Aerobic treatment of dairy wastewater with sequencing batch reactor systems
Xiujin Li,Ruihong Zhang
Abstract Performances of single-stage and two-stage se-quencing batch reactor(SBR)systems were investigated for treating dairy wastewater.A single-stage SBR system was tested with10,000mg/l chemical oxygen demand (COD)influent at three hydraulic retention times(HRTs) of1,2,and3days and20,000mg/l COD influent at four HRTs of1,2,3,and4days.A1-day HRT was found
sufficient for treating10,000-mg/l COD wastewater,with the removal efficiency of80.2%COD,63.4%total solids, 66.2%volatile solids,75%total Kjeldahl nitrogen,and 38.3%total nitrogen from the liquid effluent.Two-day HRT was believed sufficient for treating20,000-mg/l COD dairy wastewater if complete ammonia oxidation is not desired.However,4-day HRT needs to be used for achieving complete ammonia oxidation.A two-stage sys-tem consisting of an SBR and a complete-mix biofilm re-actor was capable of achieving complete ammonia oxidation and comparable carbon,solids,and nitrogen removal while using at least1/3less HRT as compared to the single SBR system.
Keywords Aerobic,dairy,wastewater,sequencing batch reactor
1
reactor thenIntroduction
Dairy wastewater is currently disposed of mainly through land application with little or no pretreatment in Califor-nia in the United States.Due to increasing awareness of the general public about potential adverse impact of ani-mal wastes on environmental quality and recent develop-ments in environmental regulations for gaseous-emission control and nutrient management,alternative wastewater treatment methods become attractive options for dairy producers.A sequencing batch reactor(SBR)is a biolog-
ical treatment reactor that uses aerobic bacteria to degrade organic carbon and remove nitrogen present in the wastewater.If designed and operated properly,it may
become a promising alternative for treating animal wastewater to control odors and reduce solids and nutrient contents.
The SBR treats wastewater in small batches andfits well
with most animal wastewater collection systems.It is a
time-oriented system and operates over repeated cycles of
five phases–fill,react,settle,decant,and idle.The major factors that control the performance of SBRs include or-
ganic loading rate,hydraulic retention time(HRT),solids retention time(SRT),dissolved oxygen(DO),and influent characteristics such as chemical oxygen demand(COD),
solids content,and carbon-to-nitrogen ratio(C/N),etc. Depending on how these parameters are controlled,the
SBR can be designed to have one or more of these func-tions:carbon oxidation,nitrification,and denitrification
[1,2].Carbon oxidation and denitrification are carried out
by heterotrophic bacteria and nitrification is by auto-
trophic bacteria.The SBR has been successfully used in the treatment of municipal and industrial wastewater,where
the high treatment performance resulted in excellent ef-
fluent quality[3,4].It is considered to be a suitable system
for wastewater treatment applications in small communi-
ties[5].The SBR is a relatively new technology for agri-cultural applications.Previous research on the SBR for
animal waste was primarily concentrated on swine wastewater treatment.Several researchers[6,7,8]re-
ported the performance of SBR in treating swine waste-
water with COD and suspended solids(SS)in the range of
1,614–2,826mg/l and175–3,824mg/l,respectively.Satis-factory removal of COD and SS from the wastewater was achieved with HRTs of22–30h.Fernades et al.[9]studied
the SBR for treating highly concentrated swine manure
with about4%total solids(TS).The influent COD,NH3-N,
and total Kjeldahl nitrogen(TKN)were as high as
31,175mg/l,1,265mg/l,and2,580mg/l,respectively.
