Moving bed biofilm reactor for treatment of Kraft pulp effluent with high organic load rate

The pulp industry uses more than 40 m of water per ton of pulp produced, generating high effluent flows. In general, it presents high concentrations of organic matter, color and ecotoxicity. The most widely used effluent treatment systems in the pulp industry are biological, including moving bed biofilm systems that are efficient in removing biodegradable organic matter. This work evaluated the removal of organic matter, total phenolic compounds, color and lignin derivatives in the treatment of Kraft cellulose effluent using the moving bed biofilm reactor (MBBR), and also evaluated the support media biofilm development by solid analysis and scanning electron microscopy. The parameters evaluated during treatment were: BOD5, COD, color, total phenolic compounds, lignin derivatives, and solids, with tests performed on organic loads from 0.7 and 8.9 kgCOD m -3 d. Organic matter removal remained stable, being over 80% to BOD5 and over 42% to COD. The color and the total phenolic compounds were removed up to approximately 7 and 28%, respectively. Over 19% removal of derivatives of lignin compounds was observed in both organic load rates. In the MBBR, biofilm was confirmed and enabled this biological system to treat the cellulose effluent in a stable way.


INTRODUCTION
The Kraft pulping process is the most widely used pulp production in the world. There is high water consumption in the pulp industry, as more than 40 m 3 of water is consumed per ton of pulp produced, which in turn generates large volumes of effluent. In general, this effluent is rich in organic matter, suspended solids, resin acids, lignin, color and ecotoxicity (Bachmann, 2009;Bakraoui et al., 2020;Furley et al., 2015;2018;Hossain and Ismail, 2015;Toczyłowska-Mamińska, 2017).
Most industrial effluent treatment systems are biological, and these are also the most used in the pulp and paper industry (Cabrera, 2017). In these industrial sectors, activated sludge, biological filters, aerated lagoons, moving bed biofilm reactors (MBBR), etc. are most often used. (Hubbe et al., 2016).
The MBBR system consists in keeping part of the mixed liquor in suspension in the liquid mass, and part adhered to the support media. It uses hydraulic retention times of 3 to 48 hours. The support media provide adhesion surface for microbial biomass and present high mobility (Leyva-Díaz et al., 2017). The support media applied in this type of biological treatment are generally made of high-density polyethylene or polypropylene. They are inert and have a high specific surface area (>1000 m 2 m -3 ) (Leyva-Díaz et al., 2017).
Among them, there is AQUAPOROUSGEL ® (APG) from Nisshinbo Chemical Inc. It presents specific surface area greater than 3000 m 2 m -3 , specific mass of 30 drykg/wetm 3 , multiporous, and its material is based on polyethylene-glycol (Sakuma, 2004). According to the manufacturer, the filling ratio for an efficient treatment is 10% of the reactor volume, a value lower than other support media, which are in the range of 25 to 70% (Haandel and Lubbe, 2012;Sakuma, 2004).
This work evaluated the performance of a moving bed biofilm reactor with spongy support media with regard to organic matter, total phenolic compounds, color, and derivatives lignin compounds removal, and also evaluated biofilm development in the support media during treatment.

MATERIALS AND METHODS
The effluent used for the treatment with MBBR (Moving Bed Biofilm Reactor) was obtained from an industry of unbleached Kraft pulp in Curitiba region, Brazil. The effluent was collected twice at different times before biological treatment, transported and stored at 4°C protected from light in 30 L vessels. The treatment occurred in a continuous flow reactor, in lab scale, with 1 L of volume, filled with 10% of spongy support AQUAPOROUSGEL ® (APG), 8 cm 3 per unit, by Nisshinbo Inc. The MBBR reactor operation was at two different organic load rates (OLR) over a total of 48 days of operation.
The operation of the reactor occurred at two different organic loading rates (OLR): 0.7 kgCOD m -3 d -1 and 8.9 kgCOD m -3 d -1 (Phases I and II, respectively). The hydraulic retention times (HRT) were 13.3 and 2.8 h to achieve those two loads worked. Those organic load rates were based on Peitz and Xavier (2017) and Vanzetto et al. (2014), who applied a load of 9.0 kgCOD m -3 d -1 and HRT between 2 and 3 h with real Kraft pulp effluent and AMB media support.
The treatment lasted 48 days, 30 days in Phase I, as an adaptation step for growing biomass by itself without inoculation, and 18 days in Phase II until steady state. The steady state was reached when the system got as far as organic matter removal variation (COD and BOD5) of up to 10%.
Adhered solids (AS) were considered as the biofilm formed in the spongy support during biological treatment. The biofilm adhered to the spongy carrier was quantified by suitability of the APHA et al. (2012), in which the APG samples were subjected to a 100 mL volume ultrasonic bath in deionized water for one (1) hour, being sequentially conducted according to the same method. Suspended solids (SS) were considered as the biomass present in the mixed liquor and were measured according to APHA et al. (2012). Both determinations were performed in triplicate during the steady state of each applied organic load rate (Phase I and II). The results were expressed by total and volatile adhered solids (TAS and VAS) and total and volatile suspended solids (TSS and VSS), as mg L -1 (APHA et al., 2012).
Scanning electron microscopy (SEM) analyses were performed at the Multiuser Materials Characterization Center of the Federal Technological University of Paraná (UTFPR). For the analyses, samples of the spongy support medium were collected at the end of each applied organic load rate (Phase I and Phase II), dried in a lyophilizer and stored in a desiccator for further analysis of SEM.

