An innovative integrated multistage bioreactor (IMBR) system, which was augmented with three predominant bacterial strains (Lactobacillus paracasei CL1107, Pichia jadinii CL1705, and Serratia marcescens CL1502) isolated from marine sediments, was developed to treat real tannery wastewater without performing physicochemical pretreatment, with the potential to reduce the generation of waste sludge and odors. The performance of the IMBR treatment system, with and without the inclusion of the predominant bacterial strains, was compared. The results indicated that the performance of the IMBR system without bioaugmentation by the predominant bacterial strains was poor. However, when in the presence of the predominant bacterial strains, the IMBR system exhibited high removal efficiencies of chemical oxygen demand (COD) (97%), NH4+-N (97.7%), and total nitrogen (TN) (90%). In addition, the system had the capacity for the simultaneous removal of organics and nitrogen, heterotrophic nitrification and denitrification being carried out concurrently, thereby avoiding the strong inhibition of high concentrations of COD on nitrification. The system possessed excellent adaptability and ability to resist influent loading fluctuations, and had a good alkalinity balance such that it could achieve a high NH4+-N, and TN removal efficiency without a supplement of external alkalinity. In addition, an empirical performance modeling of the IMBR system was analyzed.

INTRODUCTION

The tannery industry is one of the most polluting industries and is characterized by considerable water consumption associated with the extensive use of myriad chemicals (Iaconi et al. 2003). Tanneries create heavy pollution from effluents that contain high levels of salinity, organic loads, inorganic and nitrogenous compounds, chromium and sulfides, as well as other dissolved and suspended solids (Song et al. 2003). They are thus considered to have a major negative environmental impact. Therefore, there is an urgent need to control the environmental contamination from tannery wastewater. Conventional tannery wastewater treatment systems are comprised of a combination of various physicochemical pretreatments and biological processes. Among these techniques, available biological tannery treatment technologies include the activated sludge process (ASP), sequential batch reactor, biological contact oxidation process, and upflow anaerobic sludge blanket (Yusuf et al. 2013). Nevertheless, as tannery wastewaters are highly complex mixtures that contain high contaminant concentrations and toxicity, and present large fluctuations in water quality and flow rate, conventional treatment technologies suffer from a number of drawbacks, such as the large consumption of chemicals in physicochemical pretreatments, significant chemical sludge generation, and high operational costs. In addition, the presence of salinity, chromium and sulfides has an inhibitory effect on biodegradation. Therefore, it is critical to develop effective treatment systems that are both economical and environmentally compatible.

In recent years, bioaugmentation (the introduction of specialized microorganisms into polluted sites and/or bioreactors) has been recognized as a promising and attractive means of accelerating the removal of target pollutants and improving wastewater treatment performance (Leta et al. 2005; Ma et al. 2009). This technique has been applied to enhance the biological treatment of tannery wastewater. Leta et al. (2005) inoculated an efficient Brachymonas denitrificans into pilot plant reactors to augment biological nitrogen removal from a tannery effluent. Kim et al. (2013) employed a novel microbial consortium (BM-S-1) for tannery wastewater treatment bioaugmentation, allowing it to be treated without chemical pretreatment. The removal efficiency of chemical oxygen demand (COD) was 91.4%; however, the removal efficiency reached only 50.9%, whereas the concentration of the effluent remained at a high level (21.2 mg/L). This research was conducted under low levels of salinity and inhibitor (chromium, sulfide) in the influent, which had an influence on the physiological and biochemical characteristics of the microorganisms (He et al. 2012). In addition, since the halotolerant bacteria are less sensitive to high salinity conditions, some studies have been conducted on the treatment of tannery wastewater using halotolerant bacteria. Durai et al. (2011) utilized salt-tolerant bacterial strains (P. aeruginosa, B. flexus, E. homiense, and S. aureus) for the treatment of tannery wastewater in a bench-scale aerobic sequencing batch reactor. In their study, they found that the salt-tolerant microorganisms improved the reduction efficiencies of COD and the colour of the tannery wastewater. Sivaprakasam et al. (2008) reported that tannery wastewater could be treated with four salt tolerant bacterial strains, viz., P. aeruginosa, B. flexus, E. homiense and S. aureus, which were isolated from saline marine and tannery wastewater samples.

In this study, three predominant bacterial strains isolated from marine sediments were introduced into an integrated multistage bioreactor (IMBR) for the treatment of tannery wastewater. Thus, the specific aim of this study was to determine whether the IMBR system could run efficiently and steadily for the treatment of tannery wastewater without physicochemical pretreatment. In such a treatment system, it was possible to achieve economic savings and reduced sludge production, due to the absence of the physicochemical pretreatment.

