Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “testing data”

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 397 records · Page 22
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

The ASHRAE Great Energy Predictor III competition: Overview and results

In late 2019, ASHRAE hosted the Great Energy Predictor III (GEPIII) machine learning competition on the Kaggle platform. This launch marked the third energy prediction competition from ASHRAE and the first since the mid-1990s. In this updated version, the competitors were provided with over 20 million points of training data from 2,380 energy meters collected for 1,448 buildings from 16 sources. This competition’s overall objective was to find the most accurate modeling solutions for the prediction of over 41 million private and public test data points. Furthermore, the competition had 4,370 participants, split across 3,614 teams from 94 countries who submitted 39,403 predictions. In addition to the top five winning workflows, the competitors publicly shared 415 reproducible online machine learning workflow examples (notebooks), including over 40 additional, full solutions. This paper gives a high-level overview of the competition preparation and dataset, competitors and their discussions, machine learning workflows and models generated, winners and their submissions, discussion of lessons learned, and competition outputs and next steps. The most popular and accurate machine learning workflows used large ensembles of mostly gradient boosting tree models, such as LightGBM. Similar to the first predictor competition, preprocessing of the data sets emerged as a key differentiator.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2021

Bayesian calibration of strength model parameters from Taylor impact data

Materials strength plays a key role in determining the mechanical response of engineered structures. As such, accurate strength models are crucial in simulations involving complex loading conditions, particularly when deformation in the plastic regime is deemed important. In this work, a Gaussian process based surrogate for a finite element simulation of a Taylor impact test is developed and used for Bayesian calibration of the Preston–Tonks–Wallace strength model. Here, the surrogate model is shown to closely approximate the salient features of the Taylor cylinder deformation and is validated against simulation output before being used in the strength model calibration routine. The results show that Taylor impact test data can be used in the calibration of constitutive equations through the use of a combination of data science techniques, namely Gaussian processes and Bayesian inference.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Validation of numerical models for seismic fluid-structure-interaction analysis of nuclear, safety-related equipment

Seismic design, qualification, and risk assessment of nuclear safety-related vessels filled with liquid will have to consider the interaction between the vessel (tank), the contained liquid, and submerged components, if any. Seismic fluid-structure-interaction (FSI) analysis of nuclear vessels will rely on numerical models, which are required to be verified and validated. This study validates previously-verified numerical models using test data generated from earthquake-simulator experiments involving a cylindrical tank. Two solvers in LS-DYNA for FSI simulations are used for the numerical analysis: Arbitrary-Lagrangian-Eulerian (ALE) and Incompressible Computational Fluid Dynamics (ICFD). Numerical and test results are compared for responses critical to seismic design of advanced reactor vessels: hydrodynamic pressures on the tank wall, reactions at the support, and wave heights of the contained liquid. Analysis is performed for one-, two-, and three-directional seismic inputs with a range of intensity, and rocking motions. The accuracy of the numerical results and different methods for outputting wave heights are discussed. Recommendations for validation of seismic FSI numerical models of advanced reactors are provided. As a result, the validation exercise presented in this study is broadly applicable to cylindrical tanks, regardless of industry sector.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
DOE OSTI2020

Classifying and analyzing small-angle scattering data using weighted k nearest neighbors machine learning techniques

A consistent challenge for both new and expert practitioners of small-angle scattering (SAS) lies in determining how to analyze the data, given the limited information content of said data and the large number of models that can be employed. Machine learning (ML) methods are powerful tools for classifying data that have found diverse applications in many fields of science. Here, ML methods are applied to the problem of classifying SAS data for the most appropriate model to use for data analysis. The approach employed is built around the method of weighted k nearest neighbors (wKNN), and utilizes a subset of the models implemented in the SasView package (https://www.sasview.org/) for generating a well defined set of training and testing data. The prediction rate of the wKNN method implemented here using a subset of SasView models is reasonably good for many of the models, but has difficulty with others, notably those based on spherical structures. A novel expansion of the wKNN method was also developed, which uses Gaussian processes to produce local surrogate models for the classification, and this significantly improves the classification accuracy. Further, by integrating a stochastic gradient descent method during post-processing, it is possible to leverage the local surrogate model both to classify the SAS data with high accuracy and to predict the structural parameters that best describe the data. The linking of data classification and model fitting has the potential to facilitate the translation of measured data into results for both novice and expert practitioners of SAS.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Pretest Modeling A Spent Nuclear Fuel Seismic Shake Test