Characterization of the effluent
In Table 1 is shown the characterization of two samples collected from the same industry of Kraft mill effluent treated over 48 days by moving bed biofilm reactor (MBBR).  Table 1 showed that the two different samples of effluent from the same industry present organic matter, color, phenolic compounds, and derivatives of lignin compounds. During Phase I, the sample treated had the highest BOD5/COD ratio (0.34), which was favorable to the biological treatment (Hubbe et al., 2016). The effluent treated during OLR 0.7 kgCOD m -3 d -1 had total phenolic compounds, 73.1 mg L -1 , with low concentration compared to literature (Hossain and Ismail, 2015;Morales et al., 2015).
In the next Phase, with effluent from another collection from the same industry, the BOD5/COD ratio decreased to 0.21; however, it is still favorable to the treatment. According to Hubbe et al. (2016), pulp and paper effluent has BOD5/COD between 0.05 -0.50, and the closer to 1.0, the better the biological treatment. This value of 0.21 for BOD5/COD ratio indicated higher recalcitrance in the sample, and there is a greater presence of lignin derivatives compounds, aromatic and phenolic compounds, all difficult to biodegrade and with high color, as proposed by Maria et al. (2014).
In general, the characteristics of the samples in this work agree with those found by different authors (Hubbe et al., 2016). Figure 1 shows the removals calculated during treatment of the Kraft pulp mill effluent in the MBBR. Regarding organic matter, the system had an average efficiency of 42.6% of COD removal at the lowest load, remaining stable, and 41.8% removal of this parameter in the OLR of 8.9 kgCOD m -3 d -1 . Compared with the MBBR system treating Kraft pulp effluent with phytosterols, in HRT of 3h, COD removal was 40% (Peitz and Xavier, 2017). Regarding BOD5 removal, even with the difference between the BOD5/COD ratio in these 5 Moving bed biofilm reactor for treatment of … Rev. Ambient. Água vol. 15 n. 4, e2512 -Taubaté 2020 phases, (Table 1, 0.34 → 0.21), a decrease of only 7.7% in the efficiency average removal was found. In the load of 0.7 kgCOD m -3 d -1 , 88.8% of BOD5 removal was obtained and in the next phase it was 81.9%, showing how stable the MBBR is with high organic loads and shock loads, as was observed by other authors with different media support (Huang et al., 2015;Leyva-Díaz et al., 2017;Qiqi et al., 2012). The treatment of Kraft pulp effluent by MBBR with APG presented stable removal of organic matter even with low value hydraulic retention time (2.8 h).

Organic matter, color, total phenolic, and lignin derivatives compounds
Regarding color, removal of this parameter was observed in both organic loads, between 22.1% and 40.7% on average. These results were better than those of Peitz and Xavier (2017), who obtained 3% color removal at a HRT of 3 h. This was unexpected in a biological treatment, in which an increase in the effluent color was observed in reactors similar to the moving bed biofilm reactor (Cabrera, 2017;Kamali and Khodaparast, 2015;Wahyudiono et al., 2008).
This behavior can be explained by the fact that the spongy support has adsorbed the effluent color during treatment by MBBR, i.e., the lignin compounds present in the effluent of the Kraft pulp industry, which are mainly responsible for the color, were adsorbed on the sponge material of the APG, which was initially white and in the early stages of the reactor operation became brown. This color changed from brown to black at around 45 days of operation. It is also possible that the MBBR-treated samples were predominantly built of low molecular weight compounds that were metabolized due to their size (Cabrera, 2017).
Regarding the removal of total phenolic compounds (TPC), also shown in Figure 1, the efficiency of TPC removal was about 21% in Phase I and 36.6% in the next load. Peitz and Xavier (2017) obtained 36.4% of TPC removal in OLR 9.0 kgCOD m -3 d -1 , in a similar condition with real Kraft pulp effluent. These removal levels could be explained because of the action of certain bacterial communities. There are some effective species in breaking down recalcitrant compounds like lignin derivatives and phenolic compounds, reducing the color and phenolic compounds of pulp mill effluent this way, even if the removal of color and phenolic compounds in aerobic treatments is not common, as observed by others authors (Cabrera, 2017;Hubbe et al., 2016;Muñoz et al., 2019;Singh et al., 2019).
During the treatment, the removal of the aromatic (UV254nm), ligninic (UV280nm) and lignosulfonic (UV346nm) compounds was observed in all organic loading. Removals were greater than 19% in all parameters in both organic loads applied. The highest removal level was during OLR 8.9 kgCOD m -3 d -1 , 22.4% -24.3% for lignin derivatives compounds.
Regarding UV254nm/UV280nm ratio, the reduction of the values occurred during the biological treatment of 1.18 to 1.14, during ORL 0.7 kgCOD m -3 d -1 and 1.04 to 1.01 in Phase II. These values indicated higher amounts of aromatic compounds (UV245nm) in all samples compared to the lignin compounds (UV280nm), which is in agreement with other studies with cellulose effluent (Çeçen, 2003;Morales et al., 2015;Peitz and Xavier, 2017;Villamar et al., 2009;Machado, 2017).

