Final Technical Report - Hydropower Decision-Support with Machine Learning and Satellite Driven Forecasts
Explore the source record for details and available documents.
13 HYDRO ENERGY↗
SEARCH · Engineering Papers
Results for “energy forecasting”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
From RNNs to Foundation Models: An Empirical Study on Commercial Building Energy Consumption
Accurate short-term energy consumption forecasting for commercial buildings is crucial for smart grid operations. While smart meters and deep learning models enable forecasting using past data from multiple buildings, data heterogeneity from diverse buildings can reduce model performance. The impact of increasing dataset heterogeneity in time series forecasting, while keeping size and model constant, is understudied. We tackle this issue using the ComStock dataset, which provides synthetic energy consumption data for U.S. commercial buildings. Two curated subsets, identical in size and region but differing in building type diversity, are used to assess the performance of various time series forecasting models, including finetuned open-source foundation models (FMs). The results show that dataset heterogeneity and model architecture have a greater impact on post-training forecasting performance than the parameter count. Moreover, despite the higher computational cost, finetuned FMs demonstrate competitive performance compared to base models trained from scratch.
commercial buildings↗
OrbNet Denali: A machine learning potential for biological and organic chemistry with semi-empirical cost and DFT accuracy
We present OrbNet Denali, a machine learning model for an electronic structure that is designed as a drop-in replacement for groundstate density functional theory (DFT) energy calculations. The model is a message-passing graph neural network that uses symmetryadapted atomic orbital features from a low-cost quantum calculation to predict the energy of a molecule. OrbNet Denali is trained on a vast dataset of 2.3 × 10 6 DFT calculations on molecules and geometries. This dataset covers the most common elements in biochemistry and organic chemistry (H, Li, B, C, N, O, F, Na, Mg, Si, P, S, Cl, K, Ca, Br, and I) and charged molecules. OrbNet Denali is demonstrated on several well-established benchmark datasets, and we find that it provides accuracy that is on par with modern DFT methods while offering a speedup of up to three orders of magnitude. For the GMTKN55 benchmark set, OrbNet Denali achieves WTMAD-1 and WTMAD-2 scores of 7.19 and 9.84, on par with modern DFT functionals. For several GMTKN55 subsets, which contain chemical problems that are not present in the training set, OrbNet Denali produces a mean absolute error comparable to those of DFT methods. For the Hutchison conformer benchmark set, OrbNet Denali has a median correlation coefficient of R 2 = 0.90 compared to the reference DLPNO-CCSD(T) calculation and R 2 = 0.97 compared to the method used to generate the training data (ωB97X-D3/def2- TZVP), exceeding the performance of any other method with a similar cost. Similarly, the model reaches chemical accuracy for noncovalent interactions in the S66x10 dataset. For torsional profiles, OrbNet Denali reproduces the torsion profiles of ωB97X-D3/def2-TZVP with an average mean absolute error of 0.12 kcal/mol for the potential energy surfaces of the diverse fragments in the TorsionNet500 dataset.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Enhancing high-fidelity neural network potentials through low-fidelity sampling
The efficacy of neural network potentials (NNPs) critically depends on the quality of the configurational datasets used for training. Prior research using empirical potentials has shown that well-selected liquid–solid transitional configurations of a metallic system can be translated to other metallic systems. This study demonstrates that such validated configurations can be relabeled using density functional theory (DFT) calculations, thereby enhancing the development of high-fidelity NNPs. Training strategies and sampling approaches are efficiently assessed using empirical potentials and subsequently relabeled via DFT in a highly parallelized fashion for high-fidelity NNP training. Our results reveal that relying solely on energy and force for NNP training is inadequate to prevent overfitting, highlighting the necessity of incorporating stress terms into the loss functions. To optimize training involving force and stress terms, we propose employing transfer learning to fine-tune the weights, ensuring that the potential surface is smooth for these quantities composed of energy derivatives. This approach markedly improves the accuracy of elastic constants derived from simulations in both empirical potential-based NNPs and relabeled DFT-based NNPs. Overall, this study offers significant insights into leveraging empirical potentials to expedite the development of reliable and robust NNPs at the DFT level.
