Optimizing geographic allotment of photovoltaic capacity in a distributed generation setting
Explore the source record for details and available documents.
SEARCH · Engineering Papers
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.
Explore the source record for details and available documents.
Abstract Advancing the field of photocatalysis requires the elucidation of structural properties that underpin the photocatalytic properties of promising materials. The focus of the present study is layered, Bi‐rich bismuth oxyhalides, which are widely studied for photocatalytic applications yet poorly structurally understood, due to high levels of disorder, nano‐sized domains, and the large number of structurally similar compounds. By connecting insights from multiple scattering techniques, utilizing electron‐, X‐ray‐ and neutron probes, the crystal phase of the synthesized materials is allocated as layered Bi 24 O 31 X 10 (X = Cl, Br), albeit with significant deviation from the reported 3D crystalline model. The materials comprise anisotropic platelet‐shaped crystalline domains, exhibiting significant in‐plane ordering in two dimensions but disorder and an ultra‐thin morphology in the layer stacking direction. Increased synthesis pH tailored larger, more ordered crystalline domains, leading to longer excited state lifetimes determined via femtosecond transient absorption spectroscopy (fs‐TAS). Although this likely contributes to improved photocatalytic properties, assessed via the photooxidation of benzylamine, increasing the overall surface area facilitated the most significant improvement in photocatalytic performance. This study, therefore, enabled both phase allocation and a nuanced discussion of the structure‐property relationship for complicated, ultra‐thin photocatalysts.
One of the promising solutions to achieve sustainable energy systems in future smart cities is to deploy microgrids in local energy networks. Due to the decentralized nature of microgrids, large-scale utilization of these resources will increase the reliability of the energy systems as well as facilitating the integration of renewable energy resources to enable more environmentally-friendly energy network. This chapter presents development of a highly reconfigurable, laboratory-scale microgrid testbed. The proposed microgrid system integrates several distributed generation (DG) units representing photovoltaic panel, wind turbine, battery energy storage and etc. The schematic of the proposed system can be customized to investigate numerous grid architectures and control scenarios. In this study, the loads and energy resource allocation are realized by assessment of the loads and resources priority. Therefore, a hierarchical procedure will take place to allocate the power to different buses in microgrid. In the next chapter, we describe the test-bed implementation of the designed microgrid.
Collective operations in the OpenSHMEM programming model are defined over an active set, which is a grouping of Processing Elements (PEs) based on a triple of information including the starting PE, a log2 stride, and the size of the active set. In addition to the active set, collectives require Users to allocate and initialize synchronization (i.e., pSync) and scratchpad (i.e., pWrk) buffers for use by the collective operations. While active sets and the user-defined buffers were previously useful based on hardware and algorithmic considerations, future systems and applications require us to re-evaluate these concepts. In this paper, we propose Sets and Groups as abstractions to create persistent, flexible groupings of PEs (i.e., Sets) and couple these groups of PEs with memory spaces (i.e., Groups), which remove the allocation and initialization burden from the User. To evaluate Sets and Groups, we perform multiple micro-benchmarks to determine the overhead of these abstractions and demonstrate their utility by implementing a distributed All-Pairs Shortest Path (APSP) application, which we evaluate using multiple synthetic and real-world graphs.
Wide-area data transfer is central to geographically distributed scientific workflows. Faster delivery of data is important for these workflows. Predictability is equally (or even more) important. With the goal of providing a reasonably accurate estimate of data transfer time to improve resource allocation & scheduling for workflows and enable end-to-end data transfer optimization, we apply machine learning methods to develop predictive models for data transfer times over a variety of wide area networks. To build and evaluate these models, we use 201,388 transfers, involving 759 million files totaling 9 PB transferred, over 115 heavily used source-destination pairs (“edges”) between 135 unique endpoints. We evaluate the models for different retraining frequencies and different window size of history data. In the best case, the resulting models have a median prediction error of ≤21% for 50% of the edges, and ≤32% for 75% of the edges. We present a detailed analysis of these results that provides insights into the cause of some of the high errors. We envision that the performance predictor will be informative for scheduling geo-distributed workflows. The insights also suggest obvious directions for both further analysis and transfer service optimization.
