Search

The 2008 Climate Change Act set legally-binding carbon reduction targets. Demand side management (DSM) includes energy use reduction and peak shaving and offers significant potential to reduce the amount of carbon used by the electricity grid. The demand side management (DSM) schemes that have tried to meet this challenge have been dominated by engineering-based approaches and so favour tools like automation (which aims to make shifting invisible) and pricing (which requires customer response) to shift demand. These approaches tend to focus on the tools for change and take little account of people and energy-use practices. This thesis argues that these approaches are limited and therefore unlikely to produce the level of response that will be needed in future. The thesis therefore investigates the potential for time-shifting domestic energy demand but takes a different angle by trying to understand how people use energy in their daily lives, whether this use can be shifted and some of the implications of shifting it. The centrepiece of the work is an empirical study of eleven households energy-use practices. The interdisciplinary methodology involved in-house observations, interviews, photographs, metered energy data and disruptive interventions. The data was collected in two phases. Initially, a twenty-four hour observation was carried out in each household to find out how energy was implicated in everyday practices. Next, a series of three challenges were carried out, aimed at assessing the implications of disrupting practices by time-shifting food preparation, laundry and work/ leisure. A practice theory approach is used to shift the focus of attention from appliances, tools for change, behaviour or even people, to practices. The central finding of this work is that practices were flexible. This finding is nuanced, in the light of the empirical research, by an extended discussion on the nature of practices; in particular, the relationship between practices and agency and the temporal-spatial locatedness of practices. The findings demonstrate that, in this study at least, expanding the range of demand response options was possible. The research suggests numerous possibilities for extending the potential of practices to shift in time and space, shift the energy used in practices or substitute practices for other non-energy-using practices, though there are no simple technological or behavioural fixes . More profoundly, however, the thesis concludes that infrastructures of provision , such as the electricity grid and the companies that run it, underpin and facilitate energy-use practices irrespective of the time of day and year. In this context technology-led demand response schemes may ultimately contribute to the problem they purport to solve. A more fundamental interrogation of demand and the infrastructures that serve it is therefore necessary and is almost entirely absent from the demand response debate.

Electric vehicles (EVs) are becoming a more prominent portion of Texas’s light duty vehicle (LDV) fleet as they become more attractive to consumers and as relevant governing bodies work to incentivize further adoption rates in an effort to reduce emissions within the transportation industry. Meanwhile, the Electric Reliability Council of Texas (ERCOT), the electric grid that services about 90% of the state’s residents, has been seeing increases in power demand. This has been due to factors such as a growing population, increased air conditioning use, and pushes for electrification across other industries, all factors that are expected to continue contributing to power demand increases within the foreseeable future. As more vehicles in the state are electrified, they will add further power demand increases on top of the existing contributing factors. This work focuses on evaluating different EV adoption, charging management, and policy scenarios in order to then evaluate how they may be expected to impact ERCOT, particularly regarding peak demand increase within two time horizons: one into 2030 and one into 2050. Peak demand is an important consideration because as it increases, it presents challenges for maintaining the electrical grid’s reliability when it exceeds generation capacity. After constructing and refining models which predict EV presence at the county level based on socio-economic and infrastructure related feature variables, the anticipated impacts of a growing EV fleet are quantified using historical data from ERCOT, planned installations and reserve margins, EV charging patterns, and travel patterns. Additionally, this work includes results from a collaboration which culminated in applying relevant EV fleet growth predictions to a DC-OPF simulation for Austin Energy, serving as a case study for the future of EV fleet impacts on relatively small-scale utilities. The results of this study showcase the fact that it is possible to accurately predict EV presence at the county level with 6 publicly available feature variables. In examining adoption pathways, it importantly finds that current incentives will very likely be insufficient for the achievement of the Biden Administration’s 50% market share goal by 2030. Should this market share goal come to fruition, however, it is expected to be manageable at the state level and within the Austin Energy case study regarding electricity supply in 2030. Sustained growth from this scenario, however, will necessitate ambitious charging management strategies in order to limit the potentially heavy impact of a growing EV fleet on peak demand looking forward into and beyond 2050.

