Membrane system technology has been developed to a point where one or more of its forms can be applied to extract most of the contaminants of concern found in many raw water sources. Recognizing the importance of the latest findings to the drinking water supply community, the International Water Supply Association recently conducted a conference on the subject in Paris, France. The IWSA's choice of this beautiful capital city for the site of the meeting appears to have been a logical one, for if the French are not the outright leaders in applying membranes to water treatment (and to certain wastewater processes too), then certainly they are in the vanguard.
Entitled "International Workshop--Membranes in Drinking Water Production," the gathering was attended by almost 300 experts from 25 countries around the world. As one would expect, the host country registered the largest number of delegates. But strong representation also was evident from Holland, The United Kingdom, and the United States.
Reverse osmosis and electrodialysis, both membrane-based, have been applied for about 40 years in desalination systems to produce drinking water from sea water and brackish water sources. But in the last decade, membrane technology has been developed to new levels of capability. For instance, microfiltration and ultrafiltration have been found suitable for particulate removal. Reverse osmosis and nanofiltration are being applied in surface water treatment for softening and polishing steps. And specific contaminants sometimes can be removed efficiently with advanced types of electrodialysis systems.
The conference agenda was designed to review the latest information on the application of membranes, and to ask some important questions about their potential, and their possible problems, as practical components of water treatment facilities in the future. Six consecutive technical sessions over two and a half days addressed the following topics.
- Removal of specific inorganic contaminants by membrane processes.
- Characteristics of and disposal methods for the concentrates.
- Determining the best membrane system for removal of natural organics and micropollutants.
- Efficiency and reliability of membranes as physical disinfectant devices.
- Membrane fouling problems and their solutions.
- Scaling up for application in larger water treatment plants.
Dissolved Inorganics Removal
Introduced as the "godfather" of membrane technology, and known to many of the experts in the field, Professor James Taylor of the University of Central Florida in Orlando, opened the conference with a discussion of reverse osmosis (RO) and nanofiltration (NF) specifically. But in the course of his presentation he offered a concise review of the five membrane systems, including RO and NF, which have promising application potential for drinking water treatment. The others are ultrafiltration (UF), microfiltration (MF), and electrodialysis reversal (EDR). Table 1 on page 30 illustrates in simple form what each of the five can accomplish in terms of their ability to retain contaminants in different size ranges and with certain characteristics. The table categorizes regulated drinking water solutes as pathogens, organics and inorganics. Associating minimum size of solute rejection (retention) with membrane type and contaminants to be regulated gives a good indication of which application is appropriate for a drinking water treatment installation.
As Professor Taylor explained, the regulatory requirements of the Safe Drinking Water Act Amendments of 1986 have accelerated research, development, and application interest in membrane technology. Similarly, drinking water standards in Europe have become more demanding, and just as in the States, some regulatory decision making is being driven by vastly improved water quality analysis capabilities, health effects research, and membrane manufacturing and materials know-how. What this adds up to is a need in the U.S. and Europe for the development and application of better methods for reducing organic contamination of drinking water. Taylor went on to explain how membranes can be used for total removal of pathogens, and high levels of inorganic and organic compound removal, and consequently contribute to maintaining distribution system integrity and high quality delivered water.
The European community's drinking water standard currently calls for treatment techniques to meet maximum contaminant levels (MCLs) of 0.1 µg/l for individual pesticides, and 0.5 µg/l for total pesticides. Anticipated are disinfection byproduct (DBP) MCLs of 10 µg/l for individual THMs (trihalomethanes), 30 µg/l for total THMs, and 25 µg/l for total organic halogens (TOX).
Recent changes sought by the USEPA, said Professor Taylor, include decreasing the DBP MCL for THMs from 100 to 80 µg/l, and adding a 60 µg/l MCL for haloacetic acids (HAA). He also said the agency wants to reduce the MCLs for THMs and HAAs to 40 and 30 µg/l respectively, and that all water treatment plants may have to meet an MCL of 2.0 mg/l for total organic carbon (TOC) by the late '90s. Taylor believes the USEPA's regulations now recognize the relationship between TOC, DBPs, chlorine demand, chlorine residual, and disinfection. He sees a clear-cut regulatory direction in the U.S. whose aim is to reduce disinfectant demand and byproduct formation, yet improve disinfection and disinfectant residuals in water treatment plants and distribution systems.
