This article was downloaded by: [Laurentian University]On: 08 October 2014, At: 00:50Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Journal of Environmental Science and Health, Part B:Pesticides, Food Contaminants, and Agricultural WastesPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/lesb20

Piggery waste treatment by using down-flow anaerobicfixed bed reactorsE. Sánchez a , L. Travieso b , R. Borja b , M. F. Colmenarejo a , S. Nikolaeva c , F. Raposo b &B. Rincón ba Centro de Ciencias Medioambientales (CSIC) , Madrid, Spainb Instituto de la Grasa (CSIC) , Sevilla, Spainc Departamento de Física , Universidad Nacional de Costa Rica , Heredia, Costa RicaPublished online: 10 Sep 2007.

To cite this article: E. Sánchez , L. Travieso , R. Borja , M. F. Colmenarejo , S. Nikolaeva , F. Raposo & B. Rincón (2007)Piggery waste treatment by using down-flow anaerobic fixed bed reactors, Journal of Environmental Science and Health, PartB: Pesticides, Food Contaminants, and Agricultural Wastes, 42:6, 727-734, DOI: 10.1080/03601230701465999

To link to this article: http://dx.doi.org/10.1080/03601230701465999

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Journal of Environmental Science and Health Part B (2007) 42, 727–734Copyright C© Taylor & Francis Group, LLCISSN: 0360-1234 (Print); 1532-4109 (Online)DOI: 10.1080/03601230701465999

Piggery waste treatment by using down-flow anaerobic fixedbed reactors

E. SANCHEZ1, L. TRAVIESO2, R. BORJA2, M. F. COLMENAREJO1, S. NIKOLAEVA3, F. RAPOSO2

and B. RINCON2

1Centro de Ciencias Medioambientales (CSIC), Madrid, Spain2Instituto de la Grasa (CSIC), Sevilla, Spain3Departamento de Fısica, Universidad Nacional de Costa Rica, Heredia, Costa Rica

A study of the role of the depth in the performance of laboratory-scale down-flow anaerobic fixed-bed reactors (DFAFBR) was carriedout at different nominal hydraulic retention times (HRTN) using piggery waste as substrate at different influent concentrations (2,4, 6 and 8 g COD/L). The profiles of soluble chemical oxygen demand (COD) (SCOD), organic nitrogen (O.N.), ammonia nitrogen(A.N.), pH and electrical conductivity (E.C.) through the reactor depths showed an initial highly active zone, which was locatedaround the first half of the reactor depth, and a second zone with a lower biological activity. It was found that the depth of the activezone decreased as the HRTN increased and that the slopes of the profiles obtained increased with the rise in the influent concentration.A hydraulic test showed an increase in the dispersion number when the HRTN increased. The reactors showed a hydraulic patternbetween plug-flow and back-mix. The real values of HRT (�) also defined as real contact times were determined to be 0.7, 2.1, 3.4,4.7, 6.4 and 8 days for values of HRTN of 1, 2, 3, 4, 5 and 6 days, respectively. It was found that the concentration of SCOD withinthe reactor decreased exponentially with the increase in the value of θ . Additionally, the influent concentration had a strong influenceon the SCOD variation concentration, mainly at values of θ under 1.5 days, which corresponded to the first part of the reactors.

Keywords: Depth; anaerobic fixed-bed reactor; down-flow; clay; rasching rings; performance, piggery waste.

Introduction

The treatment of piggery waste by using anaerobic fixed-bed reactors has been widely studied by many authors dueto the clear advantages of this process, which is capableof retaining great concentrations of microorganisms withinthe reactor.[1−10] The increase in the microorganism con-centration within the reactor either attached to the supportor in suspension at the interstices between the pieces of sup-port improves the COD removal rate and reduces the HRTNrequired, allowing for a high process performance.[1−10] Foran optimum design of the process, different aspects must beevaluated and taken into account: the characteristics of thesupport media and its distribution, the reactor geometry,the depth of the bed, the temperature, the effluent recir-culation and the mode of wastewater circulation throughthe reactor (up-flow, down-flow or horizontal-flow). Ahn

Address correspondence to R. Borja, Instituto de la Grasa(CSIC), Avda. Padre Garcıa Tejero 4, E-41012-Sevilla, Spain, Tel.:34 95 4689654; Fax: 34 95 4691262; E-mail: rborja@cica.esReceived April 2, 2007.

