Bioresource Technology 99 (2008) 1169–1176

Available online at www.sciencedirect.com

Real-time treatment of dairy manure: Implications of oxidation reduction potential regimes to nutrient management strategies

Asif Qureshi, K. Victor Lo ¤, Ping H. Liao, Donald S. Mavinic

Department of Civil Engineering, 6250 Applied Science Lane, University of British Columbia, Vancouver, BC, Canada V6T1Z4

Received 24 July 2006; received in revised form 16 February 2007; accepted 16 February 2007Available online 30 April 2007

Abstract

A pilot-scale sequencing batch reactor (SBR) was operated at a dairy farm to test real-time based control in winter operation condi-tions. A combination of high loading and low oxidation reduction potential (ORP) conditions in the aerobic stage of SBR treatment (anend value of ¡50 to ¡150 mV) inhibited nitriWcation while maintaining carbon removal. After a period of over-aeration over severalcycles, the ORP at the end of the aerobic stage increased to values of 50–75 mV. Subsequently, nitriWcation was observed, accompanied byhigher total cycle times. SigniWcant increase in removal eYciencies of ammonical nitrogen (� < 0.0001) and chemical oxygen demand(� < 0.001) were observed for the high ORP phase. It is postulated that higher ORP regimes are needed for nitriWcation. In low ORPregimes, nitriWcation is absent or occurs at an extremely low rate. It is also noted that nitrifying systems treating high strength animalmanure can possibly lead to unacceptably high levels of eZuent nitrate + nitrite nitrogen (NOx-N). Two manure management schemes areproposed that give the farmer an option to either retain the nutrients, or remove them from the wastewater. Some advantages and disad-vantages of the schemes are also discussed.© 2007 Elsevier Ltd. All rights reserved.

Keywords: Dairy manure; Oxidation reduction potential; Real-time control; Sequencing batch reactor; Treatment

1. Introduction

Animal manure is rich in nutrients and land applicationof manure has been the traditional standard practice onmany farms. However, in recent years, many environmentalissues associated with this practice have emerged. Theseissues are related to odor emissions, nutrient build-up insoil, contamination of groundwater and eutrophication inreceiving waters. While odor emissions are due to volatili-zation and organic degradation, nutrient build-up, ground-water pollution and eutrophication are due to nutrientloading in excess to the nutrient utilization/assimilationcapacity of the soil–water–plant system.

Biological treatment of manure has been reported toalleviate the above concerns (Obaja et al., 2005). It has alsobeen found that batch treatment consisting of alternate

* Corresponding author. Tel.: +1 604 822 4880; fax: +1 604 822 6901.E-mail address: kvlo@apsc.ubc.ca (K.V. Lo).

0960-8524/$ - see front matter © 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.biortech.2007.02.046

anoxic/anaerobic and aerobic phases, as in a sequencingbatch reactor, is probably the best way to achieve carbon,as well as nitrogen and phosphorus removals (Chang et al.,2000). Further, it has been reported that real-time controlbased strategies for anoxic/aerobic time length control arebetter than Wxed-time control based strategies (Qureshi,2006).

Fixed-time control of aeration might result in under-aer-ation, or over-aeration. While under-aeration might lead topartially treated manure, over-aeration will mean thathigher than necessary oxygen is provided. This might leadto destabilization of sludge and, since oxygen is providedby means of blowers, higher electricity and maintenancecosts. Real-time control is believed to be useful in optimiz-ing the energy requirements of biological treatment, partic-ularly during the aeration phase – as soon as the targetedcompound is treated, aeration is ceased (Gao et al., 2003).Thus, application of real-time control to treatment of agri-cultural manure will be particularly useful in providing the

1170 A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176

farmer with a cost-eVective manure management strategy.In addition, consistent and excellent removals have beenachieved from real-time controlled SBRs, with >90–95%removals of ammonical nitrogen (NH4-N) and orthophos-phates (PO4-P), accompanied by >94% total organiccarbon (TOC) removal (Ra et al., 1999; Kim et al., 2004).

