The assessment of the coke wastewater treatment efficacy in rotating biological contractor

The state of the art wastewater treatment trends are focused on biological methods. They are effective in municipal wastewater treatment but, in the case of industrial sewage, biology-based methods can be difficult or, in some cases, impossible to use. The substances in industrial wastewater can be harmful to the bacteria performing biological treatment, causing bacterial community impoverishment and even its total destruction. Coke wastewater belongs to the industrial wastewater which is difficult for biological treatment due to its complex and changeable composition and the presence of such toxic substances as ammonia, thiocyanides, cyanides, sulfides, phenols and polycyclic aromatic hydrocarbons (Park et al. 2008; Chu et al. 2014). In Poland, especially in areas with heavy industry, development in coke production, as well as coke wastewater volume, is increasing. To fulfill European Union regulations towards high quality effluent obtainment, modern technological systems are required. This has led to the search for effective wastewater treatment technologies.

When considering wastewater treatment, nitrogen is one of two main biogens which should be eliminated from the wastewater effectively. In a typical wastewater treatment plant (WWTP) system a combination of nitrification and denitrification is used. However, in the case of coke wastewater, the high concentration of toxic pollutants can heavily inhibit the biological activity of nitrifying and denitrifying bacteria (Amor et al. 2005; Eiroa et al. 2005; Kim et al. 2008). Nevertheless, there is an effort to develop biological treatment methods for such streams as they are cheaper and more environmentally friendly than chemical methods. For biological treatment of coke wastewater, various reactor types and configurations have been suggested, such as anoxic–aerobic reactors, anaerobic–anoxic reactors, sequencing batch reactors, fixed bed biofilm reactors, membrane-based reactors and others (Zhao et al. 2009; Gu et al. 2014). Among various proposed processes, the pre-denitrification process deserves particular attention due to its simplicity and economic benefits (Kim et al. 2008; Park et al. 2008). More recently, the ammonia oxidation (nitritation)–Anammox process has been developed for treatment of nitrogen rich streams (Strous et al. 1997). The main advantages of this process compared to the conventional nitrification/denitrification are: low sludge production, a decrease in aeration costs by almost 60% (only half of the ammonia is oxidized to nitrite in the nitritation process without further oxidation to nitrate), and no need for an external organic carbon source addition (in the Anammox process) (De Clippeleir et al. 2011). Additionally, Anammox enables a lowered CO₂ emission in comparison to nitrification/denitrification. The disadvantage of the Anammox process is slow biomass growth and difficulties in creating favorable technological conditions. The Anammox process is also sensitive to various factors such as substrates (ammonia and nitrite), organic matter, salts, heavy metals and many others (Jin et al. 2012). There are also reports showing that phenol, which is present in coke wastewater, inhibits the Anammox process (Toh & Ashbolt 2002; Yang et al. 2013; Pereira et al. 2014). The addition of phenol not only suppresses the Anammox activity but also changes the stoichiometrics ratios and the microbial community structure and composition (Yang et al. 2013; Pereira et al. 2014). What is more, due to the presence of refractory and toxic compounds in coke wastewater, the nitrifying bacteria and other specialized microbes (as Anammox) can be easily washed out of the system due to their inhibited growth. Moreover, when the activated sludge is operated under high loading rates, problems with poor settleability were reported (Gu et al. 2014). As it has been previously described (Xiao et al. 2009; Langone et al. 2014), industrial wastewater such as coke wastewater can cause damage to activated sludge – a mixture of bacteria, metazoa and protozoa functioning in WWTP's bioreactors as suspended flocs. That is why biofilms, three dimensional structures containing extrapolimeric substances as a protective matrix, can be useful in difficult wastewater treatment without harming the bacterial community (Hall-Stoodley et al. 2004). The technological system containing biofilm which could be used in coke wastewater treatment is the rotating biological contractor (RBC).

The aim of this work was to estimate the possibility and the effectiveness of phenol and nitrogen rich wastewater treatment in laboratory-scale rotatory biological contractor. The experiment was divided into two parts: with synthetic phenolic wastewater and with real coke wastewater as an example of the industrial wastewater containing phenol and high concentration of nitrogen. The biofilm bacterial community was monitored with polymerase chain reaction–denaturing gradient gel electrophoresis (PCR-DGGE) in order to present the impact of the medium change on the bacteria performing the removal processes.

