The solar industry has grown considerably in the last few years. This larger scale has introduced more problems as well as possibilities. One of those possibilities is analyzing the data coming from the sites that are now being monitored, and using the information to answer a variety of questions. We have four questions which are of prime importance identified in this thesis: 1. Can data from customers be trusted? 2. Can we use data from existing sites to determine which sites need the most improvement? 3. Can we implement a location-based algorithm to reduce the amount of false positives for performance, or other alarms? 4. Can we improve upon the current predicted power algorithm? We find that not only can we answer these questions definitively, but the improvements found are of significant value. Each of these items represents an important question that either directly or indirectly translates into increased revenue and engineering improvements for the solar industry as a whole.

Accommodating the continued increase in energy demand in the face of global climate change has been a worldwide concern. With buildings in the US consuming nearly 40% of national energy, a concerted effort must be given to reduce building energy consumption. As new buildings continue to improve their efficiency through more restrictive energy codes, the other 76.9 billion square feet of current building stock falls further behind. The rate at which current buildings are being retrofit is not enough and better tools are needed to access the benefits of retrofits and the uncertainties associated with them. This study proposes a stochastic method of building energy model calibration coupled with a monthly normative building simulation addressed in ISO 13890. This approach takes advantage of the great efficiency of Latin Hypercube Sampling and the lightweight normative building simulation method, to deliver a set of calibrated solutions to assess the effectiveness of energy conservation measure, making uncertainty a part of the modeling process. A case study on a mixed-use university building is conducted to show the strength and performance of this simple method. Limitations and future concerns are also addressed.

The increased penetration of renewable energy sources poses new challenges for grid stability. The stochastic and uncontrollable generation of solar and wind power cannot be adjusted to match the load profile, and the transition away from traditional synchronous generators is reducing the grid capacity to arrest and recover from frequency disturbances. Additionally, the distributed nature of many renewable energy sources makes centralized control of generation more complicated. The traditional power system paradigm balances the supply and demand of electricity on the grid by regulating generation. As this becomes more difficult, one alternative is to adjust the load instead. This is not entirely novel, and utilities have incentivized large industrial customers to reduce consumption during peak hours for years. However, the residential sector, which constitutes 37% of electricity consumption in the U.S., currently has very little capacity for load control. Smart electric water heaters provide utilities with an appliance that can be remotely controlled and serves as a form of energy storage. They have very fast response times and make up a large amount of residential energy consumption, making them useful for load peak shifting as well as other ancillary grid services. As smart appliances become increasingly widespread, more and more devices can be brought into the utility's control network and aggregated into a flexible resource on a megawatt scale. This work demonstrates the usefulness of aggregated electric water heaters for peak shifting and frequency response. Because a large number of assets are required, emulators are developed based on observations of real devices. Emulated water heaters are then connected to an energy resource aggregator using an internet-of-things network. The aggregator successfully uses these assets to shift consumption away from peak hours. An algorithm was developed for detecting upward frequency disturbances in real-time. The aggregator uses this algorithm to show that an aggregation of water heaters is well-suited to respond to these frequency disturbances by quickly adding a large amount of load to the grid.