Their results indicated that above97%COD,99%NH3-N,
and93%TKN removal efficiencies were achieved in the
liquid effluent at HRTs of6and9days and SRT of over
20days.Tam et al.[10]researched SBR for treatment of wastewater from a milking center and reported that the wastewater with919–1,330mg/l COD and15–37mg/l
NH3-N could be successfully treated with a HRT of20h. Bioprocess Biosyst Eng25(2002)103–109
DOI10.1007/s00449-002-0286-9
103
Received:2October2001/Accepted:6February2002 Published online:5April2002
ÓSpringer-Verlag2002
X.Li(&)
Department of Environmental Engineering,
Beijing University of Chemical Technology,
100029,Beijing,China
E-mail:lxiujin@hotmail
Tel.:+86-010-********
Fax:+86-010-********
R.Zhang
Biological and Agricultural Engineering Department, University of California at Davis,CA95616,USA
This research was supported in part by the California Energy Commission and the Agricultural Experiment Station of the University of California,Davis,USA.
Studies on the SBR for treating dairy manure are not well documented in the literature.Previous researchfindings about the SBR for treatment of swine manure and other types of wastewater provide
valuable references for the treatment of dairy wastewater.However,due to the dif-ferences in the characteristics of dairy wastewater from other types of wastewater,research is needed to develop design and operational guidelines for the SBR in treating dairy wastewater of various characteristics.
The objectives of this study are to investigate the effects of wastewater characteristics,HRT,SRT,and organic loading rate on the performance of the SBR system in treating dairy wastewater for carbon and solids removal and nitrogen conversion,and develop design and opera-tional guidelines for the SBR system in single-and mul-tiple-stage configurations.
2
Materials and methods
2.1
Dairy manure collection and preparation
Dairy manure was collected on the Dairy Research Farm of the University of California at Davis.Due to runoff of urine on the feedlot,the collected manure was mainly feces and contained a relatively low content of ammonia nitro-gen.The manure was slurried with addition of water and then screened twice
with two sieves with openings of4·4 and2·2mm,respectively,to remove large particles.The screened manure was transported immediately to the laboratory and stored in a freezer at–20°C until use.The TS and COD of the screened manure were30,000–
40,000mg/l and35,000–50,000mg/l,respectively.When needed,the stored manure was thawed and then diluted with tap water to obtain a desired COD concentration.Due to relatively low ammonia content of the raw manure as compared to typical levels in the manure collected on dairy farms,urea was added to increase the NH3-N in the prepared manure from100–125mg/l to500–550mg/l.The prepared manure was then put into a50-l feeding tank housed in a refrigerator at4°C for daily use.The feeding tank had an agitator to mix the wastewater during the feeding of the reactors.
2.2
Experimental setup and operation
Both single-stage and two-stage treatment systems were tested.The single-stage SBR system consisted of an SBR and a solids-settling tank in series.The wastewater wasfirst fed into the SBR for treatment and the effluent of the SBR, including both sludge and liquid,was then discharged into a settling tank,where liquid was separated from sludge by gravity settling and characterized as liquid effl
uent of the system.The two-stage system consisted of an SBR(first-stage reactor),a solids-settling tank,and a complete-mix biofilm reactor(CMBR)(second-stage reactor)connected in series.The liquid effluent obtained from the solids-set-tling tank was used as influent of CMBR and further treated in the CMBR for achieving complete nitrification.The two-stage SBR-CMBR system is shown in Fig.1.
Each system was fed and decanted twice a day for12h in each treatment cycle.All the peristaltic pumps used for feeding and decanting were operated automatically with a digital time controller.The time sequence for different operations during each treatment cycle of the SBR was
1–3minfill,11h and4–8min react,40min settle,
1–3min decant,and10min idle.The CMBR was operated as a complete-mix reactor and had long SRT provided by the attached growth on the polyethylene pellets placed in the reactor.The plastic pellets had light density(920kg/ m3)and were keptfluidized with the airflow.Each pellet was10mm in diameter and10mm in height,with a cross inside the cylinder and longitudinalfins on the outside, providing a large surface area for bacterial attachment. Thefilling volume of the pellets in total occupied ap-proximately18%of liquid volume(3l)in the reactor.
