Specific air demand per permeate volume (SADp)

SADp varies not only by membrane configuration but also by influent flow rate. It tends to be lower with hollow fiber membranes than flat sheet membranes due to the higher packing density as discussed here. In addition, it decreases as hydraulic flow rate increases for a given membrane configuration since the scouring air needs to increase less than proportionally to the hydraulic flow. The survey of SADp in various MBR plants at various flux is summarized in Fig. 1.

It has been reported that SADp can be as low as 7.3 with intermittent aeration, but it is possible only under a favorable condition as discussed here. There are two reasons.

  • The low SADp claimed by manufacturers are valid only when hydraulic flow rate is at a design level. If the hydraulic flow rate (or flux) is below the design level, SADp increases since air scouring rate does not proportionally decrease. This trend is apparent in Fig. 1.
  • The low SADp advertised is typically based on the well maintained pilot system, where everything is under watch closely to maintain good mixed liquor quality. But, full-scale MBR inevitably suffers from varying hydraulic and organic loading (or F/M), varying dissolved oxygen level, MLSS fluctuation, potential nutritional imbalance, etc.

Therefore, much more scouring air is necessary in real world than expected based on pilot tests. According to Fig. 1, realisticSADp in field is 14-30 m3/m3.

The lowest field SADp found in full-scale MBR appears <10 m3/m3 obtained in MBR plant in Ulu Pandan, Singapore, when the influent is at a daily average design flow rate. The specific energy demand (SED) for membrane system excluding sludge circulation also appears as low as <0.1 kWh/m3 permeate at an intermittent aeration cycle of 10-30 (10 sec ON and 30 sec OFF).

The newly introduced GE’s LEAPmbr in summer 2011 employs irregular slug air flow as discussed here. This system aims SADp of around 6 by using air chambers underneath the membrane module similarly to Semens MemPulseTM system. However, Barillon et al. (2013) argued LEAPmbr tends to increase energy consumption relative to the intermittent aeration with defined cycle time, but no data was presented.

Specif33Fig. 1. SADp as a function of net flux in various MBR plants (Judd, 2007)


© Seong Hoon Yoon