and Foster[11,12] compared the performance of two up-flowanaerobic filters, one operating at thermophilic tempera-ture (55◦C) and the other one at mesophilic temperature(35◦C) treating starch wastewater. The authors found thatthe substrate utilization rate was 15 times higher in the caseof thermophilic temperature. The behavior of an upflowanaerobic filter at ambient temperature (9◦C–20◦C) wasevaluated with sewage waste at HRTN in the range of 6–46 hours.[13] The highest performance was achieved at thehighest temperature values. Show and Tay[14] studied theinfluence of the support media in laboratory scale up-flowanaerobic filters. They used three different rasching ringssimilar in size but with a different porosity and found thatthe process was favored by the porosity and roughness of thematerial while smooth materials caused a lower retentionof biomass and as a consequence, a poorer performance.Other authors[15] have reported the use of backed clay me-dia with significant advantages in comparison with othermedia. Other materials have also been successfully used assupport medium in anaerobic fixed-bed reactors, as for ex-ample, tyre-rubber applied to sewage, piggery and distillerywastewaters, with COD removal efficiencies comparable tothose obtained with other ceramic materials.[16−18]

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

728 Sanchez et al.

The effect of effluent recycling on anaerobic fixed-bed performance has been widely evaluated by differentauthors.[19,20] Ruiz et al.[19] successfully used a recycling ra-tio of 6:1 in an anaerobic fixed-bed reactor treating slaugh-terhouse wastewater. Yu et al.[20] studied the influence of therecycling ratio (1.0–6.0) at three different Organic LoadingRates (OLRs) (8.2, 11.3 and 13.5 g COD/(L d)). They foundan increase in the COD removal rate with an increase in therecycle ratio, the optimum ratio depending on the OLRapplied: for example for a value of 8.2 g COD/(L d) theoptimum recycle ratio was 2:1, while at 13.5 g COD/(L d)the optimum ratio was 5:1.

Jawed and Tare[21] compared the performance of down-flow and upflow anaerobic filters that operated under sim-ilar conditions even though the organic loads varied formore than 20 months. The authors observed that CODremoval efficiency, methane gas production and CODmethanisation did not differ significantly at 95% confidencelevel.

Borja and Banks[22] studied the treatment of palm oilmill effluent by an up-flow anaerobic fixed-bed reactor atorganic loading rates ranging from 1.2–11.4 g COD/(L d)and HRTN from 15 to 6 days. The reactor had a volume of23 litres and was 2.5 m in height. The packing material wasgranitic gravel, and was 1–3 cm size. These authors deter-mined the COD removal efficiency, the pH, the volatile acidconcentration and the alkalinity at different heights of thereactor. It was found that maximum COD removal occurredin the first 0.5 m, coinciding with a maximum increase inpH and alkalinity. The concentration of volatile acids de-creased considerably at the first 0.5 m of reactor in a rangeof OLR between 1.2–2.1 g COD/(L d). For higher OLRsthe concentration of volatile acids increased at a height of0.2 m but decreased at a height of 0.5 m and remained vir-tually constant up to the end of the reactor.[22]

Di Berardino et al.[23] studied the treatment of confec-tionery industry wastewater by an upflow anaerobic filter.COD removals higher than 80% were obtained, and thereactor demonstrated a great ability to assimilate changesin the waste characteristics and in the load fluctuations.Sampling at different heights demonstrated that the biogasgenerated ensured mixing within the reactor and that mostof the organic substances were used at the bottom of thereactor-located in the first 25% of the total height.

The subject of the present work was to study the effect ofthe reactor depth on the performance of downflow anaer-obic fixed-bed reactors (DFAFBRs) treating piggery wasteat different influent COD concentrations and HRTN . In ad-dition, the hydraulic characteristics of the system were alsostudied.

Materials and methods

Piggery waste

The piggery waste used in the experiments was collectedfrom a farm situated 20 km west of Havana City. After

collection, the waste was immediately screened through a2-mm sieve to remove coarse particles and was then al-lowed to settle for 1 hour to remove settleable solid parti-cles. The supernatant obtained was diluted using tap-waterat four different concentrations, giving total COD concen-trations of 2, 4, 6 and 8 g/L. The features and full char-acteristics of the raw piggery waste used are given in detailelsewhere.[24]