The SBR and CMBR reactors were made from trans-parent acrylic and had a total volume of6l each,wi
th51cm height and12cm diameter.During testing,the liquid vol-ume of each reactor was3l.Each reactor was aerated using pressurized air at a controlledflow rate.In order to mini-mize the water evaporation in the reactor,the air was hu-midified by traveling through water contained in a15-l jar prior to entering the reactor.The air was evenly distributed into the wastewater through four air stone diffusers in-stalled near the bottom of the reactor.All the reactors were initially seeded with the activated sludge obtained from the UC Davis Wastewater Treatment Plant and allowed to ac-climate for about2months before formal experiments were started.It normally took about4weeks for each SBR re-actor to reach a steady state when a new operating condi-tion was introduced.The steady state was defined to be a state when the weekly variations of effluent COD,TS,
NH3-N,and pH were less than5%.These parameters were monitored twice a week.The CMBR had been fully accli-mated with dilute dairy wastewater for about6months and had nitrification bacteria well established before being connected with the SBR.The mixed liquor suspended solids (MLSS)in the CMBR was about10,000mg/l,which was calculated from both suspended growth and attached growth solids.In order to determine the ammonia emission from SBR due to aeration,ammonia in the exiting air of SBR was collected by absorbing it in0.3N boric acid solution for 24h under each testing condition.
2.3
Experimental plan and system performance evaluation The experiment was carried out in two phases.Thefirst phase was for studying the effects of influent characteris-tics,HRT,and corresponding SRT and loading rate on the performance of the single-stage SBR system.The second phase was to evaluate the performance of a two-stage SBR-CMBR system.The two systems were then compared in terms of carbon and solids removal and nitrogen conver-sion efficiencies.
With the single-stage SBR system,three HRTs(1,2and 3days)were tested for wastewater of10,000mg/l COD and four HRTs(1,2,3and4days)for wastewater of
20,000mg/l COD.For the wastewater of10,000mg/l COD, the corresponding loading rate and SRT for the three HRTs were10,5,and3.3g COD/l/day and8,12,and
Bioprocess Biosyst Eng25(2002) 104
15days,respectively.For the wastewater of 20,000mg/l COD,the corresponding loading rate and SRT for the four HRTs were 20,10,6.7,and 5g COD/l/day and 1.5,3,4,and 6days,respectively.With the two-stage SBR system,
2days was used first as the system HRT,with 1day for the first-stage and 1day for the second-stage for both in flu-ents,and then 2.5days was used with 2days for the first stage and 0.5days for the second stage.An air flow rate of 4l/min was applied for all runs,which was able to main-tain dissolved oxygen (DO)in the SBR and CMBR above 3mg/l.
The performance of the treatment systems was evalu-ated in terms of carbon and solids removal and nitrogen conversion ef ficiencies.The parameters analyzed included TS,volatile solids (VS),COD,SCOD (soluble COD),TKN,NH 3-N,NO 2-N,and NO 3-N.Two kinds of removal/con-version ef ficiencies were used to interpret the results for carbon and solids removal and nitrogen oxidation.One ef ficiency,E t ,is based on the removal from total ef fluent (including both sludge and liquid ef fluent generated),re-flecting the removal ef ficiency through biological process alone.The other ef ficiency,E l ,was based on the removal from liquid ef fl,supernatant,representing the removal ef ficiency through both biological process and sludge separation.For the single-stage SBR system,the total ef fluent was the ef fluent from the SBR and the liquid ef fluent was the supernatant decanted from the solids settling tank.For the two-stage SBR-CMBR system,the total ef fluent was the combination of sludge from the settling tank and the final ef fluent from CMBR,and the liquid ef fluent was the liquid ef fluent of CMBR.Most of previous research only reports removal ef ficiency from liquid
ef fluent (E l ).Actually,E l does not re flect the real capability of a system for removing various constituents from wastewater,because part of these constituents are contained in the sludge that is separated from the liquid ef fluent and discharged as a separate sludge stream.
Therefore,E t needs to be used in order to assess the real capability of a system for removing various constituents from wastewater.