The U.S. Department of Energy Spent Fuel and Waste Science and Technology (SFWST) program is planning to conduct a series of full-scale shake table tests to simulate hypothetical earthquake conditions and record the response of surrogate spent nuclear fuel (SNF) assemblies in a dry canister storage system mockup. The shake table motions will represent a range of hypothetical earthquake conditions at hypothetical locations in the continental U.S. to generally define the range of mechanical loads that SNF can be expected to experience during extended dry storage periods. This paper describes the pretest predictions made with LS-DYNA models of the mockup storage systems. The test will use two dry storage system configurations, a mockup vertical concrete cask (VCC) and a mockup horizontal storage module (HSM). The test will use a production-quality canister and basket. Within the canister will be four instrumented fuel assemblies with fuel rods containing surrogate mass and 28 instrumented dummy assemblies that are intended to match the mass and outer dimension of a fuel assembly. The finite element models include models of the VCC and HSM on the shake table to calculate the system level dynamic responses and separate single fuel assembly models to calculate stress and strain in fuel assembly components. Both types of models include nonlinear behavior like rod-to-rod contact and the ability for VCC’s to rock and slide. This paper presents the expected response of the VCC, HSM, and fuel assemblies to the shake table testing that is planned to start in April of 2024. The earthquake conditions represent seismic hazards in the 2,000-to-20,000-year return period range. The test data is expected to confirm the expectation that fuel rod cladding will remain intact, fuel assembly structural components like guide tubes will remain intact, and no significant VCC sliding or tipping will occur in the range of conditions to be tested with the shake table.

Klymyshyn, Nicholas A.↗
DOE OSTI2021

Discovery of AMPX Thermal Scattering Law Processing Issue for Solid Moderators

The 2020 edition of The International Criticality Safety Benchmark Evaluation Project (ICSBEP) Handbook includes a newly produced plastic-moderated evaluation, identified as PU-MET-MIXED-002 and referred to as PMM002. The 2021 edition should include another new plastic-moderated experiment, HEU-MET-THERM-004, referred to as HMT-004. These evaluations are of particular interest, as PMM-002 is moderated with polyethylene, and HMT-004 is moderated with polymethyl methacrylate (Lucite), allowing for investigation of the differences in the thermal scattering laws (TSLs) available for these materials. ENDF/B-VIII.0 includes newly produced data for 1 H-based scattering materials, including Lucite and yttrium hydride. Other available 1 H TSLs include polyethylene, light water, zirconium hydride. The unbound 1 H (free-gas) cross sections were also used. ENDF/B-VIII.0 also includes revisions made to the ENDF/B-VII.1 light water and polyethylene evaluations. Zirconium hydride was unchanged from ENDF/B-VII.1. As of this writing, there are no benchmarks included in the Verified, Archived Library of Inputs and Data (VALID) that are primarily moderated with any solid moderator, so validation that includes the PMM-002 and HMT-004 evaluations would expand the coverage to two new moderators. A study was undertaken at Oak Ridge National Laboratory with two purposes: to provide validation data based on systems that are primarily moderated with polyethylene and Lucite, and to demonstrate the reactivity changes that can result from use of different 1 H TSLs in these plastic-moderated systems. The first objective was mainly to test data and code for SCALE and AMPX, whereas the second objective was to provide useful data for practitioners on the potential variability of predicted $k_{eff}$ based on TSL changes. The use of an exactly correct TSL is often not possible given the materials involved (e.g., lubricants), but this study was intended to provide some indication of the magnitude of the changes among similar materials that can manifest through application of different TSLs in hydrogenous systems over a range of different neutron energy spectra. Unfortunately, the nominal $k_{eff}$ results exposed deviations between SCALE and Monte Carlo N-Particle (MCNP). Some of the results using different TSLs were also unexpected and difficult to explain. An investigation revealed a processing issue in AMPX that was caused by an ambiguous description of the data for incoherent elastic scattering in the ENDF manual. This issue is discussed in detail in a SCALE User Notification, and it affects some solid moderators, but it does not apply to 1 H bound in water. This issue highlights the importance of validating all TSLs used in safety analysis calculations. The remainder of this paper provides a more detailed examination of the results that triggered the investigation and a summary of the findings of that investigation.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

