The Architecture of Green Energy Systems: The Underlying Problem and Its Challenges

Wholesale electricity markets with a significant share of intermittent wind and solar generation resources face reliability challenges relative to markets composed of dispatchable generation capacity. Reliable operation of wholesale markets with significant intermittent renewables requires a closer match between the market mechanism that sets prices and generation unit output levels and the physics of power systems operation. Simplified wholesale market models that ignore the configuration of the transmission network and operating constraints on generation units that must be respected in real-time have become increasingly costly as the share of intermittent generation in both the transmission and distribution network increases.

This talk will present several empirical analyses of the consequences of failing to adapt wholesale and retail electricity market designs to the growing amount of intermittent wind and solar generation resources. Topics considered include the design of long-term resource adequacy mechanisms, distributed versus grid-scale renewables investments, transmission and distribution network planning and expansion policies, and the design of retail pricing plans to avoid costly bypass of the grid supplied electricity. All these empirical analyses emphasize the importance of wholesale and retail market designs that account for both the physics of power system operation and the economic incentives for market participant behavior created by the market design and constraints implied by power systems operation.

Rapidly growing renewable generations and peak loads pose a serious threat to the security and reliability of modern power and energy systems. A critical question is “how to accommodate a high penetration of renewable generation and deep electrification?”. This talk studies coordinating large-scale distributed energy resources and focuses on two key challenges, including unknown information and the scalability issue. Two examples will be presented to develop distributed learning-assisted decision-making methods to address these challenges. One example is zeroth-order optimization with application to model-free optimal voltage control for handling unknown physical system models, and the other is online learning and human-in-the-loop decision-making for residential demand response to deal with unknown human user behaviors.

Amazon’s sustainability journey is fueled by science and innovation. In this talk, I’ll share our progress, challenges, and lessons learned as we tackle complex sustainability issues. From leveraging cutting-edge technologies to investing and collaborating with experts, discover how we’re driving meaningful outcomes and unlocking solutions to some of the world’s most pressing environmental and social challenges.

The management of biodiversity and climate change is delicate, due to conflicts with economic development, to inertia and delayed impacts of decisions on long horizon times, and to pervasive uncertainties. This makes it a relevant area for decision under uncertainty methods. First, I will provide an overview of the Intergovernmental Panel on Climate Change (IPCC) and Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) reports. Second, I will present theory and examples related to sustainability and resilience. Third, I will discuss opportunities to introduce risk measures as a way to handle riskand to design robust policies. Keyword list: sustainability, resilience, viability, stochastic optimization, risk

Energy systems massively powered by renewable energy sources raise fundamentally novel challenges for planning and operating electrical power systems. This talk reviews how the fusion of AI and Optimization provides novel pathways to address these challenges. It will review how trustworthy AI technologies can improve forecasting and uncertainty quantification, provide order of magnitude improvements in efficiency for market operations, and enable real-time risk assessment, all which are of great interest to system operators. The talk will also report some open challenges in resource adequacy and expansion planning.

In two-stage stochastic programs, a first-stage decision precedes a realization of a random outcome, which is followed by solving a recourse linear program to determine the second-stage decision. This structure is powerful, but it can be limiting. I will consider the use of (stochastic) simulation to replace the random realization and recourse linear program. The tractability of the resulting problems partly relies on the path-wise properties of the simulation model and the ability to obtain suitable gradient estimates from them. I’ll use examples to explore these issues.

Decarbonization policies will require terawatts of new distributed generation, e.g., solar PV and wind, to be installed across high-voltage transmission and low-voltage distribution (T&D) systems. At times, these weather-driven, variable sources will traverse electricity networks to fully supply fleets of billions of electrified loads and distributed energy resources (DERs), including electric heat-pumps, electric vehicle chargers, and various form of energy storage. At other times, the loads and DERs will need to adapt to the available supply or regional grid bottlenecks. This adaptation will require intelligent coordination of electric DERs to defer the right amount of demand, at the right times, and in the right locations in the grid. To facilitate such intelligent electrification while guaranteeing grid reliability at all times will require combining data available to (T&D) grid operators and independent DER operators with scalable control architectures. This talk will present different architectures within the context of distributed control and optimization that facilitate integration of DER flexibility across T&D interfaces. Furthermore, since many controllable loads will reside in low-voltage networks (i.e., homes and businesses), we will discuss appropriate architectures for aggregating the capability of a fleet of DERs into responsive virtual power plants (VPPs) that are also cognizant of grid bottlenecks.

 To combat global warming, energy systems need to go through broad and deep decarbonization in the coming decades. Electric power system plays a crucial role in leading the global energy system transition. In this talk, we will address some foundational issues in the planning and operation of electric power systems and showcase how operations research (OR) and optimization can make significant impacts by providing critical decision tools and insights. We will also discuss challenges and progresses in sustainable transitions of broader infrastructure systems, such as electrification of public transportation and capacity planning of battery recycling facilities. Finally, we will reflect on our past efforts in bridging OR to the broader scientific communities and share our outlook on future directions.

• What a sub-field of security might look like
• What the goals of that field would look like
• What new challenges it presents
• What technologies will likely be useful
• How game theory will play a role