The operation of large US Department of Energy (DOE) research facilities, like the DIII-D National Fusion Facility, results in the collection of complex multi-dimensional scientific datasets, both experimental and model-generated. In the future, it is envisioned that integrated data analysis coupled with large-scale high performance computing (HPC) simulations will be used to improve experimental planning and operation. Practically, massive data sets from these simulations provide the physics basis for generation of both reduced semi-analytic and machine-learning-based models. Storage of both HPC simulation datasets (generated from US DOE leadership computing facilities) and experimental datasets presents significant challenges. In this paper, we present a vision for a DOE-wide data management workflow that integrates US DOE fusion facilities with leadership computing facilities. Data persistence and long-term availability beyond the length of allocated projects is essential, particularly for verification and recalibration of artificial intelligence and machine learning (AI/ML) models. Because these data sets are often generated and shared among hundreds of users across multiple leadership computing facility centers, they would benefit from cross-platform accessibility, persistent identifiers (e.g. DOI, or digital object identifier), and provenance tracking. Here, the ability to handle different data access patterns suggests that a combination of low cost, high latency (e.g. for storing ML training sets) and high cost, low latency systems (e.g. for real-time, integrated machine control feedback) may be needed.
We present an overview of results from 11 integrated assessment models (IAMs) that participated in the 33rd study of the Stanford Energy Modeling Forum (EMF-33) on the viability of large-scale deployment of bioenergy for achieving long-run climate goals. The study explores future bioenergy use across models under harmonized scenarios for future climate policies, availability of bioenergy technologies, and constraints on biomass supply. This paper provides a more transparent description of IAMs that span a broad range of assumptions regarding model structures, energy sectors, and bioenergy conversion chains. Without emission constraints, we find vastly different CO 2 emission and bioenergy deployment patterns across models due to differences in competition with fossil fuels, the possibility to produce large-scale bio-liquids, and the flexibility of energy systems. Imposing increasingly stringent carbon budgets mostly increases bioenergy use. A diverse set of available bioenergy technology portfolios provides flexibility to allocate bioenergy to supply different final energy as well as remove carbon dioxide from the atmosphere by combining bioenergy with carbon capture and sequestration (BECCS). Sector and regional bioenergy allocation varies dramatically across models mainly due to bioenergy technology availability and costs, final energy patterns, and availability of alternative decarbonization options. Although much bioenergy is used in combination with CCS, BECCS is not necessarily the driver of bioenergy use. We find that the flexibility to use biomass feedstocks in different energy sub-sectors makes large-scale bioenergy deployment a robust strategy in mitigation scenarios that is surprisingly insensitive with respect to reduced technology availability. However, the achievability of stringent carbon budgets and associated carbon prices is sensitive. Constraints on biomass feedstock supply increase the carbon price less significantly than excluding BECCS because carbon removals are still realized and valued. Incremental sensitivity tests find that delayed readiness of bioenergy technologies until 2050 is more important than potentially higher investment costs.
This paper introduces the Yet Another Kernel Launcher (YAKL) C++ portability library, which strives to enable user-level code with the look and feel of Fortran code. The intended audience includes both C++ developers and Fortran developers unfamiliar with C++. The C++ portability approach is briefly explained, YAKL’s main features are described, and code examples are given that demonstrate YAKL’s usage. YAKL fills a niche capability important particularly to scientific applications seeking to port Fortran code quickly to a portable C++ library. YAKL places heavy emphasis on simplicity, readability, and productivity with performance mainly emphasizing Graphics Processing Units (GPUs). Central to YAKL’s ability to allow Fortran-like user-level code are three features: (1) a multi-dimensional Array class that allows Fortran behavior; (2) a limited library of Fortran intrinsic functions; and (3) an efficient pool allocator that transparently enables cheap frequent allocations and deallocations of YAKL Arrays. While YAKL allows Fortran-style code, it also allows Arrays that exhibit C-like behavior as well, including row-major index ordering and lower bounds of “0”. YAKL currently supports CPUs, CPU threading, and Nvidia, AMD, and Intel GPUs.
