Date of Award

2010

Degree Type

Report

Degree Name

Master of Science (MS)

Department

Mathematics and Statistics

First Advisor

Richard Cutler

Second Advisor

Ronald C. Sims

Third Advisor

Daniel C. Coster

Abstract

The Logan city wastewater treatment system consists of a series of seven large aerated ponds (460 acres) that biologically treats 15 million gallons per day of wastewater from Logan city and six other communities. Tighter regulations of allowed phosphorus levels in the effluent have recently been implemented due to environmental concerns of a downstream reservoir. The Biological Engineering program at Utah State University, the Bio-fuels Center, the Utah Water Research Laboratory (UWRL) and the city of Logan are working together to remediate the wastewater treatment system using microalgae. Algal growth requires the uptake of phosphorus. Thus, phosphorus in the effluent can be removed by encouraging algal growth and then removing the algae from the system. The harvested algae can then be used by the Bio-fuels Center to conduct research related to algal bio-fuels. The work reported here concerns the construction and analysis of experiments that were used by the Bio-fuels Center and UWRL to study growth and harvest methods for algae, and the design of a survey of the wastewater ponds to spatially characterize concentrations of algae. The experiments described herein demonstrate that, by encouraging algal growth in the ponds, phosphorous may be removed from the system and effective methods for growing and harvesting algae are proposed. Because a manuallysampled survey of the ponds is expensive and requires a great deal of resources, an alternative approach using aerial image data of the ponds was used to predict the manually sampled data of algae concentrations. Predicting algal concentrations using image data would mitigate the cost required of determining a location to harvest. Two prediction methods - linear models and random forests - are investigated and compared. The methods are compared on the basis of the amount of variability in the data they are able to explain. Random forests outperform linear models in predictive ability when using only the image data for prediction. The results of this project provide a proof of concept for the Bio-fuel Center, UWRL and Logan city initiative, that phosphorus can be removed from the effluent of the Logan lagoons by algae that can be grown and harvested efficiently.

Share

COinS