Considering these advantages a real-time controlled,pilot-scale sequencing batch reactor (SBR) was operated onthe site of the Dairy Education and Research Centre, Uni-versity of British Columbia, at Agassiz, British Columbia(BC), Canada. The SBR was designed to take advantage ofthe end points/break points in the oxidation reductionpotential (ORP) and dissolved oxygen (DO) curves, whichsignify the end of nitriWcation (Cheng et al., 2000). Opera-tion on real-time control ensures lower energy consumptionand optimum aeration, i.e. the mixed liquor will rarely beover- or under-aerated (Battistoni et al., 2003). Further-more, the test location (Agassiz, BC) experiences low tem-peratures in the winter months. The monthly averagetemperature from December 2005 to March 2006 was 4–7 °C, and for April 2006 was 11 °C. Winter operation of bio-logical processes poses an additional hurdle as the processwill be operated at lower kinetic rates. Therefore, testing thepilot-scale SBR under winter conditions provided a conser-vative scenario.

Following successful testing on municipal wastewater,and also on low strength dairy and municipal wastewatermix (chemical oxygen demand, COD < 800 mg/L, fromQureshi, 2006), the dairy farm application was consideredto be an extension of the previous research. The SBRresults may have a signiWcant impact as far as animal nutri-ent management schemes are concerned, they provide uswith an opportunity to either retain or remove the nutri-ents, as desired.

2. Methods

2.1. Experimental set-up

The operating set-up of the SBR consisted of a series ofphases: Wll–react–draw, with a working volume of 2760–2800 L. The dairy manure was Wrst diluted approximately10 times with tap water and then pumped into an equaliza-tion/fermentation tank. The wastewater from this tank wasthen fed to the SBR, whenever the feed phase was ON. Theoperation of the SBR and its various phases is described asfollows: (1) Fill: The SBR was fed with 110–150 L of waste-water per cycle, through control by level sensors. The maxi-mum feed time was limited to 3 min as a safety measure, incase the level sensors failed to work. (2) React: The Fillphase was followed by a 75-min anoxic period, to allow fordenitriWcation and possible phosphorus (P) release. In thesubsequent aerobic period, it was expected that anyreleased P would be taken up, and that nitriWcation wouldtake place. DiVused aeration was accomplished by air pro-vided from an air pump. The end of the aerobic period wasdetermined using a programmable controller. This end-

point was assumed to be reached once the “DO elbow” or akink in the ORP curve was observed. This point was desig-nated to be the “nitrogen break point”, signaling the end ofnitriWcation, as has been reported in numerous studies onreal-time controlled SBRs (Ra et al., 1999; Kim et al., 2004).An aeration extension of 5 min was provided to allow forresidual DO build-up. Solids were wasted to a collectiontank on the farm, and their processing was not in the scopeof this study. Solids volume was approximately 6–7 L/day.(3) Draw: After 1 h of settling, 110–150 L of the supernatantwas withdrawn as eZuent from the SBR and the cycle ofthe sequential phases repeated again. A track study wasperformed to study the nitriWcation behavior of the SBRwhen desired.

2.2. Analytical methods

Chemical oxygen demand (COD) was analyzed accord-ing to the method prescribed in Standard Methods (APHA,1998). Nitrogen in the forms of ammonical nitrogen (NH4-N), nitrate + nitrite nitrogen (NOx-N), and orthophosphate(PO4-P) were analyzed using an automated ion analyzer(Lachat QuickChem series 8000, Zellweger Analytics, Inc.),after Wltration through 1.2�m membrane Wlters. For thetrack study, PO4-P, NOx-N and NH4-N were analyzed afterthe 1.2 �m Wltration, while COD was analyzed after a0.45 �m Wltration.