High Performance Computing (HPC) systems are critical enablers for groundbreaking scientific research across various domains. Efficient resource allocation, facilitated by job scheduling, is paramount for maximizing the utilization of HPC systems. However, the variability in wait times for queued jobs poses challenges for users, necessitating accurate job wait time estimation. This paper explores the influence of job characteristics, including job size (the number of nodes requested and walltime), the queue to which the job is submitted and other resource requirements, on job wait times in leadership-class HPC systems. Focusing on the Theta Cray XC40 and Polaris machines at Argonne National Laboratory, the study evaluates the performance of different supervised learning algorithms in predicting job wait times. It also evaluates the impact of data preprocessing, including outlier detection, Principal Component Analysis (PCA), and feature selection, on the performance of wait time prediction models. The findings reveal insights into the relationship between job characteristics and wait times, offering a foundation for optimizing resource allocation and enhancing user experience. The methodologies and tools developed in this study are adaptable to other leadership-class HPC systems, providing a valuable contribution to the broader HPC community aiming to improve job scheduling efficiency and user satisfaction.
This phase II clinical trial compared the efficacy and safety of second-line irinotecan and panitumumab treatment (IRI + Pmab) with that of irinotecan, fluoropyrimidines and panitumumab treatment (control) in patients with KRAS wild-type mCRC. The primary endpoint was progression-free survival. In addition, early predictive markers of treatment efficacy were explored. Eighty patients were planned to be recruited. Due to a slow accrual rate, only 48 patients were recruited from 2012 to 2016, of which 23 were allocated to the control group and 25 were allocated to the IRI + Pmab group. The median progression-free survival was 254 days (95% confidence interval, 159–306) for control, and 190 days (95% confidence interval, 159–213) for IRI + Pmab (log-rank test, P = 0.26). The response rate without confirmation was 21.7% (5/23) for control and 40.0% (10/25) for IRI + Pmab. Neutropenia, leukopenia, and anorexia were the most common Grade 3/4 adverse events, and several early drop-outs from the treatment protocol were observed in the control group. As for the biomarkers, carcinoembryonic antigen and lactate dehydrogenase (LDH) smoothly declined immediately after the initial dosing in patients with a partial response or stable disease. After starting treatment, LDH-1 and − 2 increased, while LDH-4 and − 5 decreased, irrespective of tumor response. However, exceptions were frequent. In conclusion, this study failed to prove the safety and efficacy of irinotecan and panitumumab treatment due to insufficient patient accrual. Although LDH and its isozymes changed after initiation of treatment, their ability to predict the tumor response may not surpass that of carcinoembryonic antigen levels. The University Hospital Medical Information Network Clinical Trial Registry: UMIN000007658.
Abstract With machine learning applications now spanning a variety of computational tasks, multi-user shared computing facilities are devoting a rapidly increasing proportion of their resources to such algorithms. Graph neural networks (GNNs), for example, have provided astounding improvements in extracting complex signatures from data and are now widely used in a variety of applications, such as particle jet classification in high energy physics (HEP). However, GNNs also come with an enormous computational penalty that requires the use of GPUs to maintain reasonable throughput. At shared computing facilities, such as those used by physicists at Fermi National Accelerator Laboratory (Fermilab), methodical resource allocation and high throughput at the many-user scale are key to ensuring that resources are being used as efficiently as possible. These facilities, however, primarily provide CPU-only nodes, which proves detrimental to time-to-insight and computational throughput for workflows that include machine learning inference. In this work, we describe how a shared computing facility can use the NVIDIA Triton Inference Server to optimize its resource allocation and computing structure, recovering high throughput while scaling out to multiple users by massively parallelizing their machine learning inference. To demonstrate the effectiveness of this system in a realistic multi-user environment, we use the Fermilab Elastic Analysis Facility augmented with the Triton Inference Server to provide scalable and high-throughput access to a HEP-specific GNN and report on the outcome.
