Trace organic removal by MBR

Removal efficiency

As a result of the use of various chemicals for pharmaceutical or personal care purposes, any municipal wastewater stream carries many of the following chemicals that can potentially harm environment by affecting the matabolism of aquatic lives.

  • Steroidal hormons – Androstenedione, Androsterone, Etiocholanolone, Dihydrotestosterone, Testosterone, 17b-estradiol, Estriol, Estrone
  • Xenoestrogens -Bisphenol A, Propylparaben
  • Pharmaceutical and personal care products (PPCPs) – Amtriptyline, Atenolol, Atorvastatin, Carbamazepine, Diazepam, DEET, Dichlofenac, Fluoxetine, Gemfibrozil, Ibuprofen, Ketoprofen, Metformain, Naroxen, Omeprazol, o-hydroxyAtrovastatin, p-hydroxyAtrovastatin, Paracetamol, Sulfamethoxazole, Trichlocarban, Triclosan, Trimethoprim, Caffeine

The removal efficiency of above compounds were tested using a packaged MBR with 800 p.e. capacity (Trinh, 2012). In this study SRT was maintained at 10-15 days and MLSS at 7.5-8.5 g/L. MF membranes with 0.1-0,2 micron pore (Koch Puron) were used at 10 s on and 10 s off mode.

Fig. 1 and 2 show the concentrations of above compounds in wastewater and MBR permeate. In general, steroidal hormans and xenoestrogens were removed nearly 100%, but some PPCP had substantial amount of residuals in the permeate. The removal efficiencies are summarized in Fig. 3.

It is interesting that the removal efficiency of bisphenol A is dependant on initial concentration. In above study, bisphenol A was removed nearly 100% with an initial concentration of ~1 mg/L, but it was removed only ~90% in other study with an initial concentration of ~500 mg/L (Nghiem, 2009). It is difficult to say whether the lower removal efficiency was due to incomplete biodegradation or saturation of the adsorption capacity.

Fig. 1. Concentration of trace organic contaminants in raw sewage (Trinh, 2012).

Trace_34Fig. 2. Concentration of trace organic contaminants in MBR effluent (Trinh, 2012).

Trace_35Fig. 3. Removal of trace organic compounds through the packaged MBR (Trinh, 2012).

In other study, following 22 compounds were added to synthetic MBR feed at 5  mg/L each and the removal efficiencies in MBR and subsequent GAC (granular activated carbon) column were monitored. No sludge was withdrawn from the MBR during the 6 week degradability test. Dissolved oxygen and pH were maintained at 2-4 mg/L and 7.2-7.5, respectively. Empty bed contact time (EBCT) of GAC column was 7 minutes.

  • Salicylic acid, Metronidazole, Fenoprop, Ketoprofen, Acetaminophen, Naproxen, Primidone, Ibuprofen, Diclofenac, Carbamazepine, Gemifibrozil, Estriol, Pentachlorophenol, 4-tert-butylphenol, Estrone, Bisphenol A, 17-a-ethynylestradiol, 17-b-estradiol, 17-b-estrodiol-17-acetate, 4-tert-occtylphenol, Triclosan, 4-n-nonylphenol

As shown in Fig. 4, overall removal efficiency of MBR-GAC was mostly over 95%., but the biological removal efficiency by MBR varied between 20% and 100%. It appears that for relatively hydrophilic compounds with KOW of less than 3.2, removal efficiency is largely dominated by biodegradability with some contribution from biological adsorption. For hydrophobic compounds with KOW of higher than 3.2, removal efficiency is high whether it is contributed by biodegradability or adsorption on microorganisms.

Trace_39Fig. 4. Removal efficiency of trace organics by MBR and GAC (granular activated carbon), where initial concentrations were 5mg/L each (Nguyen, 2012)

Using a lab scale MBR with synthetic feed, trace organic removal efficiency was investigated (Tadkaew, 2011). No sludge was removed during the 4 week experimental period. Trace compounds were dosed at 2 mg/L each. Table 2 summarizes the result with literature data.

Table. 1. Removal efficiencies of the trace organic contaminants (n=16) by MBR (Tadkaew, 2011)


Trace_41Effect of hydrophobicity of compounds on removal efficiency

Especially for refractory trace organics, hydrophobicity of the compound has been known to be determining factor affecting removal efficiency. In one study, six compounds (Sulfamethoxazole, Diclofenac, Ibuprofen, Ketoprofen, Bisphenol A, Carbamazepine) were tested, where the hydrophobicity of the first four compounds were affected by pH while the last two are not as shown in Fig. 5. As a result, the removal efficiency of the first four compounds were clearly affected by pH. On the contrary, the removal efficiencies of the last two compounds (Bisphenol A and Carbamazepine) were not affected by pH changes as shown in Fig. 6.

Trace_30Fig. 5. Effective octanol-water districution coefficient (log D) as a function of pH (Tadkaew, 2010)

Trace_31Fig. 6. Removal efficiencies of ionizable (A) and non-ionizable (B) trace contaminants as a function of the mixed liquor pH  (Tadkaew, 2010).

In other study, the removal efficiency of trace organics was tested using a MBR with anoxic and aerobic tanks (Boonyaroj, 2012). Leachate with about 9,400 ppm COD was fed for 200 days after 100 days of acclimation period. While HRT in aerobic tank was maintained at 1 day, MLSS stayed largely between 10 g/L and 20 g/L without excess sludge removal. (Sludge production appears extremely low for the organic loading rate)

As summarized in Table 2, the compounds listed were removed between 50% and 76%. When the removal efficiencies of compounds were plotted against octanol-water partition coefficient (Kow), positive correlation was observed as shown in Fig. 7, where the higher the Kow is the higher the hydrophobicity is .

Table 2. Removal efficiencies of trace organics by MBR (Boonyaroj, 2012)


Trace_29Fig. 7. Relationship between log Kow of organic micro-pollutants and their removal during first and second stage operation (Boonyaroj, 2012).

Effect of temperature

In a study performed with pilot MBRs with anoxic tank in a municipal WWTP, treatment efficiency of trace organics appeared strongly correlated with water temperature. Fig. 1 shows that the removal efficiencies of AAA (acetylaminoantipyrine) by conventional activated sludge process (KW) and two pilot MBRs (PP1 and PP2) are fairly proportional to water temperature. However, this observation cannot be generalized since the trend was less apparent for other trace organic species.

Trace_36Fig. 8. Removal efficiencies of AAA (acetylaminoantipyrine) by conventional activated sludge process (KW) and two pilot MBRs (PP1 and PP2) with anoxic tanks treating municipal wastewater (Zeuhike, 2007)

Effect of SRT

Removal efficiency of certain pharmaceutical compounds improves as SRT increases as a result of microbial adaptation. According to a long-term lab study using synthetic feed (Maeng et al., 2013), the removal efficiency of gemfibrozil, ketoprofen, and clofibric acid increased substantially as SRT increases from 8 days to 20 days to 80 days as shown in Fig. 9.

Bezafibrate (Bezalip), ibuprofen (Advil), fenoprofen (Fenopron), phenacetine (Vicks), acetaminophen (Tylenol), phentoxifyline (Trental), and caffeine are removed at nearly 90% regardless of the SRT, which may suggest they are readily biodgradable or easily adsorbed on biosolids. On the contrary, diclofenac, naproxen, and carbamazepine are not removed well at all SRTs, which suggests these compounds are refractory under the condition.

Trace_2Fig. 8. Removal efficiencies of selected pharmaceuticals in MBRs (Maeng et al., 2013).


© Seong Hoon Yoon