2.3. Statistical analysis

Means and standard deviations of results from the twophases were computed using MS Excel software. To deter-mine whether Phase 2 was better than Phase 1 in terms ofNH4-N and COD removal, student’s t-score was calculatedusing the “Small Sampling Theory”. Test of hypothesis wasconducted using the DiVerences of Means method. One-tailed test was done to determine the level of signiWcance �from the obtained t value.

3. Results and discussion

The results from the entire study indicated that two dis-tinct processes can be devised: (1) a process in which nutri-ents, especially nitrogen, are retained (Phase 1 of the study);and (2) a process in which complete NH4-N removal can beachieved (Phase 2 of the study). The main indicator of thesetwo types of treatment was identiWed to be the aerobic-stage, mixed liquor ORP. While the loading was relativelystable for most of the study, the eZuent characteristics interms of nutrient concentration were distinctly diVerent.These characteristics depended on whether the mixedliquor ORP was predominantly negative or positive.

3.1. Analytical results

In the early part of the study (Phase 1), until days 81–83,there was essentially no NH4-N removal taking place

A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176 1171

(Fig. 1). The Wnal ORP at the end of the aerobic period wasconsistently negative (Fig. 2). The ORP at the end of theaerobic period was in the range ¡50 to ¡150 mV, while theORP at the end of anoxic period was of the order of ¡350to –400 mV (Fig. 2). Even though a break point wasobserved in the ORP proWle, the eZuent NH4-N valueswere, for the most part, greater than the inXuent values.This indicated that the break point observed was not the“nitrogen break point”. Frequently, this break point wasalso observed in the DO proWle. The increase in NH4-Nmight be a result of simple ammonia accumulation, or of aconversion of organic nitrogen to ammonical nitrogen. Thetotal cycle time in the low ORP study phase (Phase 1), was»170–180 min, with approximately 60 min of aerobicperiod. It was concluded that the break-point on the ORP

curve corresponded to the end of readily degradable car-bon, as has been noted by Holman and Wareham (2003)(the COD elbow in Fig. 3). This point was termed as the“carbon break point” (CBP, Fig. 2). It is also noteworthythat the SBR system of Holman and Wareham (2003) wasoperated under low DO conditions. The high ORP valuesin their system were probably a result of a comparativelylower COD loading.

On days 83–87, the level sensors were severely compro-mised, due to false triggering by the mechanical Xoats, andas a result, there was no feed to the SBR. Consequently, theSBR became over-aerated over several cycles, leading to anincrease in the overall ORP of the SBR. The lowest ORP inthe anoxic stage was ¡170 to ¡200 mV. The ORP rose to>50–75 mV within the Wrst 60 min of aerobic cycle, and

Fig. 1. InXuent and eZuent NH4-N values and corresponding removal eYciencies for the study period.

0

50

100

150

200

250

300

42 46 47 55 56 61 63 67 68 69 73 75 76 89 92 98 102

107

Day

Influ

ent

and

Eff

luen

t N

H4-

N(m

g/L

)

-100

-80

-60

-40

-20

0

20

40

60

80

100

Rem

ova

l Eff

icie

ncy

(%)

Influent

Effluent

Removal Efficiency

Fig. 2. Typical DO and ORP proWles during the low ORP Phase 1 (CBPD carbon break point).

-500.00

-450.00

-400.00

-350.00

-300.00

-250.00

-200.00

-150.00

-100.00

-50.00

0.00

4:48 7:12 9:36 12:00 14:24 16:48 19:v

Time (hr:min)

OR

P (

mV

)

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

DO

(mg

/L)

ORP

DO

(CBP)

