10:00 - 11:00 ET, April 24th

1. Cong Chen (PhD candidate, Electrical and Computer Engineering, Cornell University)

   • Talk title: Battery Storage SoC-Dependent Bids in Multi-Interval Dispatch: Convexification and Energy-Reserve Co-Optimization

   • Abstract: Battery energy storage systems (BESS) can “buy-low-sell-high” in the energy market and quickly respond to power imbalances in the regulation reserve service. Unlike traditional generators, BESS has State-of-Charge (SoC) dependent costs including battery lifetime degradation cost and opportunity cost. Therefore, new SoC-dependent bidding formats are proposed for BESS in the bid-based electricity market. In this talk, we will show that SoC-dependent bid causes nonconvex multi-interval dispatch, limiting its implementation in real-time dispatch and market clearing. We will present a simple restriction on SoC-dependent bid to remove the nonconvexity, making the multi-interval energy-reserve co-optimization a standard convex piece-wise linear program. Finally, under reasonable assumptions, we will show that SoC-dependent bids yield higher profits for BESS than those from SoC-independent bids.

   • Speaker contact: Cong's website 

2. Wenqi Cui (PhD candidate, Electrical and Computer Engineering, University of Washington)

   • Talk title: Leveraging Predictions in Power System Frequency Control: an Adaptive Approach

   • Abstract: This talk will describe an adaptive approach for frequency control in power systems with significant time-varying net load. We leverage the advances in short-term load forecasting, where the net load in the system can be accurately predicted using weather and other features. We integrate these predictions into the design of adaptive controllers, which can be seamlessly combined with most existing controllers including conventional droop control and emerging neural network-based controllers. We prove that the overall control architecture ensures input-to-state stability. 

   • Speaker contact: Wenqi's website 

10:00 - 11:00 ET, May 29th

1. Liudong Chen (PhD student, earth and environmental engineering, Columbia University)

2. • Talk title: Prudent Price-Responsive Demands

   • Abstract: We investigate a flexible demand with a risk-neutral cost-saving objective in response to volatile electricity prices. We introduce the concept of prudent demand, which states that future price uncertainties will affect immediate consumption patterns, despite the price expectations remaining unchanged. We develop a theoretical framework and prove that demand exhibits prudence when the third-order derivative of its utility cost function is positive, and show a prudent demand demonstrates risk-averse behaviors despite the objective being risk-neutral. Our analysis further reveals that, for a prudent demand, predictions of future price skewness significantly impact immediate energy consumption. Prudent demands exhibit skewness aversion, with increased price skewness elevating the cost associated with prudence. We validate our theoretical findings through numerical simulations and conclude their implications for demand response modeling and the future design of incentive-based demand response mechanisms.

   • Speaker contact: lc3671@columbia.edu

2. Seyedsoroush Karimi Madahi (PhD Student, Information Technology, Ghent University)

2. • Talk title: Learning for Energy Arbitrage in Imbalance Settlement Mechanism - Simulation to Real-world Deployment

   • Abstract: A continuous rise in the penetration of renewable energy sources, along with the use of the single imbalance pricing, provides a new opportunity for balance responsible parties to reduce their cost through energy arbitrage in the imbalance settlement mechanism. This talk outlines our reinforcement learning-based control framework for battery energy storage systems (BESS), to perform energy arbitrage in the imbalance settlement mechanism. Our proposed framework takes a risk-sensitive perspective, while constraining the daily number of cycles for BESS. Moreover, we introduced a new policy correction step using a knowledge distillation process to ensure the correctness and safety of the final policy for real-world deployment. We will present the performance of the proposed framework on the Belgian imbalance price, in both simulation and real-world implementation.

   • Speaker contact: seyedsoroush.karimimadahi@ugent.be

10:00 - 11:00 ET, June 26th (Zoom link)

1. Attila Balint (PhD Student, electrical engineering, KU Leuven)

   • Talk title: EbFoBench: A community-driven energy forecast benchmark

   • Abstract: This talk outlines the design principles underpinning EnFoBench and details its evaluation process, which focuses on the entire forecasting pipeline rather than just the model itself. We also present the list of available metrics, allowing for the evaluation of model performance, computational efficiency, and relative performance compared to baseline models in greater detail than in previous works. Finally, this talk also present results from the first round of evaluated models, sourced from our previous work, popular forecasting frameworks, previous competitions and recent foundational models.

