By David H. Paul

Summary: In the first installment of a three-part series, the author discusses important considerations when embarking into the choppy seas of the membrane water treatment industry. This article will expound on the first three of 10 tips for water treatment dealers to keep in mind when selecting the appropriate membrane system.  

If you’re like many utilities, looking at the possibility of upgrading and/or expanding your water treatment capabilities in the not-too-distant future, you may be evaluating the option of membrane water treatment technologies such as microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and reverse osmosis (RO). These technologies are significantly different from conventional water treatment technologies.

If you’re fortunate enough to have someone on staff with membrane water treatment experience, then the evaluation process will be easier. If you don’t, this three-part series of tips may be of value to you.

Part 1 here will cover the first three of 10 tips. They are:

  1. Identify your feed water for fouling and scaling threats;
  2. Select a proven membrane product;
  3. Be conservative in the design;
  4. Conduct a well-designed 2,000-hour (minimum) pilot test;
  5. Critically evaluate the pilot test data;
  6. Select an experienced designer/contractor;
  7. Train personnel very well;
  8. Don’t wet the membrane until you’re ready to run continuously;
  9. Take complete start-up data, and
  10. Implement and monitor performance trends.

We’ll address the balance of the tips in Parts 2 and 3 of this series.

Analyze feed water
One of the biggest problems in the membrane water treatment business is clients’ providing insufficient and/or inaccurate feed water analyses to the consultant or company performing a membrane water treatment feasibility study. Whatever water analysis provided is likely to later become the water analysis upon which the entire new water treatment system is designed.

Some consultants/companies don’t ask for additional feed water analysis, even if more is needed. If the full-scale system subsequently doesn’t perform as desired, the designer/contractor may say that the process was adequately designed for the feed water analysis provided. The client may have little recourse. If a small-scale, short-term pilot study was performed successfully, this would further bolster the designer/contractor’s claim that the system was designed properly for the water analysis provided.

As an example, the author was recently sent a lake water analysis report and asked to perform an initial feasibility study to determine which membrane technologies would produce drinking water that meets primary and secondary standards. The client sent a table that provided the ranges and averages of many inorganic and organic contaminants measured over a two-year period and at several depths and locations.

In short, a tremendous amount of analysis was performed. The client tested for typical finished water (drinking water) constituents. There was essentially no data, however, concerning the fouling and scaling constituents required for membrane water treatment. Additional analyses were recommended below…

Weighing the scale
Scaling causes a loss in performance of a membrane system due to dissolved solids that precipitate due to concentration of feed water within the units. MF and UF units don’t concentrate scaling compounds because the pore size of the membranes allows dissolved substances to readily pass through. Scaling, therefore, isn’t a problem with MF and UF units.

NF and RO units, however, reject the bulk of dissolved substances including scaling compounds. As water (permeate) passes through the membrane and dissolved substances are rejected, the feed water scaling compounds are concentrated. Above a certain point (100 percent saturation), scaling will occur.
Scaling is understood fairly well. There are many scale-predicting software programs for membrane water treatment. With good feed water analyses, a system is typically designed in which scaling is controlled. For surface water, “good” feed water analyses refer to multiple analyses taken at different times of the year under different circumstances, or the “worst case” analysis. For NF and RO units, Table 1 shows recommended inorganic analyses as a minimum.

While scaling is caused by dissolved substances that turn into solids (precipitate) within RO and NF units, fouling causes a loss in performance of a membrane system due primarily to suspended solids. Fouling is typically caused by living or dead particles such as bacteria and algae or non-living particles such as silt, clay and sand.

Unfortunately, we don’t have software programs that can predict fouling of membrane units based upon feed water analyses. The presence or absence of certain suspended contaminants doesn’t equate to the presence or absence of problems. There are many factors, including design ones (see below), that affect the fouling potential of a membrane unit.