97 MATHEMATICS AND COMPUTING↗
Day-ahead photovoltaic power production forecasting methodology based on machine learning and statistical post-processing
A main challenge towards ensuring large-scale and seamless integration of photovoltaic systems is to improve the accuracy of energy yield forecasts, especially in grid areas of high photovoltaic shares. The scope of this paper is to address this issue by presenting a unified methodology for hourly-averaged day-ahead photovoltaic power forecasts with improved accuracy, based on data-driven machine learning techniques and statistical post-processing. More specifically, the proposed forecasting methodology framework comprised of a data quality stage, data-driven power output machine learning model development (artificial neural networks), weather clustering assessment (K-means clustering), post-processing output optimisation (linear regressive correction method) and the final performance accuracy evaluation. The results showed that the application of linear regression coefficients to the forecasted outputs of the developed day-ahead photovoltaic power production neural network improved the performance accuracy by further correcting solar irradiance forecasting biases. The resulting optimised model provided a mean absolute percentage error of 4.7% when applied to historical system datasets. Finally, the model was validated both, at a hot as well as a cold semi-arid climatic location, and the obtained results demonstrated close agreement by yielding forecasting accuracies of mean absolute percentage error of 4.7% and 6.3%, respectively. Finally, the validation analysis provides evidence that the proposed model exhibits high performance in both forecasting accuracy and stability.
42 ENGINEERING↗
Impedance Methods for Analyzing Stability Impacts of Inverter-Based Resources: Stability Analysis Tools for Modern Power Systems
Power systems around the world are undergoing a major transformation because of the increasing shares of renewable energy, growing deployment of energy storage systems, proliferation of distributed energy resources, electrification of other sectors, and so on. In the United States, wind and solar provided almost 10% of electricity in 2019. The U.S. Energy Information Administration, in its 2020 Annual Energy Outlook, forecasted that the share of electricity from renewables will reach 38% by 2050, of which more than 80% will come from wind and solar. Wind and solar, along with battery energy storage systems, interface with the grid using power electronic inverters; hence, they are collectively referred to as inverter-based resources (IBRs). The increasing annual share of electricity from IBRs in a power system means that during more times of the year, the system will operate at a much higher concentration of IBRs. Figure 1 presents the hourly share of wind and solar generation in the Electric Reliability Council of Texas (ERCOT) system in Texas in 2019. While the annual wind share was at 20%, the instantaneous percentage share was much higher. Moments of high shares of IBRs (>50%) will continue to grow as more IBRs will be deployed in a power system.
24 POWER TRANSMISSION AND DISTRIBUTION↗
GridDS: Data Science Toolkit for Energy Grid Data
According to the U.S. Energy Information Administration (EIA), the demand for energy is expected to increase 50% by the year 20501. While energy standards, such as the Institute of Electrical and Electronics Engineers (IEEE) Standard 1547, (Basso 2015) and monitoring with wide area management systems (WAMS) (Liu 2017, Zhou 2016) have enabled large scale data collection and storage, the application of this data in mitigating costs associated with increased consumer demand is an ongoing focus for energy research. This ubiquitous data collection presents a promising opportunity for machine learning and data science to improve efficiency of distributed energy resources (DERs). The GridDS software toolkit is designed to leverage advanced metering infrastructure (AMI), outage management systems data (OMS), Supervisory control Data Acquisition (SCADA), and geographic information systems (GIS) to forecast future energy demands and detect incipient grid failures. GridDS is a python software library designed to be modular and generalizable to data recorded by DERs. In adapting to disparate datasets recorded by various WAMS, GridDS provides a range of unique functionality not presently implemented in current WAMS which have highly specific software infrastructure by design. GridDS functionality ranges from data specification and preparation, to training and validation for state of the art machine learning, to interactive data visualization. For data intake, GridDS combines: Pandera: a library for creating data specifications. TimeScaleDB: a postgresSQL database infrastructure for efficient storage of timeseries data. Dataset class: A custom dataset class / interface that ensures modularity between a range of synthetic and live recorded datasets. Is
Ladd, Alexander↗
Carbon‐Neutral Pathways for the United States
