Intelligent computational infrastructures for optimized autonomous distributed energy generation in remote communities
| dc.contributor.author | KRAJ, ANDREA | |
| dc.contributor.examiningcommittee | Kuhn, David (Mechanical Engineering) Thompson, Shirley (Natural Resources Institute) Dincer, Ibrahim (Automotive, Mechanical and Manufacturing Engineering, University of Ontario Institute of Technology) | en_US |
| dc.contributor.supervisor | Bibeau, Eric (Mechanical Engineering) Feitosa, Everaldo (Mechanical Engineering) | en_US |
| dc.date.accessioned | 2015-04-09T17:21:19Z | |
| dc.date.available | 2015-04-09T17:21:19Z | |
| dc.date.issued | 2015-04-09 | |
| dc.degree.discipline | Mechanical Engineering | en_US |
| dc.degree.level | Doctor of Philosophy (Ph.D.) | en_US |
| dc.description.abstract | Distributed generation along with smart grid applications are poised to make important contributions to the clean-tech sector and remote communities. The dependence on one source for energy supply does not prove reliable enough when the renewable resource, such as wind or solar, is variable, creating a dependence on external fuel supply and a vulnerability to foreign control. Developing an energy strategy through intelligent energy system simulation and optimization can help communities make informed decisions about their energy investments. This dissertation reasons that distributed renewable energy systems without operative computational infrastructures face a fundamental economic challenge derived from their ad-hoc design and implementation. To address this, it proposes the method of Optimal Operational Awareness (OOA)—a feedback mechanism on the state of, and changes in, the properties of the implemented subsystems and their behaviour, to meet users objectives. Despite many applications of hybrid renewable energy systems, and reputable multi-objective evolutionary algorithms (MOEAs) for optimization, no one has applied MOEAs to dynamic system operation for optimized engagement of system components. This thesis describes an application of the NSGA-II algorithm to the multi-objective optimization of the operation of a stand-alone wind-PV-biomass-diesel system with batteries and CAES storage and a central controller. The simultaneous objectives are to minimize the levelized cost of energy (LCOE), and unmet load (UL) while maximizing the renewable energy ratio (RER). This work provides a case-study evaluation from data collected on-site at the island of Fernando de Noronha (FDN), Brazil. The results show that FDN could move from an annual average of 33% RER and LCOE range of $0.26 - $0.36 per kWh to an increased RER range of 60% - 100% and LCOE of $0.10 - $0.50 per kWh, while maintaining UL of 0%, by increasing its renewable energy generation and storage capacity approximately five times. Furthermore, optimal operational awareness for this configuration shows that despite 100% RER, certain periods experience a high LCOE of $2.00 per kWh, resulting from energy spillage due to oversupply, indicating sub-optimal system sizing and wasted energy to trim by improving system configuration. This work concludes that it is possible to achieve 100% RER, but storage and/or backup diesel generation are important to include in systems with highly variable supply. The cost of electricity decreases as renewable energy penetration increases, but is configuration dependent as well dependent on storage state of charge. Oversizing storage can be just as costly, if not more costly, than supplying energy with diesel generation, thus proper sizing and dispatch strategy are critical to achieve economic electricity supply. Furthermore, the role of multiple renewable energy generators in providing autonomous supply can be more valuable to the user than increased supply cost. | en_US |
| dc.description.note | May 2015 | en_US |
| dc.identifier.uri | http://hdl.handle.net/1993/30370 | |
| dc.language.iso | eng | en_US |
| dc.rights | open access | en_US |
| dc.subject | renewable | en_US |
| dc.subject | energy | en_US |
| dc.subject | distributed | en_US |
| dc.subject | generation | en_US |
| dc.subject | autonomous | en_US |
| dc.subject | communities | en_US |
| dc.title | Intelligent computational infrastructures for optimized autonomous distributed energy generation in remote communities | en_US |
| dc.type | doctoral thesis | en_US |
| local.subject.manitoba | yes | en_US |