Search NASA⌕ Search

SEARCH · Search NASA

Results for “Operator learning”

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 847 records · Page 47
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Advancing Energy Efficiency in Developing Countries: Lessons Learned from Low-Income Residential Experiences in Industrialized Countries

Energy efficiency has great potential to address economic development, energy security, and environmental protection around the world and in developing countries in particular. Low-income residents of developing countries stand to gain significantly from energy efficiency investments. This report presents lessons learned in delivering energy-efficiency programs for low-income households that are taken from experiences in the United States and other member countries of the Organisation for Economic Co-operation and Development (OECD). Lessons learned in delivering energy-efficiency programs to low-income households of industrial countries are summarized from the perspectives of program design, program delivery, and institutions in order to provide best practices that are applicable to countries in different development stages. This report discusses who to target, how to reduce energy burden, coordinating institutional stakeholders, which efficiency tools to use, how to fund low-income energy efficiency programs, screening tools to prioritize projects, outreach strategies, and evaluation methods. The report concludes with a framework for understanding different levels of energy-efficiency interventions based on institutional capacity, political willpower, and resources.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2020

Deep anomaly detection for industrial systems: a case study

We explore the use of deep neural networks for anomaly detection of industrial systems where the data are multivariate time series measurements. We formulate the problem as a self-supervised learning where data under normal operation are used to train a deep neural network autoregressive model, i.e., use a window of time series data to predict future data values. The aim of such a model is to learn to represent the system dynamic behavior under normal conditions, while expect higher model vs. measurement discrepancies under faulty conditions. In real world applications, many control settings are discrete in nature. In this paper, vector embedding and joint losses are employed to deal with such situations. Both LSTM and CNN based deep neural network backbones are studied on the Secure Water Treatment (SWaT) testbed datasets. Also, Support Vector Data Description (SVDD) method is adapted to such anomaly detection settings with deep neural networks. Evaluation methods and results are discussed based on the SWaT dataset along with potential pitfalls.

anomaly detection, deep neural network↗
DOE OSTI2023

Machine Learning–Adjusted WRF Forecasts to Support Wind Energy Needs in Black Start Operations

Abstract The push for increased capacity of renewable sources of electricity has led to the growth of wind-power generation, with a need for accurate forecasts of winds at hub height. Forecasts for these levels were uncommon until recently, and that, combined with the nocturnal collapse of the well-mixed boundary layer and daytime growth of the boundary layer through the levels important for energy generation, has contributed to errors in numerical modeling of wind generation resources. The present study explores several machine learning algorithms to both forecast and correct standard WRF Model forecasts of winds and temperature at hub height within wind turbine plants over several different time periods that are critical for the anticipation of potential blackouts and aiding in black start operations on the power grid. It was found that mean square error for day-2 wind forecasts from the WRF Model can be improved by over 90% with the use of a multioutput neural network, and that 60-min forecasts of WRF error, which can then be used to adjust forecasts, can be made with an LSTM with great accuracy. Nowcasting of temperature and wind speed over a 10-min period using an LSTM produced very low error and especially skillful forecasts of maximum and minimum values over the turbine plant area.

