MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
JOURNAL OF ENVIRONMENTAL SCIENCE AND HEALTH Part A—Toxic/Hazardous Substances & Environmental Engineering Vol. A38, No. 10, pp. 2169–2177, 2003
BIOLOGICAL NUTRIENT REMOVAL
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
Denitrification in Tertiary Filtration: Application of an Up-Flow Filter G. Farabegoli,1,* R. Gavasci,2 F. Lombardi,2 and F. Romani2 1
Department of Hydraulics, Transportations and Roads, Faculty of Engineering, University of Rome ‘‘La Sapienza,’’ Rome, Italy 2 Department of Civil Engineering, Faculty of Engineering, University of Rome ‘‘Tor Vergata,’’ Rome, Italy
ABSTRACT The present paper shows the results obtained through an experimental work performed at the wastewater treatment plant of Rome, aimed at studying the performances of a tertiary filter regarding combined removal of suspended solids, COD, and nitrates. The up-flow sand filter was fed by the effluent coming from the secondary settling tank of the plant. The filter bed height was of 80 cm of silica sand. After a start up period, a study of particulate and soluble COD removal process was made, to establish the need of methanol in the denitrification process. Total COD removal efficiency was 60% on average, 55% due by soluble COD removal and 5% by particulate one. In the last phase of the experimental activity methanol was fed as carbon source, sodium sulfite was supplied to produce anoxic environment within the filter and the denitrification efficiency was studied. Nitrates removal rates after an acclimation period of 10 days increased up to 60%, with an effluent NO3-N of 8 mg/L. Denitrification rate was 2.4 kg/m3 d for water temperatures of 25! C. Regarding methanol demand and biologic kinetics, the biomass yield coefficient was 0.3 kgCOD-X/kgme.
*Correspondence: G. Farabegoli, Department of Hydraulics, Transportations and Roads, Faculty of Engineering, University of Rome ‘‘La Sapienza,’’ 18 Via Eudossiana, 00184 Rome, Italy; E-mail:
[email protected]. 2169 DOI: 10.1081/ESE-120023349 Copyright & 2003 by Marcel Dekker, Inc.
1093-4529 (Print); 1532-4117 (Online) www.dekker.com
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
2170
Farabegoli et al. Consequently 2.7 kg of methanol was required per kilogram of denitrified nitrogen. Key Words: filtration.
Denitrification; Head loss; Methanol; Nitrogen removal; Tertiary
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
INTRODUCTION In the last years, the attached biomass biological processes have been diffused in order to reduce plant areas and reactors volume, to decrease the management costs for wastewater treatment and to improve the performances of the existing plants. Reactors that remove simultaneously organic matter, organic and ammonia nitrogen have just been used in plant practice. Denitrification in tertiary filtration can be easily realized and it is a valid alternative to extended denitrification in the activated sludge systems. Full-scale experiments carried out at several wastewater treatment plants have shown the efficiency of this process.[1–3] The aim of this study was to investigate the performance of a tertiary filtration system by means of it is possible to obtain a highly polished effluent in terms of carbon, nitrates, and suspended solids removal. Denitrification rates and efficiency, biomass yield coefficient and biologic kinetic, dependence of the process on temperature (comparing results with those of a previous campaign on a down-flow filter setup in the same wastewater treatment plant during winter months) were determined.
MATERIALS AND METHODS Description of Pilot Plant and Filter Runs The pilot up-flow filter is made of 2 cylindrical reactors of 8 cm external diameter and 3.3 m height, joint on the base with 2 right angle connections. The influent, coming from the secondary settling tank of the industrial wastewater treatment facility at the AMA plant in Rome, and the backwashing water are fed by the first column. The second column contains the filtration bed made of 80 cm of silica sand with grains of dimensions between 0.8 and 1.5 mm, density of 2.65 kg/dm3 and uniformity coefficient 1.18. Head-loss at different depths can be measured by means of plastic pipes placed every 20 cm along the filter bed height. The constant influent flow rate is 41 L/h corresponding to a filtration rate of 9 m/h, while the variable head-loss is equal to the difference between the level of water in the first column and the height of the outlet valve. Filtration process ends when the value of head-loss or effluent turbidity is unacceptable. Once a day, at the end of the filtration process, the backwashing water is fed by the first column and allows the bed to be expanded and solids accumulated within the filter bed to be detached. During the start-up phase, low backwashing rate and time were adopted to allow the biomass development within the filter bed. When biomass is completely attached, the backwashing water flow-rate has been increased to 2 L/min in order to improve the detachment process during the backwashing phase.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Denitrification in Tertiary Filtration
2171
Measurements and Analysis
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
The following measurements were made daily every hour: water flow, head loss over the whole filter bed, influent and effluent turbidity, DO, Temperature, and pH. 24-h composite samples from influent and effluent were taken daily and analyzed for Total and Volatile Suspended Solids (TSS,VSS), total and soluble COD, nitrates and nitrites. All the measurements were made according to the Standard Methods for the Examination of Water and Wastewater.[4] TSS and VSS of backwashing water were analyzed as well.
