Jan Kleissl

Public Safety Power Shutoffs (PSPS) are a proactive strategy to mitigate wildfire risks by preemptively de-energizing power lines and redispatching generation. However, wildfire risk quantification is critical for the operational effectiveness of PSPS. Many existing PSPS formulations rely on the Wildland Fire Potential Index (WFPI) to relate wildfire risk to power system operations. However, this flammability-based wildfire risk correlates less strongly with observed wildfire ignition probabilities (OWIP) than the Large Fire Probability (WLFP). This wildfire modeling discrepancy can distort generation commitments, misinform line de-energizations, and increase real-time (RT) costs. Prior work avoided incorporating wildfire ignition probability (WIP) due to the complexity of modeling wildfire-driven failures as Bernoulli random variables, which introduce non-linear constraints. By leveraging the cross-entropy between WIP and true outages, we represent wildfire risk as the sum of each energized line’s wildfire ignition log probability (log(WIP)), rather than relying on a WFPI proxy. A cross-entropy constraint models joint line failures in a tractable manner without enumerating all failure scenarios. A stochastic day-ahead (DA) unit commitment with the PSPS framework assesses the cost impact of mapping WFPI- or WLFP-based risk metrics to WIP on the IEEE RTS 24-bus and RTS-GMLC systems. Out-of-sample results show that mapping WLFP to log(WIP) in the PSPS optimization leads to more risk-aware decisions and reduces expected and worst-case out-of-sample costs. These findings underscore the benefits of incorporating probabilistic wildfire risk metrics to improve PSPS decision-making for wildfire-resilient power systems.

Electrical load time series exhibit strong heterogeneity across daily patterns driven by calendar effects and behavioral variability, leading many forecasting models to favor dominant weekday profiles while degrading on weekends, holidays, and transition days. This paper proposes an imbalance-aware spatiotemporal forecasting framework via a cluster-conditioned state space model. Daily load patterns are identified via time-series clustering and incorporated as conditioning covariates within a sequence-continuous selective state space models (Mamba), preserving temporal coherence without explicit sequence partitioning. A cluster-weighted training objective further mitigates pattern imbalance while avoiding future-information leakage. The resulting cluster-conditioned Time Series Mamba (TSMamba) consistently improves forecasting robustness across both frequent and infrequent profiles, achieving weighted absolute percentage error (WAPE) reductions of approximately 15% on weekdays, 42% on weekends, and 39% on holidays relative to the vanilla TSMamba, with similar gains in mean absolute error (MAE) and coefficient of variation of the root mean square error (CVRMSE). These results demonstrate that conditioning state dynamics on latent load patterns yields stable and computationally efficient short-term load forecasts under profile transitions.

This paper presents a novel optimization algorithm for electric vehicle (EV) aggregators aiming to maximize net revenue in demand response markets. Aggregated EV charging stations are modeled as a battery with time-varying capacity, enabling participation in these markets. Due to uncertainties in EV plug-in duration and energy demand, it is challenging for aggregators to fulfill bid capacities in real-time (RT). To address this, EV users specify minimum acceptable service levels, allowing aggregators to optimize both charging timing and energy demand in RT. The model is composed of two layers: (1) a Day-Ahead (DA) optimizer that determines optimal EV scheduling and DA demand response market bidding, and (2) a two-stage RT optimizer that fine-tunes the charging schedule using real-time flexibility to mitigate forecast errors. The RT optimizer leverages Model Predictive Control (MPC) in a two-stage structure to address the problem’s non-convexity, which arises from two coupled unknowns: the charging time and the charging energy demand. In the first stage, it determines a cost-optimal charging schedule that ensures full service levels. In the second stage, it optimizes the charging energy demand within a feasible range, bounded above by the first-stage trajectory and below by user-defined minimum service levels, to maximize demand response market revenue. A realistic baseline and a penalty term are integrated into the demand response market revenue term of the cost function to more accurately reflect real-world conditions. Simulation results demonstrate that the proposed method yields a net economic profit at least five times higher than that of immediate (or ‘dumb’) charging. During one month of simulations, the aggregator achieves revenue equivalent to $0.21 per kWh of demand reduction under forecast uncertainty, totaling $3441.