MATERIALS AND METHODS

Tannery wastewater

The tannery wastewater utilized in this experiment was taken from a tannery factory in Quanzhou, China. The wastewater was stored in a tank prior to being dosed, as influent, into the IMBR treatment system. The chemical characteristics of the tannery wastewater are illustrated in Table 1.

Table 1

Chemical characteristics of the tannery wastewater used in this study

Parameters Units Values Parameters Units Values 
pH – 8.0–8.7 Total chromium mg/L 30–37 
COD mg/L 5,200–5,500 Chromium(VI) mg/L 0.5–0.75 
BOD5 mg/L 2,700–3,200 Sulfide (S2−mg/L 45–60 
SS mg/L 2,400–2,700 TDS mg/L 3,200–3,600 
Ammonia nitrogen (mg/L 270–300 Cl mg/L 2,670–3,150 
TN mg/L 370–400 Alkalinity mg/L 630–800 
Parameters Units Values Parameters Units Values 
pH – 8.0–8.7 Total chromium mg/L 30–37 
COD mg/L 5,200–5,500 Chromium(VI) mg/L 0.5–0.75 
BOD5 mg/L 2,700–3,200 Sulfide (S2−mg/L 45–60 
SS mg/L 2,400–2,700 TDS mg/L 3,200–3,600 
Ammonia nitrogen (mg/L 270–300 Cl mg/L 2,670–3,150 
TN mg/L 370–400 Alkalinity mg/L 630–800 

COD: chemical oxygen demand; BOD5: biochemical oxygen demand; SS: suspended solids; TDS: total dissolved solids.

Microorganisms

The predominant bacterial strains used in this work were Lactobacillus paracasei CL1107, Pichia jadinii CL1705, and Serratia marcescens CL1502, respectively. They were isolated from deep-sea sediments (North Atlantic, 3,667 m water depth) and acclimated with tannery wastewater. Previous studies have shown that these predominant bacterial strains not only had the capacity for the removal of organic substrates and nitrogen, but also had the advantage of chromium and sulfide resistance. Chromium and sulfide could be adsorbed and/or precipitated by these bacterial strains. In addition, Serratia marcescens CL1502 is a heterotrophic nitrifying bacterium that can synchronously remove organic substrates and ammonia nitrogen (data not published).

The biological sludge obtained from the anoxic–oxic (A/O) ASP of a tannery wastewater treatment plant (Quanzhou, China) was added into the IMBR system.

Experimental set-up

Figure 1 illustrates the design of the IMBR treatment system, which consisted of an influent feeding tank, a biological adsorption (BA) tank, a primary inclined settling (PIS) tank, an anoxic biofilter (AF), a biological aerated filter (BAF), a secondary inclined settling (SIS) tank, a nitrifying biofilter (NBF), a tertiary inclined settling (TIS) tank, and a sludge regeneration (SR) tank. In the IMBR system, the BA unit was primarily designed to adsorb or remove pollutants (i.e., chromium, sulfide, organic matters, etc.) using microorganisms; the function of the AF unit was realization of the denitrification process and hydrolysis of complex high molecular weight organic matters; the main function of the BAF unit was to further remove organic matter; the main function of the NBF unit was to remove . The dimensions of the primary reaction tank are shown in Table 2.
Table 2

Dimensions of primary reaction tank in the IMBR treatment system

Units Configurations Working volumes (L) Hydraulic retention time (h) 
BA L × B × H = 40 × 40 × 40 cm 50 10 
PIS L × B × H = 45 × 20 × 150 cm 70 
AF Φ × H = 35 × 100 cm 50 10 
BAF Φ × H = 35 × 100 cm 50 10 
SIS L × B × H = 45 × 20 × 150 cm 70 
NBF Φ × H = 35 × 100 cm 50 10 
TIS L × B × H = 45 × 20 × 150 cm 70 
Units Configurations Working volumes (L) Hydraulic retention time (h) 
BA L × B × H = 40 × 40 × 40 cm 50 10 
PIS L × B × H = 45 × 20 × 150 cm 70 
AF Φ × H = 35 × 100 cm 50 10 
BAF Φ × H = 35 × 100 cm 50 10 
SIS L × B × H = 45 × 20 × 150 cm 70 
NBF Φ × H = 35 × 100 cm 50 10 
TIS L × B × H = 45 × 20 × 150 cm 70 

Note, L: length of reaction tank.

B: width of reaction tank.

H: height of reaction tank.

Φ: diameter of reaction tank.

Figure 1

Flow diagram of the IMBR system.

Figure 1

Flow diagram of the IMBR system.

