The Effects of Temporality and Spatiality Due to Aged Data in LIfe Cycle Assessments
thesisposted on 22.05.2021, 17:51 by Jonathan Pryshlakivsky
Life cycle assessment is a relatively new—although decades old—method for assessing the environmental impacts of goods and services. It seeks to quantify these impacts in such a manner as to facilitate informed decisions regarding different, yet equally viable, options. However, this aim must be conditional on the notion that these impacts are measured with a number of associated qualifications or caveats, two of which is subject of this work. As subject matter, temporality and spatiality in life cycle assessment are both very broad, although this dissertation focuses specifically on temporality and spatiality due to age of data. The structure of the dissertation follows three distinct phases. The first phase contextualized the subject matter and its relation towards standardization of life cycle assessment methods. In doing so, it identifies and contextualizes the subject matter. The second phase identified Greenhouse gases, Regulatory Emissions, and Energy use in Transportation 2 as an ideal model on which to assess temporality and spatiality due to age of data since it models the life cycle assessment of an assortment of different vehicles. This phase also involved data collection, and uses a platform of assessment tools including Monte Carlo simulations, analysis of variance, F tests, regression analysis, and tests for non-normality (kurtosis and skewness). Building on the second phase, the third phase moved beyond the original phases by more than doubling the amount of materials of manufacture to be studied and adding further tools for assessment, the mainstay of which are regression analyses. Overall, this study found that the use of Monte Carlo simulations and analysis of variance are useful for identifying long term variation in energy intensity of materials. F-tests were useful in identifying which materials showed effects owing to spatiality. Although not in all instances, tests for non-normality identified which circumstances merit log transformation to bring about more accurate results. Linear regression techniques were used as a posterior test to confirm the origins of the variation seen in the Monte Carlo simulations and the analysis of variation. Moving ahead, this study pointed to the need for more concerted efforts in data promulgation.