• Sizing of stationary energy storage systems for EV charging plazas was studied. • The study was based on two years of real data from eight DC fast charging stations. • ESSs were used simultaneously for connection power reduction and power smoothing. • Use of the ESSs for both purposes simultaneously increased their degradation. The increasing number and charging power of electric vehicles might cause severe troubles for electrical grids motivating a need for investments in grid upgrades. The issues caused by high and intermittent energy demand of electric vehicle charging plazas can be solved with energy storage systems. This article covers sizing of stationary energy storage systems for electric vehicle charging plazas. Energy storage systems are sized to constrain the highest amount of power that the charging plaza takes from the grid below a particular power limit and restrict the rate of change in the grid power draw below a particular ramp rate limit. This is the first study concerning sizing of energy storage systems to smooth variations in the power that the charging plazas take from the grid. The power limit was altered from 10% to 100% and the ramp rate limit from 1 to 25 %/min in relation to the rated charging power of the charging plaza. This study was based on two years of measurement data compiled from eight direct current fast charging stations located in southern California, USA. The results show that the power and energy capacity requirements for an energy storage system used solely for reducing grid connection power below 50% increase only a little if the same energy storage system is used simultaneously also for reducing the rate of change in the grid power draw to comply with a ramp rate limit of 5 %/min or higher. However, use of the energy storage system for both purposes simultaneously increased the share of total electric vehicle charging energy that ran through the energy storage system increasing the degradation of the energy storage system.

Monthly demand charges form a significant portion of the electric bill for microgrids with variable renewable energy generation. A battery energy storage system (BESS) is commonly used to manage these demand charges. Economic model predictive control (EMPC) with a reference trajectory can be used to dispatch the BESS to optimize the microgrid operating cost. Since demand charges are incurred monthly, EMPC requires a fullmonth reference trajectory for asymptotic stability guarantees that result in optimal operating costs. However, a full-month reference trajectory is unrealistic from a renewable generation forecast perspective. Therefore, to construct a practical EMPC with a reference trajectory, an EMPC formulation considering both non-coincident demand and on-peak demand charges is designed in this work for 24 to 48 h prediction horizons. The corresponding reference trajectory is computed at each EMPC step by solving an optimal control problem over 24 to 48 h reference (trajectory) horizon. Furthermore, BESS state of charge regulation constraints are incorporated to guarantee the BESS energy level in the long term. Multiple reference and prediction horizon lengths are compared for both shrinking and rolling horizons with real-world data. The proposed EMPC with 48 hour rolling reference and prediction horizons outperforms the traditional EMPC benchmark with a 2% reduction in the annual cost, proving its economic benefits.

This paper presents a fast-acting, decentralized dynamic virtual inertial (VI) allocation method for frequency control in power grids with high penetration of inverter-connected resources under low and spatio-temporally varying inertia. The proposed decentralized method involves solving a local constrained convex optimization problem with implicit local frequency dynamics, enabling the use of projected gradient descent. The proposed controller stabilizes post-contingency frequency transients in milliseconds with less than 0.01% convergence error. Extensive simulations investigate the sensitivity of the cumulative and maximum frequency deviations, cumulative VI allocation, transient time and convergence error in frequency stabilization to varying objective function weights on phase angle and frequency deviation, penalty on control effort, and gradient step sizes. The impact of this work is to propose decentralized control schemes as new mechanisms that act before or in alignment with primary and secondary control to safely regulate the frequency in future power grids dominated by inverter-connected resources.

As opposed to stabilizing to a reference trajectory or state, Economic Model Predictive Control (EMPC) optimizes economic performance over a prediction horizon, making it particularly attractive for economic microgrid (MG) dispatch. However, as load and generation forecasts are only known 24-48 h in advance, economically optimal steady states or periodic trajectories are not available and the EMPC-based works that rely on these signals are inadequate. In addition, demand charges, based on maximum monthly grid import power of the MG, cannot be easily casted as an additive cost, which prevents the application of the principle of optimality if introduced naively. In this work, we propose to close this mismatch between the EMPC prediction horizon and existing monthly timescales by means of an appropriately generated baseline reference trajectory. To do this, we first propose an EMPC formulation for a generic deterministic discrete non-linear time-varying system subject to hard state and input constraints. We then show that, under mild assumptions on the terminal cost and region, the asymptotic average economic cost of the proposed method is no worse than a baseline given by any arbitrary reference trajectory that is only known online. In particular, this results into a practical, finite-time upper bound on the average economic cost difference with the baseline that decreases linearly to zero as time goes to infinity. We then show how the proposed EMPC framework can be used to solve optimal MG dispatch problems, introducing various costs and constraints that conform to the required assumptions. By means of this framework, we conduct realistic simulations with data from the Port of San Diego MG, which demonstrate that the proposed method can reduce monthly electricity costs in closed-loop with respect to established baseline reference trajectories.