As depicted in Figure 1, the tannery wastewater was initially pumped from the influent feeding tank into the BA tank, within which aeration with a high intensity (20 m3/(m2 h)) was supplied to ensure that the microorganisms and wastewater were well mixed, and was thus favorable for the adsorption of chromium and a portion of the organic carbon within the influent by the microorganisms. The adsorption of pollutants by microorganisms was because of the special surface properties of microorganisms, such as adhesion, complexation, ion exchange, coordination, chelation, microprecipitation, and flocculation abilities. Furthermore, influent resident sulfides could be removed by these microorganisms. As a result, toxic inhibitors (chromium and sulfide) could be removed at this stage without physicochemical pretreatment, such that the toxic effects of chromium and sulfide on subsequent biological processes could be avoided. The wastewater then flowed into the PIS tank for biomass settling, and the sludge that was generated by the PIS unit proceeded into the SR tank to regeneration by anaerobic digestion. A portion of the regeneration sludge was recycled into the BA unit, and the remainder was discharged periodically. The wastewater was then introduced into the AF, which was packed with polyurethane fibers (specific area: 2 × 104 m2/m3; specific density: 0.7–0.95 g/cm3) that served as biological carriers. At this stage, the heterotrophic nitrifying bacterium employed the wastewater organic matter as a carbon source for denitrification, and the high molecular weight organic matter was hydrolyzed into entities with lower molecular weights. The effluent of the AF unit entered into the BAF, within which most organic pollutants and a portion of the would be removed. The wastewater then entered the SIS tank for biomass settling. Finally, the effluent of the SIS unit flowed into the NBF, which was packed with polyurethane fibers that served as a biological carrier. At this stage, most of the would be removed via nitrification. Moreover, the wastewater with a low C/N ratio benefited the growth of the nitrifying bacteria, resulting in the efficient removal of . A portion of the effluent from the NBF unit was transferred to the BA unit, whereas the remaining portion was transferred to the TIS for biomass settling. The sludge that was generated from the SIS and TIS units was recycled into the BA unit.

Process startup

Twenty liters biological sludge (mixed liquor volatile suspended solids = 5,000 mg/L) and 20% of the predominant bacterial strains were initially inoculated into the BA, AF, BAF, and NBF units. These microorganisms were subsequently activated by aeration for 48 h, after which the IMBR system was filled with water up to the design volume. The raw wastewater was then diluted four times and added into the influent feeding tank, after which it was pumped into the IMBR treatment system at a flow rate of 5 L/h. Thereafter, the water from each reaction tank was replaced with diluted raw tannery wastewater after 5 days of continuous running. Hence, the startup of this system was completed, and the changes in the physical and chemical parameters of the influent and effluent were analyzed every 24 h. The hydraulic retention time for the whole system was 2.2 days. According to our pre-experiment test, it was found that the IMBR treatment system could achieve better performance when the dissolved oxygen (DO) concentrations of BAF and NBF units were 2.5 mg/L and 3.5 mg/L, respectively (data not shown). Thus, the DO concentrations in BAF and NBF units were set at 2.5 mg/L and 3.5 mg/L, respectively.

Experimental procedure

In order to determine whether bioaugmentation with predominant bacterial strains might treat tannery wastewater without any physicochemical pretreatment in a tannery wastewater treatment system, both the predominant bacterial strains and biological sludge were inoculated into the treatment system. The experiment was carried out at an influent flow rate of 5 L/h, with a mixed-liquor recycle ratio of 250% of the influent flow rate. A control experiment was conducted with the biological sludge but without the addition of the predominant bacterial strains. When the IMBR system achieved steady-state, the performance of the IMBR system was investigated by varying the quality of the influent (raw wastewater diluted four times, two times, and no dilution, respectively), and influent flow rate (5 L/h, 3 L/h, and 1.5 L/h, respectively).

Analytical methods

Physical and chemical parameters, encompassing pH, COD, biochemical oxygen demand, , total nitrogen (TN), suspended solids, total dissolved solids, Cl, S2−, Cr(VI), total chromium (Tot-Cr) and alkalinity were carried out according to standard methods issued by the Ministry of Environmental Protection Agency, China (China Environmental Protection Bureau 1989).

RESULTS AND DISCUSSION

A comparison of tannery wastewater treatment with and without bioaugmentation using predominant bacterial strains