Switchgrass (Panicum virgatum) is a perennial, warm-season grass that has been identified as a potential biofuel feedstock over a large part of North America. We examined above- and belowground responses to nitrogen fertilization in “Alamo” switchgrass grown in West Tennessee, USA. The fertilizer study included a spring and fall sampling of 5-year old switchgrass grown under annual applications of 0, 67, and 202 kg N ha -1 (as ammonium nitrate). Fertilization changed switchgrass biomass allocation as indicated by root:shoot ratios. End-of-growing season root:shoot ratios (mean ± SE) declined significantly (P ≤ 0.05) at the highest fertilizer nitrogen treatment (2.16 ± 0.08, 2.02 ± 0.18, and 0.88 ± 0.14, respectively, at 0, 67, and 202 kg N ha -1 ). Fertilization also significantly increased above- and belowground nitrogen concentrations and decreased plant C:N ratios. Data are presented for coarse live roots, fine live roots, coarse dead roots, fine dead roots, and rhizomes. At the end of the growing season, there was more carbon and nitrogen stored in belowground biomass than aboveground biomass. Finally, fertilization impacted switchgrass tissue chemistry and biomass allocation in ways that potentially impact soil carbon cycle processes and soil carbon storage.
We describe a model-data fusion (MDF) inter-comparison project (REFLEX), which compared various algorithms for estimating carbon (C) model parameters consistent with both measured carbon fluxes and states and a simple C model. Participants were provided with the model and with both synthetic net ecosystem exchange (NEE) ofCO2 and leaf area index (LAI) data, generated from the model with added noise, and observed NEE and LAI data from two eddy covariance sites. Participants endeavoured to estimate model parameters and states consistent with the model for all cases over the two years for which data were provided, and generate predictions for one additional year without observations. Nine participants contributed results using Metropolis algorithms, Kalman filters and a genetic algorithm. For the synthetic data case, parameter estimates compared well with the true values. The results of the analyses indicated that parameters linked directly to gross primary production (GPP) and ecosystem respiration, such as those related to foliage allocation and turnover, or temperature sensitivity of heterotrophic respiration,were best constrained and characterised. Poorly estimated parameters were those related to the allocation to and turnover of fine root/wood pools. Estimates of confidence intervals varied among algorithms, but several algorithms successfully located the true values of annual fluxes from synthetic experiments within relatively narrow 90% confidence intervals, achieving>80% success rate and mean NEE confidence intervals <110 gCm-2 year-1 for the synthetic case. Annual C flux estimates generated by participants generally agreed with gap-filling approaches using half-hourly data. The estimation of ecosystem respiration and GPP through MDF agreed well with outputs from partitioning studies using half-hourly data. Confidence limits on annual NEE increased by an average of 88% in the prediction year compared to the previous year, when data were available. Confidence intervals on annual NEE increased by 30% when observed data were used instead of synthetic data, reflecting and quantifying the addition of model error. Finally, our analyses indicated that incorporating additional constraints, using data on C pools (wood, soil and fine roots) would help to reduce uncertainties for model parameters poorly served by eddy covariance data.
