Development of a life-cycle model using possibilistic uncertainty propagation and compromise programming for the evaluation of alternative motor vehicle fuels

• Main Author: Tan, Raymond Girard R.
• Format: text
• Published: Animo Repository 2002
• Online Access: https://animorepository.dlsu.edu.ph/etd_doctoral/1358
• Institution: De La Salle University

The emergence of life cycle assessment (LCA) in the past decade has led to its extensive use as an environmental decision support tool in various contexts. Two areas for improvement in LCA methodology were identified in this work: modeling of data uncertainty, and decision analysis in the presence of multiple environmental performance criteria. Possibility theory was applied in this study to address both of these procedural issues in LCA. The methods developed were implemented in an embryonic software code to compare life cycles of eight different fuels and energy carriers for the Philippine automotive transport sector: electricity, liquid hydrogen, gaseous hydrogen, bioethanol, biodiesel, liquified natural gas, compressed natural gas and methanol. Conventional gasoline and diesel were also modeled to provide baseline scenarios.The composite model has the capability to perform full LCA, in contrast to many existing models that are limited to simply accounting for material and energy flows. Impact classification, characterization and valuation can be performed in the software code in addition to comprehensive uncertainty assessment. Its main features, possibilistic uncertainty propagation (PUP) and possibilistic compromise programming (PCP) capabilities, allow multiple-criteria analysis and ranking of the ten alternative fuels even in the presence of data uncertainty. Comprehensive sensitivity analysis with respect to electrical power generation mix, coproduct allocation method, and environmental impact valuation can also be performed in the model. The GREET and EDIP models are used as the inventory and impact analysis modules, respectively. The embryonic software is coded in Microsoft Excel and Visual Basic. Implementation of uncertainty analysis techniques in the model allows confidence levels to be quantified in the model outputs in a transparent manner. This feature minimizes the risk of misinterpretation of spurious or illusory results, which is a typical occurrence in conventional LCAs. The model represents a significant improvement of LCA methodology.Outputs specific to the model's decision domain indicate that the long-term technologies of electrolytic hydrogen and electricity are the environmentally oprtimal options, particularly if grid power is derived from clean energy sources. In the interim, potential environmental gains can also be realized by substituting biofuels and natural gas derivatives for petroleum-based fuels in internal-combustion engines.