PRELIMINARY TREATABILITY TEST OF A NON-CONVENTIONAL INDUSTRIAL WAS TEW ATER IN THE WOOD SECTOR: COD AND FORMALIN REDUCTION

Industrial activities commonly discharge a broad range of synthetic compounds directly into water recipients without previous treatment. Despite the existence of available technologies for industrial wastewater treatment, better understanding of treatment processes is still required, since these waters are relatively complex and usually contain either persistent or recalcitrant compounds. Treatment systems with low costs of implementation, operation and maintenance as well as energy and labor saving processes ought to be developed. Biological treatment systems are potentially good options to meet these requirements. In this study, a preliminary investigation in lab scale was carried out with a Sequencing Batch Reactor -SBR used to treat a non-conventional industrial wastewater generated by a wood-floor industry, located in Nybro, Sweden. The study focused on: (i) formalin reduction in aqueous phase and; (ii) COD reduction. The proposed SBR reached a high efficiency in reducing formalin within the aqueous phase (from 53% to 98%) suggesting the use of formalin by the microorganisms as a primary carbon source. On the other hand, COD reduction (-34% to 73%) was not satisfactory, which is probably related with the presence of polymeric compounds with high molecular weight in the urea-formaldehyde resin, Recommendations for the system improvement are: (i) effluent recirculation; (ii) longer filling periods during each batch and; (iii) both primary and secondary settling/ sedimentation. As a result of very high initial COD contents (ranging between 736-5608 mg L·\ even though a high percentage of reduction is achieved, the final effluent would still not meet the threshold limits for effluent discharges in water bodies. Additional treatment options could be advanced oxidative processes such as ozonation and Fenton.


synthetic compounds are used by different manufacturing processes being discharged afterwards in water bodies without previous treatment [ 1], Despite the availability of technologies to treat industrial wastewaters, there is a need for efficient systems with low costs of implementation, maintenance and operation.
In terms of economical feasibility, when compared to other technologies, biological systems are in the frontline, being much less costly [2]. Biological treatment has been widely accepted and used for either partial or complete stabilization of biologically degradable substances, present in wastewaters [2]. It is widely reported in literature the use of anaerobic processes in many industries for the treatment of wastewater with relatively high content of organic pollutants due to its performance, cost and energy saving advantages [3, 4, 5, 6], Industrial wastewaters, particularly those generated after washing/cleaning of machineries and equipment vary both quantitatively and qualitatively in time, which threat the proper function of biological treatment systems. Therefore, there is still a demand for knowledge development regarding operational parameters and how to achieve optimum conditions and improved treatment efficiency [7,8]. In the present study, the biological approach for treating a wastewater stream generated in a wood floor industry located at Nybro, Sweden is preliminarily evaluated. The wastewater is generated at the cleaning/washing procedures of machineries that continuously apply urea formaldehyde resins on the wood particle boards. The main characteristics of such wastewaters are the high contents of fonnalin and very high content of COD. The objective in this stage of the investigation was to assess the treatability of this wastewater through biological processes and the performance of the treatment system, with focus on: reduction of formalin in aqueous phase and; reduction of COD.

MATERIAL AND METHODS
The bench-scale experiment was divided in two stages: Stage I: Microbial growth and attachment onto an inorganic filter media; Stage 2: Treatability experiment.
Due to the intermittent generation of the wastewater to be treated, a Sequencing Batch Reactor (SBR) was proposed (see Figure 1). The system, specifically designed and built up for this experiment consisted of two cylindrical reactors ( external diameter 10 cm and height 21 cm) functioning in sequence. The first phase consisted in anaerobic conditions and the second in intermittent aeration to establish alternated aerobic-anoxic conditions (see Table 1). The anaerobic phase included the following steps: (i) filling; (ii) anaerobic degradation sedimentation and (iii) discharge. The aerobic-anoxic phase included: (i) filling; (ii) aeration aerobic degradation; (iii) aeration interruption-anoxic phase and; (iv) discharge.

Figure 1: SBR lab-scale constructed f or glue wastewater treatment: (an) anaerobic phase; (ae) Intermittent Aeration (aerobiclanoxic phase),
The temperature during the anaerobic and aerobic-anoxic phase ranged between 25-32 °C (kept through a thermal plate underneath the reactor) and 20-23 °C respectively (room temperature), Whereas the filling procedure was done with a peristaltic pump (Watson Marlow-Ali tea M20 10-100 rpm Dual channel) in an up-flow mode, the discharge was done in a down-flow by gravity, Before filling the anaerobic phase, the influent had the pH adjusted to the range between pH 6-7 either with a NaOH (IM) or HCI (0. 1 M), except for those batches where the intention was to investigate the influence of the pH over the system efficiency.
Stage I -Microbial growth and attachment phase: Activated sludge taken from the Kalmar Municipal Wastewater Treatment Plant (K WWTP) was used as inoculum to promote microbial growth and attachment onto the filter carriers, The sludge was stored at 4 °C, In order to put the microorganisms in contact with the carriers, 600 ml beackers were filled with 100 ml (4.e16 g ML VSS/L) of activated sludge plus synthetic medium (see Table 2) and the filter carriers (ceramic and polymers), The beackers were immersed in a water bath, kept at constant temperature of 30 °C, During a one-month period, every second day the activated sludge within the beackers was carefully disposed off and replaced with a fresh batch prepared with the same material previously mentioned (activated sludge+ synthetic medium). The procedure was handled carefully in order to avoid biofilm slough off After 30 days, filter carriers were put into the cylindrical reactors and the treatability experiment had commenced,   Table 3.