CBP

CBPCBP

CBP

CBP CBP CBP

SettleAnoxicAerobic

1172 A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176

remained so (or marginally increased) during the rest of theaerobic cycle (Fig. 4). The DO increased steadily during thisperiod. The DO increase at the break point (NBP, Fig. 4)was not as intense, as in the Wrst phase of the study. It wasfound that, from day 89 (Phase 2), the SBR was nitrifying(Fig. 1). The end of the aerobic period corresponded to theend of NH4-N consumption, and NH4-N removal eYcien-cies in excess of 97% were being achieved. Even anincreased inXuent NH4-N concentration towards the end of

study period, days 102–107, did not reduce the nitriWcationcapacity of the SBR. Appearance of NOx-N from day 89(Fig. 5) indicated that nitrogen was getting nitriWed, i.e.nitriWcation was taking place. As there was no reduction inNH4-N removal in the subsequent days, the decrease ineZuent NOx-N indicated amelioration of conditions suitedfor simultaneous nitriWcation–denitriWcation. In this phase,the end of the cycle corresponded to the end of nitriWcation.However, as in Phase 1, only one break point was being

Fig. 3. DO, COD, ORP and NH4-N proWles obtained by Holman and Wareham (2003).

Fig. 4. Typical DO and ORP proWles during the high ORP Phase 2.

-250.00

-200.00

-150.00

-100.00

-50.00

0.00

50.00

100.00

14:24 15:36 16:48 18:00 19:12 20:24 21:36 22:48

Time (hr:min)

OR

P (

mV

)

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

DO

(mg

/L)

ORP

DO

Nitrogen

Break Point

Aerobic Anoxic Settle

A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176 1173

detected by the control system. It is speculated that thehigher ORP regime initiated the activity of the nitriWers, orraised their activity signiWcantly (NH4-N removal eYcien-cies were signiWcantly better, with � value <0.0001) to com-fortably oVset the organic nitrogen solubilization rate. Thetotal cycle time increased to 7–12 h (depending on the feedstrength and loading), with an aerobic period of 5–9 h.

Fig. 6 indicates that, for both phases, enhanced biologi-cal phosphorus removal did not occur in the SBR. The neg-ative phosphorus removal eYciency was attributed to there-release of phosphorus from the sludge to the mixedliquor, while settling. The SBR showed 30–50% CODremoval during the Wrst phase; later, the COD removaleYciency increased signiWcantly (COD removal eYciencieswere better with � value <0.001), during the high ORPphase (Phase 2) of the study (Fig. 7). This was attributed toincreased removal due to enhanced aeration compared to

Phase 1 and/or greater bacterial activity in higher ORPregimes.

3.2. Track study results

The results from a track study on day 50 (Fig. 8) con-Wrmed the absence of nitriWcation in the SBR in Phase 1. Adecrease of NH4-N in the anoxic stage can be attributed toassimilation, while the reduction in soluble COD (s-COD)can be attributed to anaerobic stabilization (Chudoba et al.,1991; Randall et al., 1991), intracellular inclusions andassimilation. It was found that, even after the onset of theaerobic stage, the NH4-N concentration continued todecrease. This might have been due to some ammonia strip-ping, or further assimilation by the biomass. However, theNH4-N concentration started to increase after 20 min intothe aerobic stage. The likely reason was the solubilization

Fig. 5. InXuent and eZuent NOx-N values during the study period.

0

10

20

30

40

50

60

70

80

90

100

61 63 67 68 69 73 75 76 89 92 98 102

107

Day

Influ

ent

and

Eff

luen

t N

Ox-

N(m

g/L

)

Influent

Effluent

Fig. 6. InXuent and eZuent PO4-P values and corresponding removal eYciencies for the study period.

0

4

8

12

16

20

24

42 46 47 55 56 61 63 67 68 69 73 75 76 89 92 98 102

107

Day

Influ

ent

and

Eff

luen

t P

O4-

P

(mg

/L)

-80

-60

-40

-20

0

20

40

60

80

100

Rem

ova

l Eff

icie

ncy

(%)

Influent

EffluentRemoval Efficiency

1174 A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176

of organic nitrogen to ammonical nitrogen. Concurrentincreases in s-COD and PO4-P conWrmed this reasoning. Itwas also seen that the DO in the SBR started to increasefrom time 1 h:32 min, and reached a high of >5 mg/L in thenext 27 min (Fig. 9). It is reasoned that, as soon as the read-ily biodegradable carbon was exhausted, the DO increasedsharply. Even if there was any nitriWer activity in this lowORP regime, it was so low that it was easily oVset by therate of organic nitrogen solubilization. Thus, there was nonet NH4-N removal. Had the nitriWers been fully function-ing, the DO would have been lower, the mixed liquor NH4-N would have started to decrease, and mixed liquor NOx-Nwould have started to increase.