17 WIND ENERGY↗
NASA NTRS2001

Automating Trend Analysis for Spacecraft Constellations

Spacecraft trend analysis is a vital mission operations function performed by satellite controllers and engineers, who perform detailed analyses of engineering telemetry data to diagnose subsystem faults and to detect trends that may potentially lead to degraded subsystem performance or failure in the future. It is this latter function that is of greatest importance, for careful trending can often predict or detect events that may lead to a spacecraft's entry into safe-hold. Early prediction and detection of such events could result in the avoidance of, or rapid return to service from, spacecraft safing, which not only results in reduced recovery costs but also in a higher overall level of service for the satellite system. Contemporary spacecraft trending activities are manually intensive and are primarily performed diagnostically after a fault occurs, rather than proactively to predict its occurrence. They also tend to rely on information systems and software that are oudated when compared to current technologies. When coupled with the fact that flight operations teams often have limited resources, proactive trending opportunities are limited, and detailed trend analysis is often reserved for critical responses to safe holds or other on-orbit events such as maneuvers. While the contemporary trend analysis approach has sufficed for current single-spacecraft operations, it will be unfeasible for NASA's planned and proposed space science constellations. Missions such as the Dynamics, Reconnection and Configuration Observatory (DRACO), for example, are planning to launch as many as 100 'nanospacecraft' to form a homogenous constellation. A simple extrapolation of resources and manpower based on single-spacecraft operations suggests that trending for such a large spacecraft fleet will be unmanageable, unwieldy, and cost-prohibitive. It is therefore imperative that an approach to automating the spacecraft trend analysis function be studied, developed, and applied to missions such as DRACO with the intent that mission operations costs be significantly reduced. The goal of the Constellation Spacecraft Trend Analysis Toolkit (CSTAT) project is to serve as the pathfinder for a fully automated trending system to support spacecraft constellations. The development approach to be taken is evolutionary. In the first year of the project, the intent is to significantly advance the state of the art in current trending systems through improved functionality and increased automation. In the second year, the intent is to add an expert system shell, likely through the adaptation of an existing commercial-off-the-shelf (COTS) or government-off-the-shelf (GOTS) tool to implement some level of the trending intelligence that humans currently provide in manual operations. In the third year, the intent is to infuse the resulting technology into a near-term constellation or formation-flying mission to test it and gain experience in automated trending. The lessons learned from the real missions operations experience will then be used to improve the system, and to ultimately incorporate it into a fully autonomous, closed-loop mission operations system that is truly capable of supporting large constellations. In this paper, the process of automating trend analysis for spacecraft constellations will be addressed. First, the results of a survey on automation in spacecraft mission operations in general, and in trending systems in particular will be presented to provide an overview of the current state of the art. Next, a rule-based model for implementing intelligent spacecraft subsystem trending will be then presented, followed by a survey of existing COTS/GOTS tools that could be adapted for implementing such a model. The baseline design and architecture of the CSTAT system will be presented. Finally, some results obtained from initial software tests and demonstrations will be presented.

Davis, George↗
NASA NTRS2000

The Advanced Communications Technology Satellite - Performance, Reliability and Lessons Learned

The Advanced Communications Satellite (ACTS) was conceived and developed in the mid- 1980s as an experimental satellite to demonstrate unproven Ka-band technology, and potential new commercial applications and services. Since launch into geostationary orbit in September 1993. ACTS has accumulated almost seven years of essentially trouble-free operation and met all program objectives. The unique technology, service experiments. and system level demonstrations accomplished by ACTS have been reported in many forums over the past several years. As ACTS completes its final experiments activity, this paper will relate the top-level program goals that have been achieved in the design, operation, and performance of the particular satellite subsystems. Pre-launch decisions to ensure satellite reliability and the subsequent operational experiences contribute to lessons learned that may be applicable to other comsat programs.

Krawczyk, Richard J.↗
NASA NTRS2010

The Applied Meteorology Unit: Nineteen Years Successfully Transitioning Research into Operations for America's Space Program

The Applied Meteorology Unit (AMU) provides technology transition and technique development to improve operational weather support to the Space Shuttle and the entire American space program. The AMU is funded and managed by NASA and operated by a contractor that provides five meteorologists with a diverse mix of advanced degrees, operational experience, and associated skills including data processing, statistics, and the development of graphical user interfaces. The AMU's primary customers are the U.S. Air Force 45th Weather Squadron at Patrick Air Force Base, the National Weather Service Spaceflight Meteorology Group at NASA Johnson Space Center, and the National Weather Service Melbourne FL Forecast Office. The AMU has transitioned research into operations for nineteen years and worked on a wide range of topics, including new forecasting techniques for lightning probability, synoptic peak winds,.convective winds, and summer severe weather; satellite tools to predict anvil cloud trajectories and evaluate camera line of sight for Space Shuttle launch; optimized radar scan strategies; evaluated and implemented local numerical models; evaluated weather sensors; and many more. The AMU has completed 113 projects with 5 more scheduled to be completed by the end of 2010. During this rich history, the AMU and its customers have learned many lessons on how to effectively transition research into operations. Some of these lessons learned include collocating with the operational customer and periodically visiting geographically separated customers, operator submitted projects, consensus tasking process, use of operator primary advocates for each project, customer AMU liaisons with experience in both operations and research, flexibility in adapting the project plan based on lessons learned during the project, and incorporating training and other transition assistance into the project plans. Operator involvement has been critical to the AMU's remarkable success and many awards from NASA, the National Weather Association, and two citations from the Navy's Center of Excellence for Best Manufacturing Practices. This paper will present the AMU's proven methods and explain how they may be applied by other organizations to effectively transition research into operations.