A machine learning aided interpretable model for rupture strength prediction in Fe-based martensitic and austenitic alloys

Abstract The class of 9–12% Cr ferritic-martensitic alloys (FMA) and austenitic stainless steels have received considerable attention due to their numerous applications in high temperature power generation industries. To design high strength steels with prolonged service life requires a thorough understanding of the long-term properties, e.g., creep rupture strength, rupture life, etc., as a function of the chemical composition and processing parameters that govern the microstructural characteristics. In this article, the creep rupture strength of both 9–12% Cr FMA and austenitic stainless steel has been parameterized using curated experimental datasets with a gradient boosting machine. The trained model has been cross validated against unseen test data and achieved high predictive performance in terms of correlation coefficient ( $$R^{2} > 0.98 $$ R 2 > 0.98 for 9–12% Cr FMA and $$R^{2} > 0.95 $$ R 2 > 0.95 for austenitic stainless steel) thus bypassing the need for additional comprehensive tensile test campaigns or physical theoretical calculations. Furthermore, the feature importance has been computed using the Shapley value analysis to understand the complex interplay of different features.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Power generation forecasting for solar plants based on Dynamic Bayesian networks by fusing multi-source information

A Dynamic Bayesian network (DBN) model for solar power generation forecasting in solar plants is proposed in this paper. The key idea is to fuse sensor data, operational indicators, meteorological data, lagged output power information, and model errors for more accurate short-term (e.g., hours) and mid-term (e.g., days to weeks) power generation forecasting. The proposed DBN augments automated data-driven structure learning with expert knowledge encoding using continuous and categorical data given constraints to represent causal relationships within a solar inverter system. Additionally, an error compensation mechanism is proposed to capture temporal fluctuation. The effectiveness of the DBN on solar power generation forecasting was evaluated by rolling window analysis with one-year testing data collected from a local solar plant. The proposed DBN is compared with four state-of-art methods including support-vector regression (SVR), k-nearest neighbors (kNN), artificial neural network (ANN), and long short-term memory (LSTM) models. The result show that the proposed DBN achieves better accuracy in general, and it is not as data-hungry as some neural network-based models. The proposed DBN is also shown to have robust and consistent forecasting power with different forecasting horizons. The accuracy is 92% - 95% from one hour to one week ahead forecasting.

14 SOLAR ENERGY↗
DOE OSTI2020

Pressure falloff testing to characterize CO 2 plume and dry-out zone during CO 2 injection in saline aquifers

The fluid mobilities and the lateral extent of the CO 2 plume and the dry-out zone are among important unknowns required to effectively manage CO 2 storage in saline aquifers projects. These unknowns can be derived from pressure falloff testing which has been widely used to characterize the subsurface. This paper provides a three-region analytical solution for falloff pressure during the infinite-acting period of CO 2 injection in saline aquifers. The solution is derived using Laplace transformation technique of the governing diffusivity equations and the accompanying initial and boundary conditions. The exact solution is validated against numerical results obtained from a black-oil reservoir simulator for two cases. The solution is approximated into simple forms in real-time domain considering different time periods. Then, a graphical semi-log analysis technique based on the approximate solutions is provided as an interpretation tool of falloff tests where the fluid mobilities and the extents of dry-out zone and CO 2 plume can be inferred. The interpretation technique is applied to synthetic falloff test data to investigate its potential and reliability.

