Document Type

Thesis - University Access Only

Award Date

2006

Degree Name

Master of Science (MS)

Department / School

Civil and Environmental Engineering

Abstract

The United States Environmental Protection Agency (USEPA) has set a Secondary Maximum Contaminant Level (SMCL) of 0.05 mg/L for manganese and 0.30 mg/L for iron to address aesthetic and operational problems. However, water utilities are receiving complaints from customers even when the distribution system iron and manganese concentrations are below the SMCL. With the conventional process for manganese removal, the manganese particles formed are so small that they can escape from the filter media, reach the distribution system and create aesthetic problems for customers. The objective of this study was to examine the effect of mixing on agglomeration of manganese when potassium permanganate is used as the oxidant prior to the filtration process. The study was conducted with mixing speeds of 0 rpm (without mixing), 30 rpm and 80 rpm on water containing nominal manganese concentrations of 0.1 0 mg/L, 0.50 mg/L and 1.00 mg/L. The mixing effect was examined at detention times of 5 minutes and 25 minutes and filter depths of 12 inches and 24 inches. The study was first conducted at the bench-scale level using jar tests. The bench scale tests showed differences in residual manganese concentration when mixing speeds were altered. Then the study was extended to continuous run trial tests, which yielded similar results to the jar tests. The study was then further conducted at the pilot scale level with addition of a granular media filter and continuous operation to mimic a full scale treatment plant. The pilot scale tests yielded mixed results regarding the effect of mixing on removal of manganese. In many cases, the statistical analysis of results showed the effect was not significant in terms of manganese removal, indicating alteration of mixing speed has no impact on concentration of residual manganese. However, the effect of mixing was shown to significantly impact head loss through the filter media. The full scale test results revealed the head loss across the filter accumulated faster when no mixing was provided as compared to when mixing was provided. Thus, mixing (flocculation) prior to filtration could provide longer filter runs as compared to no mixing. Also the effect of mixing was more visible with higher initial concentrations of manganese as compared to lower initial concentrations indicating that a lower initial concentration of manganese is relatively difficult to remove as compared to a higher initial concentration.

Library of Congress Subject Headings

Water -- Purification -- Manganese removal

Water -- Purification -- Filtration

Water -- Purification -- Oxidation

Format

application/pdf

Number of Pages

166

Publisher

South Dakota State University

Download

Share

COinS