Two experiments in the IMBR treatment system were carried out with the aim of assessing whether bioaugmentation with predominant bacterial strains might successfully treat tannery wastewater without the requirement of physicochemical pretreatment. It could be seen from Table 3 that, after 25 d of continuous running, the removal efficiency of the IMBR treatment system without the addition of the predominant bacterial strains (IMBR-IP1) for COD, , and TN, were 88.3%, 30.8%, and 24.2%, respectively, when the influent COD concentrations averaged at about 5,624 mg/L, at 276 mg/L, TN at 392 mg/L, S2− at 51 mg/L, Cr(VI) at 0.67 mg/L, and Tot-Cr at 33.9 mg/L. The effluent COD, , TN, S2−, Cr(VI), and Tot-Cr of the IMBR-IP1 system were 656 mg/L, 191 mg/L, 297 mg/L, 26 mg/L, 0.17 mg/L, and 19.6 mg/L, respectively, which did not attain the set discharge limit in China. However, when combined with physicochemical pretreatment, the effluent COD, , and TN, were 256 mg/L, 12 mg/L and 156 mg/L, respectively, in a tannery wastewater treatment plant (Quanzhou, China) using an A/O biological treatment process. This implied that tannery wastewaters were hard to biologically remediate without toxic substances (e.g., chromium, sulfide, etc.) pretreatment. A number of researchers reported that the presence of chromium and sulfide in tannery wastewater had a toxic effect on bacterial growth, and resulted in lower COD, , and TN removal (Weimann et al. 1998; Wiegant et al. 1999; Stasinakis et al. 2002). In addition, when the physicochemical treatment technology was applied, a large amount of sodium hydroxide, ferrous sulfate and coagulant polymers were required to precipitate the pollutants, which consequently generated large quantities of sludge. This indicated that the physicochemical treatment technology was environmentally incompatible and costly. In our study, the experimental results also showed that after 25 d of continuous running, the removal efficiency of the IMBR treatment system with bioaugmentation using predominant bacterial strains (IMBR-IP2) for COD, , and TN, were 97.5%, 98.9%, and 88.7%, respectively, when the influent COD concentration averaged at about 5,361 mg/L, at 290 mg/L, TN at 391 mg/L, S2− at 50 mg/L, Cr(VI) at 0.61 mg/L, and Tot-Cr at 31.2 mg/L. The effluent of the IMBR-IP2 system had the capacity for attaining the discharge limits in China. In addition, the effluent resident S2− and Tot-Cr were removed to 3.1 and 0.5 mg/L, at the BA unit in the IMBR-IP2. This suggested that the S2− and Tot-Cr were almost completely removed at the BA stage. The proposed explanation for these phenomena was that the chromium and sulfide might be adsorbed and precipitated by the predominant bacterial strains, which was favorable for the following biological treatment process. In addition, it was found that there was no obvious odor observed in the system with the addition of the predominant bacterial strains. These results implied that bioaugmentation with predominant bacterial strains could successfully treat tannery wastewater without the use of physicochemical pretreatment.

Table 3

The treatment performance of bioaugmentation system and non-bioaugmentation system

Parameter COD (mg/L) NH4+-N (mg/L) TN (mg/L) S2− (g/L) Cr(VI) (mg/L) Tot-Cr (mg/L) 
Non-bioaugementation system (IMBR-IP1) Influent 5,624 276 392 51 0.67 33.9 
Effluent 656 191 297 26 0.17 19.6 
Bioaugmentation system (IMBR-IP2) Influent 5,361 290 391 50 0.61 31.2 
Effluent 135 3.2 44 0.2 0.07 0.63 
Parameter COD (mg/L) NH4+-N (mg/L) TN (mg/L) S2− (g/L) Cr(VI) (mg/L) Tot-Cr (mg/L) 
Non-bioaugementation system (IMBR-IP1) Influent 5,624 276 392 51 0.67 33.9 
Effluent 656 191 297 26 0.17 19.6 
Bioaugmentation system (IMBR-IP2) Influent 5,361 290 391 50 0.61 31.2 
Effluent 135 3.2 44 0.2 0.07 0.63 

COD removal

The removal of COD by the IMBR treatment system is reported in Figure 2(a). As can be seen in the figure, the COD concentration decreased after the treatment in each unit of the IMBR system. When the IMBR system operated under the initial concentration gradient (raw tannery wastewater diluted four times, influent COD: 1,352 mg/L), the COD removal efficiency of the IMBR system attained 95.7% after 10 d of continuous operation; the effluent COD of this system was 57 mg/L. The COD concentration was decreased from 1,352 to 291 mg/L after treatment by the BA unit, which was due to the BA by the predominant bacterial strains (11.3%) and dilution with the reflux of the treated water (88.7%). When the influent of the system was increased (raw tannery wastewater diluted two times, influent COD: 2,652 mg/L), the COD removal efficiency of the system was initially decreased, which was due to the fact that the microbes required some acclimatization time. Following acclimatization, the efficiency increased rapidly. The COD removal efficiency was increased, from 84.6 to 96.3%, during 10 d of operation, and the effluent COD of this system was 81 mg/L. When the raw tannery wastewater was employed as the influent of the system (influent COD: 5,359 mg/L), the COD removal efficiency fell to 77.2%. The COD removal then increased again, once the adaptation to the new influent quality was completed. It is worth noting that, although there was an instantaneous increase in the influent COD concentration, a low effluent COD level was recovered in a short period of time (5 days). This indicated that the IMBR system possessed an excellent ability to overcome sudden disturbances in input organic loading. After 20 d of operation, the COD concentration in the treated effluent attained 142 mg/L (97.4% COD removal efficiency), which could meet the discharge standard of tannery wastewater (COD < 150 mg/L) in China. Combined with the data in Table 4, it might be observed that the COD removal occurred mainly at the BAF stage, with a removal rate of 9.44 kg COD/(m3 d), which accounted for 56.06% of the COD removal. The probable reason for this phenomenon was that after treatment by the AF stage (in front of BAF stage), the high molecular weight organic matter was hydrolyzed into entities with lower molecular weights, which was more favorable for organic pollutant degradation in the BAF stage. On the other hand, 18.89%, 20.49%, and 4.56% of the influent COD might be removed at the BA, BAF, and NBF units, respectively. In addition, Figure 2(a) shows the effect of influent flow rates on the removal of COD. It could be seen from the figure that the COD removal efficiency did not noticeably change when the influent flow rate was decreased from 5 to 1.5 L/h. The average COD removal efficiency was maintained at about 97%. This phenomenon might be due to the solid retention time (25 days) being much higher than the hydraulic retention time (2.2 days), which could be attributable to the biomass that was retained by the microcarriers within the reactor (Song et al. 2003).
Table 4

