Ind. Eng. Chem. Res. 2006, 45, 7661-7666
7661
GENERAL RESEARCH Use of Hydrolyzed Primary Sludge as Internal Carbon Source for Denitrification in a SBR Treating Reject Water via Nitrite Alexandre Galı´, Joan Dosta, and J. Mata-A Ä lvarez* Department of Chemical Engineering, UniVersity of Barcelona, Martı´ i Franque` s, 1, 6th Floor, 08028 Barcelona, Spain
Different internal flows from a wastewater treatment plant were investigated for their usefulness as internal organic carbon source in the treatment of reject water with high ammonium loading, such as reject water from anaerobic digestion of sewage sludge (700-800 mg of NH4+-N L-1). This was done using a nitrification/ denitrification process via nitrite on a laboratory scale sequencing batch reactor working at 32 °C and 8 h cycle length. The hydrolyzed primary sludge was found to be the only one feasible for denitrification. When using the primary sludge, the reactor worked with an average biomass concentration of 2700 ( 250 mg of VSS L-1, obtaining a specific ammonium uptake rate of 17 mg of NH4+-N g-1 of VSS h-1, a specific nitrite uptake rate of 38 mg of NO2--N g-1 of VSS h-1, and a total nitrogen removal of 0.7 kg of N (day m3)-1. The use of that internal organic carbon source leads to a cost reduction of 0.2-0.3 euro kg-1 of N removed. Introduction The liquid effluent from anaerobic sludge digesters (reject water) in a biological wastewater treatment plant (WWTP) is characterized by a very low flow rate (0.5-2% of the total WWTP stream) and a high ammonium concentration (8001200 mg L-1). Its biological nitrogen removal through nitrification/denitrification (N/DN) in a sequencing batch reactor (SBR) or a chemostat reactor has been presented as a successful solution to reduce its discharge in the plant exit.1-7 The treatment efficiency varies depending on the operational conditions (temperature, dissolved oxygen (DO), pH, and alkalinity availability) and the strategy implemented. The cost of the process is a main important factor to consider. The development of N/DN via nitrite implies a reduction of 25% of oxygen consumption and 40% of organic carbon source in denitrification when compared with the conventional biological nitrogen removal (BNR) process.2,3,5,6 The nitrification via nitrite can be achieved working at temperatures over 20 °C and a sludge retention time (SRT) below 2 days,8,9 maintaining the pH 8,10 between 8 and 9, or working with DO11 below 1 mg L-1. The combination of the last two parameters is a good way to achieve the nitrite route in a SBR.12 Water alkalinity is another important parameter in the process. The addition of an organic carbon source in denitrification leads to a partial recovery of the alkalinity consumed during nitrification. This fact makes it possible to develop the process properly considering that reject water is characterized by a low bicarbonate-to-ammonium ratio.8 The organic carbon usually used to denitrify is an external organic carbon source such as methanol or acetate12 due to the lack of available organic carbon. But the main problem of this addition resides in its cost. However, in a WWTP it would be possible to find a useful carbon source to substitute the chemicals. Different studies have been done to * To whom correspondence should be addressed. Tel: +34-934021305. Fax: +34-93-4021291. E-mail:
[email protected].