Zero energy building targets offer a pathway for significant potential emissions and cost reduction within the built environment. However, achieving zero energy building (ZEB) targets for existing office buildings can be challenging, given limitations on potential retrofits and on-site renewable energy generation opportunities. While retrofits can improve the performance of existing buildings, it is now certain that climate change will lead to more severe and frequent extreme weather events, and its impact on the optimal selection of retrofits to future-proof building’s improvements remains unexplored. This research investigates using extreme weather datasets in HVAC building simulation to test traditional HVAC system sizing methods under extreme conditions, quantify the impact of climate change on energy use, peak demand, and thermal comfort, for a typical commercial building in the four most-populated climate types. Viable retrofit selection strategies for the building envelope, on-site equipment and generation, and HVAC refurbishments are reviewed against purchasing off-site renewable electricity, to determine the viability of retrofit-only ZEBs across five building performance levels (different NABERs star ratings) in six key Australian climates under current and future conditions. The results show the extreme weather datasets have higher variability and increase peak cooling demand (35%) and unmet cooling hours (189%). A methodology including extreme hot and cold weather and typical conditions and datasets is proposed to future-proof HVAC system design against the impacts of climate change. To overcome increased energy demands because of climate change, a marginal abatement approach to retrofit selection highlights that some retrofits (HVAC refurbishment, rooftop solar, more efficient lighting and office equipment) are cost-effective compared to Renewable Energy Certificate (REC) use only and provide greater marginal abatement. A combined review of retrofit selection under existing and future conditions in thirty energy-performance/climate scenarios showed office buildings were able to achieve a 32% average energy reduction. While building energy retrofitting for commercial office buildings are unlikely to make ZEB targets achievable for most buildings, they can drive significant energy reduction (13-45%) and reduce the need for REC use.

New York State launched Master Plan in 2019 and aimed to reach the 9 GW offshore wind goal by 2035. Currently, four offshore wind farms are under planning with a total capacity of ~4.2 GW. We simulated the four offshore wind projects by using wind farm parameterization (WFP) from the weather research and forecasting (WRF) model and evaluated their performances in ten-day periods of each summer and winter peak from 2016 to 2020. The mean generation of ~1.4 GW in summer and ~2.1 GW in winter can theoretically fulfill ~17.7% of the summer peak demand and ~35.9% of the winter peak demand. Meanwhile, a strong diurnal pattern in summer shows that very limited generation in the daytime and noticeably higher generation in the nighttime is observed. The simulation results provide significant implications for power grid operating management and regulations.

As the world transitions from fossil to renewable sources of electricity, managementof the electrical grid is becoming increasingly complex. Where once generation flowedunidirectionally from large spinning generation to variable loads, in the new worldgeneration is often also stochastic and power flows in multiple directions, includingfrom sources deep within the distribution system. This necessitates changes to themanner in which the grid is controlled.This dissertation explores the possibilities of transactive energy as a means of dis-tributing control throughout this complex system. A transactive energy testbed hasbeen created on the University of Toledo campus which is used to study various con-trol strategies, both transactive and non-transactive, which are useful for integratingdistributed energy resources and bringing demand-side assets such as buildings intothe control equation. Several experiments with non-transactive control strategies forbuildings and DER, which have been performed on this testbed, are discussed. Theseinclude experiments with peak-demand management, mitigation of variability fromphotovoltaic generation using a battery energy storage system, and load followingapplications.The university is now building a transactive energy network which will allowtesting of transactive control systems. Before these experiments take place, it isnecessary to understand which use cases which can best be deployed, which will beof the greatest value to participants, and how these can be evaluated for success.The core of this work is a valuation study of transactive energy use cases whichdevelops metrics for operation and evaluation of their performance in both technicaland economic terms.Additionally, one transactive supply-side use case is discussed which seeks to builda single transactive actor from the combination of a nuclear power facility with on-siterenewable generation and hydrogen electrolysis.

The addition of solar panels to forecasting energy demand and peak energy demand presents an entirely new challenge to a facility. By having to account for the varying energy generation from the solar panels on any given day based on the weather it becomes increasingly difficult to accurately predict energy demand. With renewable energy sources becoming more prevalent, new methods to track peak energy demand are needed to account for the energy provided by renewable sources. We know from previous research that Artificial Neural Networks (ANN) and Auto Regressive Integrated Moving Average (ARIMA) models are both capable of accurately forecasting building demand and peak electric load days without the presence of solar panels. The goal of this research was to take three different approaches for both the ANN model and the ARIMA model to find the most accurate method for forecasting monthly energy demand and peak load days while considering the varying daily solar energy production. The first approach used was to forecast net demand outright based on relevant historical training data including weather information that would help the models learn how this information affected the overall net demand. The second approach was to forecast the building demand specifically based on the same relevant historical data and then use a random decision tree forest to predict the cluster of day that each day of the month would be in terms of solar production (high, medium with early peak, medium with late peak, low). After the type of day was predicted we would subtract the average solar energy production of the predicted cluster to receive our forecasted net demand for that day. The third approach was similar to the second, but instead of subtracting the average of the cluster we subtracted multiple randomly generated days from that cluster to provide multiple overlapping forecasts. This was specifically used to try and better predict peak load days by testing the hypothesis that if 80% or higher predicted a peak day it would in fact be a peak day. The ANN model outperformed the ARIMA for each approach. Forecasting multiple days was the best of the three approaches. The multiple day ANN forecast had the highest balanced accuracy and sensitivity, the net demand ANN approach was the 2nd most accurate approach and the average solar ANN forecast was the 3rd best approach in terms of balanced accuracy and sensitivity. Based on the outcomes of this study, consumers and institutions such as RIT will be better able to predict peak usage days and use preventative measures to save money by reducing their energy intake on those predicted days. Another benefit will be that energy distribution companies will be able to accurately predict the amount of energy customers with personal solar panels will need in addition to the solar energy they are using. This will allow a greater level of reliability from the providers. Being able to accurately forecast energy demand with the presence of solar energy is going to be critical with the ever-increasing usage of renewable energy.