Equipment

The experiments were carried out in four laboratory scale-DFAFBRs (R1, R2, R3 and R4). The reactors consisted ofcylindrical glass columns with conical bottoms of 110 cmin height and 9.5 cm in inside diameter. The height of theconical part was 5 cm. Sampling ports consisted of 5-mmglass pipes and were situated every 10 cm within the reactorwith a total of 10 points for sampling. The support mediumremained totally submerged in the liquor by placing the ef-fluent outlet at the same height as the liquor level insidethe reactors. The four reactors were randomly packed with150 ceramic rasching rings of 2.0–2.2 cm in external diam-eter, 2.1–2.5 cm in length, with a specific surface area of1.18 cm2/mL. The overall wet contact area was estimatedto be 1,152 cm2. The total effective volume of the reactorswas 7 litres, while the free volume was 6 litres. Therefore, theratio between the free and effective volume of the reactors,considered as the bed porosity ( ), was 0.86. The columnswere hermetically sealed and closed with rubber caps pro-vided with two 5-mm holes, one for the influent feedingand the other for the biogas outlet. The reactors were con-tinuously fed in down-flow mode with peristaltic pumpsthrough a pipe totally submerged in the liquor covering thesupport media in order to maintain anaerobic conditionsand to prevent the biogas from escaping.

Inoculum

Each DFAFBR used in the experiments was inoculatedwith 1 litre of methanogenically active biomass from ananaerobic batch reactor processing piggery sludge after 40days of digestion time. The inoculum had a total solids(TS) concentration of 5%, with a biomass concentration,estimated as volatile suspended solids (VSS), of 21.4 g/L,and a pH of 7.3.

Acclimatization stage

The acclimatization of the biomass into the reactors wasachieved with the addition of small portions of inocu-lum mixed with water in the upper parts of the reactorsuntil the working volumes were reached. When the re-actors were completely filled, increasing aliquots of rawwastewater were successively pumped until a good balanceof acid-alkali was obtained. During this stage, the OLR wasgradually increased from 0.1 to 0.4 g COD/(L d). At the

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

Piggery waste treatment anaerobic reactors 729

end of this acclimatization period, the effluent pH valueswere around 7.0, the ratios between total volatile fatty acid(TVFA) and alkalinity concentrations were lower than 0.4,and, finally, the variations of effluent COD concentrationsof three consecutive samples were lower than 10%.

Experimental procedure

After the acclimatization step, the influents were fed to thereactors as is summarized in Table 1. Therefore, influentsubstrate concentrations of 2, 4, 6 and 8 g COD/Lwere fedto reactors 1, 2, 3 and 4, respectively. Each reactor oper-ated at one fixed influent concentration and six theoreticalHRTs in a decreasing order of 6, 5, 4, 3, 2 and 1 days.Each experiment lasted one month after steady-state con-ditions were achieved. The steady-state value of a givenparameter was taken as the average of three consecutivemeasurements for that parameter when the deviations be-tween the observed values were less than 5% in all cases.Steady-state was achieved in each case after a period equiv-alent to three times the corresponding nominal hydraulicretention time (HRTN). Therefore, transient periods of 18,15, 12, 9, 6 and 3 days were required for HRTN of 6, 5, 4,3, 2 and 1 days, respectively. During the operating period,the temperature varied from 25 ◦C to 30◦C.

Table 1. Experimental program carried out

Influentstrength (d) OLR

Run Reactor (g COD/L) HRTN (d) (g COD/(L d))

1 R1 2 6 0.302 R1 2 5 0.363 R1 2 4 0.454 R1 2 3 0.605 R1 2 2 0.906 R1 2 1 1.807 R2 4 6 0.578 R2 4 5 0.699 R2 4 4 0.86

10 R2 4 3 1.1411 R2 4 2 1.7212 R2 4 1 3.4413 R3 6 6 0.8614 R3 6 5 1.0315 R3 6 4 1.2916 R3 6 3 1.7217 R3 6 2 2.5818 R3 6 1 5.1619 R4 8 6 1.1520 R4 8 5 1.3821 R4 8 4 1.7222 R4 8 3 2.3023 R4 8 2 3.4424 R4 8 1 6.88

Frequency, Procedure for sampling and analyses

Influent and samples at depths of 0.1, 0.2, 0.3, 0.4, 0.5, 0.6,0.7, 0.8, 0.9 and 1 m were taken six times during each run,once a steady-state period was achieved. During the exper-iments, the following parameters were determined in orderto obtain the profiles through the different depths of the re-actors: soluble chemical oxygen demand (SCOD), organicnitrogen (O.N.), ammonia nitrogen (A.N.), pH and electri-cal conductivity (E.C.). Each determination was carried outin triplicate in order to minimize errors and as deviationswere always lower than 3%, average values are reported inall tables and figures of the present work. All analyses werecarried out according to Standard Methods for the Exam-ination of Water and Wastewater.[25]

