Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Operating systems”

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.

At least 1,405 records · Page 78

Nuclear-Renewable-Storage Systems: Enhancing Planning and Operations of Integrated Energy Systems

Nuclear-renewable-storage integrated energy systems (IES) are multi-carrier energy systems that include not only electricity but also other forms of demands. Because individual IES components must observe their thermo-physical limits, including ramp rates, start-up, and shut-down time, we formulate operations of the IES as an optimization model by minimizing the total operations costs subject to physical limits of all constituent components. In addition, we develop a data-driven approach to improve the computational performance of the economic dispatch model by using reinforcement learning, where an agent is rewarded by meeting demands and penalized otherwise when shifting to the next state.

25 ENERGY STORAGE↗

Optimal Power Flow for Distribution Systems under Uncertain Forecasts

The paper focuses on distribution systems featuring renewable energy sources and energy storage devices, and develops an optimal power flow (OPF) approach to optimize the system operation in spite of forecasting errors. The proposed method builds on a chance-constrained multi-period AC OPF formulation, where probabilistic constraints are utilized to enforce voltage regulation with a prescribed probability. To enable a computationally affordable solution approach, a convex reformulation of the OPF task is obtained by resorting to i) pertinent linear approximations of the power flow equations, and ii) convex approximations of the chance constraints. Particularly, the approximate chance constraints provide conservative bounds that hold for arbitrary distributions of the forecasting errors. An adaptive optimization strategy is then obtained by embedding the proposed OPF task into a model predictive control framework.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Reliability Lessons Learned From GPU Experience With The Titan Supercomputer at Oak Ridge Leadership Computing Facility

The high computational capability of graphics processing units (GPUs) is enabling and driving the scientific discovery process at large-scale. The world’s second fastest supercomputer for open science, Titan, has more than 18,000 GPUs that computational scientists use to perform scientific simu- lations and data analysis. Understanding of GPU reliability characteristics, however, is still in its nascent stage since GPUs have only recently been deployed at large-scale. This paper presents a detailed study of GPU errors and their impact on system operations and applications, describing experiences with the 18,688 GPUs on the Titan supercom- puter as well as lessons learned in the process of efficient operation of GPUs at scale. These experiences are helpful to HPC sites which already have large-scale GPU clusters or plan to deploy GPUs in the future.

Tiwari, Devesh↗

Baseline and target values for regional and point PV power forecasts: Toward improved solar forecasting

Accurate solar photovoltaic (PV) power forecasting allows utilities to reliably utilize solar resources on their systems. However, to truly measure the improvements that any new solar forecasting methods provide, it is important to develop a methodology for determining baseline and target values for the accuracy of solar forecasting at different spatial and temporal scales. This paper aims at developing a framework to derive baseline and target values for a suite of generally applicable, value-based, and custom-designed solar forecasting metrics. The work was informed by close collaboration with utility and independent system operator partners. The baseline values are established based on state-of-the-art numerical weather prediction models and persistence models in combination with a radiative transfer model. The target values are determined based on the reduction in the amount of reserves that must be held to accommodate the uncertainty of PV power output. The proposed reserve-based methodology is a reasonable and practical approach that can be used to assess the economic benefits gained from improvements in accuracy of solar forecasting. Lastly, the financial baseline and targets can be translated back to forecasting accuracy metrics and requirements, which will guide research on solar forecasting improvements toward the areas that are most beneficial to power systems operations.

14 SOLAR ENERGY↗

Are Fibre Channel SANs a Commodity?

The goal of this work is to evaluate the feasibility of putting together a Fibre Channel Storage Area Network with heterogeneous hardware running both open-source and commercial operating systems. Adherence to the Fibre Channel Specification is supposed to guarantee interoperability in such an environment. We also want to evaluate how difficult it might be to put together a SAN using open-source components. While all the commercial vendors provide Fibre Channel support, this comes at a cost, e.g., not only O/S and drivers, but usually an expensive support contract. The open-source model could lower the cost of building and maintaining a SAN. Of course, for this to be the case, the open-source platforms would have to provide the functionality to construct a SAN. We are assembling a Fibre Channel SAN from heterogeneous hardware (i386, alpha, sparc) running *BSD, Linux, Tru64, NT and Solaris operating systems. We are running several tests to investigate the level of Fibre Channel support provided by each OS. Our current testbed is specified in the table below. Currently, it only contains open-source platforms. We plan to add a PC running OpenBSD, as well as the following commercial systems: Sun Ultra 1/Solaris, DEC AlphaServer 4000/Tru64 Unix, Pentium Pro PC/Windows NT.

