design.
Our approach of identifying submodular structures through physical invariants is applicable not only to
power systems, but to other CPS domains including coordinating robotics and unmanned vehicles. Thus
the results of this project address a fundamental need in the science of CPS.
- voltage stability,
- small-signal stability, and
- transient stability.
Quantifying the Resilience of Power Systems to Natural Disasters
Sponsors: National Science Foundation
Abstract: Power systems are not likely to remain unscathed by natural disasters such as hurricanes, earthquakes,
ice storms or floods. Power outages lasting days or even weeks might ensue and will affect not only
the well-being and the economy of the affected communities but could also threaten their very fabric.
Recent events, such as Super-storm Sandy and hurricane Katrina, have highlighted the need to
improve the resilience of the electricity grid. Some utility companies, such as Consolidated Edison,
have embarked on massive investment programs aimed at hardening parts of their network. The design
changes that these companies are implementing are based on observations of which components failed
during past disasters. While such measures will undoubtedly be useful, they tend to focus on the
resilience of individual components but do not necessarily represent the most effective way to enhance
the resilience of the system. Large infrastructure investments may therefore not be targeted at the most
effective solutions. To overcome this problem, electric utilities and government agencies in all areas
that could be affected by a natural disaster need a rigorous method for assessing the relative value of
This proposal describes how the overall resilience of a power system could be quantified. It also
outlines a technique to assess the relative value of measures aimed at hardening various components or
facilitating the repair and restoration of the system.Quantifying the resilience of a power system turns out to require the solution of a very complex
optimization problem. New optimization algorithms will be needed to solve this problem and hence to
answer the pressing practical issue that this project addresses.
Transactive Campus Energy Systems: An R&D Testbed for Renewables Integration, Efficiency, and Grid Services
Sponsor: Washington Clean Energy Fund, US Department of Energy
Abstract: The project team consisting of Pacific Northwest National Laboratory (PNNL), the
University of Washington (UW) and Washington State University (WSU), will
connect the PNNL, UW, and WSU campuses to form a multi-campus test bed for
transaction-based energy management. The test bed will support the integration of
renewables and other regional needs, using the flexibility provided by loads, energy
storage, and smart inverters for batteries and photovoltaic (PV) solar systems, at
four physical scales: multi-campus, campus, microgrid, and building. The multi-
campus test bed forms an R&D platform for how: i) Campus resources can be aggregated and operated to balance fluctuations in
the region's renewable generation, both up and down, ii) Self-aware buildings, smart enough to transact with the grid to provide
services, result in reduced energy consumption and increased energy
efficiency opportunities, and iii) Campuses can support the grid by reducing their impact on local and
regional peak loads and wholesale power costs.
Each campus test bed will further be specialized as a platform upon which R&D will
be conducted to advance the state of knowledge in three key areas of critical interest
to the project's DOE sponsors: i) PNNL -- transactive energy management systems for campuses and buildings, ii) UW -- smart campus and building information systems that provide energy
efficiency benefits from transactive energy management systems, and iii) WSU -- operation of campus-scale microgrids to provide services to the bulk
grid and their extension to planning and operation of resilient distribution
systems and smart cities.
The technical aim of this joint activity is to streamline the interactions between
clean energy supply, efficient buildings and the smart grid to enhance the impact of
renewable generation, energy storage, and advanced energy efficiency -- while
simultaneously improving the reliability and resilience of the electric grid.
Moreover, through the proposed series of linked investments state-wide, and
upfront collaboration with resident industry, it is the intention of the project team
to create a test bed for advanced clean technology integration that differentiates
Washington State as a leader in the growing global market for energy management
products and services.
Assessment and enhancement the smart building's flexibility and responsiveness
Abstract: Buildings represent a large share (about 40%) of the total energy consumed at national level. The majority of this energy consumption takes place when the building is being occupied during office hours. As expected, these are the periods in which the power system's generation fleet is being used most heavily, not only producing large amounts of power to meet the demand, but also greenhouse gaseous emissions. Therefore by harnessing the flexibility of the pliant appliances in the building as well as energy efficiency measures, the timing and amount of power consumption, and thus on pollution from the supply sector would be greatly reduced. This research proposes tools to optimally operate and retrofit buildings to minimize their power consumption and their carbon footprint. Go to this project's web page.
Past Research Projects
Distributed Energy Resources Management System
Sponsors: Department of Energy, Alstom Grid
Abstract: Until recently, it was possible to operate distribution networks in a "build and forget" mode because the load evolved slowly and in a predictable way and because these networks did not involve any active components. A number of factors are converging to make this operating paradigm unsustainable:
- The integration of Distributed Energy Resources (DERs) such as wind generators, PV panels, and small scale Combined Heat and Power plants in the distribution network
- The large scale deployment of electric cars, which will not only increase the overall demand for energy but also change the profile and characteristics of the loads in the distribution network
- The implementation of Demand Response (DR) schemes that will link the load to price signals that may become increasingly local.