Hydraulic tests

In order to determine the hydraulic characteristics of thereactors, the dispersion number (DN) and the real valueof the hydraulic retention time (θ ) under the experimen-tal operational conditions of feeding were calculated. Thefirst reactor (R1) was chosen for this purpose. These ex-periments were carried out by feeding a tracer (saturatedsolution of NaCl) by means of a pulse to the reactor, anddetermining the effluent response. NaCl concentration wasindirectly determined by the electrical conductivity (E.C.)of influent and effluent samples. The determination of DNand θ was carried out at the end of the experiment withthe biomass attached to the support medium at the samerange of HRTN used in the experimental runs. The pulseof sodium chloride solution was carried out over a veryshort period of time and the response of the reactor wasevaluated by the determination of the values of E.C. in theeffluent samples, according to the methodology describedby Sanchez et al.[24], Levenspiel[26] and Pena et al.[27]

Results and discussion

Profiles of SCOD at different influent strengths andnominal hydraulic retention times

Figure 1 shows the profiles of SCOD with the depth ofthe reactors R1, R2, R3 and R4. The profiles obtainedshowed a faster decrease in the substrate concentrationwith reactor depth in the first part of the reactors. Thisbehavior was most significant as the HRTN and influentstrength increased. For an influent COD of 2 g/L (reactorR1) and values of HRTN of 1, 2, 3, 4, 5 and 6 days, thepercentages of SCOD removal in the first half of the reac-tor depth (0.5 m) were 62.5%, 62.9%, 64.7%, 69.5%, 88.0%and 85.2%, respectively. The second part of the reactor lo-cated at depths higher than 0.5 m was clearly less activedue to the lower concentration of biodegradable organicmatter and the higher molecular weight of the remainingsubstrate.[22,23] In the case of reactor R2, operating with an

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

730 Sanchez et al.

Fig. 1. Experimental profiles of soluble chemical oxygen demand(SCOD) along reactor depths.

influent strength of 4 g COD/L, the two zones of SCOD re-moval appear to be more clearly established. At an influentstrength of 6 g COD/L (R3), the substrate concentrationdecreased faster at a depth of 0.3 m than in the rest of thereactor. This behavior was more noticeable as the HRTN in-creased. It was observed that more than 80% of the SCOD

removal was achieved in the zone located in the first 0.3 mof reactor depth. As can be seen, the influent concentrationinfluenced the profiles of SCOD along the reactor depth.Finally, the profiles of SCOD through the depth of reactorR4 for different values of HRTN and influent strength of8 g COD/L show the two zones of removal within the re-actor even more clearly. The most active zone was locatedat the first 0.2 m and the less active zone was located be-tween 0.3 to 1.0 m. This behavior was more noticeable asthe HRTN increased. For example at HRTN in the range of1–3 days and in the first zone of the reactor, the SCOD re-movals were in the range of 75%–85% of the total removalobtained, while in the range of 4–6 days the SCOD removalsin the active zone were in the range of 90%–93% of the totalremoval, the maximum removal value being at a HRTN of6 days. Therefore, the influent concentration influenced theprofiles of SCOD through the reactor depth because theslopes of the curves obtained were more pronounced thanthose obtained at lower influent substrate concentrations.The results obtained also show that the increase in HRTNhad a strong influence on the shape of the profiles of SCODwith the reactor depth. The increase in HRTN also favoredthe formation of two zones of removal within the reactor: Afirst more active zone whose depth decreased as the HRTNincreased and a second zone with a significantly lower re-moval where the SCOD concentration had slight variationsand equilibrium conditions may be achieved. Additionally,the increase in the influent strength also favored the masstransfer from the bulk liquid to the biofilm mainly at lowvalues of HRTN , as was previously reported.[22,28]

Profiles of organic nitrogen and ammonia at differentinfluent strengths and nominal hydraulic retention times

Figure 2 shows the profiles of organic nitrogen (O.N.) andammonia nitrogen (A.N.) in reactor depths for all the exper-iments carried out. For an influent strength of 2 g COD/L,the organic nitrogen decreased throughout the depth of thereactor but with different patterns as the HRTN increased.In the range of 1–3 days, the decrease in O.N. while therewas a change in depth was practically constant while atHRTNs in the range of 4–6 days, most of the O.N. removaltook place in the first part of the reactor (0.5 m deep). Onthe other hand, the A.N. concentration increased through-out the depth of the reactor, an obvious consequence ofthe decomposition of O.N. The increase in the HRTN alsofavored the formation of ammonia in the first part of thereactor.