Becker, Jeff↗

Scalable Memory Registration for High-Performance Networks Using Helper Threads

Remote DMA (RDMA) enables high performance networks to reduce data copying between an application and the operating system (OS). However RDMA operations in some high performance networks require communication memory explicitly registered with the network adapter and pinned by the OS. Memory registration and pinning limits the flexibility of the memory system and reduces the amount of memory that user processes can allocate. These issues become more significant on multicore platforms, since registered memory demand grows linearly with the number of processor cores. In this paper we propose a new memory registration/deregistration strategy to reduce registered memory on multicore architectures for HPC applications. We hide the cost of dynamic memory management by offloading all dynamic memory registration and deregistration requests to a dedicated memory management helper thread. We investigate design policies and performance implications of the helper thread approach. We evaluate our framework with the NAS parallel benchmarks, for which our registration scheme significantly reduces the registered memory (23.62% on average and up to 49.39%) and avoids memory registration/deregistration costs for reused communication memory. We show that our system enables the execution of problem sizes that could not complete under existing memory registration strategies.

99 GENERAL AND MISCELLANEOUS↗

Estimating the Maximum Potential Revenue for Grid Connected Electricity Storage: Arbitrage and Regulation

The valuation of an electricity storage device is based on the expected future cash flow generated by the device. Two potential sources of income for an electricity storage system are energy arbitrage and participation in the frequency regulation market. Energy arbitrage refers to purchasing (storing) energy when electricity prices are low, and selling (discharging) energy when electricity prices are high. Frequency regulation is an ancillary service geared towards maintaining system frequency, and is typically procured by the independent system operator in some type of market. This paper outlines the calculations required to estimate the maximum potential revenue from participating in these two activities. First, a mathematical model is presented for the state of charge as a function of the storage device parameters and the quantities of electricity purchased/sold as well as the quantities offered into the regulation market. Using this mathematical model, we present a linear programming optimization approach to calculating the maximum potential revenue from an electricity storage device. The calculation of the maximum potential revenue is critical in developing an upper bound on the value of storage, as a benchmark for evaluating potential trading strategies, and a tool for capital finance risk assessment. Then, we use historical California Independent System Operator (CAISO) data from 2010-2011 to evaluate the maximum potential revenue from the Tehachapi wind energy storage project, an American Recovery and Reinvestment Act of 2009 (ARRA) energy storage demonstration project. We investigate the maximum potential revenue from two different scenarios: arbitrage only and arbitrage combined with the regulation market. Our analysis shows that participation in the regulation market produces four times the revenue compared to arbitrage in the CAISO market using 2010 and 2011 data. Then we evaluate several trading strategies to illustrate how they compare to the maximum potential revenue benchmark. We conclude with a sensitivity analysis with respect to key parameters.

electricity storage↗

Stochastic Look-Ahead Commitment: A Case Study in MISO

This paper introduces the Stochastic Look Ahead Commitment (SLAC) software prototyped and tested for the Midcontinent Independent System Operator (MISO) look ahead commitment process. SLAC can incorporate hundreds of wind, load, and net scheduled interchange (NSI) uncertainty scenarios. It uses a progressive hedging method to solve a novel two-stage stochastic unit commitment. The first stage commitment decisions, made only for those generators whose decision to commit or not in each time period cannot be deferred, can cover the uncertainties within the next three hours. The second stage includes both the dispatch for each of the scenarios and the commitment decisions that can be deferred. Study results on 15 MISO production days show that SLAC may bring economic and reliability benefits under uncertainty.

MATHEMATICS AND COMPUTING↗

DLMP of Competitive Markets in Active Distribution Networks: Models, Solutions, Applications, and Visions

Traditionally, the electric distribution system operates with uniform energy prices across all system nodes. However, as the adoption of distributed energy resources (DERs) propels a shift from passive to active distribution network (ADN) operation, a distribution-level electricity market has been proposed to manage new complexities efficiently. In addition, distribution locational marginal price (DLMP) has been established in the literature as the primary pricing mechanism. The DLMP inherits the LMP concept in the transmission-level wholesale market but incorporates characteristics of the distribution system, such as high $R/X$ ratios and power losses, system imbalance, and voltage regulation needs. The DLMP provides a solution that can be essential for competitive market operation in future distribution systems. This article first provides an overview of the current distribution-level market architectures and their early implementations. Next, the general clearing model, model relaxations, and DLMP formulation are comprehensively reviewed. The state-of-the-art solution methods for distribution market clearing are summarized and categorized into centralized, distributed, and decentralized methods. Then, DLMP applications for the operation and planning of DERs and distribution system operators (DSOs) are discussed in detail. Finally, visions of future research directions and possible barriers and challenges are presented.