The removal rate of COD, NH4+-N and TN in the each unit of the IMBR system

Parameter BA AF BAF NBF 
COD removal rate (kg COD/(m3 d)) 2.40 2.60 9.44 0.66 
The percent COD removal for each reactor (%) 18.89 20.49 56.06 4.56 
removal rate  0.11 None 0.19 0.37 
The percent removal for each reactor (%) 16.42 None 28.36 55.22 
TN removal rate (kg TN/(m3 d)) 0.16 0.36 0.11 0.21 
The percent TN removal for each reactor (%) 19.05 42.85 13.10 25.00 
Parameter BA AF BAF NBF 
COD removal rate (kg COD/(m3 d)) 2.40 2.60 9.44 0.66 
The percent COD removal for each reactor (%) 18.89 20.49 56.06 4.56 
removal rate  0.11 None 0.19 0.37 
The percent removal for each reactor (%) 16.42 None 28.36 55.22 
TN removal rate (kg TN/(m3 d)) 0.16 0.36 0.11 0.21 
The percent TN removal for each reactor (%) 19.05 42.85 13.10 25.00 
Figure 2

Performance of COD removal in the IMBR system. (a) Effect of the influent COD concentration and the influent flow rate on COD removal; (b) effect of organic loading rate on the COD removal rate.

Figure 2

Performance of COD removal in the IMBR system. (a) Effect of the influent COD concentration and the influent flow rate on COD removal; (b) effect of organic loading rate on the COD removal rate.

Figure 2(b) shows the relationship between the COD removal rate and the influent COD loading rate. It can be observed that the COD removal rate was linearly increased with the increase of the influent COD loading rate. The linear regression analysis of the COD removal rate and the COD loading rate returned a coefficient of determination of 0.9997 and slope of 0.9852 (i.e., 98.52% COD average removal efficiency). The COD removal rates increased from 3.13 to 12.65 kg COD/(m3 d), with an increase in the COD influent loading rate, from 3.21 to 12.86 kg COD/(m3 d), which were obtained at steady-state by varying the influent COD concentrations. Song et al. (2003) employed an upflow anaerobic fixed biofilm reactor for the treatment of tannery wastewater. It was found that the COD removal efficiency averaged ∼71%, and the COD removal rates varied between 0.121 and 2.15 kg COD/(m3 d) when the influent COD loading rate was varied between 0.16 and 3.0 kg COD/(m3 d). The COD removal efficiency and the removal rate of the IMBR system achieved in our study were superior when compared with the previous literature.