In the case of R2, the shapes of the profiles of organicnitrogen and ammonia through the reactor depth were alsoa function of the HRTN . The slopes of the curves of theprofiles of O.N. also increased with the increase in HRTN .In the range of 1–3 days, the major decrease in O.N. con-centration occurred in the first half of the reactor depth(0.5 m) but at HRTN in the range of 4–6 days, the moreactive zone displaced from 0.5 m to 0.3 m deep. On the

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

Piggery waste treatment anaerobic reactors 731

Fig. 2. Experimental profiles of organic nitrogen (O.N.) alongreactor depth for HRTN of: 1 d (�), 2 d (�), 3 d (�), 4 d (•), 5 d(×) and 6 d (+) and of ammonia nitrogen (A.N.) along reactordepth for HRTN of: 1 d (Π), 2 d (�), 3 d (�), 4 d (◦), 5 d (�) and6 d (�) for the reactors R1, R2, R3 and R4, respectively.

other hand, the A.N. concentration increased through thedepth of the reactor due to the decomposition of organicnitrogen to ammonia under anaerobic conditions. As theHRTN increased, a greater quantity of A.N. was producedand higher concentrations were achieved at the end of thereactor. Therefore, an increase in the slopes in the curves ofA.N. formation with the increase in HRTN was observed.

Additionally, the depth of the more active zone decreased asthe HRTN increased. These results demonstrated the strongdependence of the process on the time of reaction and hy-draulic conditions.[28]

In reactor R3, organic nitrogen also decreased through-out the depth and the slopes of the curves obtained in-creased with the increase in the HRTN . The depth of themore active zone of the reactor also depended on theHRTN, being lower, as this parameter increased. For exam-ple, at HRTNs in the range of 1–3 days, the major decreasein O.N. concentration was located in the first half of the re-actor depth (0.5 m) but at higher values of HRTN, the moreactive zone was displaced at depths of between 0.3 m–0.4m. The ammonia concentration increased throughout thedepth of the reactor due to the decomposition of organicnitrogen to ammonia under anaerobic conditions. As theHRTN increased, a greater quantity of ammonia was pro-duced and higher concentrations of ammonia were achievedat the end of the reactor with respect to the initial concen-tration. Therefore, an increase in the slopes of the profilesof ammonia nitrogen concentration with the reactor depthand with the increase in HRTN was also observed. Addi-tionally, the depth of the more active zone decreased as theHRTN increased, which was located about halfway downthe reactor for 1 and 2 days and between 0.3 m–0.4 m forthe higher values.

Finally, for reactor R4, organic nitrogen decreasedthrough the depth of the reactor but the slopes of the curvesobtained increased with the rise in the HRTN . The depthof the more active zone of the reactor also depended on theHRTN , which was lower as this parameter increased. Forexample at HRTN of 1 and 2 days, the two zones of organicnitrogen removal were not well-defined, but at higher val-ues of HRTN , the two zones were better-defined, the mostactive one being around 0.3 m–0.5 m. It was observed thatat HRTNs of 5 and 6 days, the active zone decreased toa depth of 0.3 m. In the case of ammonia, the concentra-tion increased through the reactor depth due to the decom-position of organic nitrogen to ammonia under anaerobicconditions. As the HRTN increased, a greater quantity ofammonia was produced and higher concentrations of am-monia were achieved at the end of the reactor with respectto the initial concentration. As a consequence, an increasein the slopes in the profiles of ammonia nitrogen concentra-tion with the reactor depth and with the increase in HRTNwas also appreciated. Additionally, the depth of the moreactive zone decreased as the HRTN increased, which wasaround half the reactor for 1 and 2 days and between 0.3 m–0.4 m for higher values. These results again demonstratedthe strong dependence of the process with respect to thetime of reaction and the hydraulic conditions.[24,28]

Therefore, all the results obtained show that the increasein HRTN has a strong influence on the shape of the profilesof all parameters evaluated. The increase in HRTN favoredthe formation of two zones of removal within the reactor: Afirst, more active zone, whose depth decreased as the HRTN

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

732 Sanchez et al.

increased and a second zone, with a significantly lower re-moval rate, where the parameters had a slight variation andwhere equilibrium conditions may be achieved. Addition-ally, the increase in the influent strength also favored themass transfer from the bulk liquid to the biofilm, mainly atlow values of HRTN .[22,28]