42 ENGINEERING↗

Data-driven cyber attack detection and mitigation for decentralized wide-area protection and control in smart grids

Modern power systems have already evolved into complicated cyber physical systems (CPS), often referred to as smart grids, due to the continuous expansion of the electrical infrastructure, the augmentation of the number of heterogeneous system components and players, and the consequential application of a diversity of information and telecommunication technologies to facilitate the Wide Area Monitoring, Protection and Control (WAMPAC) of the day-to-day power system operation. Because of the reliance on cyber technologies, WAMPAC, among other critical functions, is prone to various malicious cyber attacks. Successful cyber attacks, especially those sabotage the operation of Bulk Electric System (BES), can cause great financial losses and social panics. Application of conventional IT security solutions is indispensable, but it often turns out to be insufficient to mitigate sophisticated attacks that deploy zero-day vulnerabilities or social engineering tactics. To further improve the resilience of the operation of smart grids when facing cyber attacks, it is desirable to make the WAMPAC functions per se capable of detecting various anomalies automatically, carrying out adaptive activity adjustments in time and thus staying unimpaired even under attack. Most of the existing research efforts attempt to achieve this by adding novel functional modules, such as model-based anomaly detectors, to the legacy centralized WAMPAC functions. In contrast, this dissertation investigates the application of data-driven algorithms in cyber attack detection and mitigation within a decentralized architecture aiming at improving the situational awareness and self-adaptiveness of WAMPAC. First part of the research focuses on the decentralization of System Integrity Protection Scheme (SIPS) with Multi-Agent System (MAS), within which the data-driven anomaly detection and optimal adaptive load shedding are further explored. An algorithm named as Support Vector Machine embedded Layered Decision Tree (SVMLDT) is proposed for the anomaly detection, which provides satisfactory detection accuracy as well as decision-making interpretability. The adaptive load shedding is carried out by every agent individually with dynamic programming. The load shedding relies on the load profile propagation among peer agents and the attack adaptiveness is accomplished by maintaining the historical mean of load shedding proportion. Load shedding only takes place after the consensus pertaining to the anomaly detection is achieved among all interconnected agents and it serves the purpose of mitigating certain cyber attacks. The attack resilience of the decentralized SIPS is evaluated using IEEE 39 bus model. It is shown that, unlike the traditional centralized SIPS, the proposed solution is able to carry out the remedial actions under most Denial of Service (DoS) attacks. The second part investigates the clustering based anomalous behavior detection and peer-assisted mitigation for power system generation control. To reduce the dimensionality of the data, three metrics are designed to interpret the behavior conformity of generator within the same balancing area. Semi-supervised K-means clustering and a density sensitive clustering algorithm based on Hierarchical DBSCAN (HDBSCAN) are both applied in clustering in the 3D feature space. Aiming to mitigate the cyber attacks targeting the generation control commands, a peer-assisted strategy is proposed. When the control commands from control center is detected as anomalous, i.e. either missing or the payload of which have been manipulated, the generating unit utilizes the peer data to infer and estimate a new generation adjustment value as replacement. Linear regression is utilized to obtain the relation of control values received by different generating units, Moving Target Defense (MTD) is adopted during the peer selection and 1-dimensional clustering is performed with the inferred control values, which are followed by the final control value estimation. The mitigation strategy proposed requires that generating units can communicate with each other in a peer-to-peer manner. Evaluation results suggest the efficacy of the proposed solution in counteracting data availability and data integrity attacks targeting the generation controls. However, the strategy stays effective only if less than half of the generating units are compromised and it is not able to mitigate cyber attacks targeting the measurements involved in the generation control.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Using Synchrophasor Status Word as Data Quality Indicator: What to Expect in the Field?