Madura, John T.↗
DOE OSTI2024

Transient anisotropic kernel for probabilistic learning on manifolds

PLoM (Probabilistic Learning on Manifolds) is a method introduced in 2016 for handling small training datasets by projecting an Itô equation from a stochastic dissipative Hamiltonian dynamical system, acting as the MCMC generator, for which the KDE-estimated probability measure with the training dataset is the invariant measure. PLoM performs a projection on a reduced-order vector basis related to the training dataset, using the diffusion maps (DMAPS) basis constructed with a time-independent isotropic kernel. In this paper, we propose a new ISDE projection vector basis built from a transient anisotropic kernel, providing an alternative to the DMAPS basis to improve statistical surrogates for stochastic manifolds with heterogeneous data. The construction ensures that for times near the initial time, the DMAPS basis coincides with the transient basis. For larger times, the differences between the two bases are characterized by the angle of their spanned vector subspaces. The optimal instant yielding the optimal transient basis is determined using an estimation of mutual information from Information Theory, which is normalized by the entropy estimation to account for the effects of the number of realizations used in the estimations. Consequently, this new vector basis better represents statistical dependencies in the learned probability measure for any dimension. Three applications with varying levels of statistical complexity and data heterogeneity validate the proposed theory, showing that the transient anisotropic kernel improves the learned probability measure.

Diffusion maps↗
DOE OSTI2022

Randomized probe imaging through deep k-learning

Randomized probe imaging (RPI) is a single-frame diffractive imaging method that uses highly randomized light to reconstruct the spatial features of a scattering object. The reconstruction process, known as phase retrieval, aims to recover a unique solution for the object without measuring the far-field phase information. Typically, reconstruction is done via time-consuming iterative algorithms. In this work, we propose a fast and efficient deep learning based method to reconstruct phase objects from RPI data. The method, which we call deep k-learning, applies the physical propagation operator to generate an approximation of the object as an input to the neural network. This way, the network no longer needs to parametrize the far-field diffraction physics, dramatically improving the results. Deep k-learning is shown to be computationally efficient and robust to Poisson noise. The advantages provided by our method may enable the analysis of far larger datasets in photon starved conditions, with important applications to the study of dynamic phenomena in physical science and biological engineering.

Guo, Zhen (ORCID:0000000213473451)↗
NASA NTRS2011

Human Spaceflight Conjunction Assessment: Lessons Learned

This viewgraph presentation reviews the process of a human space flight conjunction assessment and lessons learned from the more than twelve years of International Space Station (ISS) operations. Also, the application of these lessons learned to a recent ISS conjunction assessment with object 84180 on July 16, 2009 is also presented.

Smith, Jason T.↗
NASA NTRS2018

Human-Centered Operations

Efforts to improve operational safety often focus on preventing human error. But humans don't just make mistakes. They do, in fact, make a tremendous contribution to operational safety, and there is much to learn from what goes right. To support people in their role, the operation should be human-centered. To make the operation human-centered, the framework of the 4Ps can be used to create a clear, coherent, consistent and comprehensive guidance.