Abstract The Intergovernmental Panel on Climate Change (IPCC) Special Report on Global Warming of 1.5°C points to the need for carbon neutrality by mid‐century. Achieving this in the United States in only 30 years will be challenging, and practical pathways detailing the technologies, infrastructure, costs, and tradeoffs involved are needed. Modeling the entire U.S. energy and industrial system with new analysis tools that capture synergies not represented in sector‐specific or integrated assessment models, we created multiple pathways to net zero and net negative CO 2 emissions by 2050. They met all forecast U.S. energy needs at a net cost of 0.2–1.2% of GDP in 2050, using only commercial or near‐commercial technologies, and requiring no early retirement of existing infrastructure. Pathways with constraints on consumer behavior, land use, biomass use, and technology choices (e.g., no nuclear) met the target but at higher cost. All pathways employed four basic strategies: energy efficiency, decarbonized electricity, electrification, and carbon capture. Least‐cost pathways were based on >80% wind and solar electricity plus thermal generation for reliability. A 100% renewable primary energy system was feasible but had higher cost and land use. We found multiple feasible options for supplying low‐carbon fuels for non‐electrifiable end uses in industry, freight, and aviation, which were not required in bulk until after 2035. In the next decade, the actions required in all pathways were similar: expand renewable capacity 3.5 fold, retire coal, maintain existing gas generating capacity, and increase electric vehicle and heat pump sales to >50% of market share. This study provides a playbook for carbon neutrality policy with concrete near‐term priorities.
Williams, James H.↗
Estimating the Drivers of the Cost of Saved Electricity in Utility Customer-Funded Energy Efficiency Programs
Energy efficiency programs funded by utility customers provide an electricity resource in most U.S. states, but their scale and cost of saving electricity varies significantly by state. In this paper, we explore the drivers of the cost of saved electricity in these programs with an econometric model and nearly a decade of data reported by efficiency program administrators. We found strong evidence for economies of scale and weak evidence for diseconomies of scale, which suggests that states with low levels of efficiency savings relative to retail sales can increase the size of their efficiency programs without large increases to the cost of saved electricity. We discuss examples of energy efficiency forecasting and potential modeling in light our econometric analysis and identify methodological improvements relevant to utilities and grid operators. This paper provides insights into the economics of customer-funded efficiency programs that will support regulators, utilities, and policymakers to utilize energy efficiency as a resource.
32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
Impacts of Dispatch Strategies and Forecast Errors on the Economics of Behind-the-Meter PV-Battery Systems
To assess the economic value of batteries in hybrid PV-battery systems, one must create a dispatch profile for the battery. Many analyses of battery value assume perfect forecasts of PV generation and load, determining an upper limit on the value of the battery. Prior work that accounts for forecast uncertainty often does so in the context of a single dispatch algorithm, which does not provide a baseline for comparison. Furthermore, when multiple dispatch algorithms are assessed with uncertainty, the benefits considered are for diesel generation in a microgrid, not retail rate savings. This work addresses the gaps in the literature by comparing the performance of both heuristic and optimal dispatch algorithms for retail rate savings under forecast uncertainty, and provides comparisons of the robustness of these algorithms and their associated estimates of economic value. We find that using a perfect forecast can overestimate the value of hybrid PV-battery systems between 1% and 8% compared to the reality of using a day-ahead forecast, depending on the dispatch algorithm used. Thus, accounting for forecast uncertainty in system design and analysis will significantly improve the accuracy of modeled system values.
batteries↗
High-Frequency Forecasting of Solar Radiation Using a Hybrid Numerical Weather Prediction and Stochastic Downscaling Model
This poster presents an overview of high-frequency forecasting of solar radiation using a hybrid numerical weather prediction and stochastic downscaling model.
14 SOLAR ENERGY↗
Comparing micro climate weather conditions from the weather research and forecasting (wrf) model with typical meteorological year (tmy) data for building energy simulations
Conference paper titled, Comparing microclimate weather conditions from the weather research and forecasting (WRF) model with typical meteorological year (TMY) data for building energy simulations accepted for Building Simulation 2025 Conference in Brisbane, Australia in Aug 2025.