find out a useful C-source to denitrify the main line of a WWTP13,14 showing the hydrolyzed primary sludge as the most appropriate. Although there is not a specific work focused on the direct application of these sources for reject water treatment, their conclusions can be taken into account. The VFA found in the soluble chemical oxygen demand (COD) of that hydrolyzed primary sludge is the powerful fraction to be used for denitrification.15 Different methods to hydrolyze the primary sludge are widely known, such as thermal hydrolysis,16 biological hydrolysis,13,14 or chemical hydrolysis with acid and alkali.17 In biological hydrolysis the fraction of readily biodegradable COD is higher than in the other two but the degree of solubilization is lower.13 The characteristics of the hidrolysate can vary considerably, but in a mixed chemical/biological sludge from a WWTP the degree of solubilization of biological hydrolyzed sludge is around 11%,15 where the fraction of VFA of the soluble COD is between 60 and 70%.13,14 Using the hidrolysate of the primary sludge in denitrification is also positive if there is an anaerobic zone for phosphorus removal in the reactor,15 but one also has to consider that the hydrolyzed sludge contains a slightly high concentration of ammonium13,14 which would be discharged in the reactor. If the WWTP has a two-phase anaerobic sludge digester, another option would be the use of the supernatant from the acid-phase digester.18 The aim of this paper is to study the denitrification capacity of several organic internal C-sources (including hydrolyzed primary sludge) from a real WWTP with anaerobic sludge treatment. Then, the biological treatment of real reject water in a SBR working with the selected C-source in the denitrification step will be optimized. Methods Experimental Devices. Treatment of reject water was carried out at laboratory scale, where the nitrification/denitrification
10.1021/ie060409i CCC: $33.50 © 2006 American Chemical Society Published on Web 10/04/2006
7662
Ind. Eng. Chem. Res., Vol. 45, No. 22, 2006
process was developed in one SBR of 3 L such as the one used by Galı´ et al.12 To operate the system, three pumps (two ColeParmer instruments 7553-85 and one EYELA microtube pump MP-3), one oxygen valve, and one mechanical stirrer were necessary. Moreover, the system was controlled and monitored by a computer with an acquisition data card (PCL-812PG), a control box, and an interphase card (PCL-743/745) connecting both systems. The computer worked with Bioexpert version 1.1 x. Temperature was maintained at 30 ( 0.5 °C by means of a thermostatic bath (RM6 Lauda), and pH was monitored with an electrode (Crison Rocon 18). pH and DO profiles were monitored, and these data were then exported and represented in each cycle. A closed intermittent-flow respirometer was used to characterize the different COD fractions and to perform the batch tests for denitrification. This device consisted of an aeration vessel (3 L) and a stirred watertight closed respiration chamber (0.250 L). A heating system (Polystat, Bioblock Scientific) was used to maintain the temperature at 30 ( 0.5 °C in the whole system. The respiration chamber was equipped with a dissolved oxygen electrode (Oxi 34v0i, WTW), and the pH in the aeration vessel was measured with a Crison pH 28 electrode. When the oxygen level dropped below 2 mg of O2 L-1 or the measuring period lasted more than 100 s, the mixed liquor in the measurement vessel was replaced by pumping aerated mixed liquor from the aeration vessel into the measuring chamber for 75 s, time enough to renew for three times the volume of the respiration chamber. The 4-20 mA signals of both oxygen and pH probes were collected and logged on a PC equipped with the Advantech Genie software package and combined A/D I/O modules (Adam 4050/Adam 4520/Adam 4018). The pH was controlled within a narrow pH setpoint ( ∆pH region. When the measured pH value did not lie inside the desired region, acid (HCl, 0.2 N) or base (NaOH, 0.2 N) was added by opening an electromagnetic valve for a very short period of time to adjust the pH. Substrate and Inoculum. Reject water was obtained from a mesophilic anaerobic digester of a WWTP situated in the Barcelona metropolitan area (Spain). This effluent was centrifuged to remove suspended solids before its recirculation to the plant head. This stream was collected and kept at 4 °C in the laboratory until its treatment. The hydrolyzed primary sludge and the other possible internal organic C-sources for denitrification were taken from the same WWTP and kept in the same conditions as reject water. Microorganisms to do the respirometric and batch tests were taken from the withdrawals of different SBRs that were operating with nitrification/denitrification via nitrite at the laboratory treating reject water. Analytical Methods. Analyses of chemical oxygen demand (COD), biological oxygen demand (BOD5), alkalinity, suspended solids (SS), and volatile suspended solids (VSS) were performed according to the Standard Methods for the Examination of Water and Wastewater.19 Ammonium was determined by an ammonia-specific electrode (Crison, model pH 2002). Nitrogen compounds, such as nitrates and nitrites, were analyzed by capillary electrophoresis (Hewlett-Packard 3D). Volatile fatty acids (VFA) were analyzed by gas chromatography (HewlettPackard 5890). Once the samples were withdrawn from the reactor, they were immediately centrifuged at 10 000 rpm for 10 min, and, for capillary electrophoresis analysis and chromatography; they were filtered through 0.45 µm cellulose paper filters. Results and Discussion Internal Organic C-Sources Selection. Figure 1 shows a scheme of the biological WWTP under study. First of all, the