2.4
Sampling and analytical methods
After each reactor reached steady state under testing conditions,samples were taken from the in fluent,mixed liquor,total ef fluent,and liquid ef fluent of the reactor three times a week (every other day)for analyses of COD,SCOD,TS,VS,NH 3-N,NO 2-N,NO 3-N,and TKN.The re-moval ef ficiencies,E l and E t ,were calculated based on the data from in fluent,liquid ef fluent,and total ef fluent of the systems.The separation of sludge and liquid in the total ef fluent of the SBR was performed by settling the ef fluent in a 1-l graduated cylinder for 2h and then decanting the liquid fraction above the sludge-liquid interface line.The COD,SCOD,TS,VS,and TKN were measured according to APHA standard methods [11].The COD measured in this study was COD Cr .The pH was measured with an Accumet p
H meter (Fisher Scienti fic,Pittsburgh,Pa.).The NH 3-N was measured with a gas-sensing elec-trode and the pH meter.The DO in the reactors was
monitored on a daily basis with a DO meter (YSI Mode158,Fisher Scienti fic,Pittsburgh,Pa.).The NO 2-N was analyzed with the HACH method,using a DR/2000spectropho-tometer [12].The NO 3-N was measured with a diffusion –conductivity analyzer [13].3
Results and discussion
3.1
Performance of the single-stage SBR system
3.1.1
Removal of carbon and solids
The performance data of the SBR for 10,000mg/l COD in fluent COD of 10,000are shown in Table 1.With the increase of HRT from 1to 3days,the COD,SCOD,TS,and VS in the liquid ef fluent became lower,yielding better ef fluent quality due to increased biological conversion and improved sludge settleability,as indicated by the increased removal ef ficiencies (E l and E t ).However,there was
no
Fig.1.Laboratory setup for a two-stage SBR-CMBR system for dairy wastewater treatment
X.Li,R.Zhang:Aerobic treatment of dairy wastewater with sequencing batch reactor systems
105
signi ficant difference in terms of carbon and solids rem-ovals and liquid ef fluent quality for the three H
RTs.For example,the increase of COD and TS removal ef ficiency E l was 5.1%and 0.3%,and E t was 5.7%and 2.0%,respec-tively,when HRT increased from 1to 3days.Therefore,1-day HRT was believed to be suf ficient for treating the dairy wastewater with 10,000mg/l COD for its satisfactory removal ef ficiency and relatively short HRT.At 1-day HRT,the removal ef ficiency from the liquid ef fluent (E l )was 80.2%for COD,63.4%for TS,and 66.2%for VS.These removals were due to both biological conversion in the SBR and sludge separation in the solids-settling tank.The removal due to biological conversion alone in the SBR,as measured by E t ,was 45.0%for COD,21.4%for TS,and 34.2%for VS.E t was signi ficantly greater than E l ,sug-gesting that the sludge separation after SBR treatment is necessary for achieving signi ficant carbon and solids re-moval from the dairy wastewater.It was found that aerobic treatment greatly enhanced the flocculation and settlea-bility of the solids in the wastewater.Good settleability of sludge was important for achieving high carbon and solids removal ef ficiency.
The performance data of the SBR for 20,000mg/l COD in fluent are shown in Table 2.The 1-day HRT was tested first.It was found that it was impossible to control the SRT at a desired level due to fast solids buildup in the reactor and poor solids settleability.When the HRT was increased to 2days,there was signi ficant improvement in the ef fluent quality and increase of removal ef ficiencies.However,when the HRT was further increased to 3days,the changes in the ef fluent quality,C
OD,and solids removals were not signif-icant.Therefore,2-day HRT was considered enough for COD and solids removal for 20,000mg/l COD in fluent due to its relatively short retention time and high removal ef fi-ciency.At 2-day HRT,the removal ef ficiency E l of COD,SCOD,TS,and VS was 85.7%,67.1%,71.0%,and 70.6%,respectively,and E t was 35.9%,67.1%,22.8%,and 25.6%,respectively.The 4-day HRT was tested for achieving com-plete ammonia conversion.Since ammonia was not com-pletely converted at 2-day and 3-day HRT,longer HRT was needed when complete nitri fication was desired.This will be further discussed in the following nitrogen removal section.The sludge separated from the ef fluent of the SBR contained 4.1–5.9%TS.The lower in fluent COD
(10,000mg/l)resulted in better sludge settleability than the higher in fluent COD (20,000mg/l).The sludge volume as the fraction of total ef fluent volume was 5–6%and 13–16%for the lower and higher levels of in fluent COD,respectively.The sludge was composed of not-degraded solids in the wastewater and newly formed bacterial cells.It can be further processed into organic soil amendment through dewatering and composting.