Residential electricity consumption in the United States has many adverse impacts, such as greenhouse gas emissions, dependence on fossil fuels, and costs. Efficient and renewable energy technologies have the potential to help mitigate some of these impacts, but appear to be under-utilized in the United States. One major barrier to expanding the deployment of these kinds of technologies and maximizing the benefits they can provide is a lack of consumer engagement. The overall purpose of this thesis is to better understand the extent to which efficient and renewable energy technologies are being engaged with and what factors may influence such engagement (or lack thereof) through case studies on smart meters and a community anaerobic digester system (CADS) in Vermont. In this thesis, engagement involves awareness, support, and utilization. Additionally, a subset of awareness (a precursor to awareness for many) was examined in each of these studies, which is interest in receiving additional information on the technology. While each case study focuses on different aspects of engagement that are unique to each smart meters and CADS, there is some overlap on the topics explored, especially when it comes to awareness of the technology, potential concerns about the technology, and interest in receiving additional information on it. The focus of the first study is on how efficiently smart meters have been utilized by residential electricity customers in Vermont and what factors may influence this. This study was conducted via a statewide telephone survey in Vermont and involved a sample that was statistically representative of the state. These data were analyzed via quantitative analysis. The focus of the second study is on local support of a CADS in Vermont and what factors may influence this. This study was conducted via a mailout survey to houses located in or near the area where the community anaerobic digester was located, and the data were analyzed via quantitative and qualitative analysis. In both studies, limitations to engagement with the technologies were found. In the smart meter study, less than 50% of the surveyed customers reported having a smart meter and, for those who did report having a smart meter, less than 20% of them thought that the smart meter had reduced their electricity use. In the CADS study, 52.1% of respondents reported being familiar with the CADS project, and 69.8% reported support for the project. However, other forms of support for the project, such as WTP for the Cow Power program or willingness to drop of food scraps to the CADS, were more limited. Additionally, a variety of demographic and other factors were found to have a statistically significant impact on or relationship to consumer engagement with these technologies. Overall, the results show that there is some engagement with these technologies, but more can be done to bolster engagement with them. One potential strategy to increase engagement with these technologies may be to tailor outreach according to factors that correspond to different levels of engagement. It is hoped that the results from these studies can be used to help improve consumer engagement with these and other efficient and renewable energy technologies, thus hopefully expanding their utilization and benefits they can provide in the process.

Our energy system is rapidly transforming, partially due to advances in internet and communications technologies that leverage an unprecedented amount of data. Industry proponents of the so-called “smart grid” suggest these technologies facilitate deeper engagement with end-users of energy (utility customers) that can in turn drive behavior-based changes and accelerate a renewable energy transition. While there has been progress in understanding how these technologies change consumer behavior using, for example, real-time feedback, it’s unclear how specific segments (e.g., renters) respond to these interventions; it’s also unclear why feedback is, or is not, producing changes in energy consumption. The literature suggests that behavioral strategies (e.g. information feedback, competitions, incentives) coupled with technology may present a way for utilities and efficiency programs to create savings—expanding opportunities for those often underserved by traditional approaches, such as renters—yet this coupling is not well understood, neither broadly (for all end users) nor specifically (for renters). This dissertation builds upon that literature and explores a human side of the smart grid, using a field experiment in renter households to test the interacting effects of real-time energy feedback and a novel form of financial incentive, referred to here as a competitive performance-based incentive. The experiment had two phases: phase one tested the feedback against a control group; phase two tested feedback, the incentive, and a combined treatment, against a control group. Results of these interventions were measured with pre- and post-treatment surveys as well as observed electricity consumption data from each household’s smart meter. The results of this experiment are described in three papers. Paper one examines the interventions’ individual and combined effectiveness at motivating renters to reduce or shift timing of electricity consumption. Feedback alone produced a significant savings effect in phase one. In phase two, the effect of the feedback wore off; the incentive alone had no significant effect; and the group that received feedback and the incentive experienced a doubling of savings relative to the effect of feedback alone, as observed in phase one. Paper two uses pre- and post-intervention survey data to examine how individual perceptions of energy change as a result of the interventions. Perception of large energy-using appliances changed the most in households that received feedback, suggesting that better information may lead to more effective behavior changes. Paper three leverages the results of the first two components to evaluate the policy implications and impacts on demand side management for utilities, efficiency programs, and the potential for behavior-based energy efficiency programs. Advocates of the smart grid must recognize the technology alone cannot produce savings without better engagement of end-users. Utility rate designers must carefully consider how time-based rates alone may over-burden those without the enabling technology to understand the impact of their energy choices.