Public Safety Power Shutoffs (PSPS) are a pre-emptive strategy to mitigate the wildfires caused by power system malfunction. System operators implement PSPS to balance wildfire mitigation efforts through de-energization of transmission lines against the risk of widespread blackouts modeled with load shedding. Existing approaches do not incorporate decision-dependent wildfire-driven failure probabilities, as modeling outage scenario probabilities requires incorporating high-order polynomial terms in the objective. This paper uses distribution shaping to develop an efficient MILP problem representation of the distributionally robust PSPS problem. Building upon the author's prior work, the wildfire risk of operating a transmission line is a function of the probability of a wildfire-driven outage and its subsequent expected impact in acres burned. A day-ahead unit commitment and line de-energization PSPS framework is used to assess the trade-off between total cost and wildfire risk at different levels of distributional robustness, parameterized by a level of distributional dissimilarity $κ$. We perform simulations on the IEEE RTS 24-bus test system.

Chance constrained stochastic model predictive controllers (CC-SMPCs) tradeoff full constraint satisfaction for economical plant performance under uncertainty. Previous CC-SMPC works are over-conservative in constraint violations leading to worse economic performance. Other past works require a priori information about the uncertainty set, limiting their application. This article considers a discrete linear time-invariant (LTI) system with hard constraints on inputs and chance constraints on states, with unknown uncertainty distribution, statistics, or samples. This work proposes a novel adaptive online update rule to relax the state constraints based on the time average of past constraint violations, to achieve reduced conservativeness in closed-loop. Under an ideal control policy assumption, it is proven that the time average of constraint violations asymptotically converges to the maximum allowed violation probability. The method is applied for optimal battery energy storage system (BESS) dispatch in a grid-connected microgrid (MG) with photovoltaic (PV) generation and load demand, with chance constraints on BESS state of charge (SOC). Realistic simulations show the superior electricity cost-saving potential of the proposed method as compared with the traditional economic model predictive control (EMPC) without chance constraints, and a state-of-the-art approach with chance constraints. We satisfy the chance constraints nonconservatively in closed-loop, effectively trading off increased cost savings with minimal adverse effects on BESS lifetime.

PV trackers are among the most effective technological advancements to increase photovoltaic (PV) electricity generation and yearly energy yield without additional PV panels. In this publication, a novel dual axis PV tracking device with face-to-face solar panel protection is described. The analysis of the proposed system addresses three scientific fields and explains their correlations and mutual influences. First, PV tracking technology and its potential are reviewed in the global techno-economic context of the renewable energy transition and characteristics of the PV and tracker market. The current situation and trends in energy production, manufacturing and supply chain issues, as well as PV module life cycle assessment and recycling are discussed. Second, engineering aspects of PV trackers are presented with a focus on the novel design for a low-cost PV tracker incorporating a patented folding and tracking mechanism with a dual-use actuating system. Third, a PV yield simulation quantifies snow shedding and soiling reduction benefits of the proposed PV tracker, as well as the optional utilization of bi-facial PV modules. The proposed PV tracker is shown to be particularly beneficial in high latitude, high elevation, and high albedo environments. Other benefits include longer cleaning intervals in dusty regions and resistance to extreme weather events. Potential drawbacks of the design approach and next steps for commercialization are also discussed. • Interdisciplinary assessment of the role and potential of PV trackers. • Analyses of PV tracker energy yield, soiling reductions, snow shedding and resilience. • Comprehensive and comparative state of the art analyses of available PV trackers. • Patented PV tracker featuring a dual-use actuator and face-to-face panel protection. • Comparison of PV yield of various tracker topologies in simulations.