Fouling potential is quantified most accurately during pilot testing. Examples of some valuable feed water analyses concerning fouling, however, are in Table 2. Total organic carbon (TOC) is important because most microorganisms “eat” pre-existing carbon compounds. TOC is a measurement of the concentration of dissolved organic compounds. Assimilable organic carbon (AOC) is better than TOC to determine how much of the dissolved organic carbon can be used (assimilated) by bacteria. Bacteria cannot assimilate all organic compounds. It’s the assimilable organic constituents that promote growth.

SDI test
For NF and RO units, there’s a test that’s a standard in the industry for measuring the fouling potential of a feed water. This is called the Silt Density Index (SDI). Feed water is passed through a 0.45-micron (µm) filter pad at 30 pounds per square inch (psi), or 2 bar of pressure for 15 minutes. The time it takes to pass 500 milliliters (ml) of feed water through the filter pad at the beginning and end of 15 minutes are recorded. The longer it takes for the second 500 ml to pass through, the more fouling particles that are present in the feed water. An equation converts the time difference into a plugging factor and then an SDI number. The lower the SDI number, the better. For NF or RO feasibility studies, the SDI of the feed water should be measured enough times to accurately categorize the fouling potential.

Sulfate-reducing bacteria (SRB) and iron-related bacteria can cause fouling problems. They’re commonly found in certain well water. They’re sometimes found in membrane units operating on surface water depending upon upstream materials of construction and the process.

Slime-forming bacteria can cause rapid deterioration of performance. They can be found in many applications. Again, the presence or absence of fouling particles doesn’t necessarily mean that a membrane unit will experience unacceptably high fouling. Comprehensive pilot testing is the most important step in quantifying fouling potential.

Select a proven product
The membrane water treatment field is expanding rapidly. New products are being introduced, and more will be introduced in the future. Whatever membrane product or products your consultant or engineering company recommends, ask for references and performance guarantees. You usually don’t want to have the dubious honor of having a “guinea pig”-like serial number like 0001.

When evaluating references, be sure to find out the feed water and operating parameters. The ideal, though rare, is to find a nearby plant that has somewhat similar feed water and operating characteristics and note the performance of the membrane system.

Be conservative in design
One of the most important considerations in controlling fouling is to have a conservative water flux rate. Water flux refers to the water that passes through a given area of membrane in a given amount of time. In general, the higher the water flux rate, the higher the fouling rate, no matter which membrane technology is considered.

Water flux is measured as gallons of permeate per square foot of membrane per day (gfd) or liters per square meter of membrane per hour (l/m2/hr). Generally, the more membrane in a unit—given constant permeate flow rate—the lower the fouling rate. Pilot testing determines the fouling rate based upon a selected water flux rate. Higher fouling rates require more backwashing (MF and UF) and chemical cleaning (MF, UF, NF and RO).

If one bidder bids a system that has less membrane than other bidders, the projected capital cost for the proposed system may be less. The bid may be accepted based on lowest price. The operation and maintenance (O&M) costs, however, may be considerably higher due to increased backwashing and/or chemical cleaning requirements. Other important design considerations are flow velocity and demand variations. These are important considerations for all membrane technologies, but especially for NF and RO systems.

Fouling, especially biofouling, will typically occur faster at lower flow rates. Equipment that’s out of service for several hours or more at a time (especially on warm surface water) may experience high biofouling rates. Installing a lot of excess capacity needed for future use or for short-term, high-demand periods may require rotating operating units frequently to minimize fouling.

If you’re considering membrane water treatment for the first time, there are things to watch out for and consider. This series of three articles will cover 10 important tips. In this article, the most important tip discussed was to ensure an accurate characterization of the fouling and scaling characteristics of the feed water.

About the author
David Paul is president of David H. Paul Inc., an advanced water treatment training and technical services firm in Farmington, N.M.  He has over 25 years of experience in advanced water treatment. Paul has published over 100 technical articles and papers. He created and administers a 4,000-page, college-accredited correspondence training program plus three on-campus college degree programs in advanced water treatment. He holds a bachelor’s degree in biology and a master’s degree in microbiology from New Mexico State University. Paul can be reached at (877) 711-4347, (505) 327-2934 (fax), email: or website:


Comments are closed.