The experiment was performed for 719 days with synthetic medium (Table 2), then for 192 days with real coke wastewater from Jadwiga coke plant in Zabrze Poland (Table 3).

During the research, samples were collected from the influent, effluent and each stage of the contactor. The efficiency of the biological treatment was followed in terms of the general parameters such as: chemical oxygen demand (COD) (dichromate method), phenol concentration (Merck tests), ammonia nitrite, and nitrate nitrogen forms (Merck tests). The process was monitored by measuring other parameters: flow rate temperature, pH (WTW pH 340i), and dissolved oxygen (WTW Oxi 340i). No specific heating was applied and the temperature was kept at the level of 20.2 ± 2.2 °C.

The RBC used in this experiment was previously performing nitrogen removal from landfill leachate (Cema 2010) and then reject water derived from municipal WWTP in Zabrze, Poland was treated (data not published). The influent NH₄-N and COD concentrations were on the level of 1,010 mg/L and 1,427 mg/L, respectively. Microbial analysis (fluorescence in situ hybridization (FISH)) confirmed the coexistence of nitrifiers and the Anammox bacteria belonging to Candidatus Brocadia anammoxidans and/or Candidatus Kuenenia stuttgartiensis in the RBC.

Biofilm samples (volume of 50 ml) were collected from all three of the RBC chambers, vortexed and stored at −20 °C until DNA isolation.

Total genomic DNA was extracted from 0.2 g of the biofilm samples using the mechanical method. The samples were washed three times with 1× phosphate-buffered saline (PBS) buffer (Sigma) and disintegrated with bead beating (Roth, Germany) in lysis buffer (Tris–HCl 100 mM, EDTA 100 mM, NaCl 1.5 M; pH = 8.0). The samples were incubated 20 minutes in 1,400 rpm and 200 μl 10% sodium dodecyl sulphate (SDS) was added. After 30 minutes of incubation in 65 °C samples were centrifuged twice at 13,000 rpm and placed on spin filters (A&A Biotechnology). DNA attached to the filter was washed twice with 70% ethanol solution (A&A Biotechnology). The amount of DNA was measured spectrophotometrically using Qubit (Invitrogen) and stored at −20 °C until PCR amplification.

Partial 16S rRNA gene amplification of all the bacteria was performed using primers: 338f-GC and 518r gene fragment (Muyzer et al. 1993). PCR reaction was performed in 30 μl mixture and the amplification was performed in thermocycler T-1000 (Bio-Rad) as previously described (Ziembińska et al. 2009).

The DGGE of the PCR products obtained in reactions with 338F-GC/518R primers underwent electrophoretic separation in the DCode Universal Mutation Detection System (BioRad). Polyacrylamide gel (8%, 37:1 acrylamide–bisacrylamide, Fluka) with a gradient of 30–60% denaturant was prepared according to the manufacturer's instruction. The gel was run for 17 h at 40 V in a 1× TAE buffer at a constant temperature of 60 °C and stained as previously described (Ziembińska et al. 2009).

The analysis of DGGE fingerprints was performed using Quantity One 1D software (BioRad). Bacterial biodiversity was estimated on the basis of densitometric measurements and Shannon diversity index as previously described (Ziembińska et al. 2009). Dendrogram was constructed on the basis of the neighbor-joining algorithm with Dice coefficient.

In this experiment, the RBC was fed for 719 days with synthetic phenolic wastewater (period I) and for 192 days with real coke wastewater (period II) derived from Jadwiga coke plant in Zabrze, Poland. The temperature during the experiment was 20.2 ± 2.2 °C and was much lower than the temperature of 37 °C which is usually reported as an optimum value for the Anammox process (Schmidt et al. 2003). The pH value in the inflow did not exceed 8.3 ± 0.3. During the operational period I, a slight increase in the pH value between the influent and chamber I was observed (to 8.6 on average) and then a slight decrease to 8.0 on average. This phenomenon can be explained by three different processes overlapped in the contactor. During the nitrification process, a decrease in pH value is normally observed due to alkalinity reduction. The alkalinity depletion is partially recovered during denitrification process by production of one equivalent of alkalinity for one equivalent of NO₃-N reduced. Additionally, an increase in pH value is expected in the Anammox process due to consumption of hydrogen ions during cell synthesis. In the part of the experiment utilizing real coke wastewater treatment, the pH value dropped to 6.5 in the first chamber. The average hydraulic retention time was 4 days and the average wastewater inflow was 3.5 L/d.