The transportation sector is a largest emitter of greenhouse gases in the U.S., accounting for 28.6% of all 2016 emissions, the majority of which come from the passenger vehicle fleet [1,2]. One major technology that is being investigated by researchers, planners, and policy makers to help lower the emissions from the transportation sector is the plug-in electric vehicle (PEV). The focus of this work is to investigate and model the impacts of increased levels of PEVs on the regional electric power grid and on the net change in CO2 emissions due to the decrease tailpipe emissions and the increase in electricity generation under current emissions caps. The study scope includes all of New England and New York state, modeled as one system of electricity supply and demand, which includes the estimated 2030 baseline demand and the cur- rent generation capacity plus increased renewable capacity to meet state Renewable Portfolio Standard targets for 2030. The models presented here include fully electric vehicles and plug-in hybrids, public charging infrastructure scenarios, hourly charging demand, solar and wind generation and capacity factors, and real-world travel derived from the 2016-2017 National Household Travel Survey. We make certain assumptions, informed by the literature, with the goal of creating a modeling methodology to improve the estimation of hourly PEV charging demand for input into regional electric sector dispatch models. The methodology included novel stochastic processes, considered seasonal and weekday versus weekend differences in travel, and did not force the PEV battery state-of-charge to be full at any specific time of day. The results support the need for public charging infrastructure, specifically at workplaces, with the “work” infrastructure scenario shifting more of the unmanaged charging demand to daylight hours when solar generation could be utilized. Workplace charging accounted for 40% of all non-home charging demand in the scenario where charging infrastructure was “universally” available. Under the increased renewable fuel portfolio, the reduction in average CO2 emissions ranged from 90 to 92% for the vehicles converted from ICEV to PEV. The total emissions reduced for 15% PEV penetration and universally available charging infrastructure was 5.85 million metric tons, 5.27% of system-wide emissions. The results support the premise of plug-in electric vehicles being an important strategy for the reduction of CO2 emissions in our study region. Future investigation into the extent of reductions possible with both the optimization of charging schedules through pricing or other mechanisms and the modeling of grid level energy storage is warranted. Additional model development should include a sensitivity analysis of the PEV charging demand model parameters, and better data on the charging behavior of PEV owners as they continue to penetrate the market at higher rates.

The surge of mobile devices, such as smartphones, and tables, demands additional capacity. On the other hand, Internet-of-Things (IoT) and smart grid, which connects numerous sensors, devices, and machines require ubiquitous connectivity and data security. Additionally, some use cases, such as automated manufacturing process, automated transportation, and smart grid, require latency as low as 1 ms, and reliability as high as 99.99\%. To enhance throughput and support massive connectivity, sharing of the unlicensed spectrum (3.5 GHz, 5GHz, and mmWave) is a potential solution. On the other hand, to address the latency, drastic changes in the network architecture is required. The fifth generation (5G) cellular networks will embrace the spectrum sharing and network architecture modifications to address the throughput enhancement, massive connectivity, and low latency. To utilize the unlicensed spectrum, we propose a fixed duty cycle based coexistence of LTE and WiFi, in which the duty cycle of LTE transmission can be adjusted based on the amount of data. In the second approach, a multi-arm bandit learning based coexistence of LTE and WiFi has been developed. The duty cycle of transmission and downlink power are adapted through the exploration and exploitation. This approach improves the aggregated capacity by 33\%, along with cell edge and energy efficiency enhancement. We also investigate the performance of LTE and ZigBee coexistence using smart grid as a scenario. In case of low latency, we summarize the existing works into three domains in the context of 5G networks: core, radio and caching networks. Along with this, fundamental constraints for achieving low latency are identified followed by a general overview of exemplary 5G networks. Besides that, a loop-free, low latency and local-decision based routing protocol is derived in the context of smart grid. This approach ensures low latency and reliable data communication for stationary devices. To address data security in wireless communication, we introduce a geo-location based data encryption, along with node authentication by k-nearest neighbor algorithm. In the second approach, node authentication by the support vector machine, along with public-private key management, is proposed. Both approaches ensure data security without increasing the packet overhead compared to the existing approaches.