02 PETROLEUM↗
DOE OSTI2021

Quantitative Risk Analysis of High Safety Significant Safety-related Digital Instrumentation and Control Systems in Nuclear Power Plants using IRADIC Technology

This report documents the activities performed by Idaho National Laboratory (INL) during fiscal year (FY) 2021 for the U.S. Department of Energy (DOE) Light Water Reactor Sustainability (LWRS) Program, Risk Informed Systems Analysis (RISA) Pathway, digital instrumentation and control (DI&C) Risk Assessment project. In FY-2019, the RISA Pathway initiated a project to develop a risk assessment strategy for delivering a strong technical basis to support effective, licensable, and secure DI&C technologies for digital upgrades/designs. An integrated risk assessment technology for the DI&C systems (IRADIC technology) was proposed for this strategy, which aims to (1) provide a best-estimate risk-informed capability to quantitatively and accurately estimate the safety margin obtained from plant modernization, especially for the High Safety Significant Safety-related (HSSSR) DI&C systems, (2) develop an advanced risk assessment technology to support transition from analog to DI&C technologies for nuclear industry, (3) assure the long-term safety and reliability of vital HSSSR DI&C systems, (4) reduce uncertainty in costs and support integration of DI&C systems in the plant. To achieve these technical goals and deal with the expensive licensing justifications from regulatory insights, the IRADIC technology is instructive for nuclear vendors and utilities to effectively lower the costs associated with digital compliance and speed industry advances by: (1) defining an integrated risk-informed analysis process for DI&C upgrade, including hazard analysis, reliability analysis, and consequence analysis, (2) applying systematic and risk-informed tools to address common cause failures (CCFs) and quantify responding failure probabilities for DI&C technologies, particularly software CCFs, (3) evaluating the impact of digital failures at the individual level, system level, and plant level, (4) providing insights and suggestions on designs to manage the risks; thus, to support the development, licensing, and deployment of advanced DI&C technologies on nuclear power plant (NPPs). In this report, an approach for performing software CCF analysis, given limited data, is developed and demonstrated using a case study of a highly redundant digital reactor trip system. Consequence analysis is also performed based on different accident scenarios. Results indicate that plant modernization including the improvement of HSSSR DI&C systems will make great benefits to plant safety by providing more safety margins to accident management. In addition, a novel approach is proposed in this report for the quantification of software hazards when sufficient operational and testing data available. The method incorporates software development quality as well as strong analysis techniques to identify and link software defects to potential failure modes. The approach includes both semantic and test-based analysis to detect failures that can exist in different stages of the software development life cycle. This method is applied to an advanced human system interface relevant to reactor trip safety developed from the APR 1400 design.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2023

An Explainable Machine-Learning Model for Compensatory Reserve Measurement: Methods for Feature Selection and the Effects of Subject Variability

Tracking vital signs accurately is critical for triaging a patient and ensuring timely therapeutic intervention. The patient’s status is often clouded by compensatory mechanisms that can mask injury severity. The compensatory reserve measurement (CRM) is a triaging tool derived from an arterial waveform that has been shown to allow for earlier detection of hemorrhagic shock. However, the deep-learning artificial neural networks developed for its estimation do not explain how specific arterial waveform elements lead to predicting CRM due to the large number of parameters needed to tune these models. Alternatively, we investigate how classical machine-learning models driven by specific features extracted from the arterial waveform can be used to estimate CRM. More than 50 features were extracted from human arterial blood pressure data sets collected during simulated hypovolemic shock resulting from exposure to progressive levels of lower body negative pressure. A bagged decision tree design using the ten most significant features was selected as optimal for CRM estimation. This resulted in an average root mean squared error in all test data of 0.171, similar to the error for a deep-learning CRM algorithm at 0.159. By separating the dataset into sub-groups based on the severity of simulated hypovolemic shock withstood, large subject variability was observed, and the key features identified for these sub-groups differed. This methodology could allow for the identification of unique features and machine-learning models to differentiate individuals with good compensatory mechanisms against hypovolemia from those that might be poor compensators, leading to improved triage of trauma patients and ultimately enhancing military and emergency medicine.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Degradation and Modeling of Large-Format Commercial Lithium-Ion Cells as a Function of Chemistry, Design, and Aging Conditions