3.3. Discussion

From the diVerent nature and extent of treatment inthe two phases of this study, it appears that a low, sub-zero ORP in the aerobic phase of an SBR treatment pro-cess can selectively inhibit nitriWcation (as had occurred inthe Phase 1 of this study), while maintaining carbonremoval. When the ORP level increased in the secondphase of the study, nitriWcation rates increased signiW-cantly. It may be noted that the low ORP phenomenon isgenerally not observed in municipal wastewaters. Thestrength of municipal wastewater is generally very low(COD » 300 mg/L), and it is as such, diYcult to reach low

Fig. 7. InXuent and eZuent COD values and corresponding removal eYciencies for the study period.

0

1000

2000

3000

4000

5000

6000

61 63 67 68 69 73 75 76 89 92 9810

210

7

Day

Infl

uen

t an

d E

fflu

ent

CO

D

(mg

/L)

-20

0

20

40

60

80

100

Rem

ova

l Eff

icie

ncy

(%

)

Influent

Effluent

Removal Efficiency

Fig. 8. Track study proWle on day 50 of the study period.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0:00

0:32

0:55

1:15

1:27

1:36

1:47

1:59 2:

242:

59

Time (hr:min)

s-C

OD

(m

g/L

)

0

10

20

30

40

50

60

70

80

90

100

NH

4-N

, P

O4-

P,

NO

x-N

(m

g/L

)

CODNH4-NPO4-PNOx-N

Anoxic Aerobic Settle

A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176 1175

ORP levels. The wastewater used in this study had a CODstrength of 3500–5000 mg/L.

3.4. Proposed treatment schemes

3.4.1. Low ORP treatmentIt is hypothesized that the nitriWcation capacity of a

reactor could be restrained, if the mixed liquor ORP in theaerobic stage was maintained to sub-zero levels, likely inthe range ¡50 to ¡150 mV. This would ensure nitrogenretention, and organic carbon removal. PO4-P would alsobe retained. It is also postulated from observations in thisresearch that, if aeration is continued ahead of the s-CODutilization point (CBP in Fig. 2), eventually, the ORP willrise above zero, and so will the activity of nitriWers.

No nitriWcation translates into lesser energy require-ments (low overall oxygen demand), and will add to thecost eVectiveness of the proposed low ORP treatmentscheme. It can also be noted that the total processed volumeper day for Phase 1 was about 950–1130 L (approximately38% of the working volume) whereas for Phase 2, it wasabout 225–300 L (approximately 10% of working volume).Thus, the no nitriWcation system had a higher volumetriccapacity, along with lower energy requirements.

Thus, SBR systems can be designed to control the ORP ofthe mixed liquor, and selectively control the nitriWcationcapacity. Such systems will be especially useful in cases ofagricultural manures which contain a high organic loading,as well as nutrient content. The proposed low ORP treat-ment will retain the nutrient content but remove the organiccarbon part. In such a treatment scheme, an anoxic/anaero-bic stage will be followed by a low DO – low ORP aerobicstage. DO and ORP can be controlled by maintaining a highstrength loading in conjunction with low air supply. It isexpected that, if the DO is kept suYciently low, or the feedstrength is kept suYciently high, while keeping the aerobicORP below a maximum value, say ¡50 mV, it can be assuredthat carbon will be removed, while NH4-N, and probablyPO4-P, will not. The eZuent of such a system will then be rel-

atively free of odor problems (due to degasiWcation in theanoxic and aerobic stages, Zhang et al., 2006) and free ofdegradable carbon, while maintaining the nutrient content.It is hypothesized that this eZuent can then be sprayeddirectly to land. However, the eZuent will contain substan-tial quantities of NH4-N. This must be kept in mind by thedecision maker when devising a manure management plan,and excess nitrogen loading to the land should be avoided.