Many workplaces and other institutions are grappling with how to plan and deploy electric vehicle (EV) charging networks to support their employees and other constituents who drive EVs. We develop a novel approach for designing EV charging networks that is driver-centric: it estimates drivers' charging needs based on their driving and charging habits and determines the optimal number and type of chargers to install to meet those needs. Unlike prior literature, our framework establishes a unique individual profile for each driver based on their observed behaviors and habits. We demonstrate our approach at the University of California San Diego (UCSD) EV network of 439 charging ports using behavioral data derived from 800 EV drivers. We find that using unique driver profiles significantly increases expected network usage and size—in some cases requiring fivefold more workplace charging sessions and a threefold larger network compared to the same analysis based on regionally-averaged driver data. These increases are driven by drivers' tendency to recharge with high battery state-of-charge, which increases network size by 50 % alone and implies less value in using high-power chargers. An institution's goals for supporting drivers, which are important for equity, also significantly affect network size.

Public safety power shutoffs (PSPS) are a common pre-emptive measure to reduce wildfire risk due to power system equipment failure. System operators use PSPS to de-energize electric grid elements that are either prone to failure or located in regions at a high risk of experiencing a wildfire. Successful power system operation during PSPS involves coordination across different time scales. Adjustments to generator commitments and transmission line de-energizations occur at day-ahead intervals while adjustments to load servicing occur at hourly intervals. Generator commitments and operational decisions have to be made under uncertainty in electric grid demand and wildfire potential forecasts. This paper presents deterministic and twostage mean-CVaR stochastic frameworks to show how the likelihood of large wildfires near transmission lines affects generator commitment and transmission line de-energization strategies. The optimal costs of commitment, operation, and lost load on the IEEE 14-bus and 24-bus test systems are compared to the costs generated from prior optimal power shut-off (OPS) formulations. The proposed mean-CVaR stochastic program generates less total expected costs evaluated with respect to higher demand scenarios than costs generated by risk-neutral and deterministic methods.

Resilience is increasingly important in microgrid design as extreme weather and grid disturbances become more frequent, yet there is no widely accepted, operationally interpretable metric for microgrids. This paper defines a planning-oriented resilience metric—the mean survivable outage duration, i.e., the average number of hours a microgrid can sustain critical load during islanded operation under outages initiated at each hour of a representative year. We contribute: (i) a transparent simulation workflow that couples NREL’s REopt dispatch with an open-source Excel/VBA calculator to compute survivable outage duration without relying on historical outage data; and (ii) a sensitivity framework that maps resilience across solar PV and battery energy storage (BESS) capacities to quantify how distributed energy resource (DER) sizing affects sustained islanded operation. In a microgrid test case, resilience is highly sensitive to DER sizing: with PV fixed at the REopt nominal size, reducing BESS capacity by 10–50% lowers mean resilience by 7.4–94.1%; with BESS fixed at nominal, reducing PV by 10–50% lowers resilience by 6.2–87.1%. Conversely, increasing PV capacity by 50% raises mean resilience by 148.5%, compared with a 112.4% increase from a 50% BESS increase. These results show that the proposed metric and tool provide direct, decision-relevant guidance for DER sizing to enhance sustained islanded operation.

Monthly demand charges form a significant portion of the electric bill for microgrids with variable renewable energy generation. A battery energy storage system (BESS) is commonly used to manage these demand charges. Economic model predictive control (EMPC) with a reference trajectory can be used to dispatch the BESS to optimize the microgrid operating cost. Since demand charges are incurred monthly, EMPC requires a full-month reference trajectory for asymptotic stability guarantees that result in optimal operating costs. However, a full-month reference trajectory is unrealistic from a renewable generation forecast perspective. Therefore, to construct a practical EMPC with a reference trajectory, an EMPC formulation considering both non-coincident demand and on-peak demand charges is designed in this work for 24 to 48 h prediction horizons. The corresponding reference trajectory is computed at each EMPC step by solving an optimal control problem over 24 to 48 h reference (trajectory) horizon. Furthermore, BESS state of charge regulation constraints are incorporated to guarantee the BESS energy level in the long term. Multiple reference and prediction horizon lengths are compared for both shrinking and rolling horizons with real-world data. The proposed EMPC with 48 h rolling reference and prediction horizons outperforms the traditional EMPC benchmark with a 2% reduction in the annual cost, proving its economic benefits.