Nitrogen removal

The capacity for nitrogen removal by the IMBR system is presented in Figure 3. Generally, it may be seen that the concentration was decreased in the BA, BAF and NBF reactors. The TN concentration was decreased in the BA, AF, BAF and NBF reactors. This indicated that and TN might be removed and accompanied by COD removal under various tannery wastewater strengths. When the influent strength was raw tannery wastewater diluted four times (67 mg/L influent , 82.7 mg/L influent TN), the effluent resident and TN of the system were 1.2 mg/L (98.3%) and 12 mg/L (86.1%), respectively, after 10 d of continuous operation. When the influent strength was increased to raw tannery wastewater diluted two times (122.1 mg/L influent , 195.8 mg/L influent TN), the and TN removal efficiency of the system initially rapidly decreased. After 10 d of operation, the removal efficiency increased, from 85.4 to 99.2%, and the TN removal efficiency increased, from 76.7 to 88.8%. The effluent and TN of the system were 1.2 mg/L and 22 mg/L, respectively. Subsequently, when the raw tannery wastewater was employed as the influent of the system (286.7 mg/L influent , 389.9 mg/L influent TN), the and TN removal efficiencies were immediately reduced to 71.7% and 64.5%, respectively. The IMBR system attained a steady state following 10 d of operation (running from day 21 to 31). At this stage, the and TN removal efficiencies of the system were 97.7% and 87.2%, respectively. On day 50, the final and TN concentrations in the treated effluent were 2.4 mg/L and 39 mg/L, respectively, which attained the discharge limit in China. These experimental results indicated that the IMBR system showed a high stability. In addition, combined with the data of Table 4 and Figure 4, it was found that the and TN concentrations, which were decreased significantly in the BA unit, were due primarily to the dilution that was initiated by the recycled effluent, and microbial adsorption. Most of the was removed in the NBF unit, at a removal rate of 0.37 kg , which accounted for 55.22% of the influent removal. On the other hand, about 24.68% of influent was removed in the BAF unit. This was because the predominant bacterial strains had the ability to simultaneously remove COD and . Lu et al. (2014) discovered that a heterotrophic nitrifying–denitrifying bacterium, Serratia sp. LJ-1, was capable of simultaneously removing the organic toxicant (phenol) and ammonium, and this bacterium had a high tolerance to phenol toxicity. Gupta & Gupta (2001) reported that a predominant bacterium, Thiosphaera pantotropha, exhibited simultaneous carbon removal and nitrification in a fixed biofilm system. Additionally, wastewater treatment plants have combined carbon oxidation and nitrification in the same reactor since the 1970s, although heterotrophs compete with nitrifying bacteria for oxygen and space (Nogueira et al. 2002). Due to the capacity of anoxic denitrification to play a prominent role in the removal of TN, the elimination of TN in the IMBR system occurred primarily in the AF unit (42.85% of influent TN removal). Moreover, the NO3-N and NO2-N concentrations were minimized in the AF unit, which was due to the majority of NO3-N and NO2-N being transformed to nitrogen gas or other volatile nitrogen oxides, via the denitrification process. Furthermore, it should be noted that 25% of the TN in the influent was reduced in the NBF unit, and 13.10% of the TN in the influent was reduced in the BAF unit. According to the analysis of the nitrogen species balance, the simultaneous nitrification and denitrification phenomenon took place in the NBF unit. This observation might possibly be because: (1) the predominant bacterial strains were able to simultaneously induce nitrification and denitrification under aerobic conditions; (2) the polyurethane fibers were used as carrier media for biomass immobilization, in which the outer biofilm layer was aerobic, and the inner biofilm was anoxic. That is, nitrification proceeded at the carrier interface, whereas the anoxic microzones existed in the deeper layers of the biofilm, which allowed heterotrophic denitrifiers to complete the denitrification process in a conventional manner. Polyurethane materials have previously been reported as carriers for enhancing simultaneous nitrification–denitrification processes (Daniel et al. 2009). Kim et al. (2005) reported that some bacterial strains had the ability to initiate simultaneous aerobic nitrification and denitrification, such as B. cereus, B. subtilis, B. licheniformis, A. faecalis, H. X, P. denitrificans, P. stutzeri. In addition, the effects of influent flow rates on and TN removal in this system are also presented in Figure 3. As can be seen, the removal efficiency of showed no obvious variation when the influent flow rate was varied between 1.5 and 5 L/h. The removal efficiency remained stable at ∼99%, and the corresponding concentration in the treated effluent was ∼1.2 mg/L, whereas the TN removal efficiency was increased slightly from 90 to 93%, and the concentration of TN in the treated effluent was reduced from 39 to 25 mg/L, when the influent flow rate was decreased from 5 to 1.5 L/h.
Figure 3

Performance of (a) NH4+-N, (b) TN removal and effect of influent flow rates in the IMBR system.

Figure 3

Performance of (a) NH4+-N, (b) TN removal and effect of influent flow rates in the IMBR system.

Figure 4

Nitrogen variations in the IMBR system flow path.

Figure 4

Nitrogen variations in the IMBR system flow path.