The growing computational demand of scientific applications has greatly motivated the development of large-scale high-performance computing (HPC) systems in the past decade. To accommodate the increasing demand of applications, HPC systems have been going through dramatic architectural changes (e.g., introduction of many-core and multi-core systems, rapid growth of complex interconnection network for efficient communication between thousands of nodes), as well as significant increase in size (e.g., modern supercomputers consist of hundreds of thousands of nodes). With such changes in architecture and size, the energy consumption by these systems has increased significantly. With the advent of exascale supercomputers in the next few years, power consumption of the HPC systems will surely increase; some systems may even consume hundreds of megawatts of electricity. Demand response programs are designed to help the energy service providers to stabilize the power system by reducing the energy consumption of participating systems during the time periods of high demand power usage or temporary shortage in power supply. This dissertation focuses on developing energy-efficient demand-response models and algorithms to enable HPC system's demand response participation. In the first part, we present interconnection network models for performance prediction of large-scale HPC applications. They are based on interconnected topologies widely used in HPC systems: dragonfly, torus, and fat-tree. Our interconnect models are fully integrated with an implementation of message-passing interface (MPI) that can mimic most of its functions with packet-level accuracy. Extensive experiments show that our integrated models provide good accuracy for predicting the network behavior, while at the same time allowing for good parallel scaling performance. In the second part, we present an energy-efficient demand-response model to reduce HPC systems' energy consumption during demand response periods. We propose HPC job scheduling and resource provisioning schemes to enable HPC system's emergency demand response participation. In the final part, we propose an economic demand-response model to allow both HPC operator and HPC users to jointly reduce HPC system's energy cost. Our proposed model allows the participation of HPC systems in economic demand-response programs through a contract-based rewarding scheme that can incentivize HPC users to participate in demand response.

Wireless power transfer (WPT) technologies have become important for our everyday life. The most commonly used near-field WPT method is inductive coupling, which suffers from low efficiency and small range. The Strongly Coupled Magnetic Resonance (SCMR) method was developed recently, and it can be used to wirelessly transfer power with higher efficiency over a longer distance than the inductive coupling method. This dissertation develops new SCMR systems that have better performance compared to standard SCMR systems. Specifically, two new 3-D SCMR systems are designed to improve the angular misalignment sensitivity of WPT systems. Their power transfer efficiency for different angular misalignment positions are studied and analyzed. Prototypes are built for both systems and their performance is validated through measurement. Furthermore, new planar broadband conformal SCMR (CSCMR) systems are developed that maintain high efficiency while providing significantly larger bandwidth than standard CSCMR systems. Such broadband CSCMR systems are used here for the first time to simultaneously accomplish highly efficient wireless power transfer and high data rate communication through the same wireless link. These systems that combine wireless power and communication are expected to enable next-generation applications with battery-less and “power-hungry” sensors. Example applications include implantable and wearable sensors as well as embedded sensors for structural health monitoring.

With the growth of renewable energy usage and energy storage adoption in recent decades, people have started to reevaluate the possible roles of dc systems in current and future electrical systems. The dc voltage distribution has been applied in various applications, such as data centers and aircraft industry, for high efficiency and power density. However, for some applications such as subsea gas and oil fields, and ocean observatory systems, the dc current distribution is preferred over dc voltage distribution for its low cost and robustness against cable faults. Design and control of dc power distribution systems for different applications is an emerging research area with complex technical challenges. This dissertation solves the technical challenges in analysis, design, modeling, control and protection of series resonant converters (SRCs) for dc current distribution applications. An optimum design that has high efficiency, high reliability, and minimum required control efforts for the SRC with constant input current has been achieved and demonstrated by applying the analysis and design procedures developed in this dissertation. The modeling and analysis presented in this dissertation represents an operating condition that has not been studied in the literature and could be easily extended to other resonant converter topologies. Explicit analytical expressions have been provided for all key transfer functions, including input impedance and control-to-output, offering valuable resources to design feed-back regulation and to evaluate system stability. Based on the control strategies and control design presented in this dissertation, stable and reliable operation of dc current distribution systems with long distance cable has been achieved and demonstrated. The proposed analysis, design procedure, stability evaluation, control strategy and protection techniques in this dissertation can be applied to a wide range of similar scenarios as well, which greatly increases their value.