Biomass adhered and suspended
The result of the quantification of biomass adhered to the support media (AS) and suspended in MBBR (SS) measured during steady state is presented in Figure 2. During the treatment, the system was not inoculated with sludge.
During the aerobic treatment, the biomass grew, mainly attached to APG support media. Volatile Adhered Solids (VAS) of 770.4 mg L -1 were observed in the load 0.7 kgCOD m -3 d -1 (Figure 2A). Volatile Suspended biomass (VSS) reached a concentration of 233.3 mg L -1 at load 0.7 kgCOD m -3 d -1 , lower than expected (1500-5000 mg L -1 for conventional activated sludge), and 3.3 times lower than VAS. This preference for adhered solids in support media was observed by other authors (Minegatti et al., 2012;Qiqi et al., 2012;Vanzetto et al., 2014). In the next phase, OLR 8.9 kgCOD m -3 d -1 , both concentrations grew up to 3419.2 and 1432.5 mg L -1 , respectively (Sperling, 2014;Toczyłowska-Mamińska, 2017). Regarding Total Solids, TS, Figure 2B, it arrived at 3831 mg L -1 in the second phase in the support media (TAS), in agreement with the literature, 3000 -4000 mg L -1 (Barwal and Chaudhary, 2014). Regarding Total Suspended Solids, the concentration was 2227.5 mg L -1 . Table 2 shows the ratios for the different loads applied during treatment by the APG spongy support MBBR reactor. For OLR 0.7 kgCOD m -3 d -1 , the VSS/TSS ratio was 0.38, characterized as a stabilized sludge, with load increase in Phase II, this ratio up to 0.64, as a partially stabilized sludge with high organic matter content, since the biomass was not removed from the system during the treatment (Sperling, 2014). Compared with the VAS/TAS ratio, the sludge is characterized as a partially stabilized sludge, mainly due to the increase of organic load in the system.
These results are in accordance with SEM images, Figure 3, in which were verified biofilm growth in all phases.
In Figure 3A-B, the APG support media has a high surface area with pores of different diameters for biofilm growth. During aerobic treatment, biomass grew mainly linked to support media such as biofilm. In Figure 3B-C, Phase I, with 13.3h HRT, biofilm growth did not fully cover the micropores, allowing biofilm development. At OLR 8.9 kgCOD m -3 d -1 , Figure 3E-F, growth continued and the increase was observed by VAS analysis and formation of extracellular polymeric material seen by SEM.
With the increase in organic load there was a greater development of biofilm ( Figure 3E-F), which cooperated in the removal of the parameters analyzed in this biological treatment.

CONCLUSIONS
The removal efficiency of organic matter remained stable in relation to COD, remaining above 40%, not presenting large discrepancies between the applied loads, constituting MBBR with APG as a stable system. Considering the removal of BOD5, a drop of 7.7% being justified by the low biodegradability of treated Kraft pulp effluent, showing the stability of the system in load shocks.
The values achieved for color removal were close to 40% on average for the highest load, which was not expected from a biological treatment. The removal efficiency for total phenolic compounds (TPC) was not negatively affected by the 15-times increase in load, which could be considered another favorable point for the stability of the MBBR with spongy support APG. Regarding biomass, its affinity with the support medium APG was verified because its concentration was 3.3 and 2.4 times higher than the biomass suspended in the reactor in each organic load rate applied. The VSS/TSS ratio for suspended solids was a partially stabilized sludge.
To sum up, it can be affirmed that the MBBR system based on support medium APG on OLR 8.9 kgCOD m -3 d -1 has the potential to treat Kraft cellulose.