Guest Column | June 30, 2014 | Reproduced with Permission of Water Online
By Harold G. Fravel Jr., Executive Director, American Membrane Technology Association and Karen Lindsey, VP, Avista Technologies, Inc.
Despite a global effort to standardize units of measure to the metric system, the water treatment industry still follows the U.S. trend in resisting that directive by referring to water volumes in treatment plants as gallons per day (GPD). While achieving a specified permeate GPD is certainly important in overall membrane plant design, production capacity has absolutely nothing to do with determining optimal methods of treating site-specific feedwaters and offers no indication of potential membrane fouling rates and related maintenance costs. To evaluate these factors, knowledgeable professionals rely on membrane flux rates as the key consideration in achieving optimal system design and operation.
Calculating Flux Rates
A reverse osmosis (RO) element has a specific membrane surface area derived from each of the membrane leaves wound around a common permeate tube. Depending on the membrane model or manufacturer, an element may contain as many as 34 individual membrane leaves, each glued on three sides so that the section within the glue lines is considered the “active membrane area.” It is this active area that is used to calculate the RO element flux rate. The active membrane area of a contemporary 8” diameter RO and NF element can range from 365 to 440 sq. ft. with a variance of ±2 percent to 5 percent. For example, If an RO train then consists of 14 pressure vessels and each vessel contains 6 RO elements, the active membrane surface area would be derived from a total of 84 elements.
Membrane and element manufacturers issue specific guidelines on optimal flux rates which are typically expressed as volume per area per unit of time. Flux is used to express the rate at which water permeates a reverse osmosis membrane. In the U.S., the typical unit of measure is gallons per square foot per day (i.e., GFD or GSFD). The flux calculation is illustrated below:
Flux Rate = GPD / square feet of active membrane surface area
To determine the flux rate of a single RO spiral wound element, we would calculate the amount of permeate (filtrate) produced by that element in one day divided by the total active membrane surface area. Determining the flux rate for an RO system would consider the total permeate flow produced in a system in one day divided by the total active membrane surface area in the entire system.
Calculating the flux rate of an individual stage of an RO system would include the total permeate produced by the RO stage divided by the number of elements times the active area of an element. For instance, if a stage has 10 pressure vessels and each vessel holds six elements with 400 square feet of active area each, then the total active area for that stage would be 10 x 6 x 400 = 24,000 sq. ft. If that stage produced 240,000 gallons of permeate each day, the flux would be 10 GFD.
It should be noted that RO operations are dynamic with varying feed flows, pressures and recoveries so individual element flux rates can differ greatly from the overall system flux rate.
The flux rate specified for a full scale system is actually based on the average performance of individual elements, so it’s critical for operators to know and adhere to upper flux limitations during daily plant operation.
Flux And RO Membrane Performance
RO system feedwaters contain a variety of organic and inorganic constituents that can potentially foul or scale membranes. RO flux decline typically results from an accumulation of foulants near the membrane surface, also known as the boundary layer. The thickness of the boundary layer can be related to the membrane flux and the feed flow velocity. As the foulant layer continues to build, it becomes increasingly difficult for water to permeate through the membrane. The quality of the permeate declines and the production capacity of the system decreases. This trend requires system operators to increase feed pressures in order to maintain desired permeate flows. Operators who rely on this technique to stabilize flow rates expose the plant to increased energy costs and potential membrane damage and replacement costs.
Higher flux rates also result in greater concentration polarization as the convective forces applied by the permeate stream draws more salts to the membrane surface. Concentration polarization is typically equalized by the sheer force of the feed-concentrate stream which ensures that accumulated salts are continually removed from the membrane surface. However, high flux operations work against this equilibrium.
Suggested flux rates are heavily influenced by the quality of the raw water to be treated and the corresponding fouling rate potential. Increased flux rates mean more water, and presumably more fouling constituents, are flowing toward the membrane surface. As a result, fouling is accelerated due to the increased volume of water passing through the membrane. An acceptable flux rate would be higher for a brackish water with a low fouling potential and a Silt Density Index (SDI) value less than 3 than a brackish water with a SDI value between 3 and 5. For this reason, systems treating seawater or surface water have lower acceptable flux rates because those sources typically have an increased fouling potential. Flux also influences RO system recovery rates, though the quality of the site specific feedwater is also a significant factor.
Flux rates are proportional to the feedwater temperature and the pressure or vacuum applied within the membrane system. Water viscosity and related forces applied against the membrane barrier layer are critical to overall performance. It is estimated that every 1 degrees F (-17.22 degrees C) drop in feedwater temperature, there is a corresponding 1.5 percent decrease in membrane flux. Below are examples of some RO manufacturers suggested flux rates:
Flux And Microfiltration (MF) And Ultrafiltration (UF) Performance
The filtrate flow through a MF or UF membrane is also called flux and achieving optimal rates in these processes is just as important as it is in RO. The flux rate in MF/UF is calculated using the area of the membrane surface and the amount of filtrate
There are significant variances in flux rates between RO and MF/UF processes due to the different barrier layer mechanics and the method of separation. Flux rates for MF/UF are typically 20 to 80 GFD at pressures in the 10 to 50 psi while BWRO flux rates vary from 10 to 30 GFD at 125 to 400 psi. Most MF/UF systems operate in a dead-end filtration pattern where feedwater is slightly pressurized against a membrane surface and distinct pores remove suspended solids by size differentiation. When the accumulation of solids on the membrane surface begins to impact the desired volume of filtrate, the membranes are backwashed to remove deposits and improve performance. When a simple backwash no longer gives the desired results, a periodic cleaning solution is passed through the system. Some systems have a recirculation stream to keep membranes clean.
The final recovery rate for a MF/UF plant is calculated by subtracting the volume of water required for the backwash and intermittent cleanings from the overall system production.
System Design and Flux
Membrane flux has a significant influence on initial capital costs and a direct effect on short and long term operation and maintenance (O&M) costs. Suppliers who knowingly overstate flux rates can decrease capital costs in a system bid by reducing the required number of elements or modules and the overall system footprint to their competitive advantage. However, the long term consequences of this strategy for system end users are significant and can quickly negate any initial advantage. These include significantly higher fouling rates, higher feed pressures and related power costs, more frequent cleaning, increased system downtime, reduced productive membrane life and premature element replacement.
Understanding the importance of membrane flux rates is critical for everyone involved in membrane treatment including: manufacturers who want to meet contractual obligations, engineering firms designing efficient membrane plants, system operators hoping to achieve optimal membrane performance, and end-users reviewing competitive proposals.
Harold Fravel accepted the position of Executive Director for the American Membrane Technology Association (AMTA) after working for Dow Chemical /FilmTec Corporation for 36 years. He has a PhD in Organic Chemistry from the University of North Carolina and a BS in Chemistry from Florida State University. He resides in Jupiter, FL.
Karen Lindsey is an Executive Member of the American Membrane Technology Association (AMTA) Board of Directors. She is the V.P. and co-founder of Avista Technologies and has almost 30 years’ experience in the water treatment industry, working with companies that cast cellulose acetate membrane, produced polyamide elements and formulated specialty chemicals.