Demand for large-format (>10 Ah) lithium-ion batteries has increased substantially in recent years, due to the growth of both electric vehicle and stationary energy storage markets. The economics of these applications is sensitive to the lifetime of the batteries, and end-of-life can either be due to energy or power limitations. Despite this, there is little information from cell manufacturers on the sensitivity of cell degradation to environmental conditions or battery use. This work reports accelerated aging test data from four commercial large-format lithium-ion batteries from three manufacturers, with varying design (thickness, casings, ...), chemistry (lithium-iron-phosphate (LFP) or lithium-nickel-manganese-cobalt-oxide positive electrodes (NMC), with graphite (Gr) negative electrodes), and capacity (50 to 250 Amp hours). The tested LFP|Gr cell is found to be relatively insensitive to cycling conditions like temperature or voltage window, while NMC|Gr cells have varying sensitivity. Degradation trends are further investigated by training predictive models: simple polynomial trend lines, a semi-empirical reduced-order model, and an empirical reduced-order model identified using machine-learning based on symbolic regression. Calendar and cycle life are simulated over a variety of conditions to directly compare the various batteries. Cell size and thickness are found to substantially impact sensitivity to temperature during cycle aging, while electrode chemistry impacts depth-of-discharge sensitivity. Real-world battery lifetime is evaluated by simulating residential energy storage and commercial frequency containment reserve systems in several U.S. climate regions. Predicted lifetime across cell types varies from 7 years to 20+ years, though all cells are predicted to have at least 10 year life in certain conditions.

battery lifetime↗
DOE OSTI2020

Calibrating Synchronous-Generator-Interfaced DG Models in Microgrids Using Multiple Event Data

Microgrids, consisting of distributed generators (DGs), loads, and energy storage systems, play an important role in future smart grids. In order to evaluate how microgrids operate and their dynamic response, it is necessary to develop accurate dynamic models of DGs in microgrids. This paper proposes a systematic method to calibrate synchronous-generator-interfaced DG models. Challenges associated with model validation in microgrids are analyzed. Underlying parameters are categorized into two groups, i.e., the steady-state parameters and time constants, which are estimated in two successive stages using multiple event data. This methodology ensures that the validated model with the estimated parameters are applicable for all events. A modified unscented Kalman filter (UKF) is proposed to deal with the cases in which measurement of power angle is unavailable. The effectiveness of the proposed method is validated using field test data of a 2.6 MVA diesel generator in a real microgrid.

Wang, Zhiwen↗
DOE OSTI2022

Balancing Impedance and Controllability in Response Reconstruction

One concept in smart dynamic testing is to match the impedance that a component experiences between test and the environment of interest, but this begs the question: how much of an impedance match is needed and could there be too much? In a prior work, the authors performed MIMO testing with a small component connected to various assemblies, each of which had a differing degree of similarity to the actual flight boundary conditions. The results showed that the fidelity of the response at locations away from the control accelerometers was highly sensitive to the impedance. This work presents further case studies to explore these ideas. Subsequent tests are presented for an assembly that presumably matched the impedance even better, and which was also much more flexible, and the results obtained are even worse than when no attention was given to the impedance. Hence, the work presented here suggests that one should seek a balance between 1.) matching the impedance and 2.) improving the controllability of the component of interest. The concepts are explored using both test data of a benchmark component, for which the environment of interest was recorded as the component flew on a sounding rocket.

Shaker Test, Operational Vibration Environment, Su↗
DOE OSTI2023

Long–short-term memory encoder–decoder with regularized hidden dynamics for fault detection in industrial processes

The ability of recurrent neural networks (RNN) to model nonlinear dynamics of high dimensional process data has enabled data-driven RNN-based fault detection algorithms. Previous studies have focused on detecting faults by identifying the discrepancies in data distribution between the faulty and normal data, as reflected in prediction errors generated by RNN models. However, in industrial processes, variations in data distribution can also result from changes in normal control setpoints and compensatory control adjustments in response to disturbances, making it hard to differentiate between normal and faulty conditions. This paper proposes a fault detection method utilizing a long short-term memory (LSTM) encoder–decoder structure with regularized hidden dynamics and reversible instance normalization (RevIN) to compactly represent high-dimensional measurements for effective monitoring. During training, the hidden states of the model are regularized to form a low-dimensional latent space representation of the original multivariate time series data. As a result, the prediction errors of the latent states can be used to monitor the abnormal dynamic variations, while the reconstruction errors of the measured variables are used to monitor the abnormal static variations. Furthermore, the proposed indices can reflect operating conditions, even when the distribution of test data changes, which helps distinguish faults from normal adjustments and disturbances that controllers can settle. Here, data from numerical simulation and the Tennessee Eastman process are used to illustrate the effectiveness of the proposed fault detection method.