flight safety↗
DOE OSTI2020

Semisupervised Learning for Seismic Monitoring Applications

The impressive performance that deep neural networks demonstrate on a range of seismic monitoring tasks depends largely on the availability of event catalogs that have been manually curated over many years or decades. However, the quality, duration, and availability of seismic event catalogs vary significantly across the range of monitoring operations, regions, and objectives. Semisupervised learning (SSL) enables learning from both labeled and unlabeled data and provides a framework to leverage the abundance of unreviewed seismic data for training deep neural networks on a variety of target tasks. We apply two SSL algorithms (mean-teacher and virtual adversarial training) as well as a novel hybrid technique (exponential average adversarial training) to seismic event classification to examine how unlabeled data with SSL can enhance model performance. In general, we find that SSL can perform as well as supervised learning with fewer labels. We also observe in some scenarios that almost half of the benefits of SSL are the result of the meaningful regularization enforced through SSL techniques and may not be attributable to unlabeled data directly. Lastly, the benefits from unlabeled data scale with the difficulty of the predictive task when we evaluate the use of unlabeled data to characterize sources in new geographic regions. Finally, in geographic areas where supervised model performance is low, SSL significantly increases the accuracy of source-type classification using unlabeled data.

58 GEOSCIENCES↗
NASA NTRS2011

Acquisition of Skill Proficiency Over Multiple Sessions of a Novel Rover Simulation

Following long-duration exploration transits, adaptive changes in sensorimotor function may impair the crew's ability to safely perform manual control tasks such as operating pressurized rovers. Postflight performance will also be influenced by the level of preflight skill proficiency they have attained. The purpose of this study was to characterize the acquisition of skills in a motion-based rover simulation over multiple sessions, and to investigate the effects of varying the simulation scenarios. METHODS: Twenty healthy subjects were tested in 5 sessions, with 1-3 days between sessions. Each session consisted of a serial presentation of 8 discrete tasks to be completed as quickly and accurately as possible. Each task consisted of 1) perspective-taking, using a map that defined a docking target, 2) navigation toward the target around a Martian outpost, and 3) docking a side hatch of the rover to a visually guided target. The simulator utilized a Stewart-type motion base (CKAS, Australia), single-seat cabin with triple scene projection covering 150 deg horizontal by 50 deg vertical, and joystick controller. Subjects were randomly assigned to a control group (tasks identical in the first 4 sessions) or a varied-practice group. The dependent variables for each task included accuracy toward the target and time to completion. RESULTS: The greatest improvements in time to completion occurred during the docking phase. The varied-practice group showed more improvement in perspective-taking accuracy. Perspective-taking accuracy was also affected by the relative orientation of the rover to the docking target. Skill acquisition was correlated with self-ratings of previous gaming experience. DISCUSSION: Varying task selection and difficulty will optimize the preflight acquisition of skills when performing novel operational tasks. Simulation of operational manual control will provide functionally relevant evidence regarding the impact of sensorimotor adaptation on early surface operations and what countermeasures are needed. Learning Objective: The use of a motion-based simulation to investigate decrements in the proficiency to operate pressurized rovers during early surface operations of space exploration missions, along with the acquisition of skill proficiency needed during the preflight phase of the mission.

Dean, S. L.↗
DOE OSTI2020

Lessons Learned from Demolition of Hanford's Plutonium Finishing Plant - 20507

U.S. Department of Energy (DOE) contractor CH2M Hill Plateau Remediation Company (CHPRC), a Jacobs owned company, is making significant progress demolishing the Plutonium Finishing Plant (PFP), which has long been known as one of the most hazardous buildings in the DOE Environmental Management Complex (DOE Complex). The facility is located on the Hanford Site and produced plutonium metal during the Cold War. Production was stopped in 1989, the facility was formally shut down in 1996, and material processing was completed in 2004. After approximately 25 years of demolition preparations, open-air demolition of the main portions of the Hazard Category II nuclear facility began in November 2016. To prepare the PFP for demolition, CHPRC employees performed some of the most hazardous work across the DOE Complex. In addition to extensive demolition preparations, including a wide-reaching communication strategy across the Hanford Site, CHPRC implemented robust controls and monitoring during demolition, and when necessary, adjusted demolition practices and sequence to maintain employee safety and project efficiency. Demolition of the main PFP building began in November 2016. The technical complexity, high hazards and radiological concerns have all led to a difficult demolition environment. Following a spread of contamination at the project site beyond the radiological boundaries in December 2017, work was halted and a recovery plan was generated to allow the demolition work to continue. After the implementation of additional controls and concurrence from regulatory agencies and DOE, the demolition was ready to resume work and reinitiated in August 2018. There are a number of lessons learned from the demolition activities and recovery plan process. This paper allows CHPRC and DoE's Richland Operations Office (RL) to share lessons learned and progress to date with other challenging and hazardous projects across the DOE Complex. Understanding the risk acceptance level is critical to execution of any open-air demolition project and is unique to each facility. Four key lessons learned from this project include: 1) plan execution issues, 2) stakeholder communication issues, 3) watching for summit fever and 4) minimizing exposed contaminated surfaces during demolition. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2023