Figure 5 illustrates the effects of organic loading rates on removal in the IMBR system. From the figure, it is observed that, when the influent COD loading rate was in the range of 3.21–12.86 kg COD/(m3 d), the removal efficiency was about 99%, which showed the change of influent COD loading rate would not affect the removal. However, it is commonly known that high organic loading rates can seriously inhibit nitrification (Bernet et al. 2001; Im et al. 2001; Iaconi et al. 2003; Zhu et al. 2014). This is because high organic loading rates suppress the activity and growth of autotrophic nitrifiers, whereas, in contrary, the growth of heterotrophic bacteria is promoted. Furthermore, the heterotrophic bacteria competed with nitrifiers for oxygen, which further inhibited nitrification. Leta et al. (2004) utilized a predenitrification–nitrification pilot process for the treatment of tannery wastewater. They observed that the removal efficiency for declined from 70 to 40% when the organic loading rate was increased from 1.5 to 2.0 kg COD/(m3 d). Consequently, the results of our study clearly indicated that the IMBR system could achieve a high removal efficiency through nitrification, even at high organic loading rates and could endure higher loading perturbations. These phenomena might be because: (1) the nitrification unit was placed at the end of the IMBR system; hence, most of the carbon in the wastewater would be removed in the BA and BAF units, which led to lower influent COD loading to the nitrification unit; thus the nitrifying bacteria might be more competitive in growth and function; (2) the predominant bacterial strains used in this study were of heterotrophic nitrifying bacteria, which have been demonstrated by our previous study (data not published). In our previous study, we found that the heterotrophic nitrifying bacteria were capable of simultaneous heterotrophic nitrification and aerobic denitrification compared to the autotrophic nitrifying bacteria; so the organic matter might be used as electron donor and carbon and energy source for denitrification by the heterotrophic nitrifying bacteria. Hence, the COD and might be simultaneously removed by the heterotrophic nitrifying bacteria. In addition, the heterotrophic nitrifying bacteria had the potency to resist high organic loading rates (Yang et al. 2011).
Figure 5

The effect of organic loading rates on the NH4+-N removal efficiency of the IMBR system.

Figure 5

The effect of organic loading rates on the NH4+-N removal efficiency of the IMBR system.

Effect of alkalinity

Biological nitrogen removal was typically accompanied by alkalinity consumption or production. Theoretically, alkalinity decreases by 7.14 mg (as CaCO3) for each milligram of nitrified, and increases by 3.57 mg for each milligram of NO3-N denitrified. Figure 6 indicates the variations in alkalinity, , and TN in each reactor of the IMBR system. Due to the recycled effluent and BA, the alkalinity of the BA unit was decreased in comparison with that of the influent. The alkalinity of influent was ∼639 mg/L (as CaCO3), whereas the alkalinity of the effluent of the BA unit was 216.7 mg/L. In the AF unit, the concentration of was not noticeably altered, whereas the concentration of TN was reduced by up to 43 mg/L, which implied that the theoretical value of alkalinity production during denitrification was 153.5 mg/L. However, the actual increase of alkalinity was approximately 160.5 mg/L in the AF stage (from 216.7 to 377.2 mg/L). The actual increase of alkalinity was similar to that of the theoretical value. The alkalinity in the BAF unit was decreased, from 377.2 to 269.9 mg/L, which was attributable to the consumption of alkalinity via nitrification. In this stage, the concentrations of and TN were decreased by up to 23 mg/L and 13 mg/L, respectively, which implied that the total alkaline requirement was ∼117.8 mg/L. Further, because the microorganisms in the NBF unit had the capacity for simultaneous nitrification and denitrification, the and TN concentrations were reduced by 44.6 mg/L and 36 mg/L, respectively. This meant that the total alkaline requirement was ∼190 mg/L. Fortunately, the influent alkalinity of the NBF unit was 269.9 mg/L, which satisfied the alkalinity requirement of this stage. In light of the above discussion, it can be known that the total alkalinity consumption could be balanced by the alkalinity that was generated via denitrification and the alkalinity from the influent during the overall operational timeline of the IMBR system. Hence, the IMBR system might achieve a high and TN removal efficiencies without the supplement of external alkalinity.
Figure 6

The variations of alkalinity, NH4+-N and TN in the IMBR system flow path.

Figure 6

The variations of alkalinity, NH4+-N and TN in the IMBR system flow path.

Empirical performance modeling of the IMBR system by equations similar to Stover-Kincannon equation and Monod equation

For the formulation of the empirical performance modeling of the IMBR system, the following hypotheses were assumed: (1) the IMBR system was operated at a steady state; (2) the IMBR system was considered as a one-stage system without considering the different biological processes at each unit; (3) the microorganism concentration was considered as constant.

Taking the above assumptions into account, the equations similar to Stover-Kincannon and Monod model equations were used to describe the substrate degradation in this system. The equation similar to the Stover-Kincannon equation is given below (Sandhya & Swaminathan 2006; Jin & Zheng 2009): 
formula
1
where θH is the hydraulic retention time (d); S0 and Se are influent and effluent substrate concentration at steady-state (mg/L); Umax and KB are the empirical constants of the IMBR system.
In addition, the equation similar to the Monod equation is given (Hooshyari et al. 2009): 
formula
2
where X is the microorganism concentration in the system (g/L); ks and vmax are the empirical constants of the IMBR system.

From Table 5, it can be seen that the equation similar to the Stover-Kincannon equation is more suitable for describing the COD, , and TN removal in the IMBR system. The empirical performance modelings of COD, , and TN removal in the IMBR system were , , and , respectively. As a result, the effluent concentrations of COD, and TN could be predicted by using the empirical equation similar to the Stover-Kincannon equation under different influent conditions.