In the first part of the experiment (period I) when the synthetic wastewater was directed to the system, the average COD value in the inflow was 1264.6 ± 454.4 mgO₂/L, COD removal was 62.5 ± 22.5%. From day 719 of the experiment real coke wastewater was directed to the system with organic load ranging between 350–675 mg O₂/L (the average COD value in the influent was 495 mg O₂/L, median was 425 mg O₂/L). During the total length of the real wastewater treatment (period II), the organic matter removal was unstable and it fluctuated between 1 and 80%, with an average value of 45.3%.

In the beginning of period II, a breakdown in ammonium nitrogen removal was observed. It seems that toxic substances present in real coke wastewater seriously affected the nitrification process. After 15 days of real wastewater dozing, the ammonium nitrogen removal dropped to only 10.9%. However, after that time, a gradual process of restoration was observed and after 50 days the ammonium oxidation exceeded 50%.

The COD requirement for nitrate denitrification is 2.9 g COD/g N-NO₃. Courtens et al. (2014) showed, that for denitrification with acetate as carbon source, the ration of 3.6 gCOD_removed/gN_removed was measured in mesophilic condition, and additionally for lower temperatures, this ratio was even higher. In our reactor, the ratio of COD_removed/N_removed was equal to 1.5 (between days 430 and 719 of the experiment) so it may be stated, that denitrification could not be the main process responsible for nitrogen removal. Especially if we consider that the main part of COD was removed in the first chamber (COD removal at the level of 55% with ca. 20% of nitrogen removal), whereas the nitrogen was removed mainly in the second chamber of RBC (effectiveness ca. 61%) (Figure 3(a) and (b)). In the first chamber, there was additionally a very high concentration of free ammonia (even over 200 mg NH₃/L) which is considered an Anammox process inhibitor even at low concentrations of 20–25 mg NH₃/L for continuous operation (Fernandez et al. 2012). High free ammonia concentration is also an inhibitor of the nitrification process, thus, the main nitrogen conversion took place in the second chamber with the Anammox process responsible for the nitrogen removal as COD being removed mainly in first chamber. Generally, according to Jenni et al. (2014) it should be stated that in a single-stage process with partial nitritation and Anammox with high COD/N ratio, it is not possible to assess the activities of the different bacterial groups based on mass balances. Additionally, in some cases the process instabilities were observed in systems, where elevated COD/N ration were in the influent to the system (Jenni et al. 2014).

The change from synthetic to real coke wastewater caused the nearly complete inhibition of nitrogen removal. Directly before medium change from synthetic to real wastewater, the nitrogen removal averaged over 75%, while in the first day after the real coke wastewater addition, it plummeted to 34%, and by day 15, a further drop to only 6% (Figures 4(a) and 5). At the end of the experiment, the nitrogen removal increased slightly to a value of 43%. This process was performed mainly in the first RBC chamber as opposed to the synthetic wastewater where nitrogen removal predominated in the second RBC chamber (Figure 4(b)). Such a decrease in the nitrogen removal effectiveness can most probably be explained by the toxic compounds present in the real coke wastewater influencing bacterial performance in RBC biofilm.

During the synthetic wastewater treatment period, ammonium removal efficiency was at the level of 96.7 ± 4.4%; however, in order to reach this high efficiency, a very long start-up period for the Anammox process (190 days) was required. Stable nitrogen removal of over 70% was achieved after more than 400 days of the experiment; however, the process could be successfully operated even at a relatively low temperature of around 20.2 ± 2.2 °C. The change of synthetic to real wastewater caused a temporary process break down but, at the end of the experiment, a process restoration was observed. The research performed on the RBC revealed that coke wastewater is a difficult type of sewage to be treated using biological methods; nonetheless, most likely due to the protective role of biofilm matrix, such a biocenosis could be a tool in such treatment. On the basis of the research performed in this experiment, we could state that the biodiversity of the specialized community for coke wastewater treatment is lower than the one treating synthetic wastewater and the community is composed mainly in GC-rich bacterial genotypes. Probably, the GC-rich group of bacteria would be the community specialized enough to treat recalcitrants (e.g. phenols, polycyclic aromatic hydrocarbons (PAHs)) present in real sewage.

This research was supported by the Polish Ministry of Science and Higher Education, grant no. N N523 56213.