Data quality plays a crucial role in successful applications of synchrophasor data in power system operation and control. This paper presents the results of a data quality analysis of a multi-year field-recorded synchrophasor dataset. The analysis has identified several typical data quality issues encountered in the field data. An examination of the PMU status words included with the dataset has revealed several inconsistent implementations and the lack of correlation between the PMU data quality and the status word, which impacts the usefulness of such information. Our investigation has concluded that the status word alone as found in the recorded field dataset could not be used as a reliable indicator of data quality for field-recorded data. Several recommendations are proposed to improve the usefulness of the PMU status word.

Cheng, Zheyuan↗

A simple way to integrate distributed storage into a wholesale electricity market

Abstract Current plans to decarbonize the electric supply system imply that the generation from wind and solar sources will grow substantially. This growth will increase the uncertainty of system operations due to the inherent variability of these renewable sources, and as a result, more reserve capacity will be required to provide the ramping (flexibility) needed for reliable operations. This paper assumes that all of the increased uncertainty comes from wind farms on the grid, and it shows how distributed storage managed locally by aggregators can provide the ramping needed without introducing a separate market for flexibility. This can be accomplished when the aggregators minimize the expected daily cost of the energy purchased from the grid for their customers by submitting optimal bids into the wholesale market with high and low price thresholds for discharging and charging the storage. This model is illustrated using a stochastic multi-period security constrained optimal power flow together with realistic data for a reduction of the network in the Northeast Power Coordinating Council region of the United States. The results show that the bidding strategy for distributed storage provides ramping to the grid just as effectively as storage managed by a system operator.

Lamadrid, Alberto J.↗

Toward Synthesis, Analysis, and Certification of Security Protocols

Implemented security protocols are basically pieces of software which are used to (a) authenticate the other communication partners, (b) establish a secure communication channel between them (using insecure communication media), and (c) transfer data between the communication partners in such a way that these data only available to the desired receiver, but not to anyone else. Such an implementation usually consists of the following components: the protocol-engine, which controls in which sequence the messages of the protocol are sent over the network, and which controls the assembly/disassembly and processing (e.g., decryption) of the data. the cryptographic routines to actually encrypt or decrypt the data (using given keys), and t,he interface to the operating system and to the application. For a correct working of such a security protocol, all of these components must work flawlessly. Many formal-methods based techniques for the analysis of a security protocols have been developed. They range from using specific logics (e.g.: BAN-logic [4], or higher order logics [12] to model checking [2] approaches. In each approach, the analysis tries to prove that no (or at least not a modeled intruder) can get access to secret data. Otherwise, a scenario illustrating the &tack may be produced. Despite the seeming simplicity of security protocols ("only" a few messages are sent between the protocol partners in order to ensure a secure communication), many flaws have been detected. Unfortunately, even a perfect protocol engine does not guarantee flawless working of a security protocol, as incidents show. Many break-ins and security vulnerabilities are caused by exploiting errors in the implementation of the protocol engine or the underlying operating system. Attacks using buffer-overflows are a very common class of such attacks. Errors in the implementation of exception or error handling can open up additional vulnerabilities. For example, on a website with a log-in screen: multiple tries with invalid passwords caused the expected error message (too many retries). but let the user nevertheless pass. Finally, security can be compromised by silly implementation bugs or design decisions. In a commercial VPN software, all calls to the encryption routines were incidentally replaced by stubs, probably during factory testing. The product worked nicely. and the error (an open VPN) would have gone undetected, if a team member had not inspected the low-level traffic out of curiosity. Also, the use secret proprietary encryption routines can backfire, because such algorithms often exhibit weaknesses which can be exploited easily (see e.g., DVD encoding). Summarizing, there is large number of possibilities to make errors which can compromise the security of a protocol. In today s world with short time-to-market and the use of security protocols in open and hostile networks for safety-critical applications (e.g., power or air-traffic control), such slips could lead to catastrophic situations. Thus, formal methods and automatic reasoning techniques should not be used just for the formal proof of absence of an attack, but they ought to be used to provide an end-to-end tool-supported framework for security software. With such an approach all required artifacts (code, documentation, test cases) , formal analyses, and reliable certification will be generated automatically, given a single, high level specification. By a combination of program synthesis, formal protocol analysis, certification; and proof-carrying code, this goal is within practical reach, since all the important technologies for such an approach actually exist and only need to be assembled in the right way.