Table 5

Summary of empirical performance modeling of the IMBR system obtained from the different models applied

Kinetic models Substrate Kinetic equation R2 
Stover-Kincannon COD  0.999 
  0.999 
TN  0.999 
Monod COD  0.955 
  0.414 
TN  0.985 
Kinetic models Substrate Kinetic equation R2 
Stover-Kincannon COD  0.999 
  0.999 
TN  0.999 
Monod COD  0.955 
  0.414 
TN  0.985 

CONCLUSIONS

In our study, bioaugmentation with predominant bacterial strains from marine sediments led to the successful treatment of tannery wastewater in an IMBR system without the use of physicochemical pretreatment. Furthermore, this system had the capacity for reducing waste sludge and odor production. In the absence of physicochemical pretreatment and predominant bacterial strains, the performance of the IMBR system was relatively poor, and the effluent did not meet the discharge standards of tannery wastewaters in China. However, when the predominant bacterial strains were present, the performance of the IMBR system was very satisfactory, and high efficiencies were recorded for the removal of COD (97%), (97.7%) and TN (90%) without the use of physicochemical pretreatment, with the result that the treated effluent met the limits fixed by the enforced Chinese regulations. This system had a higher COD removal rate (12.65 kg COD/(m3 d)) than previous literature reported, and the COD removal rate was increased linearly with increases in influent COD loading rates. High COD loading rates did not affect nitrification in this system. Influent flow rates had no significant influence on the COD, , and TN removal. In addition, the system possessed excellent adaptability and ability to resist influent loading fluctuations, and had a good alkalinity balance such that it could achieve a high and TN removal efficiency without the use of external alkalinity supplements. The result of empirical performance modeling analysis revealed that the equation similar to the Stover-Kincannon equation was very appropriate for describing the performance of the IMBR system for the removal of COD, and TN.

REFERENCES

REFERENCES
Bernet
B. N.
Dangcong
P.
Delgenes
J. P.
Moletta
R.
2001
Nitrification at low oxygen concentration in biofilm reactor
.
J. Environ. Eng
.
127
,
266
271
.
China Environmental Protection Bureau
1989
Standard Methods for Water and Wastewater Analysis
.
China Environmental Science Publishing House
,
Beijing
.
He
J. G.
Liu
F.
Han
B. P.
Zhao
B. W.
Liu
J.
2012
Treatment of tannery wastewater with salt tolerant bacteria basing on different culture mediums
.
Adv. Mater. Res
.
403
,
625
633
.
Iaconi
C. D.
Lopez
A.
Ramadori
R.
Passino
R.
2003
Tannery wastewater treatment by sequencing batch biofilm reactor
.
Environ. Sci. Technol
.
37
,
3199
3205
.
Kim
J. K.
Park
K. J.
Cho
K. S.
Nam
S. W.
Park
T. J.
Bajpai
R.
2005
Aerobic nitrification-denitrification by heterotrophic Bacillus strains
.
Bioresour. Technol
.
96
,
1897
1906
.
Kim
I. S.
Ekpeghere
K.
Ha
S. Y.
Kim
S. H.
Kim
B. S.
Song
B.
Chun
J.
Chang
J. S.
Kim
H. G.
Koh
S. C.
2013
An eco-friendly treatment of tannery wastewater using bioaugmentation with a novel microbial consortium
.
J. Environ. Sci. Health. A
.
48
,
1732
1739
.
Sivaprakasam
S.
Mahadevan
S.
Sekar
S.
Rajakumar
S.
2008
Biological treatment of tannery wastewater by using salt-tolerant bacterial strains
.
Microb. Cell Fact
.
7
,
1
7
.
Song
Z.
Williams
C. J.
Edyvean
R. G. J.
2003
Tannery wastewater treatment using an upflow anaerobic fixed biofilm reactor (UAFBR)
.
Environ. Eng. Sci
.
20
,
587
599
.
Stasinakis
A. S.
Mamais
D.
Thomaidis
N. S.
Lekkas
T. D.
2002
Effect of chromium(VI) on bacterial kinetics of heterotrophic biomass of activated sludge
.
Water Res
.
36
,
3342
3350
.
Wiegant
W. M.
Kalker
T. J. J.
Sontakke
V. N.
Zwaag
R. R.
1999
Full scale experience with tannery water management: an integrated approach
.
Water Sci. Technol
.
39
,
169
176
.
Yusuf
R. O.
Noor
Z. Z.
Hassan
M. A. A.
Agrry
S. E.
Solomon
B. O.
2013
A comparison of the efficacy of two trains of Bacillus subtilis and Pseudomonas fragii in the treatment of tannery wastewater
.
Desalin. Water Treat
.
51
,
3189
3195
.