Profiles of pH and electrical conductivity at differentinfluent strengths and nominal hydraulic retention times

Figure 3 shows the profiles of pH and electrical conductiv-ity (E.C.) through the reactor depths at different conditionsof influent strengths and HRTN . For reactor R1, the valueof pH increased with the depth as a consequence of the am-monia production due to the decomposition of the organicmatter, mainly in the case of proteinaceous compounds. Asthe HRTN increased, a higher increase in pH was appreci-ated throughout the reactor, mainly in the first half of thedepth, coinciding with the most active zone.[23] The elec-trical conductivity increased slightly in the first part of thereactor due to the liquefaction of the organic matter, whichbrought about the increase in the concentration of organicand inorganic salts. However, this phenomenon changed inthe lower part of the reactor and the concentration of sol-uble compounds appeared to decrease with a consequentreduction in the conductivity.

The profiles of pH and E.C. throughout the reactor depthat an influent strength of 4 g COD/L (R2) and HRTN from1–6 days also show an increase in pH with the depth asa consequence of the decomposition of the organic mat-ter and the reduction of organic acids concentration. Thisbehavior was strongly influenced by the HRTN . As theHRTN increased, a higher increase in pH was observedthrough the reactor, mainly in the zone located in the first0.5 m coinciding with the most active zone. The values ofE.C. also increased slightly in the first part of the reactordue to the liquefaction and decomposition of the organicmatter that determined the increase in the concentrationof organic and inorganic salts. However, this phenomenonchanged in the lower part of the reactor and the values ofE.C. decreased due to the reduction in the concentration ofsoluble compounds.

The profiles of pH and E.C. throughout the reactor depthat an influent strength of 6 g COD/L (R3) and HRTN from1–6 days show an increase in pH values with the increase inthe reactor depth as a consequence of the decomposition ofthe organic matter and the reduction of organic acids con-centration. This behavior was also strongly influenced bythe HRTN , and, in addition, the values of pH at the end ofthe reactor were higher than those achieved for lower influ-ent strengths. Moreover, as the HRTN increased, a higherincrease in pH was appreciated through the reactor mainlyin the first half of the reactor depth, coinciding with themost active zone. The values of E.C. increased in the firstpart of the reactor due to the liquefaction and decompo-sition of the organic matter which determined an increase

Fig. 3. Experimental profiles of pH along reactor depth for HRTNof: 1 d (�), 2 d (�), 3 d (�), 4 d (•), 5 d (×) and 6 d (+) and ofelectrical conductivity (E.C.) along reactor depth for HRTN of: 1d (Π), 2 d (�), 3 d (�), 4 d (◦), 5 d (�) and 6 d (�) for the reactorsR1, R2, R3 and R4, respectively.

in the concentration of organic and inorganic salts. How-ever, this phenomenon changed in the lower part of the reac-tor and the concentration of soluble compounds appearedto decrease with the subsequent reduction in the valuesof E.C.

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

Piggery waste treatment anaerobic reactors 733

The profiles of pH and E.C. throughout the reactor depthat an influent strength of 8 g COD/L (R4) and HRTN from1–6 days reveal that the values of pH through the reac-tor were strongly influenced by the HRTN . Values higherthan those obtained for lower influent strengths in spiteof the lower initial pH values were achieved at the endof the reactor.[23] Additionally, as the HRTN increased, ahigher increase in pH was appreciated throughout the re-actor, mainly, in the quarter of the reactor depth that co-incided with the most active zone. The values of E.C. alsoincreased in the active zone characterized by a higher vari-ation in substrate concentration. In the lower part of thereactor the value of E.C. decreased due to the decrease inthe concentrations of organic and inorganic salts by physi-cal and biological mechanisms.[24]

Determination of the real contact time

Based on previous results, the influence of the contact timeand the hydraulic pattern of the reactor were determined.Table 2 shows the average values of the real hydraulic re-tention time, the variance and the dispersion number (DN)for the different HRTN studied in the experiment. As canbe seen, an increase in the dispersion number was observedwhen the nominal hydraulic retention time increased as aconsequence of the rise of the axial speed of the tracer andtracer adsorbtion on the biofilm.[24,26] Therefore, the realhydraulic retention time was higher than the nominal ortheoretical hydraulic retention time in all cases except for anHRTN of one day, in which a certain short-circuiting mustoccur. Therefore, the increase in the HRTN determines theapproach of the reactor to the complete mixing conditionscausing the increase in the dispersion number, the maximumvalue being at 6 days, while at the lowest hydraulic retentiontime (1 d) the reactor approached plug-flow condition andthe dispersion number was the minimum. Therefore, the re-actors may be considered as intermediate hydraulic patternand the profiles of the different parameters evaluated cor-roborate these conditions. Based on the data in Table 2, thereal retention time for different depths and HRTN were de-termined and defined as contact time or time of reaction (θ ).