Solar forecasting is a rapidly evolving field that can substantially contribute to the effective integration of large amounts of solar photovoltaic (PV) capacity into the electricity system. However, newly developed solar forecasting models are rarely tested in an operational context considering the intended application and objective. Besides, models are typically evaluated considering only technical error metrics, disregarding their economic value. This paper proposes an operational bidding strategy that optimizes the participation of a PV power plant in the electricity spot markets. To this end, a novel multistage stochastic optimization method is developed that considers the day-ahead, intraday, and imbalance markets. As the developed method utilizes a scenario generation algorithm, the proposed method can be adopted for a wide variety of related applications. The performance of the developed method is assessed using technical and economic metrics and compared to a reference method. The results demonstrate the effectiveness of the proposed bidding strategy, as it substantially outperforms the reference market bidding strategy. The findings also provide insights into the value of a multistage bidding method, as extending market participation from the day-ahead to the intraday market increases revenues by 22%, while halving the total imbalance. Additionally, the study examines the relationship between the technical and economic performance of solar power forecasting models, revealing a non-linear correlation.

Chance constrained stochastic model predictive controllers (CC-SMPC) trade off full constraint satisfaction for economical plant performance under uncertainty. Previous CC-SMPC works are over-conservative in constraint violations leading to worse economic performance. Other past works require a-priori information about the uncertainty set, limiting their application. This paper considers a discrete LTI system with hard constraints on inputs and chance constraints on states, with unknown uncertainty distribution, statistics, or samples. This work proposes a novel adaptive online update rule to relax the state constraints based on the time-average of past constraint violations, to achieve reduced conservativeness in closed-loop. Under an ideal control policy assumption, it is proven that the time-average of constraint violations asymptotically converges to the maximum allowed violation probability. The method is applied for optimal battery energy storage system (BESS) dispatch in a grid connected microgrid with PV generation and load demand, with chance constraints on BESS state-of-charge (SOC). Realistic simulations show the superior electricity cost saving potential of the proposed method as compared to the traditional economic MPC without chance constraints, and a state-of-the-art approach with chance constraints. We satisfy the chance constraints non-conservatively in closed-loop, effectively trading off increased cost savings with minimal adverse effects on BESS lifetime.

This paper studies real-time hybrid co-simulation of voltage regulation in the University of California San Diego (UCSD) microgrid. We integrate a hardware inverter controller with two subsystems simultaneously simulated in RTDS and Typhoon HIL platforms within a single framework. Using real-time load metering data, we emulate realistic conditions of the UCSD microgrid. To enhance voltage regulation, we employ centralized and distributed algorithms based on a second-order conic relaxation of the nonlinear voltage regulation formulation. These algorithms are tested in a hybrid physical and simulation environment using the physical assets of the distributed energy resources connect (DERConnect) testbed under various realtime loading scenarios, providing a realistic testing ground. Our results demonstrate the effectiveness and suitability of these algorithms for distributed voltage regulation in real-world microgrid scenarios. This research advances practical simulation of voltage control strategies in microgrids, with potential applications in enhancing the stability, scalability, and efficiency of power distribution systems.

This work proposes a new approach for improving one day ahead point forecasting of stochastic individual household electricity consumption. The focus is tackling the double peak penalty effect and improving peak predictions. Each prediction is generated by comparing a household’s energy usage of the seven days leading up to the target day with all seven day periods from all households in the dataset. The households with the closest consumption patterns are then used to create the forecast. The proposed method selects nearest neighbors in a similar manner as in the kNN algorithm. However, it utilizes the Hungarian algorithm to extend this approach to allow for comparisons between consumption values that occur at different times. A case study using an open dataset composed of electric consumption data from 100 Irish households demonstrates that this method improves performance of RMSE over kNN by up to 4.5% and 10.6% for persistence forecasting.