Looking at Table 1 again, EDR-based processes can remove the smallest substances, but an electric charge is needed. They are ineffective for pathogen removal and most organics. UF and MF are sieving controlled, can remove pathogens, and are very effective in removing turbidity and microbiological contaminants. Membrane processes operating under pressure have a significant advantage over conventional coagulation since the membrane is a solid-film barrier to pathogens, which are simply too big to pass through. RO and NF are diffusion and sieving controlled and can remove all pathogens and many organic substances. Some or almost total removal of ionic compounds can be accomplished by diffusion. RO and NF systems have the broadest potential for treatment applications. Professor Taylor summed up this section of his talk by saying that membrane technology eventually may be applied as the surface water treatment of choice throughout the world because of these various properties.
He summarized also the abilities of the membrane systems to meet current and anticipated regulatory requirements in the U.S. (Table 2). Concluding that NF and RO are the most versatile of the membrane systems for these tasks, he pointed out that no process should be seen as universally suitable for removing all contaminants. Pilot studies were strongly recommended before design decisions are made. One thorny issue is the disposal of the unwanted concentrate from the process, which is a site-specific problem. This is seen as a major obstacle to the installation of membrane plants at certain sites, although a later speaker on the program suggested that the problem should be no more difficult to resolve than getting rid of the sludge generated in a conventional treatment facility. RO, NF and EDR systems have more serious concentrate disposal problems than their UF and MF cousins since they operate at higher levels of solute recovery. Also, membrane-treated water may have low pH values, therefore corrosivity problems and steps to recover the alkalinity should be considered.
Other presenters in Session 1 of the conference discussed specific areas in the development of membrane materials and systems. For instance, one paper covered the work being done to produce a nanofiltration membrane which will reject organics effectively, yet not affect the inorganic make-up of the water to any significant degree. The author stated that advanced membranes with such attributes are available, and the major remaining problem is to find an economically viable and efficient product suitable for application in large treatment plants to replace conventional treatment.
Another paper described the promising membrane bioreactor (MBR) technology which has gone through pilot test stages and is now in use in a small rural water treatment plant south of Paris. The combination of biological and physico-chemical processes is effective in reducing nitrates to accepted standard levels. Known as the Biocristal DN process, it uses both powered activated carbon (PAC) and ultrafiltration membranes to achieve simultaneous denitrification, disinfection, pesticide removal, and TOC reduction. The sidebar article on page 26 provides more details on this interesting system.
Membrane Concentrate Disposal
Several presentations on the disposal of concentrate streams from various membrane filtration projects in Europe were given. These concentrates can contain significant quantities of materials which must be disposed of in terms of their possible harmful effects on receiving sites or waters. Disposal options being considered are:
- Direct discharge to surface water (e.g., canal, river, lake, sea)
- Indirect discharge to surface water (e.g., via sewers and wastewater treatment plants)
- Deep well disposal
- Mixing and treating with other wastewater streams
- Land application (e.g., as irrigation water or fertilizer)
- Concentration by evaporation or crystallization
- Specific component removal (e.g., by denitrification or desulfurization)
Similarly, two French reports on studies conducted on a number of membrane plant concentrate disposal practices (one UF plant had a 15 mgd water treatment capacity), also reviewed options and basically came to the same conclusions. Gravity filters, sludge thickeners, and settling lagoons appear to be adequate for smaller plants. A two-stage UF arrangement in a larger plant scenario will reduce the treated water losses and cut the waste flow significantly.
In his paper on concentrate disposal in the United States, Mike Mickley of Mickley Associates, Boulder, Colorado, noted that as membrane technology proliferates, "There is much research, education and communication needed to ensure that appropriate levels of understanding of both technical and non-technical issues are available for the decision makers." He was referring to the fact that the interface between the industry (makers and users of membranes) and regulatory agencies is recent and at an early stage of development.