3.4.2. High ORP treatmentWhen the goal of manure management is nitrogen

removal, a high ORP scheme can be utilized. The oxygensupply rate and supply duration can be kept high, so as toobserve the nitrogen break point in the ORP or DO curves,as described in Section 3.2. With such a system, as seen inthis study, and others (Ra et al., 1999; Kim et al., 2004),complete NH4-N removal, as well as good PO4-P removalcan be achieved.

3.4.3. Trade-oVDepending on the requirements of the farm establish-

ment, either of the two treatment schemes can be employed.If the aim is nutrient retention, a low ORP treatmentscheme is suggested. If the aim is nutrient removal, a highORP treatment approach is advised.

However, a nitrifying treatment system may have poten-tial drawbacks: (1) the NH4-N is converted to NOx-N. Con-sidering the high concentration of NH4-N in agriculturalmanures, eZuent NOx-N concentrations might pose anenvironmental challenge (unless good simultaneous nitriW-cation and denitriWcation is achieved). That is, a formpotentially useful to plant crops (NH4-N rich wastewater)is converted to a form that is potentially more harmful tothe environment (NOx-N rich eZuent); (2) if enhanced bio-logical phosphorus removal takes place, the biologicallyremoved phosphorus needs further processing for recoveryin a usable form. While suitable technologies are beingdeveloped to attend to these recovery aspects (Burns et al.,2003; Burns and Moody, 2002; Greaves et al., 1999), it may

Fig. 9. ORP and DO proWles in the track study conducted on day 50.

- 400

- 350

- 300

- 250

- 200

- 150

- 100

-50

0

0:00

0:55

1:27

1:34

1:36

1:41

1:47

1:51

1:55

1:59

2:59

Time (hr:min)

OR

P (

mV

)

0

1

2

3

4

5

6

DO

(mg

/L)

ORP

DO

Anoxic Aerobic Settle

1176 A. Qureshi et al. / Bioresource Technology 99 (2008) 1169–1176

take a substantial period of time before they become cost-eVective.

In the meantime, it would be very useful to devise a pro-cess that can produce an acceptable, less concentrated eZu-ent. The low ORP treatment observed in this study can bereWned to become one such process – the eZuent from thisprocess can be discharged directly onto the receiving land.A less concentrated liquid will mean that more of thetreated eZuent can be applied to the same land mass, eVec-tively increasing its volumetric capacity.

4. Conclusions

From the results obtained from this study, the followingconclusions and recommendations are made. Low ORPconditions (<¡50 mV) in the aerobic stage of SBR treat-ment of dairy manure favored conditions that led to netNH4-N and PO4-P retention. Air supply and consequentorganic oxidation of readily soluble substrate removedmost of the readily degradable organics and alleviatedpotential odor problems. A process based on the combina-tion of a real-time SBR control and low, sub-zero ORP val-ues is proposed. This combination produces an eZuentwhich is free of odor, contains less organics, and maintainsa useful, though lower, nutrient content. Moreover, the sys-tems will have a lower energy requirement and a higher vol-umetric capacity, as compared to a nitrifying system. HighORP conditions (ORP > 50 mV consistently) in the aerobicstage of the SBR treatment led to almost complete NH4-Nremoval and higher organic removal. Depending on therequirements of any manure management scheme, nutrientscan be retained or removed by adequately designing theprocess, in terms of aerobic stage mixed liquor ORP. Themain controllable parameters are proposed to be CODloading and air supply.