Li, Hang↗
A novel improved model for building energy consumption prediction based on model integration
Building energy consumption prediction plays an irreplaceable role in energy planning, management, and conservation. Constantly improving the performance of prediction models is the key to ensuring the efficient operation of energy systems. Moreover, accuracy is no longer the only factor in revealing model performance, it is more important to evaluate the model from multiple perspectives, considering the characteristics of engineering applications. Based on the idea of model integration, this paper proposes a novel improved integration model (stacking model) that can be used to forecast building energy consumption. The stacking model combines advantages of various base prediction algorithms and forms them into “meta-features” to ensure that the final model can observe datasets from different spatial and structural angles. Two cases are used to demonstrate practical engineering applications of the stacking model. A comparative analysis is performed to evaluate the prediction performance of the stacking model in contrast with existing well-known prediction models including Random Forest, Gradient Boosted Decision Tree, Extreme Gradient Boosting, Support Vector Machine, and K-Nearest Neighbor. The results indicate that the stacking method achieves better performance than other models, regarding accuracy (improvement of 9.5%–31.6% for Case A and 16.2%–49.4% for Case B), generalization (improvement of 6.7%–29.5% for Case A and 7.1%-34.6% for Case B), and robustness (improvement of 1.5%–34.1% for Case A and 1.8%–19.3% for Case B). The proposed model enriches the diversity of algorithm libraries of empirical models.
32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
Local-thermal-gradient and large-scale-circulation impacts on turbine-height wind speed forecasting over the Columbia River Basin
Abstract. We investigate the sensitivity of turbine-height wind speed forecast to initial condition (IC) uncertainties over the Columbia River Gorge (CRG) and Columbia River Basin (CRB) for two typical weather phenomena, i.e., local-thermal-gradient-induced marine air intrusion and a cold frontal passage. Four types of turbine-height wind forecast anomalies and their associated IC uncertainties related to local thermal gradients and large-scale circulations are identified using the self-organizing map (SOM) technique. The four SOM types are categorized into two patterns, each accounting for half of the ensemble members. The first pattern corresponds to IC uncertainties that alter the wind forecast through a modulating weather system, which produces the strongest wind anomalies in the CRG and CRB. In the second pattern, the moderate uncertainties in local thermal gradient and large-scale circulation jointly contribute to wind forecast anomaly. We analyze the cross section of wind and temperature anomalies through the gorge to explore the evolution of vertical features of each SOM type. The turbine-height wind anomalies induced by large-scale IC uncertainties are more concentrated near the front. In contrast, turbine-height wind anomalies induced by the local IC thermal uncertainties are found above the surface thermal anomalies. Moreover, the wind forecast accuracy in the CRG and CRB is limited by IC uncertainties in a few specific regions, e.g., the 2 m temperature within the basin and large-scale circulation over the northeast Pacific around 140∘ W.
17 WIND ENERGY↗
Balancing renewable energy and river resources by moving from individual assessments of hydropower projects to energy system planning
As governments and non-state actors strive to minimize global warming, a primary strategy is the decarbonization of power systems which will require a massive increase in renewable electricity generation. Leading energy agencies forecast a doubling of global hydropower capacity as part of that necessary expansion of renewables. While hydropower provides generally low-carbon generation and can integrate variable renewables, such as wind and solar, into electrical grids, hydropower dams are one of the primary reasons that only one-third of the world’s major rivers remain free-flowing. This loss of free-flowing rivers has contributed to dramatic declines of migratory fish and sediment delivery to agriculturally productive deltas. Further, the reservoirs behind dams have displaced tens of millions of people. Thus, hydropower challenges the world’s efforts to meet climate targets while simultaneously achieving other Sustainable Development Goals. In this paper, we explore strategies to achieve the needed renewable energy expansion while sustaining the diverse social and environmental benefits of rivers. These strategies can be implemented at scales ranging from the individual project (environmental flows, fish passage and other site-level mitigation) to hydropower cascades to river basins and regional electrical power systems. While we review evidence that project-level management and mitigation can reduce environmental and social costs, we posit that the most effective scale for finding balanced solutions occurs at the scale of power systems. We further hypothesize that the pursuit of solutions at the system scale can also provide benefits for investors, developers and governments; evidence of benefits to these actors will be necessary for achieving broad uptake of the approaches described in this paper. We test this hypothesis through cases from Chile and Uganda that demonstrate the potential for system-scale power planning to allow countries to meet low-carbon energy targets with power systems that avoid damming high priority rivers (e.g., those that would cause conflicts with other social and environmental benefits) for a similar system cost as status quo approaches. We also show that, through reduction of risk and potential conflict, strategic planning of hydropower site selection can improve financial performance for investors and developers, with a case study from Colombia.