Table 1.Ef fluent quality and treatment ef ficiencies of SBR for 10,000mg/l COD in fluent
Parameters
In fluent (mg/l)
1-day HRT 2-day HRT
3-day HRT Liquid ef fluent Total ef fluent E l (%)E t (%)Liquid ef fluent Total ef fluent E l (%)E t (%)Liquid ef fluent Total ef fluent E l (%)E t (%)(mg/l)(mg/l)(mg/l)(mg/l)(mg/l)(mg/l)COD 10,000
1,9806,50080.245.01,5805,13084.248.71,4704,93085.350.7SCOD 2,9141,4571,45750.050.01,4511,45150.250.21,4281,42851.051.0TS 6,6562,4365,23263.421.42,4765,07962.823.72,4165,09863.723.4VS 5,1081,7243,36166.234.21,5323,24970.036.41,4003,16772.638.0TKN 7801953657553.218535476.354.616533878.856.7TN 78048160738.322.248059838.523.447459639.223.6NH 3-N 510120120
76.5
76.5
105105
79.4
79.4
7070
86.3
86.3
NO 3-N 0375545NO 2-N 024*******pH
8.1
6.8
6.7
6.7
Table 2.Ef fluent quality and treatment ef ficiencies of SBR for 20,000mg/l COD in fluent
Para-meters In fluent (mg/l)
1-day HRT 2-day HRT 3-day HRT 4-day HRT
Liquid ef fluent Total ef fluent E l (%)E t (%)Liquid ef fluent Total ef fluent E l (%)E t (%)Liquid ef fluent Total ef fluent E l (%)E t (%)Liquid ef fluent Total ef fluent E l (%)E t (%)(mg/l)(mg/l)(mg/l)(mg/l)(mg/l)(mg/l)(mg/l)(mg/l)COD 20,0004,30013,92078.530.42,87012,82085.735.92,66012,32086.738.4167010,90091.743.3SCOD 6,6603,1973,19752.052.02,1902,19067.167.12,0052,00569.969.912151,21581.881.8TS 12,4424,36710,11564.918.73,6129,60571.022.83,4989,58071.923.033509,19573.124.3VS 10,1043,1427,92268.921.62,9727,51770.625.62,8097,33672.227.424607,05375.729.5TKN 1,14054089952.621.118063884.244.017060085.147.48550092.552.3TN 1,14057393349.718.248491857.519.548891857.219.538888863.020.5NH 3-N 54031031042.6
42.6
828284.8
84.8
808085.2
85.2
00
100100
NO 3-N 020*********NO 2-N 01314013010pH
8.0
8.7
7.9
7.8
7.6
Bioprocess Biosyst Eng 25(2002)
106
3.1.2
Nitrogen conversion
With the influent of10,000mg/l COD and1-to3-day HRT,22.2–23.6%of total nitrogen(TN)was lost in the treatment process as indicated by the E t.The losses of TN for the three HRTs were not significantly different.The ammonia collection results showed that the amount of NH3-N volatilized accounted for only2–3%of TN,indi-cating that the ammonia loss through volatilization was small under these operating conditions.The rest of TN loss (approximately20%)might be due to the emission of other nitrogenous gases,such as nitrous oxides(NO and NO2)formed in the nitrification process,and nitrogen gas (N2)formed in the denitrification process.The TKN removal was53.2–56.7%from the total effluent and75–78.8%from the liquid effluent,respectively.The TKN removal mainly resulted from ammonia oxidation.