RESULTS AND DISCUSSION COD Removal Efficiency After a start up period, during which the hydraulic performance of the filter was analyzed, a study of particulate and soluble COD removal process was made, to establish the need of methanol during the denitrification process. Total COD removal efficiency was 60% on average, 55% due by soluble COD removal, and 5% by particulate one, as represented in Fig. 1. Analysis of TSS and VSS showed that 10% of influent VSS is not removed by solids capture within the filter bed but by biological degradation. This means that microorganisms use not only the soluble COD but also 10% of particulate COD as carbon source. Totally 55.5% of total COD is removed by biological degradation (mostly soluble, partly particulate), while 4.5% is removed by solids capture. Hence, considering the influent COD used by microorganisms as carbon source, it has been possible to establish a lower dosage of methanol in the denitrification process. At the beginning of the denitrification
Figure 1.
COD removal efficiency.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
2172
Farabegoli et al.
process, methanol was dosed in excess to avoid the carbon substrate to be the limiting factor.
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
Efficiency of Denitrification Dosing methanol and sodium sulfite, after an acclimation period of about 10 days, heterotrophic biomass was adapted to anoxic conditions and denitrification efficiency increased from 10% up to 60%, with values of effluent nitrates of 8 mg/L, as represented in Fig. 2 and in Fig. 3 respectively. Backwashing rate and time adopted, not causing an excessive detachment of biomass from the filter bed, allowed to maintain a high nitrate removal efficiency also during the first hours of the filtration process. Figure 4 showed the nitrate removal efficiency trend.
Figure 2.
Figure 3.
Nitrate removal efficiency.
Influent and effluent nitrates.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
Denitrification in Tertiary Filtration
Figure 4.
2173
Nitrate removal efficiency trend.
Temperature Dependency of Denitrification Under steady-state conditions, the volumetric rate of denitrification can be written as follows: rden ¼
Qin # ðNin % Nout Þ vf # ðNin % Nout Þ ¼ Vf hf
ð1Þ
where: N ¼ [NO3-N] þ [NO2-N] þ [NH4-N]; Qin ¼ inflow ¼ 984 L/d; Vf ¼ filter bed volume ¼ 3.6 L; hf ¼ depth of filter bed ¼ 0.8 m; vf ¼ filtration rate ¼ 216 m/d. During experimentation the average temperature was 23! C. Figure 5 shows denitrification rate and wastewater temperature during the campaign. After a period of bacterial population acclimation, denitrification rates increased up to 2.4 kg/m3 d for water temperatures higher than 25! C. Describing the temperature dependency of denitrification by an exponential function used by Strohmeier and Schroeter,[5] the following relationship can be written: !
rden, max ðTÞ ¼ rden, max ðT ¼ 15! CÞ # eK T ðT%15
CÞ
ð2Þ
with KT ¼ 0.07! C%1. A comparison between results obtained in this study with those obtained in a previous study during the winter period with a down-flow filter setup in the same wastewater treatment plant was carried out. In Table 1 is showed the comparison between winter and summer campaigns in terms of temperature, dissolved oxygen concentration and denitrification rate. Dissolved oxygen concentration is the same in both cases while differences in temperature in the two cases are 10! C. The increase of the denitrification rate from winter to summer has been about 1 kg/m3 d.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
2174
Farabegoli et al.
Figure 5.
Table 1.
Comparison between winter and summer campaigns.
Period 24/1/00–10/3/00
2/5/01–4/6/01
Denitrification rate and wastewater temperature.
Temperature (! C)
DO (mg/L)
rden (kg/m3 d)
12 13 14 15 22 23 24 25
2 2 2 2 2 2 2 2
1.7 1.83 1.94 2 2.04 2.46 2.24 2.26
Nitrites Accumulation Comparing results with those of the previous study during the winter period, nitrites accumulation seems to depend on water temperature (results not showed). The higher reduction rate of the first step (NO3 to NO2) compared to the second step (NO2 to N2) of the denitrification process might be due to a bacterial population shift from winter to summer. As described in Fig. 6, in the initial phase of the denitrification process, the higher production of nitrites in the effluent is caused by the particular reagent dosed in the influent flow.
Methanol Demand for Denitrification and Biomass Yield Coefficient The nitrogen balance over the filter cell and the methanol dosage allows the biomass yield coefficient to be calculated (nitrogen incorporation for biomass growth
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
Denitrification in Tertiary Filtration
Figure 6.