Mickley's comments amounted to an overview of a two-year feasibility study on membranes and concentrate disposal which was conducted for the American Water Works Association Research Foundation (AWWARF). An important observation was that while conventional wastes are characterized by additive chemicals, membrane concentrates reflect the make-up of the raw water in treatment. The survey looked at 137 membrane treatment plants in the U.S., all of them more than 25,000 gpd capacity. Combined capacity was 204 mgd, with a site average of 1.49 mgd. Nearly 80 percent of the plants studied, and of the total capacity, are located in Florida, California and Texas.
Dissolved Organics Removal
Norway and other countries in similar northern latitudes tend to have surface water sources that are heavy in dissolved organic matter, which is mostly brown plant residuals or humus, as well as bacteria. A Norwegian speaker described some of the promising membrane work that is going on to solve this problem for drinking water utilities. Fourteen membrane plants are operational in his country, one of them since 1989, and several others are under construction or on order. So far a membrane life of seven years seems achievable, but fouling is a problem and selection of the type of membrane must be considered along with the total plant concept and the cleaning chemicals and procedures.
Other papers in this session addressed the serious problems of natural organic matter (NOM), organohalide (THM) precursers, and the reactivity of these substances with the disinfectant chlorine. One group studied how NF systems behaved when fed with water high in humic content and reported that nanofiltration did remove THM precursers effectively. Another project looked at a combination of ozone, PAC and UF as a means of removing natural and synthetic organic matter. The early pilot work was successful and led to the design and present construction near Paris of what will be a 15 mgd UF treatment plant upon its completion.
Disinfection and Membranes
Several presentations covered the application of membrane filtration systems for the removal of protozoa, bacteria and viruses from water supplies. Work reported by Joseph Jacangelo of Montgomery Watson in the States, carried out with colleague Samer Adham, and with Jean-Michel Laine of Lyonnaise des Eaux in France, showed that polymeric MF and UF systems provide an absolute barrier to protozoa and selected bacteria, probably by physical sieving. Both types can meet current SWTR requirements for Guardia removal, and are expected to handle removal of Cryptosporidium and others, as will be called for in the enhanced rule.
Two disinfection studies by the research arms of the very large French water companies, Compagnie General des Eaux and Lyonnaise des Eaux, were summarized for the conference. The conclusions were virtually identical-there is a bright future ahead for membrane technology in the water industry. Compared to chemical disinfection systems they offer broad-spectrum removal and the absence of undesirable byproducts. They are not without problems and researchers and manufacturers are looking into such weaknesses as membrane surface defects, leaks arising from not-yet-mature manufacturing techniques, and membrane and seal failures. Post disinfection is recommended to protect the distribution system, and the conditions for carrying this out effectively are going to be excellent, given the performance of the various membranes upstream in removing solids and organics.
Large Scale Applications
First in line in this session was Ron Orach of Metcalf & Eddy in Sunrise, Florida. The subject of his talk was the 37.5 mgd water treatment plant in Hollywood, Florida, which is currently in the midst of an upgrade project. When finished later this year the plant will have three treatment trains-lime softening and filtration (7.5 mgd), NF (16 mgd), and RO (14 mgd). The three treated water streams will be blended for delivery to the distribution system.
Several other talks described current scale-up projects involving micro-, ultra- and nanofiltration, and the design choices made to accommodate source water characteristics and quantity fluctuations caused by, for instance, seasonal changes in climate and population.
AWWARF's Membrane Technology Activities
In his address to the conference, AWWARF Project Manager Roy Martinez explained that the American Water Works Association's Research Foundation is the primary source in North America for planning, managing, and funding cooperative research and development in drinking water. Its first membrane report was published in 1989. Martinez review briefly the topics covered in four studies completed and reported, five ongoing projects, and two funded this year. Covering a range of topics not unlike the conference program, these involved universities, consulting firms, water utilities and research groups.
An appropriate conclusion to this review of the excellent, well-managed IWSA conference is borrowed directly from one of the presentations, which detailed six years of experience in the production and distribution of ultrafiltered water.
"The future of low-pressure membrane technology is assured for drinking water treatment. Large UF plants are under construction in France and elsewhere in the world. Combined with oxidation, activated carbon, or bioreactors, completely new treatment lines are emerging. This recent progress also will soon give membrane unit operations the opportunity for penetrating the wastewater market."