Life Cycle Assessment of Solar Photovoltaic Microgrid Systems in Off-Grid Communities.

Access to a reliable source of electricity creates significant benefits for developing communities. Smaller versions of electricity grids, known as microgrids, have been developed as a solution to energy access problems. Using attributional life cycle assessment, this project evaluates the environmental and energy impacts of three photovoltiac (PV) microgrids compared to other energy options for a model village in Kenya. When normalized per kilowatt hour of electricity consumed, PV microgrids, particularly PV-battery systems, have lower impacts than other energy access solutions in climate change, particulate matter, photochemical oxidants, and terrestrial acidification. When compared to small-scale diesel generators, PV-battery systems save 94-99% in the above categories. When compared to the marginal electricity grid in Kenya, PV-battery systems save 80-88%. Contribution analysis suggests that electricity and primary metal use during component, particularly battery, manufacturing are the largest contributors to overall PV-battery microgrid impacts. Accordingly, additional savings could be seen from changing battery manufacturing location and ensuring end of life recycling. Overall, this project highlights the potential for PV microgrids to be feasible, adaptable, long-term energy access solutions, with health and environmental advantages compared to traditional electrification options.

Evaluation of a risk-based environmental hot spot delineation algorithm.

Following remedial investigations of hazardous waste sites, remedial strategies may be developed that target the removal of "hot spots," localized areas of elevated contamination. For a given exposure area, a hot spot may be defined as a sub-area that causes risks for the whole exposure area to be unacceptable. The converse of this statement may also apply: when a hot spot is removed from within an exposure area, risks for the exposure area may drop below unacceptable thresholds. The latter is the motivation for a risk-based approach to hot spot delineation, which was evaluated using Monte Carlo simulation. Random samples taken from a virtual site ("true site") were used to create an interpolated site. The latter was gridded and concentrations from the center of each grid box were used to calculate 95% upper confidence limits on the mean site contaminant concentration and corresponding hazard quotients for a potential receptor. Grid cells with the highest concentrations were removed and hazard quotients were recalculated until the site hazard quotient dropped below the threshold of 1. The grid cells removed in this way define the spatial extent of the hot spot. For each of the 100,000 Monte Carlo iterations, the delineated hot spot was compared to the hot spot in the "true site." On average, the algorithm was able to delineate hot spots that were collocated with and equal to or greater in size than the "true hot spot." When delineated hot spots were mapped onto the "true site," setting contaminant concentrations in the mapped area to zero, the hazard quotients for these "remediated true sites" were on average within 5% of the acceptable threshold of 1.

Evaluation of statistical methods for left-censored environmental data with nonuniform detection limits.

Monte Carlo simulations were used to evaluate statistical methods for estimating 95% upper confidence limits of mean constituent concentrations for left-censored data with nonuniform detection limits. Two primary scenarios were evaluated: data sets with 15 to 50% nondetected samples and data sets with 51 to 80% nondetected samples. Sample size and the percentage of nondetected samples were allowed to vary randomly to generate a variety of left-censored data sets. All statistical methods were evaluated for efficacy by comparing the 95% upper confidence limits for the left-censored data with the 95% upper confidence limits for the noncensored data and by determining percent coverage of the true mean (micro). For data sets with 15 to 50% nondetected samples, the trimmed mean, Winsorization, Aitchison's, and log-probit regression methods were evaluated. The log-probit regression was the only method that yielded sufficient coverage (99-100%) of micro, as well as a high correlation coefficient (r2 = 0.99) and small average percent residuals (-0.1%) between upper confidence limits for censored versus noncensored data sets. For data sets with 51 to 80% nondetected samples, a bounding method was effective (r2 = 0.96 - 0.99, average residual = -5% to -7%, 95-98% coverage of micro), except when applied to distributions with low coefficients of variation (standard deviation/micro < 0.5). Thus, the following recommendations are supported by this research: data sets with 15 to 50% nondetected samples--log-probit regression method and use of Chebyshev theorem to estimate 95% upper confidence limits; data sets with 51 to 80% nondetected samples-bounding method and use of Chebyshev theorem to estimate 95% upper confidence limits.