Table 2. Variation of the average real hydraulic retention time(HRTR) and the dispersion number (DN) with the theoretical ornominal hydraulic retention time (HRTN)

HRTN HRTR(d) (d) Variance DN

1 0.7 0.28 0.172 2.1 0.37 0.243 3.4 0.42 0.294 4.7 0.44 0.325 6.4 0.47 0.356 8.0 0.49 0.37

Fig. 4. Experimental profiles of soluble chemical oxygen demand(SCOD) at different influent strengths and contact times (θ )within the reactors.

Effect of the real contact time on the process

Figure 4 shows the profiles of SCOD as a function of thecontact time (θ ) within the reactor for the different influentstrengths studied (reactors: R1, R2, R3 and R4). The plotsof the experimental values obtained show that the substrateconcentration decreased exponentially with the increase inthe value of θ . It was also found that the influent strengthhad a strong influence on the SCOD concentration mainlyat values of θ under 1.5 days, which corresponded to the firstpart of the reactors and at low values of nominal hydraulicretention time.

Conclusions

The results obtained demonstrated that organic matter re-moval within the DFAFBRs occurs mainly in a zone locatedin the first part of the reactor. The slopes of the profiles ofdifferent parameters evaluated increased with the influentstrength and with the nominal hydraulic retention time. Itwas found that the depth of the active zone decreased asthe HRTN increased. The increase in the HRTN determinedthe approach of the reactor to the complete mixing condi-tions. The reactors may be considered as an intermediatehydraulic pattern between plug-flow and back-mix.

The plots of the experimental values of SCOD obtainedversus the contact time (θ ) showed that the substrate con-centration decreased exponentially with the increase in thevalue of θ . It was also found that the influent strength hada strong influence on the SCOD concentration mainly atvalues of θ under 1.5 days (which corresponded to the firstpart of the reactors) and at low values of nominal hydraulicretention time.

Acknowledgments

The authors wish to express their gratitude to Ministeriode Educacion y Ciencia and Junta de Andalucia of Spain

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4

734 Sanchez et al.

for providing financial support, to the Ministerio de Cien-cia, Tecnologıa y Medio Ambiente of Cuba and to theAlexander von Humboldt Foundation of Germany.

References

[1] Kennedy, K.J.; Van den Berg, L. Anaerobic digestion of piggery-waste using stationary fixed film reactor. Agric. Wastes 1982, 4,151–158.

[2] Kennedy, K.J.; Van den Berg, L. Continuous vs. slug loading ofdown-flow stationary fixed film reactors digesting piggery waste.Biotechnol. Lett. 1982, 4, 137–142.

[3] Wilkie, A.; Coleran, E. Start-up of anaerobic filters containing dif-ferent support materials using pig slurry supernatant. Biotechnol.Lett. 1984, 6, 735–740.

[4] Wilkie, A.; Coleran, E. Pilot-scale digestion of pig slurry anaerobicfilters containing different support materials using pig slurry super-natant. Biotechnol. Lett. 1984, 6, 741–746.

[5] Hill, D.T.; Bolte, J.P. Synthetic fixed media reactor performancetreating screened swine waste liquids. Trans ASAE 1988, 31, 1525–1531.

[6] Sanchez, E.; Monroy, L.; Canizares, R.O.; Travieso, L. Comparativestudy of piggery waste treatment by upflow sludge beds anaerobicreactors and packed bed reactors. J. Agric. Eng. Research 1995, 62,71–76.

[7] Sanchez, E.; Weiland, P.; Travieso, L. Effect of the organic volumet-ric loading rate on soluble COD removal in down-flow anaerobicfixed-bed reactors. Bioresour. Technol. 1994, 47, 173–176.

[8] Sanchez, E.; Milan, Z.; Borja, R.; Weiland, P.; Rodrıguez, X., Pig-gery waste treatment by anaerobic digestion and nutrient removalby ionic exchange. Resour. Conserv. Recy. 1995, 15, 235–244.

[9] Ng, W.J.; Chin, K.K. Treatment of piggery waste by expanded bedanaerobic filters. Biol. Wastes 1998, 26, 215–228.