DeepONet-grid-UQ: A trustworthy deep operator framework for predicting the power grid’s post-fault trajectories

This paper proposes a novel data-driven method for the reliable prediction of the power grid’s post-fault trajectories, i.e., the power grid’s dynamic response after a disturbance or fault. Here, the proposed method is based on the recently proposed concept of Deep Operator Networks (DeepONets). Unlike traditional neural networks that learn to approximate functions, DeepONets are designed to approximate nonlinear operators, i.e., mappings between infinite-dimensional spaces. Under this operator framework, we design a novel and efficient DeepONet that (i) takes as inputs the trajectories collected before and during the fault and (ii) outputs the predicted post-fault trajectories. In addition, we endow our method with the much-needed ability to balance efficiency with reliable/trustworthy predictions via uncertainty quantification. To this end, we propose and compare two novel methods that enable quantifying the predictive uncertainty. First, we propose a Bayesian DeepONet (B-DeepONet) that uses stochastic gradient Hamiltonian Monte-Carlo to sample from the posterior distribution of the DeepONet trainable parameters. Then, we design a Probabilistic DeepONet (Prob-DeepONet) that uses a probabilistic training strategy to enable quantifying uncertainty at virtually no extra computational cost. Finally, we validate the proposed methods’ predictive power and uncertainty quantification capability using the New York-New England power grid model.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Deep Koopman operators for causal discovery

Causal discovery aims to identify cause-effect mechanisms for better scientific understanding, explainable decision-making, and more accurate modeling. Standard statistical frameworks, such as Granger causality, lack the ability to quantify causal relationships in nonlinear dynamics due to the presence of complex feedback mechanisms, timescale mixing, and nonstationarity. Thus, applying these methods to study causal dynamics in real-world systems, such as the Earth, is a major challenge. Addressing this shortcoming, we leverage deep learning and a Koopman operator-theoretic formalism to present a class of causal discovery algorithms. Kausal uses deep Koopman operator methods to approximate nonlinear dynamics in a linearized vector space in which traditional causal inference methods such as Granger causality can be more easily applied. Our idealized experiments demonstrate Kausal’s superior ability in discovering and characterizing causal signals compared to existing deep learning and non-deep learning state-of-the-art approaches. Finally, the successful identification of major El Niño and La Niña events in observations showcases Kausal’s skill to handle real-world applications.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Heavy Oil Polymer EOR in the Challenging Alaskan Arctic - It Works!