water is conducted to a primary treatment (Q3 to Q6) followed by a conventional activated sludge system without the Nremoval step (between flows Q6 and Q11). Moreover, the WWTP has a one-phase anaerobic sludge digester for the treatment of primary and secondary sludge (between flows Q20 and Q21). The flow rates (Q, m3 day-1) of each WWTP stream are also detailed in Figure 1. The total flow rate in the plant is 49 000 m3 day-1 (Q3), and the supernatant of the anaerobic sludge digester (reject water) flow is 275 m3 day-1 (Q13), which represents 0.6% of the total water flow. Due to its high ammonia concentration, reject water is recirculated to the plant head with other flows (Q15) and this contributes to an increase of N concentration in the plant effluent (Q8) because the WWTP does not have a nitrogen removal unit. Considering the diagram and flow rates shown in Figure 1, there are different organic carbon sources that could be used for the denitrification step in the reject water treatment apart from the hydrolyzed primary sludge of the thickener (Q18). Table 1 shows the composition on an average basis of all the organic sources tested where the main contaminants analyzed are the COD, BOD5, VFA, N, and the solids. In the following lines there are explained the different sources: (a) Reject Water (Q13). Usually, reject water contains refractory COD, but, considering the quantity of COD shown in Table 1, it has to be verified that it is not useful for denitrification because it would be the cheapest source of organic carbon in the WWTP. However, the low values of BOD5/COD and BOD5/N that could be calculated from Table 1 and two respirograms12 demonstrated the low quantity of biodegradable COD, which leads to a very low denitrification capacity (5-6%). (b) Influent Flow Rate (Q1) and Secondary Biological Reactor Influent (Q6). Considering, as reported in Table 1, the amount of BOD5 in the influent of the plant (Q1) and also in the influent to the biological reactor (Q6), it is clear that both streams would be good for denitrification. However, the main problem does not reside in the quality of the wastewater COD but in the quantity of this water that would have to be added to the reactor treating reject water. From stoichiometric balances it is deduced that it would be necessary to add around 430 kg of biodegradable COD/day for the denitrification of reject water via nitrite in the SBR. Considering the BOD5 values of Table 1, this would lead to bypass of 950 m3 day-1 of Q1 or 1700 m3 day-1 of Q6. The latter fact would imply that this option should be discarded from the economic point of view because it would mean a reactor construction to treat reject water 4-6 times bigger than the SBR operating with methanol. Moreover when bypassing such a quantity of wastewater, the capacity for denitrification in the main line of the WWTP would be reduced. Consequently, these two flow rates are discarded. (c) Hydrolyzed Primary Sludge (Q18). Considering the concentrations presented in Table 1, the hydrolyzed primary sludge of the thickener is the source that owns more quantity of readily biodegradable COD according to the high quantity of VFA. The primary hydrolyzed sludge contains a large percentage (3-3.5%) of total and volatile solids, and it would be centrifuged at 7500 rpm when doing the different tests because the most part of COD in the solids is particulate and, therefore, slowly biodegradable. Moreover, around 60% of the soluble COD of the hidrolysate is composed of VFA, which is in concordance with the experiments done by Esoy and Odegaar.13 To asses the biodegradability of the hydrolyzed sludge, one respirometric assay has been done at 30 °C and a pH of 8.0 ( 0.1 using 12 mg L-1 of allyl-thiourea (ATU) as a nitrification inhibitor due to its considerable ammonium concentration. In
Ind. Eng. Chem. Res., Vol. 45, No. 22, 2006 7663
Figure 1. Wastewater treatment plant flow rates. Table 1. Characterization of Different C-Sources (Average Values January-April 2006) parameter
units
reject water
influent WWTP
influent biological reactor
primary hydrolysate
secondary hydrolysate
line flow rate
m3 day-1
Q13 275
Q1 46300
Q6 47200
Q18 235
Q11 900
CODtotal CODsoluble VFA (COD) BOD5 NH4+-N NO2-N NO3-N TS TVS TSS VSS pH
mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 mg L-1 -
1300 ( 100 300 ( 50 0 120 ( 20 650 ( 50