   • Speaker contact: attila.balint@kuleuven.be

2. Zhirui Liang (PhD Student, electrical engineering, Johns Hopkins University)

   • Talk title: Learning with Adaptive Conservativeness for Distributionally Robust Optimization: Incentive Design for Voltage Regulation

   • Abstract: Information asymmetry between the Distribution System Operator (DSO) and Distributed Energy Resource Aggregators (DERAs) obstructs designing effective incentives for voltage regulation. To capture this effect, we employ a Stackelberg game-theoretic framework, where the DSO seeks to overcome the information asymmetry and refine its incentive strategies by learning from DERA behavior over multiple iterations. We introduce a model-based online learning algorithm for the DSO, aimed at inferring the relationship between incentives and DERA responses. Given the uncertain nature of these responses, we also propose a distributionally robust incentive design model to control the probability of voltage regulation failure and then reformulate it into a convex problem. This model allows the DSO to periodically revise distribution assumptions on uncertain parameters in the decision model of the DERA. Finally, we present a gradient-based method that permits the DSO to adaptively modify its conservativeness level, measured by the size of a Wasserstein metric-based ambiguity set, according to historical voltage regulation performance. The effectiveness of our proposed method is demonstrated through numerical experiments.

   • Speaker contact: zliang31@jhu.edu  

10:00 - 11:00 ET, July 31st

1. Yufan Zhang (Postdoc., electrical and computer engineering, UC San Diego)

2. • Talk title: Improving Sequential Market Clearing via Value-oriented Renewable Energy Forecasting

   • Abstract: Large penetration of renewable energy sources (RESs) brings huge uncertainty into the electricity markets. While existing deterministic market clearing fails to accommodate the uncertainty, the recently proposed stochastic market clearing struggles to achieve desirable market properties. In this work, we propose a value-oriented forecasting approach, which tactically determines the RESs generation that enters the day-ahead market. With such a forecast, the existing deterministic market clearing framework can be maintained, and the day-ahead and real-time overall operation cost is reduced. At the training phase, the forecast model parameters are estimated to minimize expected day-ahead and real-time overall operation costs, instead of minimizing forecast errors in a statistical sense. Theoretically, we derive the exact form of the loss function for training the forecast model that aligns with such a goal. For market clearing modeled by linear programs, this loss function is a piecewise linear function. Additionally, we derive the analytical gradient of the loss function with respect to the forecast, which inspires an efficient training strategy. A numerical study shows our forecasts can bring significant benefits of the overall cost reduction to deterministic market clearing, compared to quality-oriented forecasting approach.

   • Speaker contact: Yufan's website

2. Zuguang Gao (Assistant Professor, School of Business, UC Irvine)

2. • Talk title: Aggregating Distributed Energy Resources: Efficiency and Market Power

   • Abstract: The rapid expansion of distributed energy resources (DERs) is one of the most significant changes to electricity systems around the world. Examples of DERs include solar panels, electric vehicles/storage, thermal storage, combined heat and power plants, etc. Due to the small supply capacities of these DERs, it is impractical for them to participate directly in the wholesale electricity market. In this talk, we discuss the question of how to integrate these DER supplies into the electricity market, with the objective of achieving full market efficiency. Specifically, we study three aggregation models, where there is an aggregator who procures electricity from DERs, and sells them in the wholesale market. In the first aggregation model, a profit-maximizing aggregator announces a differential two-part pricing policy to the DER owners. We show that this model preserves full market efficiency, i.e., the social welfare achieved by the aggregation model is the same as that when DERs participate directly in the wholesale market. In the second aggregation model, the profit-seeking aggregator is forced to impose a uniform two-part pricing policy to prosumers from the same location, and we numerically show the efficiency loss of this model. In the third aggregation model, a uniform two-part pricing policy is applied to DER owners, while the aggregator becomes fully regulated but is guaranteed positive profit. It is shown that this third model again achieves full market efficiency. Furthermore, we show that DER aggregation also leads to a reduction on the market power of conventional generators. DER aggregation via profit-seeking and/or regulated aggregators have been investigated by CAISO and NYISO, among others, and the recent FERC Order No. 2222 paved the way for aggregators to bid in the wholesale market. Our efficient aggregation models may settle the debate on how DERs should be included in the wholesale electricity market.

