Ordering Information:
ORDER NUMBER:  91208
DATE AVAILABLE: Spring 2008

Printed Report		PDF
Subscribers	Order Report
Non-Subscriber	Coming Soon	N/A

PRINCIPAL INVESTIGATORS:
Jim Lozier, Lisa Cappucci, Gary Amy, NoHwa Lee, Joe Jacangelo, Haiou Huang, Thayer Young, Jean-Philippe Croué, Chandra Mysore, Christophe Cemeraux, Jeremie Clouet, and Bas Heijmann

OBJECTIVES:
The overall goal of this project was to investigate the specific contributions of the different types of natural organic matter (NOM) to microfiltration/ultrafiltration (MF/UF) fouling. The intent was to develop a surrogate test or index that could be used to predict NOM fouling at low cost through a combination of source water characterization and rapid bench-scale testing.

BACKGROUND:
Low-pressure (LP) membrane use has increased dramatically over the last decade in response to more stringent pathogen-related drinking water regulations, water reclamation and the need for more effective reverse osmosis pretreatment, and from dramatically reduced membrane costs. More cost-effective and reliable operation of LP membrane systems is constrained, however, by fouling, in particular fouling by NOM. NOM fouling is poorly understood because of both the complexity and types of NOM that exists in natural sources and wastewater effluent and NOM-membrane interactions. NOM exists in three primary forms (allochthonous, autochthonous, and effluent-derived), with a variety of components having differing fouling tendencies. LP membranes comprise hollow fibers of differing polymeric materials, with a range of properties that likewise influence fouling propensity. Fouling management strategies (backwash, air scrub, chemical cleaning) employed with LP membrane systems differ from supplier to supplier. This, combined with a number of the methods used to reduce NOM levels prior to membrane treatment (e.g., coagulation, clarification), further complicate the understanding of NOM fouling.

HIGHLIGHTS:

1. The high molecular weight, polysaccharide (PS) fraction of NOM was found to be the major contributor to fouling of the low pressure membranes evaluated in this study, with the humic substances fraction of lesser importance.
2. A quantitative fouling index, called the unified modified fouling index (UMFI), was developed that can be used to measure the fouling potential of a source water rapidly at relatively low cost using short-term, batch-type bench tests.
3. The fouling response of different commercial hollow fiber polyvinylidene fluoride (PVDF) membranes was quite similar, despite differences in certain membrane properties (pore size and surface roughness). In contrast, fouling response varied significantly for the polyethersulphone (PES) membranes, with the difference most likely related to the presence or absence of a hydrophylizing agent (polyvinyl pyrrolidene [PVP]).
4. Fouling management strategies commonly employed with full-scale LP membrane systems, including chemical washing and coagulation, are effective in reducing NOM fouling.

APPROACH:
The research incorporated bench-, pilot- and full-scale investigations. Testing was conducted with four source waters, selected to capture the fouling characteristics of the three primary types of NOM. Bench testing use a stirred, cell apparatus and three flat sheet membrane types, representing commercially dominant MF and UF hollow fiber membranes of the same polymer types. Hollow fiber bench testing used two PVDF and two PES membranes operated in both sequential and alternating filtration/backwash mode. Pilot testing included PVDF MF and UF and PES UF systems operated on three of the four source waters and incorporated a host of fouling management strategies. Full-scale investigations captured operating data from several plants having differing levels and types
of NOM.

RESULTS/FINDINGS:
The high molecular weight, PS fraction of NOM was found to be the major contributor to fouling of the low pressure membranes evaluated in this study. The humic substances fraction, historically considered to be the primary NOM fouling component, was shown to be of secondary importance, except in cases where this fraction was present at very high levels, or where high levels of calcium were present in the source water. The PS fraction was amenable to removal from the membrane surface by backwashing.

This research led to the development of a quantitative fouling index, called the unified modified fouling index or UMFI, that can be used to measure the fouling potential of a source water rapidly at relatively low cost. The research showed that measurement of the short-term, hydraulically reverse fouling rate (e.g., UMFI₁₅₀) using a bench-scale apparatus employing hollow-fiber membranes can predict the longer-term fouling rate of similar membranes observed at pilot scale with reasonable accuracy. The fouling response of different PVDF membranes was quite similar, despite differences in certain membrane properties (pore size and surface roughness). In contrast, fouling response varied significantly for the PES membranes, with the difference most likely related to the presence or absence of a hydrophylizing agent (PVP). Flat sheet membranes used in stirred cell tests proved to be less than satisfactory as a means to predict fouling of commercial hollow fiber MF and UF membranes. Commonly-employed fouling management techniques, including caustic and chlorine-based chemical washing, as well as coagulation with both ferric and aluminum salts, proved to be of significant value in reducing NOM fouling.

IMPACT:
Water utilities using LP membranes should focus on minimizing the level of high molecular weight, PS type NOM in the feedwater through source selection, control of algal activity in the source water, and use of coagulation or chemical clarification prior to membrane use to reduce PS levels. Utilities investigating or planning for the use of LP membranes should consider employing bench-scale testing and UMFI measurements to more effectively assess the treatability and cost of membranes for their application and to screen commercially available membrane products, pretreatment alternatives, and the effectiveness of chemically-based fouling management strategies.

RESEARCH PARTNER:
U.S. Environmental Protection Agency

PARTICIPANTS:
Utilities from the United States, Canada, and the Netherlands participated in this project.