Final Report: Contaminant Removal Using Membrane Distillation for Sustainable Drinking Water Treatment

EPA Grant Number: R835333
Title: Contaminant Removal Using Membrane Distillation for Sustainable Drinking Water Treatment
Investigators: Childress, Amy E , Kolodziej, Edward P. , Park, Chanwoo
Institution: University of Southern California
EPA Project Officer: Packard, Benjamin H
Project Period: September 1, 2013 through February 15, 2017
Project Amount: $499,743
RFA: Research and Demonstration of Innovative Drinking Water Treatment Technologies in Small Systems (2011) RFA Text |  Recipients Lists
Research Category: Drinking Water , Water


The main objectives of this investigation were to quantify the range of drinking water contaminants and contaminant classes that can be removed by membrane distillation (MD) and to develop and test a small-scale pilot MD system that operates using waste heat. In the proposed investigation, the research identified contaminants and contaminant groups of interest to small water systems as well as evaluated sites that have the best opportunities to harness waste energy for drinking water treatment. The research presented in this report provides evaluation of a relatively unproven technology, MD, for the removal of contaminants from drinking water sources for small systems using available waste heat sources.

Summary/Accomplishments (Outputs/Outcomes):

Bench-scale MD experiments were performed with a broad spectrum of contaminants and contaminant classes. The bench-scale system used an innovative design with gas-tight feed and distillate reservoirs. The unique design allowed detailed analyses of rejection and recovery of volatile and semi-volatile contaminants that would otherwise escape from a traditional open system. Contaminant classes tested included inorganic salts, metals, nitrogenated disinfection by-products (nitrosamines), organics, along with pharmaceuticals and personal care products.

As expected, rejection was nearly 100% for all non-volatile contaminants. Rejection of semi-volatile and volatile contaminants varied significantly, although a general trend of decreasing rejection with increasing volatility was observed. Evaluation of the nitrosamines revealed that a relatively stable rejection was observed over the last 24 hours of the experiment. The average rejection during this time was termed the “equilibrium rejection” and was found to correlate well with Henry’s Law constant. Evaluation of volatile and semi-volatile organic contaminants data identified a number of contaminants that did not follow the correlation between volatility and rejection, and it was found that more ionizable contaminants tended to deviate more than non-ionizable contaminants. Mass recovery of volatile and semi-volatile organics was observed to be inversely related to the compound’s hydrophobicity.

These results indicate that rejection of volatiles and semi-volatiles in batch systems can be considered an equilibrium partitioning process that requires consideration of equilibrium rejection and system pH instead of only considering a single endpoint rejection. Henry’s Law was shown to predict rejection for less-volatile compounds well, but it may provide inconsistent results for more volatile compounds or more ionizable compounds. The results of the bench-scale experiments provide insight into the mechanisms that predict contaminant rejection in MD and the validity of those predictions over different ranges of contaminant volatilities. The information has highlighted the performance capabilities of MD in treating waters contaminated with compounds from a wide range of contaminant classes. This information will allow water system engineers to assess whether waste-heat driven MD could be used to remove challenging contaminants and meet regulatory requirements and public health goals.

Several sites were investigated for providing an industrial waste heat source, leading to a preliminary pilot-scale MD system being installed in a campus boiler room. A finned-tube heat exchanger was installed in the side of the flue stack of the recreation center boilers using a flanged design, allowing waste heat to be transferred from the hot flue gas into the feed solution in the MD system. Temperature data for the flue gas showed a highly variable temperature over time. Despite significant waste heat source variability, a relatively low, but constant flux (maximum of 4 LMH) was maintained throughout the experiment.

Data from the preliminary pilot system resulted in a number of important insights into waste-heat-driven MD system design and operation. The water flux and flue gas waste heat source temperature profiles demonstrated the importance of waste heat variability in MD system design and operation. Challenges with heat exchanger corrosion highlighted a need for highly corrosion-resistant heat exchanger materials like titanium to be used to overcome problems with corrosion on the feed and flue-gas sides of the heat exchanger. The low flux observed during the preliminary pilot experiments indicated the importance of heat transfer from the feed solution through the membrane during operation, which should be considered during heat exchanger sizing. There are few instances of MD systems driven by waste heat published in the scientific literature, and those that do exist provide little detail regarding important design and operational factors like heat source variability, heat transfer through the membrane, and heat exchanger corrosion. The insights gained from the preliminary pilot system are a valuable addition to the MD scientific literature, leading researchers to more practical installations of waste-heat-driven MD systems. The insights gained from the preliminary pilot system testing also led to improved design of a community water testing pilot system.

EPA’s Safe Drinking Water Information System (SDWIS) was used to identify community water systems with current or possible future water quality compliance issues. Perchlorate was selected as a contaminant that is prevalent in small water systems across the US and has not been previously tested with MD. Perchlorate has a maximum contaminant level of 6 µg/L in California and was placed on EPA’s federal Contaminant Candidate List in 2009. The ability of MD to remove perchlorate from water samples has not been studied in the scientific literature. MD was expected to provide very high rejection of perchlorate because perchlorate is a non-volatile contaminant, and indeed, 100% rejection of perchlorate was found in testing. This indicates that MD is likely to be a useful technology for helping small water systems meet perchlorate regulatory requirements, particularly when a source of waste heat is available.

The pilot system also was used to investigate the many factors affecting waste-heat-driven MD performance, including: heat source variability, heat exchanger configuration, feed-water replenishment, and feed-tank blowdown. These different factors were considered with respect to their effects on performance indicators such as water flux and thermal stability. The unique test system used a liquid heat source with recirculating heated and cooled baths to simulate variable-temperature and constant-temperature heat sources. Experiments evaluating heat exchanger configuration were also performed. A direct open-loop heat exchanger configuration was shown to provide an average of 30.1% higher flux compared to an indirect open-loop heat exchanger configuration. The impacts of feed-water replenishment and feed-tank blowdown on thermal stability and water flux were shown to be minimal, while allowing for effective long-term control of feed solution concentration when using a recirculating system design


Insights regarding waste-heat-driven MD system design and operation gained from the pilot system testing are a significant contribution to the very limited literature that currently exists regarding waste-heat-driven MD systems. These insights will allow other researchers to design more practical MD systems that will be more easily integrated into new and existing small water systems, allowing them to more easily meet existing and expected future regulations for traditional and emerging contaminants.

Journal Articles on this Report : 2 Displayed | Download in RIS Format

Other project views: All 15 publications 2 publications in selected types All 2 journal articles
Type Citation Project Document Sources
Journal Article McGaughey AL, Gustafson RD, Childress AE. Effect of long-term operation on membrane surface characteristics and performance in membrane distillation. Journal of Membrane Science 2017;543:143-150. R835333 (Final)
  • Full-text: ScienceDirect-Full Text HTML
  • Abstract: ScienceDirect-Abstract
  • Other: ScienceDirect-Full Text PDF
  • Journal Article Rao G, Hiibel SR, Achilli A, Childress AE. Factors contributing to flux improvement in vacuum-enhanced direct contact membrane distillation. Desalination 2015;367:197-205. R835333 (2015)
    R835333 (Final)
  • Full-text: ScienceDirect-Full Text HTML
  • Abstract: ScienceDirect-Abstract
  • Other: ScienceDirect-Full Text PDF
  • Supplemental Keywords:

    Chemicals, particulates, metals, heavy metals, organics, dissolved solids, sustainable development, clean technologies, innovative technology, renewable, water treatment

    Progress and Final Reports:

    Original Abstract
  • 2013 Progress Report
  • 2014 Progress Report
  • 2015 Progress Report
  • 2016 Progress Report