Under co-sponsorship of the US Department of Energy and Hilcorp Alaska LLC the first ever polymer field pilot commenced on 8/28/2018 in the Schrader Bluff heavy oil reservoir at the Milne Point Field on Alaska North Slope (ANS). The primary objective of the pilot is to prove the efficacy of polymer Enhanced Oil Recovery (EOR) to unlock the vast heavy oil resources on ANS. More than two and half years after startup, the polymer injection, supporting laboratory experiments and simulation studies steadily continue. The pilot started injecting hydrolyzed polyacrylamide (HPAM), at a concentration of 1,750 ppm to achieve a target viscosity of 45 cP, into the two horizontal injectors in the flood pattern. Production is monitored in the two horizontal producers. Based on laboratory measurements of polymer viscosity at reservoir conditions, the team decided to reduce polymer concentration to 1,200 ppm since July 2020 in an effort to control injection pressure and optimize polymer utilization. Quality control (QC) on the field ensures uniform polymer solution properties. Representative rock and fluid systems and test conditions are utilized in the corefloods on polymer retention, effect of injection water salinity, polymer loading, and their combinations on oil recovery. A history matched reservoir simulation model for forecasting oil recovery was developed on the basis of all the available field data. Field concerns related to the post-polymer breakthrough impact on flow assurance is addressed via specialized laboratory tests. Notwithstanding early operational disruptions and hydration issues, continuous polymer injection in both injectors has been achieved. To date, 950,000 lbs of polymer or 2 million barrels of polymer solution, equating to 13% of total pore volume (PV), has been placed in the flood pattern, serving as an effective indicator of adequate polymer injectivity. So far, the success of polymer EOR is evident from drastically reduced water cut in the producers, an estimated incremental 1,000 bopd over waterflood, and a favorable polymer utilization of 1.7 lbs/barrel of incremental oil. Polymer breakthrough was observed 26 months after the start of polymer injection. Main observations from corefloods are unusually high polymer retention values in some cases and a positive response to low salinity water. Although the heterogeneity in the flood pattern and exceptionally low water cut pose some challenges, persistent novel and justifiable simulation approaches have resulted in a robust history matched model. Experimental results on produced fluids treatment provide operational guidance for improved oil-water separation and mitigation of heater tube fouling. The ongoing success of the pilot is a key indicator of bringing the team ever closer to meeting the project’s primary objective. The encouraging results of the pilot is one of the drivers that has provided the impetus to apply polymer EOR throughout the Milne Point Field, which would increase oil recovery and extend the economic life of the Trans Alaska Pipeline System. The scientific knowledge, including the many lessons learned from this pilot also has referential value for other potential heavy oil EOR projects throughout the world. The collected data, operational lessons learned, and the overall success of the pilot are summarized in the paper.

None, None↗
DOE OSTI2019

Commercial Building Technology Demonstration (Final Scientific/Technical Report)