Supply-demand mismatch and demand peaks jeopardize the reliability of power systems and incur severe implications such as huge network investment, increased generation capacity and electricity cost to customers. The increased penetration of renewable energy in future will reduce the carbon footprint, however, their reduced inertia and variable output will make the power system vulnerable to disturbance, further exacerbating the problem by reducing controllability of generators. In the past, demand side management has shown great potential in reducing power mismatch as well as demand peaks. The potential of demand response in peak clipping, valley filling and reducing price volatility is already explored in the literature. Residential appliances such as refrigerators, air-conditioners, hot water systems (EWHs) have large thermal inertia, and therefore represent energy storage opportunities showing great demand response potential. However, increased penetration of renewables as well as high energy consuming residential appliances such as aircons and EWHs bring new challenges to demand response making it meaningful to discover novel demand control strategies. The thesis presents a detailed study of demand control approaches and peak demand drivers as well as their limitations. These are addressed by undertaking a study of centralized and decentralized control system topologies, modelling power network setup with conventional diesel and wind generation, modelling refrigerators as control loads for power balancing, studying large scale residential appliance level data from Australia’s largest network provider, Ausgrid and using the data to investigate major demand drivers in summer and winter demand peaks as well as studying demand response opportunities with them. Contributions of the thesis include introducing a novel hybrid control topology for supply-demand balancing by using domestic refrigerators as control loads in a hybrid power network setup, in addition to analysing Smart Grid Smart City data to study the actual contribution of aircons and other residential appliances. Clustered operational models combined with aircon load control strategies are used to study possible peak load reductions from aircons at the New South Wales state level. Also, winter and summer peak data are studied for many zone substations (ZS) and demand response opportunities with potential controllable loads at the ZS are investigated.

This work considers how various integrated energy systems can be managed in order to provide economic or energetic benefits. Energy systems can gain additional degrees of freedom by incorporating some form of energy storage (in this work, thermal energy storage), and the increasing penetration of smart grid technologies provides a wealth of data for both modeling and management. Data used for the system models here come primarily from the Pecan Street Smart Grid Demonstration Project in Austin, Texas, USA. Other data are from the Austin Energy Mueller Energy Center and the University of Texas Hal C. Weaver combined heat and power plant. Systems considered in this work include thermal energy storage, chiller plants, combined heat and power plants, turbine inlet cooling, residential air conditioning, and solar photovoltaics. These systems are modeled and controlled in integrated environments in order to provide system benefits. In a district cooling system with thermal energy storage, combined heat and power, and turbine inlet cooling, model-based optimization strategies are able to reduce peak demand and decrease cooling electricity costs by 79%. Smart grid data are employed to consider a system of 900 residential homes in Austin. In order to make the system model tractable for a model predictive controller, a reduced-order home modeling strategy is developed that maps thermostat set points to air conditioner electricity consumption. When the model predictive controller is developed for the system, the system is able to reduce total peak demand by 9%. Further work with the model of 900 residential homes presents a modified dual formulation for determining the optimal prices that produce a desired result in the residential homes. By using the modified dual formulation, it is found that the optimal pricing strategy for peak demand reduction is a critical peak pricing rate structure, and that those prices can be used in place of centralized control strategies to achieve peak reduction goals.

The peaking factor (PF) is the ratio of the maximum flow during some specified time interval to the average annual flow. The PF concept has been used in both the drinking water and wastewater industries for nearly a century. Since there is uncertainty in the magnitudes of the maximum flow and the average annual flow, both can be treated as random variables and the corresponding PF should be interpreted in a probabilistic sense. Conventional methods for estimating the PF are based on empirical deterministic expressions, derived from a combination of field measurements and engineering judgment. While existing PF expressions have performed well, they lack a firm theoretical basis. The objective of this study is to combine results from the Poisson Rectangular Pulse (PRP) model for residential water use with principles from extreme value frequency analysis to develop a theoretical reliability-based estimate of the PF. Results show the theoretical PFs agree quite well with those from conventional empirical methods. The well-documented decrease in PF with network size is readily explained. The incorporation of indoor and outdoor water use and leakage makes its general application possible. The new PRP-based theoretical results are significant because they provide a direct way to estimate the PF in instances where flow data are scarce or where service area conditions may be changing.