Acknowledgements

The authors acknowledge the Strategic Project fundingby the Natural Science and Engineering Research Council(NSERC) of Canada, the help and support of the farm per-sonnel at the Dairy Education and Research Center, Agas-siz, BC, the assistance of Mr. Bud Fraser, Vision EnvirotechInternational Ltd. with the control system, and the help byMr. Fred A. Koch in setting up the pilot plant.

References

American Public Health Association, 1998. Standard Methods for theExamination of Water and Wastewater 20th ed. American PublicHealth Association, Washington, DC.

Battistoni, P., De Angelis, A., Boccadoro, R., Bolzonella, D., 2003. Anautomatically controlled alternate oxic–anoxic process for smallmunicipal wastewater treatment plants. Ind. Eng. Chem. Res. 42, 509–515.

Burns, R.T., Moody, L.B., 2002. Phosphorus recovery from animalmanures using optimized struvite precipitation. In: Proceedings ofCoagulants and Flocculants: Global Market and Technical Opportu-nities for Water Treatment Chemicals, Chicago, IL.

Burns, R.T., Moody, L.B., Celen, I., Buchanan, J.R., 2003. Optimization ofphosphorus precipitation from swine manure slurries to enhancerecovery. Water Sci. Technol. 48, 139–146.

Chang, H.N., Moon, R.K., Park, B.G., Lim, S.-J., Choi, D.W., Lee, W.G.,Song, S.L., Ahn, Y.H., 2000. Simulation of sequential batch reactor(SBR) operation for simultaneous removal of nitrogen and phospho-rus. Bioprocess Eng. 23, 513–521.

Cheng, N., Lo, K.V., Yip, K.H., 2000. Swine wastewater treatment in atwo-stage sequencing batch reactor using real-time control. J. Environ.Sci. Heal. B 35, 379–398.

Chudoba, P., Chevalier, J.J., Chang, J., Capdeville, B., 1991. EVect ofanaerobic stabilization of activated sludge on its production underbatch conditions at various So/Xo ratios. Water Sci. Technol. 23, 917–926.

Gao, D.-W., Peng, Y.-Z., Liang, H., Wang, P., 2003. Using oxidation–reduction potential (ORP) and pH value for process control of short-cut nitriWcation–denitriWcation. J. Environ. Sci. Heal. A 38, 2933–2942.

Greaves, J., Hobbs, P., Chadwick, D., Haygarth, P., 1999. Prospects for therecovery of phosphorus from animal manures: a review. Environ.Technol. 20, 697–708.

Holman, J.B., Wareham, D.G., 2003. Oxidation–reduction potential as amonitoring tool in a low dissolved oxygen wastewater treatment pro-cess. J. Environ. Eng. 129, 52–58.

Kim, J.-H., Chem, M., Kishida, N., Sud, R., 2004. Integrated real-time con-trol strategy for nitrogen removal in swine wastewater treatment usingsequencing batch reactors. Water Res. 38, 3340–3348.

Obaja, D., Mace, S., Mata-Alvarez, J., 2005. Biological nutrient removal bya sequencing batch reactor (SBR) using an internal organic carbonsource in digested piggery wastewater. Bioresour. Technol. 96, 7–14.

Qureshi, A., 2006. Fixed- and real-time treatment of dairy manure andexperiments on digestion and nutrient recovery. M.A.Sc. Thesis, Uni-versity of British Columbia, Vancouver, Canada.

Ra, C.S., Lo, K.V., Mavinic, D.S., 1999. Control of a swine manure treat-ment process using speciWc feature of oxidation reduction potential.Bioresour. Technol. 70, 117–127.

Randall, C.W., Brannan, K.P., McClintock, S.A., Pattarkine, V.M., 1991.The case for anaerobic reduction of oxygen requirements in biologicalphosphorus removal systems. Water Environ. Res. 64, 824–833.

Zhang, Z., Zhu, J., Park, K.J., 2006. A bench-scale aeration study usingbatch reactors on swine manure stabilization to control odour in post-treatment storage. Water Res. 40, 162–174.