[10] Sanchez, E.; Borja, R.; Travieso, L.; Martın, A.; Colmenarejo, M.F.;Nikolaeva, S. An evaluation of variables affecting the stability andperformance of down-flow anaerobic fixed bed reactors treating pig-gery wastewater. Environ. Technol. 2006, 27, 85–93.

[11] Ahn, J.H.; Foster, C.F. Kinetic analyses of the operation ofmesophilic and thermophilic anaerobic filters treating a simulatedstarch wastewater. Process Biochem. 2000, 36, 19–23.

[12] Ahn, J.H.; Foster, C.F. A comparison of mesophilic and ther-mophilic anaerobic upflow filters treating paper-pulp-liquors. Pro-cess Biochem. 2002, 38, 257–262.

[13] Bodik, I.; Herdova, B.; Drtil, M. The use of upflow anaerobic fil-ter and AnsSBR for wastewater treatment at ambient temperature.Water Res. 2002, 36, 1084–1088.

[14] Show, K.Y.; Tay, J.H. Influence of the support media on biomassgrowth and retention in anaerobic filters. Water Res. 1999, 33, 1471–1480.

[15] Gouari, S.; Achkari-Begdouri, A. Use of backed clay media asbiomass supports for anaerobic filters. Appl. Clay Sci. 1997, 12,365–375.

[16] Borja, R.; Sanchez, E.; Martın, A.; Jimenez, A.M. Kinetic behaviorof waste tyre rubber as microorganisms support in an anaerobic di-gester treating cane molasses distillery slops. Bioprocess Eng. 1996,16, 17–23.

[17] Reyes, O.; Sanchez, E.; Rovirosa, N.; Borja, R.; Cruz, M.;Colmenarejo, M.F.; Escobedo, R.; Ruiz, M.; Rodrıguez, X; Correa,O. Low-strength wastewater treatment by a multistage anaerobic fil-ter packed with waste tyre rubber. Bioresour. Technol. 1999, 70, 55–60.

[18] Nikolaeva, S.; Sanchez, E.; Borja, R.; Travieso, L.; Weiland, P.;Milan, Z. Treatment of piggery waste by anaerobic fixed bed re-actor and zeolite bed filter in a tropical climate: a pilot scale study.Process Biochem. 2002, 38, 405–409.

[19] Ruiz, I.; Veiga, M.C.; de Santiago, P.; Blazquez, R. Treatment ofslaughterhouse wastewater in an UASB reactor and an anaerobicfilter. Bioresour. Technol. 1997, 60, 251–268.

[20] Yu, H.Q.; Hu, Z.H.; Hong, T.Q.; Gu, G.W. Performance ofan anaerobic filter treating soybean processing wastewater withand without effluent recycle. Process Biochem. 2002, 38, 507–513.

[21] Jawed, M.; Tare, V. Post-mortem examination and analysis of anaer-obic filters. Bioresour. Technol. 2000, 72, 75–84.

[22] Borja, R.; Banks, C.J. Treatment of palm oil mill effluent by upflowanaerobic filtration. J. Chem. Technol. Biotechnol. 1994, 61, 103–109.

[23] Di Berardino, S.; Costa, S.; Converti, A. Semi-continuous anaero-bic digestion of a food industry wastewater in an anaerobic filter.Bioresour. Technol. 2000, 71, 261–266.

[24] Sanchez, E.; Borja, R.; Travieso, L.; Colmenarejo, M.F.; Chica, A.;Martın, A. Treatment of settled piggery waste by a down-flow anaer-obic fixed bed reactor. J. Chem. Technol. Biotechnol. 2004, 79, 851–862.

[25] American Public Health Association (APHA). Standard Methodsfor the Examination of Water and Wastewater, 18th edn. APHA:Washington DC, 1992.

[26] Levenspiel, O. Chemical Reaction Engineering, 3rd edn. John Wileyand Sons: New York, 1986.

[27] Pena, M.R.; Mara, D.D.; Sanchez, A. Dispersion studies in anaer-obic ponds: implications for design and operation. Water Sci. Tech-nol. 2000, 42, 273–282.

[28] Huang, J.S.; Jih, C.G. Deep-Biofilm kinetics of substrate utilizationin anaerobic filters. Water Res. 1997, 31, 2309–2317.

Dow

nloa

ded

by [

Lau

rent

ian

Uni

vers

ity]

at 0

0:50

08

Oct

ober

201

4