   • Speaker contact: Zuguang's website

10:00 - 11:00 ET, August 28th

1. Sindhu Kanya (PhD student, Chalmers University)

2. • Talk title: Social and technical potential of single family houses in increasing the resilience of the power grid during severe disturbances

   • Abstract: Flexible resources aids in enhancing the resilience of a renewable dominated power system. Space heating systems equipped with heat pumps is one such flexible resource. With this background, the current study deals with the quantification of flexibility potential of space heating systems in houses equipped with various heat pump types. A heat pump model is represented using a vapour compression heat pump cycle. This model is integrated with a thermal model of a house to estimate electricity consumption, for maintaining the indoor temperature at a set value, as flexibility quantification depends on electricity consumption. In addition to this, flexibility potential is quantified by, analysing and incorporating the results on minimum acceptable indoor temperature from twelve interviews with households owning heat pumps, into the integrated model. The results from interviews reveal that, there is an uncertainty in minimum acceptable indoor temperature, as it is dependant on a number of factors such as frequency and duration of interruption, access to additional heating and motivation to be flexible. Hence, to quantify flexibility using thermal simulations, the indoor temperature is reduced from 20◦C to values between 18◦C and 15◦C, based on minimum acceptable temperatures stated in the interviews. The flexibility potential is quantified in terms of an instantaneous reduction in electric power and reduction in electric energy. By reducing the indoor temperature from 20◦C to the aforementioned values at an outdoor ambient temperature of -5◦C, an instantaneous reduction in electric power is estimated to be 1.6 GW, for a power system with 23 GW plannable power. Additionally, considering the recovery of the indoor temperature to 20◦C in 24 hours, electric energy reduction is found to be between 4.06 GWh and 7.4 GWh, when the reference indoor temperature is reduced to values between 18◦C and 15◦C respectively, over 17.25 hours. Furthermore, with time the amount of flexibility offered reduces, becomes negative during the recovery period and finally reaches zero, when the indoor temperature is restored. The results reveal that space heating systems in houses equipped with heat pumps have the potential to enhance the resilience of the power grid during severe grid disturbances.

   • Speaker contact: sindhukanya.naliniramakrishna@chalmers.se

2. Elizabeth Buechler (PhD candidate, Mechanical Engineering, Stanford University)

• Talk title: Design and Experimental Validation of Model Predictive Control Strategies for Grid-Interactive Water Heaters

• Abstract: Residential electric water heaters offer significant load shifting capabilities due to their thermal heat capacity and large energy consumption. Model predictive control (MPC) strategies can be effective at shifting water heating load in response to time-varying price signals while maintaining thermal comfort. In this work, we analyze how control model fidelity affects MPC performance for stratified electric water heaters under time-of-use rates. Specifically, we propose an MPC formulation based on a three-node thermal model that coarsely captures tank stratification and compare it to a one-node formulation that does not capture stratification and a standard thermostatic controller. These strategies are compared through both real-time laboratory testing and simulation-based evaluation for different water use patterns. Results show that the performance of the proposed formulation is significantly improved due to reduced plant-model mismatch. We also propose methods for control model parameter identification.

• Speaker contact: Lily's website

10:00 - 11:00 ET, September 25th

1. Nikolaus Houben (PhD student, TU Wien)

2. • Talk title: Model-Based Deep Learning for Household Load Forecasting

Abstract: The integration of behind-the-meter generation and storage systems in the residential sector has a major effect on electricity consumption profiles, motivating novel methods for net load forecasting. This paper proposes a model-based deep learning approach for forecasting the net load of households equipped with photovoltaic (PV) systems, batteries, and home energy management systems (HEMS). Our approach leverages a hybrid method combining an irradiance-power transposition model with a model-based neural network, which includes priors through a differentiable home energy management system layer.

2. • Speaker contact: nikolaus.houben@gmail.com

2. Marek Miltner (PhD candidate, Stanford University)

• Talk title: AI4Energy vs. Energy4AI: The intertwined role of sustainable power generation and compute in the dawn of the AI age

• Abstract: In recent decades, Artificial Intelligence (AI) has come from being a relatively obscure topic of research to a broadly mainstream topic driving innovation and investment in industry and academia alike. Within this expansion of interest, most attention is given to new opportunities in how AI approaches can improve the state of the art to address present challenges in various fields.  To address this research gap, we show that the near-term sustainability of the current explosion of AI applications and its tied power consumption increase is closely co-dependent on investment in power grid infrastructure, and some specific ways how AI can in turn be applied to help Power Grids.