Schumann, Johann↗

NASA Balloon Technology Activities

The National Aeronautics and Space Administration (NASA) Balloon Program technology development efforts are fundamental to improving the capabilities of the balloon systems, better understanding of the flight dynamics, and to support the science missions throughout the next decade. Building on the foundations of the 20-year research and development program, a technology roadmap has been generated which identifies specific areas of interest to NASA and the vision of future developments. The major components of the roadmap are: vehicle systems, ballooncraft systems, operational and safety support systems, and planetary vehicles. Within each of these major components, technologies are targeted that will provide both better understanding and foster advancements. The Program's technology thrust areas are directed both in broad efforts that touch on a number of the major components as well as specific tasks that address elements within a specific component. Advances in vehicle systems have focused on producing better balloon designs. This is being attempted through the use of improved inputs into the ba!loon des@ process. Central to this is an increasing the understanding of materials used to fabricate balloons. Testing techniques have been improved with better bi-axial characterization of the balloon materials. More realistic radiative properties of the balloon films and components have also been made. Analytical assessments of the balloon designs are also key in improving balloon designs. These analytical assessments have been accomplished using improving analysis tools and an increased understanding of the float environment. Details of these improvements will be presented. To help improve the operational and flight safety of the balloons, improved flight performance predictions have been made using new analytical tools as well as incorporating the increased understanding of the float environment. These two have combined to allow for improved post-flight balloon performance correlations. A potential to significantly improve the potential safety of a long duration flight can be accomplished using some form of a trajectory modification system. An overview of the development efforts for a trajectory modification system will be presented. A number of efforts will be presented that are focused on the improvement of the ballooncraft systems. Planetary ballooning activities have benefited from a number of separate but related technology advancements made by the NASA Balloon Program's technology developments. The tools, techniques, and advancements made for balloon design and balloon analysis are directly applicable to the processes to develop planetary balloons for either Mars or Venus. An overview of the NASA Balloon Program's role in these efforts and testing will be presented. The technology roadmap, as well as specific projects and recent advancements, will be presented. This overview will demonstrate the technology advancements made by the NASA Balloon Program are focusing on leading balloon developments into the future.

Fairbrother, Debora A.↗

Compact System for Operating and Decommissioned Reactor Used Fuel Management - 17243

Most nuclear power plants in operation today were designed with limited wet storage for their discharged used nuclear fuel (UNF), predicated on the assumption that either recycling or geologic storage would be available within the reactor's lifetime. While that has proven to be the case in some countries such as France with its recycling, it has not for most. In the USA, dry storage of UNF at the plant site at an Independent Spent Fuel Storage Installation (ISFSI) has since the mid-1980s been the primary mode of interim storage so that the plant can off-load its UNF from the pools at the power plants. Initially the space reserved for the ISFSI at a reactor site was based on the assumption that the Department of Energy (DOE) would be transporting that UNF to a geological repository site, such as Yucca Mountain, by 1998. However, due to the delay in the opening of any geological repository the utilities are running out of space on their existing ISFSIs for additional UNF dry storage systems. Since plants prefer to keep their ISFSIs inside the protected area of the plant, options for expanding the ISFSI are becoming limited. At the same time, increasing the capacity of the dry storage systems is also not an option in most cases as the dry storage systems have almost reached their capacity limits with the 37 PWR or 89 BWR fuel assembly canister designs. For plants that have announced plans to shut down, their main priority is to off-load the UNF from the pool to dry storage at an ISFSI so that they can proceed to 'Safe store' or other storage options, and start their decommissioning process as soon as possible. Thus, keeping the size of the ISFSI compact is very important for safety, security and economics. This paper will describe the details of a reduced footprint version of AREVA TN's NUHOMS{sup R} System that addresses the need for maximizing ISFSI capacity. This system not only offers a compact design that stores nearly double the number of UNF assemblies in the same space, it also offers superior radiation protection to the workers and public. In addition, it offers unmatched protection against beyond design basis (BDB) events like the tsunami at Fukushima in Japan, the earthquake at North Anna site, tornadoes and manmade attacks such as the 9/11 attack on the World Trade Center in New York. It also includes significant enhancement for a robust Canister Evolved Safety Program for Aging Management. The utilities in the USA are currently using dry storage systems with canisters of different diameter, different length, and different weights. The fuel assemblies that are stored in these canisters are also different in their fuel types (PWR or BWR),different fuel parameters like burnups, enrichments, and cooling times and different storage orientations in the storage overpacks like horizontal or vertical. The compact version of AREVA TN's NUHOMS{sup R} System offers a 'Universal' Storage Overpack which is designed to accept all these different design dry storage canister available in the industry. As UNF storage continues to drives the industry towards innovations that address the need for consolidated storage, compact ISFSIs and aging management, this new evolution of the AREVA TN NUHOMS{sup R} system offers a significant leap forward in addressing all of these challenges. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Generating short-term probabilistic wind power scenarios via nonparametric forecast error density estimators: Generating short-term probabilistic wind power scenarios via nonparametric forecast error density estimators