Opperman, Jeffrey J.↗
Assessing Machine Learning as a Tool to Explain Variance in Deployed Photovoltaic (PV) System Degradation
Degradation remains a large uncertainty in forecasting production for PV plants, creating significant risk for developers and financiers. This study aims to quantify the distribution and drivers of degradation across 10,000 PV systems deployed for distributed or utility generation by training a machine learning model to predict year-over-year degradation rates from metadata characteristics. A combination of K-Means clustering and random forest regressor were found to associate multiple metadata features as potential drivers of degradation, including module characteristics, system design, and climate features. From this, it is inferred that if machine learning is able to find complex patterns between metadata features and system performance loss, such methods can be employed to help developers and financiers make data-informed decisions when estimating long-term energy production forecasts in financial models.
Dunn, Jimmy C.↗
Unfounded concerns about photovoltaic module toxicity and waste are slowing decarbonization
Unsubstantiated claims that fuel growing public concern over the toxicity of photovoltaic modules and their waste are slowing their deployment. Clarifying these issues will help to facilitate the decarbonization that our world depends on. Harnessing the potential of photovoltaic (PV) electricity generation is a key part of the transition to less carbon-intensive energy sources. The most recent energy production forecasts call for a massive 75 TW of global PV capacity by 2050 to have a chance of limiting global temperature rise to 1.5 °C and minimizing the impacts of climate change. This is more than a tenfold increase in the current manufacturing and deployment rate in less than 15 years. PV modules are new to many people, so increasing PV deployment has led to growing concerns about the quantity of waste that may arise from decommissioning them (if they are not recycled), and their potential to leach toxic metals. In conclusion, debunking misinformation about PV modules and PV module waste is the first step in addressing these concerns that are unnecessarily slowing PV deployment.
14 SOLAR ENERGY↗
Machine Learning Derived Dynamic Operating Reserve Requirements in High-Renewable Power Systems
Accurately forecasting wind and solar power output poses challenges for deeply decarbonized electricity systems. Grid operators must commit resources to provide reserves to ensure reliable operations in the face of forecast errors, a process which can increase fuel consumption and emissions. We apply neural network-based machine learning to expand the usefulness of median point forecast data by creating probabilistic distributions of short-term uncertainty in demand, wind, and solar forecasts that adapt to prevailing grid conditions. Machine learning derived estimates of forecast errors compare favorably to estimates based on incumbent methods. Reserves derived from machine learning are usually smaller than values derived using incumbent methods, which enables fuel savings during most hours. Machine learning reserves are generally larger than incumbent reserves during times of higher forecast error, potentially improving system reliability. Performance is tested using multi-stage production simulation modeling of the California Independent System Operator (CAISO) system. Machine learning reserves provide production cost and greenhouse gas (GHG) emission reductions of approximately 0.3% relative to historical 2019 requirements. Savings in the 2030 timeframe are highly dependent on battery storage capacity. At lower levels of battery capacity, savings of 0.4% from machine learning reserves are shown. Significant quantities of battery storage are expected to be added to meet California's resource adequacy needs and GHG reduction targets. Addition of these batteries saturate reserve needs and results in minimal within-hour balancing costs in 2030.
24 POWER TRANSMISSION AND DISTRIBUTION↗