Earth system models (ESMs) are essential tools to study the impacts of historical and future climate on regional and global food production, as well as to assess the effectiveness of possible adaptations and their potential feedback to climate. Several current ESMs have the capabilities to simulate crop growth. However, some critical crop growth processes (e.g. flowering and other reproductive processes) and their responses to environmental extremes (e.g. heat stress) are not yet represented in most of these models. In this paper, an improved maize growth model was implemented in the Community Land Model version 4.5 (CLM4.5) by modifying the maize planting scheme, incorporating the phenology scheme adopted from the APSIM model (Agricultural Production Systems sIMulator), adding a new carbon allocation scheme into CLM4.5, and improving the estimation of canopy structure parameters including leaf area index (LAI) and canopy height. Unique features of the new model (CLM-APSIM) include more detailed phenology stages, an explicit implementation of the impacts of various abiotic environmental stresses (including nitrogen, water, temperature and heat stresses) on maize phenology and carbon allocation, as well as an explicit simulation of grain number. Evaluation of results at 7 AmeriFlux sites located in the US Corn Belt show that the CLM-APSIM model performs better than the original CLM4.5 in simulating phenology (LAI and canopy height), surface fluxes including gross primary production (GPP), net ecosystem exchange (NEE), latent heat (LH), and sensible heat (SH), and especially in simulating the biomass partition and maize yield. The CLM-APSIM model corrects a serious deficiency in CLM4.5-related to CLM4.5’s underestimation of aboveground biomass (i.e. overestimation of belowground biomass) and overestimation of Harvest Index, which lead to a reasonable yield estimation with wrong mechanisms. Moreover, 13-year simulation results from 2001 to 2013 at the three Mead sites (US-Ne1, Ne2 and Ne3) show that the CLM-APSIM model can more accurately reproduce maize yield responses to growing season climate (temperature and precipitation) than the original CLM4.5 when benchmarked with the site-based observations and USDA county-level survey statistics. The CLM-APSIM model is thus more suitable than its predecessor models in terms of simulating abiotic environmental stresses on maize yield. This new model provides an improved tool to attribute maize yield change to various processes under historical and future climate, as well as to assess and design effective climate adaptation strategies for sustainable agricultural production.
Microalgae are a promising feedstock for sustainable fuel and nutritional products. To be economically viable, algal biorefineries will need to produce multiple products, such as nutraceuticals and animal feed, in addition to fuel. Life cycle assessment (LCA) has been applied widely to algal biorefineries, but results vary across studies. Here we address two methodological challenges that can lead to variability in LCAs of multi-product algal biorefineries: the choice of functional unit and the method used to handle co-products. LCA was conducted for an algal biorefinery that produces two representative products—fuel and animal feed—using three functional units: (1) 1 MJ fuel, (2) 1 kg animal feed, and (3) 1 ha of production area, coupled with the system expansion method. For the fuel-based functional unit, the system expansion method was compared to impact allocation based on mass, energy, and market value of the biorefinery products. The choice of functional unit affects the results for several environmental indicators. For ecosystem quality, the choice of functional unit can change the overall balance of environmental impacts from harmful to beneficial. The effect of functional unit on climate change depends on the boundary considered in the model—i.e., “well to gate” or “well to wheels”. When a fuel-based functional unit is used, the choice of co-product handling method (i.e., allocation versus the system expansion method) affects environmental impacts, such as ecosystem quality and climate change, differently. For LCAs that use a fuel-based functional unit to compare algal biofuel to other fuels, multiple co-product handling methods should be analyzed and presented to stakeholders. A land based functional unit coupled with the system expansion method can be used to improve LCAs that compare among multi-product algal biorefineries.
With the increasing integration of distributed energy resources, localized peer-to-peer energy transactions are receiving more attention since they could absorb more distributed renewables locally and stabilize the power grid. To better study the trading behavior and provide guidance for a viable local market design, different game-theoretical frameworks including cooperative and non-cooperative games are established and compared here. For the cooperative grand coalition, the maximum social welfare is guaranteed, the Shapley value and bargaining based approaches are implemented for further profit redistribution. It is shown that the asymmetrical bargaining power can be critical in the final payment clearing and it is equivalent to the Shapley value allocation if prosumers bargain based on their marginal contributions. Non-cooperative trading is formulated by using a Stackelberg bilevel model in which the upper level has the privilege in internal price setting. However, system optimum is not obtained due to competition, and the selfish leader could take away 89% of the collective benefit by price-gap based arbitraging. Aiming for a long-lasting trading market, a virtual benevolent leader and a uniform price structure in which purchase price equals selling price are tested. Experimental results indicate that maximum social welfare and a fairer profit allocation can be achieved with a uniform price structure, regardless of the selfish or benevolent attitude.
