Optimal residential energy consumption, prediction, and analysis
dc.contributor.advisor | Webber, Michael E., 1971- | en |
dc.contributor.committeeMember | Blackhurst, Michael F | en |
dc.contributor.committeeMember | Edgar, Thomas F | en |
dc.contributor.committeeMember | King, Carey | en |
dc.contributor.committeeMember | Novoselac, Atila | en |
dc.creator | Rhodes, Joshua Daniel | en |
dc.date.accessioned | 2016-02-19T15:14:58Z | en |
dc.date.available | 2016-02-19T15:14:58Z | en |
dc.date.issued | 2014-05 | en |
dc.date.submitted | May 2014 | en |
dc.date.updated | 2016-02-19T15:14:58Z | en |
dc.description.abstract | In the United States, buildings are responsible for 40.36 Quads (40.36 x 10¹⁵ BTU) of total primary energy consumption per year, 22.15 of which are used in residential buildings (reference year 2010). Also, the United States residential sector is responsible for about 20% of United States carbon emissions or about 4% of the world's total. While there are over 130 million residential units in the United States, only 0.1% of R&D is spent in the residential sector. This means the residential sector represents an underinvested opportunity for energy savings. Tackling that problem, this dissertation presents work that is focused on assessing, analyzing, and optimizing how residential buildings use and generate energy. This work presents an analysis of a unique dataset of 4971 energy audits performed on homes in Austin, Texas in 2009 - 2010. The analysis quantifies the prevalence of typical air-conditioner design and installation issues such as low efficiency, oversizing, duct leakage, and low measured capacity, then estimates the impacts that resolving these issues would have on peak power demand and cooling energy consumption. It is estimated that air-conditioner use in single-family residences currently accounts for 17 - 18% of peak demand in Austin, and that improving equipment efficiency alone could save up to 205 MW, or 8%, of peak demand. It was also found that 31% of systems in this study were oversized, leading to up to 41 MW of excess peak demand. Replacing oversized systems with correctly sized higher efficiency units has the potential for further savings of up to 81 MW. Also, the mean system could achieve 18% and 20% in cooling energy savings by sealing duct leaks and servicing air-conditioning units to achieve 100% of nominal capacity, respectively. A different dataset of measured whole-home electricity consumption from 103 homes in Austin, TX was analyzed to 1) determine the shape of seasonally-resolved residential demand profiles, 2) determine the optimal number of normalized representative residential electricity use profiles within each season, and 3) draw correlations to the different profiles based on survey data from the occupants of the 103 homes. Within each season, homes with similar hourly electricity use patterns were clustered into groups using the k-means clustering algorithm. The number of groups within each season was determined by comparing 30 different optimal clustering criteria. Then probit regression was performed to determine if homeowner survey responses could serve as explanatory variables for the clustering results. This analysis found that Austin homes typically fall into one of two seasonal groups. Because these groups differ in temporal energy use and the wholesale electricity price is temporal, homes in one group use more expensive electricity than others. The probit regression results indicated that variables such as whether or not someone worked from home, the number of hours of television watched per week, and level of education have significant correlation with average profile shape, but that significant indicators of profile shape can vary across seasons. Also, these results point to markers of households that might be more impacted by time-of-use (TOU) or real time price (RTP) electricity rates and can act as predictors as to how changing local demographics can change local electricity demand patterns. This work also considers the effect of the placement (azimuth and tilt) of fixed solar PV systems on their total energy production, peak power production, and economic value given local solar radiation, weather, and electricity market prices and rate structures. This model was then used to calculate the output of solar PV systems across a range of azimuths and tilts to find the energetically and economically optimal placement. The result of this method, which concludes that the optimal placement can vary with a multitude of conditions, challenges the default due-south placement that is conventional for typical installations. For Austin, TX the optimal azimuth to maximize energy production is about 187 - 188°, or 7 - 8° west of south, while the optimal azimuth to maximize economic output based on the value of the solar energy produced is about 200 - 230° or 20 - 50° west of south. The differences between due south (which is the conventional orientation) and the optimal placement were on the order of 1 - 7%. For the rest of the US and for most locations the energetically optimal solar PV azimuth is within 10° of south. Considering the temporal value of the solar energy produced from spatially-resolved market conditions derived from local time-of-use rates, the optimal placement shifts to a west-of-south azimuth in attempt to align solar energy production with higher afternoon prices and higher grid stress times. There are some locations particularly on the west coast that have west-of-south energy optimal placements, possibly due to typical morning clouds or fog. These results have the potential to be significant for solar PV installations: utilities might alter rate structures to encourage solar generation that is more coincident with peak demand, utilities might also use west-of-south solar placements as a hedge against future wholesale electricity price volatility, building codes might encourage buildings to match roof azimuths with local optimal solar PV generation placements, and calculations of the true value of solar in optimal and non-optimal placements can be more accurate. This analysis also uses a dataset of whole home electricity consumption to consider the role of small distributed fuel cells in managing energy and thermal flows in the home. The analysis determines that the average home in Austin, TX could utilize a 5.5 kW fuel cell either for total generation or backup, and the average home could operate as its own micro-grid while not sacrificing core functionality. Matching the thermal output of a possibly smaller fuel cell, used in combined heat and power mode (CHP), to an absorption refrigeration system in place of traditional space cooling further reduces the needed capacity. Lastly, it is estimated that the system efficiency could possibly double by transporting natural gas to the end user to be converted into electricity and heat as compared with traditional methods of using natural gas for power generation followed by electricity delivery. Results from two regression analyses of total energy use and energy use reductions following energy efficiency retrofits are also presented. The first model shows that home size and age were positively correlated with total yearly energy use, but there is significant correlation of reduced yearly energy use with increased energy and water knowledge. This result might lend some support for increased energy and water education campaigns. The second model (retrofit analysis) also provided results that utilities can use to assess the value of residential retrofit rebates as compared to the cost of acquiring energy on the wholesale market. The second model indicates that the current level of rebates is cost effective for the utility (on a $ per kWh offset basis) for all three considered retrofits (air-sealing, attic insulation, and air-conditioner replacement) and the rebates could be increased while still being below the cost of acquiring electricity on the wholesale market. Considering an average of $0.113/kWh for residential electric service, both the air-sealing and increased attic insulation seem to make economic sense for the homeowner given current rebate structures. | en |
dc.description.department | Civil, Architectural, and Environmental Engineering | en |
dc.format.mimetype | application/pdf | en |
dc.identifier | doi:10.15781/T2DM5W | en |
dc.identifier.uri | http://hdl.handle.net/2152/33342 | en |
dc.language.iso | en | en |
dc.subject | Residential energy use | en |
dc.subject | Peak energy use | en |
dc.subject | HVAC | en |
dc.subject | Temporal energy use | en |
dc.title | Optimal residential energy consumption, prediction, and analysis | en |
dc.type | Thesis | en |
thesis.degree.department | Civil, Architectural, and Environmental Engineering | en |
thesis.degree.discipline | Civil Engineering | en |
thesis.degree.grantor | The University of Texas at Austin | en |
thesis.degree.level | Doctoral | en |
thesis.degree.name | Doctor of Philosophy | en |