With the influent of20,000mg/l COD and1-to4-day HRTs,the loss of TN was18.2–20.5%.For the1-day HRT, the ammonia collection results showed that ammonia volatilization accounted for16%of TN,indicating that most of TN loss was due to ammonia volatilization.This occurred with the low nitrification rate in the SBR.But ammonia volatilization was insignificant at2-to4-day HRTs,at which the
SBR had high nitrification activities. These results might imply that ammonia volatilization could be related to nitrification activity.Little nitrification occurrence at1-day HRT was due to the short SRT of 1.5days.This agrees with thefindings of Prakasam and Loehr[14],who stated that2-day SRT was the minimum for nitrification of poultry wastes.Therefore,HRT was increased to2days and3days,and corresponding SRT were3days and4days.It was found that nitrification was able to sustain in the SBR at both HRTs.At2-day and
3-day HRT,the TN and TKN removals were19.5%and44.0–47.4%from the total effluent,and57.5–57.2%and84.2–85.1%from the liquid effluent,respectively.Significant NH3-N was removed,as indicated by removal efficiency of 84.8%for2-day HRT and85.2%for3-day HRT,although there was still80–82mg/l residual NH3-N present in the effluent.It can be seen that there was no significant dif-ference between two HRTs in terms of TN,TKN,and NH3-N removal.Therefore,if complete ammonia oxidation is not required,2-day HRT would be considered efficient for treating20,000mg/l COD influent in terms of both nitrogen removal discussed here and COD and solids removal as mentioned in Sect.3.1.1.
Certain amounts of residual ammonia were present in the effluent from20,000mg/l COD influent at2-day and 3-day HRT.This indicates that the nitrification process might have been inhibited in both operation conditions. Nitrification inhibition might be due to possible inhibitions of nitrification bacteria b
y free ammonia(FA)and free nitrous acids(FNA)and suppression of nitrification bac-teria by more competitive heterotrophic bacteria[15].NH3 was undesirable because of its odor and toxicity to aquatic lives;thus,it needed to be removed from the wastewater. Shammas[16]studied the interaction of temperature,pH, and biomass on the nitrification process and concluded that high nitrification efficiency can only be obtained with either very long detention time or a combination of high
solids concentration and elevated temperature.Therefore,
HRT was further increased to4days in order to obtain complete ammonia conversion.It was found that4-day
HRT,corresponding6-day SRT,was enough for complete ammonia conversion,as indicated by zero ammonia pre-
sent in the effluent(see Table2).Therefore,it could be concluded that if complete ammonia conversion is desired,
4-day HRT would be needed for treating20,000mg/l COD wastewater with540mg/l NH3-N.
A track study was conducted in order to further un-derstand the nitrification process in the SBR.The va
ria-
tions of NH3-N,NO2-N,and NO3-N in the SBR during a
12-h operating cycle in treating the wastewater of
10,000mg/l COD at2-day HRT are shown in Fig.2.Am-monia oxidation mostly occurred in thefirst5h,as indi-
cated by the increase of NO2-N and decrease of NH3-N.
Since a large amount of ammonia was oxidized in the early
stage of one cycle with high nitrification,the amount of ammonia volatilization may be decreased in contrast to
the condition when nitrification is small as mentioned
above.The relationship between ammonia volatilization
and nitrification activity needs to be further investigated in
future study.The pH could be another factor related to ammonia volatilization.Since higher medium pH in-
creased the gas fraction of total ammonia dissolved in the medium,ammonia volatilization could have been high
when there was little nitrification and pH maintained rel-atively high(approximately8.0),but small when there was
good nitrification and the pH was decreased(Fig.2).The
NO2-N increased to the peak value about5h later after feeding and then started to decrease,while NO3-N started
to increase slightly.Generally speaking,the variations of
NH3-N,NO2-N,NO3-N,and pH in the SBR during the operating cycle depends on the bioconversion dynamics in
the reactor,initial ammonia concentration,and alkalinity
in the wastewater.
3.2
Performance of the two-stage SBR-CMBR system
As stated above,a4-day HRT is needed for achieving complete oxidation of ammonia in the dairy wastewater
in
107
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论