RESULTS
COD reduction: The total COD reduction along 26 cycles during four months (average of 6.5 cycles per month) ranged between -34% and 73% (Figure 2). The observed oscillation in efficiency can be related to the presence of polymeric compounds in the glue, with a relatively high molecular weight characterized by persistence and in the worst case, recalcitrance [9] and changes in the running conditions, such as nutrients availability. As a consequence, whereas in some cycles (5, 6, 9, 11, 22 and 23) the system achieved a satisfactory efficiency, in others, it didn't During Cycle 3, for instance, the insignificant reduction of total COD (3%) was probably due to the previous metabolic stress suffered by the system, with higher retention times than usual and a consequent starvation process. According to [IOJ the lack of nutrients within a SBR resulted in inhibition of microbial metabolism with a low rate of biological transfomiation and mineralization of organic compounds. Moreover, during Cycle 19, the total COD reduction was negative (-34%), suggesting either the production of particularly persistent or recalcitrant by-products as a result of biological transformation of the original compounds in the influent, a phenomenon already described in the literature [II]. According to [11], the majority of the organic compounds in an effluent after biological treatment was not detected in the original influent, but were actually produced by the system.

Figure 2: Total COD reduction achieved by the SRB including all phases.
In order to identify the treatment phase responsible for the COD increase, Figure 3 shows the reduction in each phase of the biological process (anaerobic-aerobic-anoxic). During Cycles 6 and 1 9, the anaerobic phase was responsible for a COD increase. It can also be observed during the anoxic phase; low negative reductions in Cycles 1 1, 1 7, 1 8, 20 and 21 (see Figure  3 ).

Figure 3: COD reduction efficiency of the different biological phases in the SBR.
An interesting fact was observed during the cycle 19, when the system performed its lowest efficiency regarding formaldehyde reduction (53%) and COD (-34%). This cycle coincided with the lowest values of formaldehyde and COD in the influent (24 mg L.e 1 and 737 mg L'e 1 respectively).
Regardless the pH variation at the anaerobic influent, the system presented a buffering capacity in the range of 6 < pH < 7, principally taking into account the cycles 9-13 and 17-24 ( Figure 6). The outlet pH in the acidic range suggests the presence of organic acids, and a formaldehyde shock loading (221 mg HCOH L' 1 da/) at cycle 25 might be the principal cause of an outlet pH of 3.9. [13], highlight that a typical response of organic shock loading is that pH decreases down to an acidic range as a consequence of organic acids accumulation, Figures 7 and 8 show the pH variation within the intermittently aerated reactor. It is clearly observed that whereas the pH is raised during the aerobic phase, when the aeration is turned off a decreasing trend for the pH occurs. No significant variation of inlet and outlet pH is observed, being reasonable to associate that with a buffering capacity of the system as a whole, both for the neutral (anaerobic phase) and alkaline range. Further investigation of the nitrification-denitrification process is expected to bring better understanding as regards pH dynamics along the treatment system, principally within the intennittently aerated phase. Similarly to the anaerobic reactor, a pH decrease after cycle 25 suggests the system was negatively affected by formaldehyde shock loading as already mentioned before.

CONCLUSIONS
According to the results, the proposed system in a bench scale showed capacity to remove with high efficiency formaldehyde (53% to 98%), suggesting the utilization of fonnaldehyde as a carbon source by microorganisms. On the other hand, the system did not achieve a satisfactory COD reduction (-34% to 73%), which might be related to the presence of either persistent or recalcitrant polymeric compounds with high molecular weight at the Urea formaldehyde resins. Recommendations for further investigation are: (i) effle uent recirculation; (ii) longer feeding periods; (iii) primary and secondary settling/sedimentation systems. It is important to emphasize that even though the treatment system might be able to achieve high efficiency in reducing COD, very high initial COD values would still result in effluents with values above the threshold limits for effluent discharge. Options for a final polishing include advanced oxidation processes such as ozonation and Fenton.

AKNOWLEDGEMENTS
The CAPES Foundation -Brazil Ministry of Education is acknowledged for the first author's PhD scholarship. The authors are thankful to the financial support from AB Gustaf Kiihr and particularly the relevant assistance from the Environmental Manager Mr Ake Erlandsson. The financial support from Akzo Nobel, Beckers Acroma, Swedwood International and Kalmar Energi and Swedish Institute is here recognized.