Intermittent aeration with undefined (or random) cycle

MemPulseTM  (Siemens)

The long-term reliability of air cycle valves has been a significant issue of intermittent aeration with defined cycle times, but published information is apparently very scarce on this issue. In fact, the air cycling valves must change their position frequently, e.g. every 10 seconds or 8,640 times per day, which causes stresses in the valves.

In order to solve this problem, a system called MemPulseTM was devised by Siemens in 2008 (Fig. 1). In this system, air chambers called MemPulseTM device in Fig. 1a are installed underneath the membrane module and compressed air is supplied to the air chamber at a constant rate. Once the air chamber is filled up with the compressed air, air slugs are discharged from the chamber quickly in a few seconds, which scour hollow fibers vigorously for short time. The frequency of air slug is dependant on the size of the air chamber and the compressed air flow rate, but the bubbling time appears only a quarter or less of the total cycle time. Overall, this method allows less freedom for air cycling time control, but provides better bubble distribution and much stronger scouring effect.

Interm1                                                                                               a) Module

Interm2c) Schematic of MemPulseTM system

Fig. 1. MemPulseTM system developed by Siemens (Woods, 2008).

   The exact structures and mechanisms of MemPulseTM device are proprietary, but it is expected to be based on the intermittent airlift pump technology as shown in Fig. 2. The structure in the figure was patented by Kondo in 1998. In this device, compressed air fills up the chamber continuously as in Fig. 2a. Once the air level reaches the outlet to the vertical airlift channel, air bubbles start to leak as in Fig. 2b. Due to the airlift pump effect of the bubbles, liquid starts to move upward in the airlift channel. The moving mixture of liquid and gas can pull more air from the chamber according to the Bernoulli’s theorem. Then the increasing air flow rate to the airlift channel enhances upflow, which again increases the air flow to the airlift channel. This synergistic interaction allows air flow to the vertical chamber quickly reaches its maximum. The cycle stops when the air chamber is mostly filled up with water as shown in Fig. 2c. The cycle time of this device is basically dependant on the chamber size and the air flow rate to the chamber.

Interm30Fig. 2. Mechanism of intermittent airlift pump with a continuous air flow (Kondo in 1998).

LEAPmbr (GE)

    Recently in July 2011, GE Power&Water introduced LEAPmbr (Fig. 3) that uses similar air cycling methods as MemPulseTM. Details would come later, but the intermittent aeration with undefined cycling time is claimed to save up to 30% of air when it is compared with the intermittent aeration with defined cycle time. Specific aeration demand per permeate volume (SADp) is aimed at around 6 m3/m3, but no experimental data is in public domain yet (as of September 2011).


   In one on-going project in City of Riverside, California, as of 2012, total net savings of $800,000 with power savings of $74,538/year are anticipated in 98,400 m3/d (or 26 MGD) project by adapting LEAPmbr instead of the old version membrane module, ZW500d. The savings were largely from the following (Ciccotelli, 2012).

  • Deletion of 2 air scour blowers
  • Deletion of 1420 feet (or 433 m) of air scouring piping
  • Deletion of 112 cassette air scour piping connections
  • Deletion of 24 air scour valves
  • Reduction in size of the main air scour header
  • Reduction in the height of the piping supports
  • Reduction in the traveling crane and canopy by 5.5 feet (or 1.7 m)

mbrlkjhFig. 3. LEAPmbr of GE Power & Water (GE, 2011).

   Although the exact structure of the air chamber underneath the membrane module is not known, it can be deduced based on manufacture’s pending patent (Cumin, 2009). As shown in Fig. 4, the air chamber of LEAPmbr appears to have slightly different structure, where the vertical airlift channel does not have an opening to the bottom. As a consequence, no additional liquid other than the liquid already filled in the airlift channel moves upward during the air purging step (Fig. 4b). To date, no experimental data comparing the efficacy of the two different air chambers (Fig. 2 and Fig. 4) are known in public domain.

Interm30Fig. 4. Mechanism of intermittent airlift pump with a continuous air flow (Cumin, 2009).

LEAPmbr presented in New Orleans in WEFTEC 2012   

© Seong Hoon Yoon