• Speaker contact: marek.miltner@stanford.edu

10:00 - 11:00 ET, October 30th

1. Ming Yi (Postdoc, Columbia University)

2. • Talk title: Perturbed Decision-Focused Learning for Modeling Strategic Energy Storage

   • Abstract: This work presents a novel decision-focused framework integrating the physical energy storage model into machine learning pipelines. Motivated by the model predictive control for energy storage, our end-to-end method incorporates the prior knowledge of the storage model and infers the hidden reward that incentivizes energy storage decisions. This is achieved through a dual-layer framework, combining a prediction layer with an optimization layer. We introduce the perturbation idea into the designed decision-focused loss function to ensure the differentiability over linear storage models, supported by a theoretical analysis of the perturbed loss function. We also develop a hybrid loss function for effective model training. We provide two challenging applications for our proposed framework: energy storage arbitrage, and energy storage behavior prediction. The numerical experiments on real price data demonstrate that our arbitrage approach achieves the highest profit against existing methods. The numerical experiments on synthetic and real-world energy storage data show that our approach achieves the best behavior prediction performance against existing benchmark methods, which shows the effectiveness of our method.

   • Speaker contact: my2826@columbia.edu

2. Grant Ruan (Postdoc, MIT)

2. • Talk title: Data-driven energy management of virtual power plants

   • Abstract: A virtual power plant (VPP) refers to an active aggregator of heterogeneous distributed energy resources (DERs), which creates a promising pathway to expand renewable energy and demand-side electrification for deep decarbonization. Currently, VPPs still face technical challenges in dealing with the inherent uncertainty of DERs, and data emerge as a promising and essential resource to handle this issue. In this talk, I will provide a thorough summary of the most recent development of VPP technologies from a data-centric perspective, and then analyze the major role of data within every decision phase. I will then present a typical case of building virtual power plants, and show how to apply adaptive piecewise linear regression to capture the flexibility bounds of building thermal dynamics. This talk will end with a detailed discussions of the future challenges and opportunities such as the needs for technical advances in data management and support systems for the growing scale of future VPP systems.

   • Speaker contact: gruan@mit.edu

10:00 - 11:00 ET, November 27th

1. Adam Lechowicz (PhD candidate, UMass Amherst)

2. • Talk title: Optimizing Individualized Incentives from Grid Measurements and Limited Knowledge of Agent Behavior

   • Abstract: During extreme events, traditional methods of distribution grid regulation (e.g., energy prices, net power injection limits) may be insufficient, making it desirable to explicitly incorporate dimensions of human behavior to ensure system stability.  Grid operators lack direct control over end-users grid interactions, such as energy usage, but incentives can influence behavior -- for example, an end-user that receives a grid-driven incentive may adjust their consumption or expose relevant control variables in response.

A key challenge in studying such incentives is the lack of data about human behavior, which usually motivates strong assumptions, such as distributional assumptions on compliance or rational utility-maximization.  In this paper, we propose a general incentive mechanism in the form of a constrained optimization problem -- our approach is distinguished from prior work by modeling human behavior (e.g., reactions to an incentive) as an arbitrary unknown function.  

2. • Speaker contact: alechowicz@umass.edu

2. Diptyaroop Maji (PhD candidate, UMass Amherst)

• Talk title: CarbonCast: Multi-Day Forecasting of Grid Carbon Intensity using Machine Learning

• Abstract: The ever-increasing demand for energy is resulting in considerable carbon emissions from the electricity grid. In recent years, there has been growing attention on demand-side optimizations to reduce carbon emissions from electricity usage. A vital component of these optimizations is short-term forecasting of the carbon intensity of the grid-supplied electricity. Many recent forecasting techniques focus on day-ahead forecasts, but obtaining such forecasts for longer periods, such as multiple days, while useful, has not gotten much attention. In this paper, we present CarbonCast, a machine-learning-based hierarchical approach that provides multi-day forecasts of the grid's carbon intensity. CarbonCast uses neural networks to first generate production forecasts for all the electricity-generating sources. It then uses a hybrid CNN-LSTM approach to combine these first-tier forecasts with historical carbon intensity data and weather forecasts to generate a carbon intensity forecast for up to four days. Our results show that such a hierarchical design improves the robustness of the predictions against the uncertainty associated with a longer multi-day forecasting period. We analyze which factors most influence the carbon intensity forecasts of any region with a specific mixture of electricity-generating sources and also show that accurate source production forecasts are vital in obtaining precise carbon intensity forecasts. CarbonCast's 4-day forecasts have a MAPE of 3.42--19.95% across more than 25 geographically distributed regions while outperforming state-of-the-art methods. Importantly, CarbonCast is the first open-sourced tool for multi-day carbon intensity forecasts where the code and data are freely available to the research community. We are also developing a UI and a service to provide CI forecasts in real-time.

• Speaker contact: dmaji@umass.edu