The background for this project was a unique and innovative concept that had been proposed and demonstrated by enVerid Systems, reducing the energy consumption in HVAC systems in commercial and public buildings by up to 50% at peak loads. The concept involves selective scrubbing of gas contaminants from indoor air as a practical substitute for outside air ventilation, the latter being the conventional practice today. This scrubbing enables recycling of most of the indoor air and significantly decreasing the amount of outside air, resulting in a reduced HVAC load and commensurate energy savings for the facility. Key to our approach is low cost scrubbing of all gas contaminants from indoor air using novel, efficient and regenerable sorbent materials. The total solution, labeled HLR (HVAC Load Reduction) is implemented by means of compact modules that can be retrofitted onto virtually any building, and utilizes novel sorbents that have been developed and brought to scale production by enVerid. Although the technology had been deployed and demonstrated, the US commercial building and HVAC markets, famously conservative with respect to innovation and new technologies, had yet to embrace the idea or the product. The overarching goal of the project was to help overcome market hesitancy and enable widespread adoption, by creating a critical mass of success stories across different regions and building types, all under oversight and verification of the DOE and NREL, to ensure credibility and accuracy. The project, overall, was a resounding success. Several installations were deployed and tested comprehensively for the impact on both energy consumption and indoor air quality, and results were not only excellent, but quantitatively in line with expectations. A total number of eight (8) installations were attempted, with 2 aborted due to building issues. Of the six successful installations, however, complete M&V was performed by NREL on 2 (Miami and NYC) and partial analysis on a third site (ArcBest), and these three, along with their results, were therefore presented from a financial and reporting standpoint. Although the number of case studies fell short of the original target of 10, we nevertheless had excellent demonstrations in vastly different climate zones: Southeast, Northeast and Midwest, and in very different building types. This included a modern high-rise office building in New York City and a University Campus health center in Miami, and a mid-rise high-density office in Arkansas. The cooling load reduction was shown to be significant, averaging 37% during the final cooling season in the Miami site. In the NYC site, the cooling energy savings were generally less, at around 12%, as expected. While this is a lower number for average/total savings, two comments are important to note: (i) the outside airflow (OA) reduction was not consistent during the data collection period, at least in part understating the attainable savings through incorrect measurement of the “high energy” reference state; and (ii) the peak impact in NYC is almost certainly as high as Miami, with critical ramifications for peak load reduction and equipment downsizing ability. Importantly, the measured energy savings in all locations were related to the outdoor climate conditions and the actual achievable reduction in outside air, very much in line with models and calculations; this leaves us with increased confidence in our understanding of the interaction between the technology and the building, and our ability to project the impact of this innovation in the future, through broad adoption. Simply put: the energy savings and load reduction impact of the HLR solution has been confirmed quantitatively. The results also confirmed the air quality objective. This may be even more significant, in that air quality is more complex and harder to measure and quantify, and the performance of the HLR as an air scrubber is the linchpin of the entire solution. In all sites, indoor air quality met or exceeded the required parameters, in terms of achieving the target (low) levels of indoor pollutants. Furthermore, the concentrations of outdoor sourced pollutants were markedly lower, as expected, mostly due to the reduction in outside air intake although in part also due to the scrubbing. The success was not immediate nor easy, and early challenges with the systems, especially (but not only) around controls and communication, caused delays in commissioning and undermined some of our M&V schedules. The good news is no fundamental issues were encountered and the number of such glitches was declining quickly by the end of the project. Other challenges had to do with coordinating the control of the building OA dampers, which the HLR needs to have in order to deliver the load reduction, but which the facility manager controls regularly and independently. This is not necessarily a problem for normal HLR operation but interferes with accurate M&V, where switching OA settings back and forth on a precise timetable is essential for clean data. The project provided valuable experience in retrofitting buildings that had been originally designed to operate with less recirculation and more outside air, and some key learnings related to the cyclical regeneration of the sorbents in the operational environment of the building. Some of these learnings gave already been incorporated into ongoing improvements in the design of the HLR module as well as the software that is used to control its operation. The cost and time of retrofit installations is much better understood; so is the test and balance procedure of the building that determines the available range of outside air reduction, which in turn determines the amount of load reduced. We had also implemented at least one annual cycle of sorbent cartridge replacement in the field, after which we tested the used sorbent to assess the rate of degradation of sorbent potency – an important factor affecting the long-term economics of HLR deployments. A notable shortfall of the project, other than the fact that the number of sites was below target, was that we had not (yet) completed testing to measure and verify (M&V) the winter energy savings. This is primarily due to late start on the NYC sites (Miami does not portend any heating savings associated with reduced outside air), combined with the issues of OA damper control. It is by no means suggested that such savings cannot be captured, only that the M&V with respect to winter in NYC has been incomplete. The value of this project and its interest to the general public are hard to overstate. The case studies created from Miami and especially from NYC have, without a doubt, opened up the interest of the broad HVAC industry to HLR in the past year. Major manufacturer reps have signed up to represent and promote the product, and dozens of top tier engineering consultants have introduced HLR as a base of design in upcoming construction and renovation projects – precisely the overarching goal of this project. These would not have happened without the highly visible installation in NYC and its ultimately satisfied customer. Furthermore, the DOE itself issued a report in late 2017 analyzing hundreds of candidate technologies for reducing energy consumption in HVAC; remarkably, the budding HLR solution ranked among the top 3(!); no doubt, its status as a viable and significant contender on this list is informed by the success of these demonstrations. And furthermore, in the wake of these demos, several major energy utilities, led by ConEdison in NYC, have endorsed the HLR solution and now routinely provide substantial cash rebates for customers choosing to incorporate the HLR in their buildings and reap the benefit of the reduced peak load as well as reduced cumulative energy demand. Last but not least, in January of 2019, as this report was being written, the HLR module won the AHR Expo Innovation Award for Green Buildings and, remarkably, the prestigious and only Product of the Year 2019, being selected by an independent professional panel of judges from AHR and ASHRAE. Given the confirmation of the enormous energy saving potential of HLR in widescale deployment, as well as ancillary but significant benefits in terms of equipment downsizing and indoor air quality, the benefit of this effort to the public speaks for itself. Accelerating adoption of HLR, without dependence on subsidies and driven by the private sector, is well under way. There is still much work to be done in terms of educating the HVAC and construction ecosystems and improving the product’s robustness and simplicity, to further facilitate the adoption of this technology and the capture of its potential value. But the American public will greatly benefit both directly and indirectly in the immediate future and beyond.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