2175
Nitrites concentration in the influent and effluent.
neglected). Ammonia was not taken into account because NH4-N concentration in the influent was below 0.3 mg/L. The same yield coefficient was assumed for aerobic methanol degradation and for anoxic growth. In steady-state conditions the volumetric rate of substrate respiration rCOD formulated in COD-equivalents (kgCOD/m3 d) is: rCOD ¼ rO2 þ rden,COD ¼
Qin ½O2,in þ 2:86ðNO3,in % NO3,out Þþ 1:72ðNO2,in % NO2,out Þ) Vf ð3Þ
where: Qin ¼ inflow ¼ 984 L/d; Vf ¼ filter volume ¼ 3.6 L; O2,in ¼ DO concentration (mg/L). The maximum volumetric rate of methanol dosage qme (kg/m3 d) is: qme ¼
Qme Vf
ð4Þ
with Qme ¼ methanol dosage (kg/d). The ratio between the volumetric rate of substrate respiration and the volumetric rate of substrate dosage can be determined: rCOD ð5Þ qme # iCOD,me with iCOD,me ¼ 1.5 kgCOD/kgme. If methanol has not been lost in the filter effluent, the biomass yield coefficient can be expressed as: rCOD YCOD ¼ 1 % ð6Þ qme # iCOD,me As can be seen in Fig. 7, in the first period of acclimation during which a substantial fraction of methanol was lost in the effluent, as confirmed by the high value of effluent soluble COD, the denitrification rate was low. In the last days of experimentation the ratio of substrate respiration to substrate dosage was 0.7
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
2176
Farabegoli et al.
Figure 7. Ratio between the volumetric rate of substrate respiration and the volumetric rate of substrate dosage.
(kgCOD-X/kgme). Therefore the biomass yield coefficient can be assumed 0.3 (kgCOD-X/kgme). To denitrify 1 kg NO% 3 -N the following methanol dosage is required: ð!NO3 -NÞ # 2:86 1:0 # 2:86 ¼ 2:7ðkgme =kg NO3 -N denitrifiedÞ ¼ ð1 % YCOD Þ # iCOD,me ð1 % 0:3Þ # 1:5
ð7Þ
2.7 kg of methanol is required per kilogram of denitrified nitrogen. Regarding the sludge production during the denitrification process, 0.94 kg SST has been produced per kilogram of denitrified nitrogen.
CONCLUSIONS The aim of this study was to determine the performances of a tertiary filter regarding combined removal of suspended solids, COD, and nitrates. Results obtained showed that it is possible to obtain a highly polished and denitrified effluent with reduced volumes and surfaces need. After a start up period, during which the hydraulic performance of the filter was analyzed, a study of particulate and soluble COD removal process was made, to establish the need of methanol during the denitrification process. Total COD removal efficiency was 60% on average, 55% due by soluble COD removal and 5% by particulate one. Nitrates removal rate, after an acclimation period of 10 days, increased up to 60%, with an effluent NO3-N of 8 mg/L. Denitrification rate was about 2.4 kg/m3 d for water temperatures of 25! C. Nitrites accumulation seems to depend on water temperature, increasing at higher temperatures. Regarding methanol demand and biologic kinetics, the biomass yield coefficient was 0.3 kgCOD-X/kgme and 2.7 kg of methanol were required per kilogram of denitrified nitrogen.
MARCEL DEKKER, INC. • 270 MADISON AVENUE • NEW YORK, NY 10016 ©2003 Marcel Dekker, Inc. All rights reserved. This material may not be used or reproduced in any form without the express written permission of Marcel Dekker, Inc.
Denitrification in Tertiary Filtration
2177
Downloaded by [Universita Studi la Sapienza] at 07:40 11 October 2012
REFERENCES 1. Eichinger, J. Upgrading of the Munich wastewater treatment plants for denitrification in effluent filters. Water Science and Technology 1994, 29 (12), 217–225. 2. Hultman, B.; Jonsson, K.; Plaza, E. Combined nitrogen and phosphorus removal in a full-scale continuous up-flow sand filter. Water Science and Technology 1994, 29 (10–11), 127–134. 3. Koch, G.; Siegrist, H. Denitrification with methanol in tertiary filtration. Water Research 1997, 31 (12), 3029–3038. 4. American Public Health Association/American Water Works Association/ Water Environment Federation. Standard Methods for the Examination of Water and Wastewater, 19th Ed.; Washington, DC, USA, 1995. 5. Strohmeier, A.; Schroeter, I. In Experience and Biological Filtration in Advanced Wastewater Treatment, Proceedings of the European Water Filtration Congress, Oostende, Belgium, 1993, Vol. 17.