In an era of increasing energy demands and environmental concerns, optimizing energy consumption within buildings is crucial. Despite the vast improvements in HVAC and lighting systems, plug loads remain an under-studied area for enhancing building energy efficiency. This paper studies smart plug active operating mode detection, plug-level load forecasting, and plug scheduling methodologies. This research leverages a unique dataset from the University of California, San Diego, consisting of readings from over 150 smart plugs in several office buildings for more than a year, notably during the post-Covid era. This dataset is made publicly available. A comprehensive literature review on plug, i.e., appliances operating mode detection is presented. Novel unsupervised learning approaches are applied to identify plug operating modes. A pipeline integrating the detected modes with forecasting and scheduling is developed, aiming at building energy consumption reduction. Our findings offer valuable insights and promising results into smart plug management for energy-efficient buildings.

In recent years, with the growing number of EV charging stations integrated into the grid, optimizing the aggregated EV load based on individual EV flexibility has drawn aggregators’ attention as a way to regulate the grid and provide grid services, such as day-ahead (DA) demand responses. Due to the forecast uncertainty of EV charging timings and charging energy demands, the actual delivered demand response is usually different from the DA bidding capacity, making it difficult for aggregators to profit from the energy market. This paper presents a two-layer online feedback control algorithm that exploits the EV flexibility with controlled EV charging timings and energy demands. Firstly, the offline model optimizes the EV dispatch considering demand charge management and energy market participation, and secondly, model predictive control is used in the online feedback model, which exploits the aggregated EV flexibility region by reducing the charging energy based on the pre-decided service level for demand response in real time (RT). The proposed algorithm is tested with one year of data for 51 EVs at a workplace charging site. The results show that with a 20% service level reduction in December 2022, the aggregated EV flexibility can be used to compensate for the cost of EV forecast errors and benefit from day-ahead energy market participation by USD 217. The proposed algorithm is proven to be economically practical and profitable.

Public safety power shutoffs (PSPS) are a common pre-emptive measure to reduce wildfire risk due to power system equipment. System operators use PSPS to de-energize electric grid elements that are either prone to failure or located in regions at a high risk of experiencing a wildfire. Successful power system operation during PSPS involves coordination across different time scales. Adjustments to generator commitments and transmission line de-energizations occur at day-ahead intervals while adjustments to load servicing occur at hourly intervals. Generator commitments and operational decisions have to be made under uncertainty in electric grid demand and wildfire potential forecasts. This paper presents deterministic and two-stage mean-CVaR stochastic frameworks to show how the likelihood of large wildfires near transmission lines affects generator commitment and transmission line de-energization strategies. The optimal costs of commitment, operation, and lost load on the IEEE 14-bus test system are compared to the costs generated from prior optimal power shut-off (OPS) formulations. The proposed mean-CVaR stochastic program generates less total expected costs evaluated with respect to higher demand scenarios than costs generated by risk-neutral and deterministic methods.

Forecasting plays an indispensable role in grid integration of solar energy, which is an important pathway toward the grand goal of achieving planetary carbon neutrality. This rather specialized field of solar forecasting constitutes both irradiance and photovoltaic power forecasting. Its dependence on atmospheric sciences and implications for power system operations and planning make the multi-disciplinary nature of solar forecasting immediately obvious. Advances in solar forecasting represent a quiet revolution, as the landscape of solar forecasting research and practice has dramatically advanced as compared to just a decade ago. Solar Irradiance and Photovoltaic Power Forecasting provides the reader with a holistic view of all major aspects of solar forecasting: the philosophy, statistical preliminaries, data and software, base forecasting methods, post-processing techniques, forecast verification tools, irradiance-to-power conversion sequences, and the hierarchical and firm forecasting framework. The book’s scope and subject matter are designed to help anyone entering the field or wishing to stay current in understanding solar forecasting theory and applications. The text provides concrete and honest advice, methodological details and algorithms, and broader perspectives for solar forecasting. Both authors are internationally recognized experts in the field, with notable accomplishments in both academia and industry. Each author has many years of experience serving as editors of top journals in solar energy meteorology. The authors, as forecasters, are concerned not merely with delivering the technical specifics through this book, but more so with the hopes of steering future solar forecasting research in a direction that can truly expand the boundary of forecasting science.