42 ENGINEERING↗
DOE OSTI2022

ORNL Neutron Cross Section Measurements of 90 Zr NCSP ND-1 Task [Slides]

Luiz Leal's idea to integrate the experimental uncertainties, covariances, and sensitivities in this workflow and centralize the process at one location led to the creation of a cohesive tool for data testing. It has been practiced for more than two decades by the NCS and ND groups at ORNL. The development and use of the ORNL codes TSUNAMI and TSURFER within SCALE made it possible. Today, it has been extended to SAMINT which enables the coupling of differential and integral data evaluation in a continuous-energy framework.

43 PARTICLE ACCELERATORS↗
DOE OSTI2022

ORNL Neutron Cross Section Measurements of 90 Zr NCSP ND-1 Task [Slides]

Luiz Leal's idea to integrate the experimental uncertainties, covariances, and sensitivities in the ND workflow and to centralize the process at one location created a cohesive tool for data testing. It has been practiced for more than two decades by the NCS and ND groups at ORNL. The development and use of the ORNL codes TSUNAMI and TSURFER within SCALE made it possible. Today, it has been extended to SAMINT, which enables the coupling of differential and integral data evaluation in a continuous-energy framework.

43 PARTICLE ACCELERATORS↗
DOE OSTI2021

High Solids Performance Testing in a Scaled TSCR System

The Tank Side Cesium Removal (TSCR) project is a technology demonstration that will pretreat Hanford tank waste supernatant in support of the Direct Feed Low-Activity Waste (DFLAW) mission. The TSCR system employs two key separation technologies: dead-end filtration (DEF) and ion exchange (IX) using crystalline silicotitanate (CST) media. DEF will be used to remove undissolved solids from tank waste to protect the functionality of the IX columns and the IX system will remove Cs-137 from tank waste. The separation technologies (DEF and IX) used in TSCR are technically mature and have also been successfully deployed at the Savannah River site in a similar facility known as the Tank Closure Cesium Removal (TCCR) system. While testing with simulants and real waste has been successfully performed under conditions expected during the initial operation of TSCR, test data is absent for assessing off normal high solids loading that may be in the TSCR waste feed. Normal TSCR treatment operations are expected to handle wastes with solids content on the order of 200 ppm, and off normal solids loading could be much larger than the nominal level. The testing program described in this report was conducted to understand the consequence of operating the TSCR system at elevated solids loadings up to the high-solids limit of 15,000 ppm [i.e., 1.5-wt%] identified in the TSCR design basis. Although the system is not required to make throughput above the nominal solids loading, the testing was intended to provide important information related to potential off normal operations. At off normal levels near the high-solids limit, there are potential implications for TSCR performance in the areas of throughput, DEF pressure drop, filter backflush frequency, and IX column pressure drop. In addition, intrusion of solids into the IX column was postulated to impact the Cs-137 loading behavior by promoting channeling or flow maldistribution in the column; since the magnitude of the postulated effect was unknown, assessing it was also of interest. The testing was performed using representative waste simulants and a prototypic, integrated TSCR system designed and assembled specifically to conduct the high solids performance assessment. Overall, the scaled TSCR testing demonstrated that full-scale unit operations can succeed in fulfilling their processing objectives in the presence of solids up to 3,000 ppm, but there are potential performance challenges to filtration operations at solids loadings as low as ~500 ppm. The severity of the challenge is likely to be dependent on the type and size distribution of solids, of which the current testing only examined a single type and size distribution. To provide some flexibility for future full-scale operations, the results of the testing suggest two possible risk reduction strategies that can be implemented without any changes in TSCR design or configuration. One option would be to enact an administrative limit on the solids loading to protect TSCR from feeds that are likely to require a high DEF swap frequency. Another option is to permit operation of the DEFs at differential pressures greater than 2 psid before swapping filters. The selection of a higher differential pressure target is not anticipated to adversely impact DEF backflushing efficacy and would reduce both swap frequency and the amount of waste sent to AP-108.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