Finding the Optimal Plant Design

Related search terms from www.waterinfolink.com: fouling, permeability, filtration, flux

Design of a membrane filtration plant for the water industry requires a careful selection of flux in order to achieve a stable cost-effective design, in which fouling can be controlled at an acceptable level.1,2 A low flux will make the plant expensive in terms of capital cost (capex) because a high membrane area will be required to achieve the output. A high flux will lead to high operating costs (opex) because fouling rates increase exponentially with flux and necessitate the use of high driving pressures and frequent chemical cleaning. The design, therefore, reflects a compromise between capex and opex.

The Influence of Module Format on Plant Design

Membrane filtration applications in the water industry use one of three module formats, namely pressure-driven inside feed (PDI), pressure-driven outside feed (PDO) and submerged vacuum-driven (SUB) (which is also an outside-feed format). The different formats have different advantages and disadvantages, with all three formats competitive in most applications. Indeed, the market is almost evenly split between the three formats. However, distinct advantages arise for some niches based on feed type, feed quality, project size, etc.

The different characteristics of the three module formats result in variation in the optimum flux for commercial plant design. For example, pressure-driven formats often use frequent chemical cleaning, which allows the use of high design flux. In contrast, the absence of containment in SUB increases chemical cleaning volumes and downtime, resulting in an optimized flux at a lower level.

Fouling Rate Case Studies

This paper looks at fouling behavior for two surface water case studies, one using an SUB format and the second a PDI format. In the first example, the PVDF submerged membrane is operating on a low-turbidity surface water source. Average turbidity for this source was around 0.8 NTU, with a normal maximum rising to 2 NTU. The membrane system was operated in three distinct phases at different fluxes, providing the opportunity to evaluate fouling rates as a function of flux.

The second example was for a PDI system using a PES membrane. This feed was also a surface water source, with a normal average turbidity of around 3 to 4 NTU peaking to 10 NTU, or 15 in one case during the winter.

Flux versus fouling rate behavior for the two case studies is shown in Figure 1. In both cases, fouling rate increases exponentially with flux.2 It is notable that the two different formats have similar fouling rates at a similar flux. However, it is apparent that for the SUB case, the slope of the fouling rate curve is less, probably due to the low turbidity of the feed. Also, note that the flux range that could be employed for the submerged system was lower, partly due to the transmembrane pressure limit of the vacuum system, and partly due to the lower permeability of PVDF compared to PES.

Monitoring Indices

Most membrane systems use a chemical-enhanced backwash (CEB) to maintain performance. This procedure can be carried out automatically and is inexpensive in terms of chemical consumption, waste produced and downtime. Because the cleaning action relies on a short soak period, it is not completely effective. Therefore, from time to time, off-line cleaning is used, sometimes termed clean in place (CIP). CIP often uses similar chemicals and concentrations to CEB, but extended soak times, together with recirculation or drain down (depending on system format) ensures greater effectiveness and normally a full recovery of permeability.

The limitation of CIP is that the volumes of chemical are greater and the loss of production due to downtime is significant. It is therefore desirable to restrict CIP frequency as much as possible and instead rely on CEB.

Thus, CEB is intended to recover permeability as much as possible. If the CEB is completely efficient, the permeability will be returned to the value after the last CEB. If there is a fouling trend, either the CEB frequency should be increased, the concentration and/or choice of chemicals should be adjusted or there will eventually be a need for a CIP to recover the permeability loss.

Permeability data can be summarized in two simple parameters which are trended with time and are termed the fouling rate index and the cleaning index. The fouling rate index measures the rate of permeability loss over the course of a CEB cycle and is a measure of the fouling propensity of the feed. The cleaning index compares permeability from one CEB cycle to the next. The two indices are illustrated in Figure 2 for a permeability trend illustrating rapid fouling.

If permeability data is more scattered than Figure 2, the indices can be expressed in terms of the loss of permeability as a ratio rather than the trend with time as measured by the slope of the curve. The indices can be used in combination with permeability measurements to control CEB frequency and anticipate the requirement for off-line cleaning. Also, the comparison of the cleaning index between racks can be assessed, highlighting differences in rack performance, perhaps due to valve or pump failure.

Evaluating the Final Data

Plant data from submerged and PDI feed formats show that for a given flux, the two formats exhibit similar fouling rates. Membrane permeability can be used to monitor plants to ensure stable long-term performance, and monitoring indices can be evaluated from permeability trends. These indices can be used together with the permeability data to adjust chemical wash frequency and procedures as well as predict CIP intervals.

Graeme Pearce is director of Membrane Consultancy Associates. Pearce can be reached at gpearce@membraneconsultancy.com.

Leave A Comment

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.