Bisphosphonates function to reduce bone turnover, which consequently increases the mean degree of tissue mineralization at an organ level. However, it is not clear if bisphosphonates alter the length of time required for an individual bone-modeling unit (BMU) to fully mineralize. We have recently demonstrated that it takes {approx}350 days (d) for normal, untreated cortical bone to fully mineralize. The aim of this study was to determine the rate at which newly formed trabecular BMUs become fully mineralized in rabbits treated for up to 414 d with clinical doses of either risedronate (RIS) or alendronate (ALN). Thirty-six, 4-month old virgin female New Zealand white rabbits were allocated to RIS (n=12; 2.4 {mu}g/kg body weight), ALN (n=12; 2.4 {mu}g/kg body weight), or volume-matched saline controls (CON; n=12). Fluorochrome labels were administered at specific time intervals to quantify the rate and level of mineralization of trabecular bone from the femoral neck (FN) by Fourier transform infrared microspectroscopy (FTIRM). The organic (collagen) and inorganic (phosphate and carbonate) IR spectral characteristics of trabecular bone from undecalcified 4 micron thick tissue sections were quantified from fluorescently labels regions that had mineralized for 1, 8, 18, 35, 70, 105, 140, 210, 280, and 385 d (4 rabbits per time point and treatment group). All groups exhibited a rapid increase in mineralization over the first 18 days, the period of primary mineralization, with no significant differences between treatments. Mineralization continued to increase, at a slower rate up, to 385 days (secondary mineralization), and was not different among treatments. There were no significant differences between treatments for the rate of mineralization within an individual BMU; however, ALN and RIS both increased global tissue mineralization as demonstrated by areal bone mineral density from DXA. We conclude that increases in tissue mineralization that occur following a period of bisphosphonate treatment is a function of the suppressed rate of remodeling that allows for a greater number of BMUs to obtain a greater degree of mineralization.
Decentralized power generation can play a significant role in contributing to renewable energy (RE) supply. Accordingly, regions can be important players in the transformation of the energy system. However, only scarce spatial capacities are available for a sustainable RE generation. Knowledge gaps exist concerning data and methods for integrating RE assessment and environmental planning methods. This paper presents a methodology for the integrated assessment of different RE potentials and their land requirements. The potential is contrasted with the actual availability of land for RE-generation considering environmental restrictions. An application in the Hanover region demonstrates that using energy capacity maps supports using the most efficient mixture in RE generation. Generally, a combination of wind and solar energies produces the highest energy yield per ha. Furthermore, relying primarily on generalized environmental restrictions for defining exclusion areas can be only a first step: On the one hand the assessment is probably underrating the potential for environmentally sound RE generation due to the undifferentiated exclusion of large areas of a certain protective status. On the other hand, some energy sources, in particular energy crops for bioenergy generation, are malpositioned due to missing spatially explicit information about ecosystem sensitivities and a lack of regulative possibilities. Further research is needed to explore synergistic combinations of energy potentials and their environmentally sound spatial allocation in more detail. - Highlights: • We combine models of RE potentials with technical and environmental restrictions. • We explore the resulting “decision-spaces” for regional RE development. • “Decision-spaces” are small and need to be developed carefully to avoid trade-offs. • Planning needs more differentiated analyses to test RE allocation-scenarios. • We recommend detailed analyses on (cumulative) impacts of (multifunctional) RE use.