Specific energy demand (SED) of immersed MBR

Table 1 summarizes the average SED of many different municipal MBR plants, which includes all energy consumptions in the plants including biological air, membrane scouring air, sludge circulation, permeate suction, office lighting, etc. In general, SED ranges 1.8-6 kWh/m3 for the plants with <5,000 P.E. capacity and 0.8-1.4 kWh/m3 for the plant with >5,000 P.E capacity (Gnirss, 2009). In a highly optimized plant with a design flow rate of 23,000 m3/d in Ulu Pandan, Singapore, where settled sewage is fed to MBR with anoxic tank, SED was observed at 0.55 kWh/m3 at an intermittent aeration mode at 10s ON/10s OFF mode in 2009 (Fig. 1). In the same plant, SED could be reduced down to 0.4 kWh/m3 by adjusting aeration mode to 10s ON/30s OFF (Chen, 2012).

Table 1. Average specific energy demand (SED) in municipal MBR (Giesen 2008)

Location Capacity

(P.E.)

Membrane Year Total average SED

(kWh/m3)

Rödingen (Germany) 3,000 GE 2001 2.03
Markranstadt (Germany) 12,000 GE 2001-2003 0.8-1.5
Knautnaundorf (Germany) 900 Huber (VRM) 2002-2003 1.3-2
Seelscheid (Germany) 11,000 Kubota 2004-2005 0.9-1.7
Monheim (Germany) 9,700 GE 2003-2005 ~1
Brescia (Italy) 46,000 GE 2003-2005 0.85
Nordkanal (Germany) 80,000 GE 2004-2005 0.90
Varsseveld (Netherlands) 23,150 GE 2005

2006

2007

1.00

0.88

0.83

Note: Population equivalent (P.E.) varies country by country but it is mostly ~60 g BOD5/day/person in European countries, which is equivalent to 0.3 m3 wastewater/day/person with a BOD of 200 mg/L.

SED in1

Fig. 1. Specific energy consumption in Ulu Pandan MBR (Chen, 2012).

SED also varies depending on influent flow and strength even in a given plant. For example, SED decreased from 1.66 kWh/m3 to 0.39 kWh/m3 when flow rate increased from 8,000 m3/day to 44,000 m3/day in Nordkanal, Germany in 2007 (Engelhardt, 2008). Meanwhile, the yearly average SED was at 0.85 kWh/m3 in the same place in 2007.

Similar results were obtained in a small MBR with 6,000 person equivalent (PE) capacity in France. SED was estimated at 0.69 kWh/m3 and 1.01 kWh/m3 for winter and summer, respectively (Barillon, 2012). The low SED in winter was attributed to higher flow rate at lower organic strength than in summer. Fig. 2 shows the relation between flow rate and SED obtained from 60,000 PE capacity MBR plant in France, where SED drops as flow rate increases.

Table 2 summarizes energy consumption breakdowns of two full-scale municipal MBR plants. It is noteworthy that the SED for membrane aeration in realistic conditions are as high as 0.23 kWh/m3 and 0.34 kWh/m3 in the two plants, which is much higher than the SED claimed by manufacturers, e.g. 0.07-0.15 kWh/m3.

SED in5

Fig. 2. SED as a function of flow rate in a municipal MBR plant with GE hollow fiber membranes with 60,000 person equivalent capacity in France (Barillon et al. 2013).

 

Table 2. Energy consumption breakdown for two full-scale municipal MBR plants (De Wever, 2009)

Energy consumption (kWh/m3) Kaarst-Nordkanal, (Germany) Varsseveld, The (Netherlands)
Membrane aeration 0.23 0.34
Bioreactor aeration 0.30 0.24
Recirculation pumps 0.03 0.11
Permeate pumps 0.04 0.12
Mixers 0.05 0.04
Others    0.25 * 0.05
Total SED 0.90 0.90

* Including sludge dewatering

SED also varies depending on membrane type. As shown in Fig. 3, hollow fiber (HF) membranes tend to consume less energy than flat sheet (FS) membranes based on the audit performed in six different MBRs (two FS and four HF). In the figure, SED includes energy to run activated sludge process and membrane system, but not includes energy for building maintenance, etc. When hydraulic and organic loadings are around 40-50% of the design values, FS consumes nearly 50-100% more energy than HF (Barillon, 2012).

For hollow fiber membranes, SED for membrane scouring was equal or less than for bio-aeration in three different MBR plants (Fig. 4). For flat sheet membranes, however, SED for membrane scouring was two to three times higher than that for bio-aeration in two different MBR plants. Overall, SED was significantly higher for flat sheet membranes than hollow fiber membranes.

SED in8Fig. 3. SED of six assessed MBR as a function of hydraulic load and organic load (Barillon et al. 2013).

SED in7Fig. 4. SED for aeration in five different MBR plants with either flat sheet (FS) membranes or hollow fiber (HF) membranes (Barillon et al. 2013).

 

© Seong Hoon Yoon