In this project we are exploring how we can leverage the various sources of data that are becoming available (e.g. smart meters, distribution automation) to make the operation of the distribution networks more efficient and more reliable and to facilitate the integration of distributed energy resources. Go to this project's web page.
Energy Positioning: Control and Economics
Sponsor: Department of Energy, ARPA-E GENI program
Collaborators: Prof. Ian Hiskens, University of Michigan
- Prof. Daniel Kirschen
- Dr. Hrvoje Pandzic
- PhD student Ting Qiu
- PhD student Yishen Wang
Abstract: The University of Washington and the University of Michigan are developing an integrated system to match well-positioned energy storage facilities with precise control technologies so the electric grid can more easily include energy from renewable power sources like wind and solar. Because renewable energy sources provide intermittent power, it is difficult for the grid to efficiently allocate those resources without developing solutions to store their energy for later use. The two universities are working with utilities, regulators, and the private sector to position renewable energy storage facilities in locations that optimize their ability to provide and transmit electricity where and when it is needed most. Expanding the network of transmission lines is prohibitively expensive, so combining well-placed storage facilities with robust control systems to efficiently route their power will save consumers money and enable the widespread use of safe, renewable sources of power. Go to this project's web page.
For a project description, click here.
For the project fact sheet, click here.
Using Distribution-Level Energy Assets to Help Optimize Regional Transmission
- Prof. Daniel Kirschen (PI)
- Dr. Ricardo Fernandez-Blanco
- M.Sc. student Kelly Kozdras
In the proposed project, Snohomish County PUD (the "District") will give BPA incentive-based access to District-owned and operated energy storage (ES) and demand response (DR) assets. The technologies deployed will let BPA send specific requests to District control software to supply or absorb energy (e.g., absorb 2MW for 2 hours, starting at HE03). Resulting energy transfers can be used to support BPA operations by, for example, reducing network congestion, mitigating energy imbalance or improving wind integration.
District assets made available in this project include 4 MW and 8 MWh of energy storage funded jointly by the District and Washington State Department of Commerce (DoC), under the Washington Clean Energy Fund.
An initial project task, with BPA input, will define the communication protocol for processing energy transfer requests. The protocol will be standards-based and may, optionally, use transactive control signal technology from the PNW Smart Grid Demonstration Project.
Software deployed for BPA will also include a transmission-oriented optimizer, based on Energy Positioning technology developed by the University of Washington (UW) and the University of Michigan (UM) under an ARPA-E grant. This technology will help BPA determine optimal requests for distributed energy services.
Once fielded and tested in this project, the proposed technologies can be further deployed as standard mechanisms, enabling BPA requests for energy services from any distribution-connected ES and DR assets, throughout the BPA grid.
Development of Tools for Analyzing the Profitability of Energy Storage in Competitive Electricity Markets
Sponsor: Sandia National Laboratory
Abstract: The objective of this project is to investigate techniques to assess the profitability of
deploying distributed energy storage systems in a power system operating in a
competitive market environment. The project will start by reviewing the rules that
have been implemented or proposed for the integration of such devices. It will then
develop optimization models of increasing accuracy and complexity to determine
the optimal location and size of these devices.
Development of aggregated dynamic models for active distribution systems using heuristic optimization techniques
Abstract: The increasing connection of distributed energy resources (DERs) to
distribution systems transforms the latter from a passive to an active
part of the power system. In bulk power system stability studies, the
dynamic behavior of these active distribution systems (ADSs) has to
be accurately accounted for in order to properly reflect their influence
on the overall dynamic performance of the system. However,
limitations in computational performance as well as in the availability
of distribution system data require a certain degree of simplification.
The objective of this project is to refine an existing aggregation
methodology that maintains the minimum level of detail necessary to
accurately model ADSs in bulk system stability studies. The
contribution of the work is expected to be the application of a selected
heuristic optimization techniques in order to determine the "equivalent
impedance" that represents the lines in the detailed North American
network of a specific voltage level as well as the creation and
extension of dynamic models. Model validation techniques will be used
to determine how accurately the aggregated ADS model represents the
response of the detailed ADS to an external network fault, e.g.
transmission system fault.
Architectural and Algorithmic Solutions for Large-Scale PEV Integration into Power Grids
Sponsors: National Science Foundation
Abstract: Electric Vehicles (EVs) are now a reality; and it is expected that in the near future large volumes of these devices will be integrated to the existing power grid. If allowed to charge in an uncontrolled manner, these devices will charge their batteries when connected to the grid circuits, increasing the already high peak-demands that are served by expensive generation sources. Furthermore these additional demands could be translated into potential over-loadings and the need to invest into wires and power generation assets in order to be successfully accommodated. On the other hand, technical and economic benefit would be attained if the charging of these devices takes place when the system is lightly loaded, and being served by cheap generation.
The research conducted under this grant proposes algorithmic solutions to accommodate the EVs demand in an optimal manner, not only to minimize costs, but also to exploit their ability to charge and discharge on command, to provide services to the power system. At the same time, we also seek to exploit this inherent flexibility to minimize not only the electricity cost for the EV owners, but also the degradation of their batteries. Go to this project's web page.