Forecasts of available wind power are critical in key electric power systems operations planning problems, including economic dispatch and unit commitment. Such forecasts are necessarily uncertain, limiting the reliability and cost effectiveness of operations planning models based on a single deterministic or “point” forecast. A common approach to address this limitation involves the use of a number of probabilistic scenarios, each specifying a possible trajectory of wind power production, with associated probability. We present and analyze a novel method for generating probabilistic wind power scenarios, leveraging available historical information in the form of forecasted and corresponding observed wind power time series. We estimate non-parametric forecast error densities, specifically using epi-spline basis functions, allowing us to capture the skewed and non-parametric nature of error densities observed in real-world data. We then describe a method to generate probabilistic scenarios from these basis functions that allows users to control for the degree to which extreme errors are captured.We compare the performance of our approach to the current state-of-the-art considering publicly available data associated with the Bonneville Power Administration, analyzing aggregate production of a number of wind farms over a large geographic region. Finally, we discuss the advantages of our approach in the context of specific power systems operations planning problems: stochastic unit commitment and economic dispatch. Here, our methodology is embodied in the joint Sandia – University of California Davis Prescient software package for assessing and analyzing stochastic operations strategies.

17 WIND ENERGY↗

DayCent data and results for "Robust paths to net greenhouse gas mitigation and negative emissions via advanced biofuels"

DayCent data and results for:J. L. Field, T. L. Richard, E. A. Smithwick, H. Cai, M. S. Laser, D. S. LeBauer, S. P. Long, K. Paustian, Z. Qin, J. J. Sheehan, P. Smith, M. Q. Wang, L. R. Lynd, Robust paths to net greenhouse gas mitigation and negative emissions via advanced biofuels. Proceedings of the National Academy of Sciences (2020). https://doi.org/10.1073/pnas.1920877117 This zip file contains a UNIX-format DayCent model executable, input files, automation code, and associated directory structure necessary to re-produce the DayCent analysis underlying the manuscript. The main script 'autodaycent.py' (written for Python 2.7) opens an interactive command line routine that facilitates:* Calibrating the DayCent pine growth model.* Initializing DayCent for a set of case studies sites.* Executing an ensemble of model runs representing case study site reforestation, grassland restoration, or conversion to switchgrass cultivation. * Results analysis & generation of manuscript Fig. 3. Note that the interactive analysis code requires that all input files to be contained in the directory structure as uploaded, without modification. Executable versions of the DayCent model (https://www.nrel.colostate.edu/projects/daycent/) compatible with other operating systems are available upon request.Please send questions/comments to John.L.Field@gmail.com

Agro-ecosystem function and prediction↗

Probabilistic Solar Power Forecasting Using Bayesian Model Averaging

There is rising interest in probabilistic forecasting to mitigate risks from solar power uncertainty, but the numerical weather prediction (NWP) ensembles readily available to system operators are often biased and underdispersed. We propose a Bayesian model averaging (BMA) post-processing method suitable for forecasting power from utility-scale photovoltaic (PV) plants at multiple time horizons up to at least the day-ahead timescale. BMA is a kernel dressing technique for NWP ensembles in which the forecast is a weighted sum of member-specific probability density functions. We tailor BMA for utility-scale PV forecasting by modeling power clipping at the AC inverter rating and advance the theory of BMA with a new beta kernel parameterization that accommodates theoretical constraints not previously addressed. BMA is demonstrated for a case study of 11 utility-scale PV plants in Texas, forecasting at hourly resolution for the complete year 2018. BMA's mixture-model approach mitigates underdispersion of the raw ensemble to significantly improve forecast calibration, while consistently outperforming an ensemble model output statistics (EMOS) parametric approach from the literature. At 4-hour lead time, the BMA post-processing achieves continuous ranked probability skill scores of 2--36% over the raw ensemble, with consistent performance at multiple lead times suitable for power system operations.

14 SOLAR ENERGY↗