anaerobic digestion of animal waste: effect of mixing

Bioresource Technology 96 (2005) 1607–1612

Anaerobic digestion of animal waste: Effect of mixing

Khursheed Karim a, K. Thomas Klasson b, Rebecca Hoffmann a, Sadie R. Drescher b,David W. DePaoli b, M.H. Al-Dahhan a,*

a Chemical Reaction Engineering Laboratory (CREL), Department of Chemical Engineering, Washington University,

1 Brookings Drive, Campus Box 1198, St Louis, MO 63130 4899, USAb Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA

Received 3 March 2004; received in revised form 2 December 2004; accepted 4 December 2004

Available online 26 February 2005

Abstract

Six laboratory scale biogas mixed anaerobic digesters were operated to study the effect of biogas recycling rates and draft tube

height on their performance. The digesters produced methane at 0.40–0.45 L per liter of digester volume per day. A higher methane

production rate was observed in unmixed digesters, while increased biogas circulation rate reduced methane production. However,

different draft tube heights caused no difference in the methane production rate. Air infiltration (up to 15% oxygen in the biogas) was

observed in the digesters mixed by biogas recirculation. Slight air permeability of tubing or leakage on the vacuum side of the air

pump may have caused the observed air infiltration. The similar performance of the mixed and unmixed digesters might be the result

of the low solids concentration (50 g dry solids per liter of slurry) in the fed animal slurry, which could be sufficiently mixed by the

naturally produced biogas.

� 2005 Elsevier Ltd. All rights reserved.

Keywords: Anaerobic; Biogas; Draft tube; Digestion; Manure; Mixing

1. Introduction

Recent growth of the livestock industries has given

rise to new requirements for safe disposal of large quan-

tities of animal waste (manure) generated at dairy,

swine, and poultry farms. In the United States approxi-

mately 230 million tons of animal waste (dry matter) are

generated every year (Sheffield, 2002). Unsafe and im-

proper disposal of decomposable animal waste causesmajor environmental pollution problems, including sur-

face and groundwater contamination, odors, dust, and

ammonia emission. There is also a concern regarding

methane emissions, which contribute to the green house

effect. Through anaerobic digestion, these large amounts

of waste can be converted to methane, a renewable en-

0960-8524/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.biortech.2004.12.021

* Corresponding author. Tel.: +1 314 935 7187; fax: +1 314 935 7211.

E-mail address: [email protected] (M.H. Al-Dahhan).

ergy source. A survey of dairy and swine farms in thecountry reaffirmed that anaerobic digestion is a technol-

ogy with considerable potential (Lusk, 1998). The sur-

vey further suggests that, ignoring caged layer poultry,

about 0.426 Tg of methane are potentially recoverable

from 3000 dairy and swine farms in 19 states of the

USA.

Over the past 25 years, anaerobic digestion processes

have been applied to a wide array of industrial and agri-cultural wastes (Speece, 1996; Ghosh, 1997). The perfor-

mance of anaerobic digesters is affected primarily by the

retention time of the substrate in the reactor and the

degree of contact between incoming substrate and the

viable bacterial population. These parameters are a

function of the flow pattern (mixing) inside the reactors.

Thorough mixing of the substrate in the digester distrib-

utes organisms uniformly and also transfers heat, andthus is regarded as essential in high-rate anaerobic

mailto:[email protected]

Draft tube (38 mm dia)140mm

152 mm

194 mm

26 mm

344 mm

Gas compositionsampling

Gas bag for gascollection

Wet test meter forperiodic gas productionmeasurements

Air pump for biogasrecirculation

Valve for feed

Valve for effluent

Fig. 1. Schematic diagram of the experimental set-up.

Table 1

Operational conditions for the digesters

Digesters Biogas recirculation

rate (L/min)

Draft tube height from

bottom (mm)

Digester 1 None (unmixed) 40

Digester 2 1 40

Digester 3 1 26

Digester 4 2 40

Digester 5 3 40

Digester 6 1 13

1608 K. Karim et al. / Bioresource Technology 96 (2005) 1607–1612

digesters (Sawyer and Grumbling, 1960; Meynell, 1976).

Furthermore, agitation helps to reduce particle size as

digestion progresses and to release biogas from the mix-

ture. The importance of mixing in achieving efficient

substrate conversion has been noted by many research-

ers (Casey, 1986; Lee et al., 1995; Smith et al., 1996),although the optimum mixing pattern is a subject of

much debate. An intermediate degree of mixing appears

to be optimal for substrate conversion (Smith et al.,

1996).

Mixing can be accomplished by mechanical mixers,

biogas recirculation, or by slurry recirculation. Mechan-

ical mixers are reported to be most efficient in terms of

power consumed per gallon mixed (Brade and Noone,1981). However, the internal fittings and equipment

are not accessible for maintenance during digester oper-

ation, and long term reliability of operation is of para-

mount importance. In general, such reliability can be

more readily attained with biogas or liquor recirculation

systems, where there are no moving parts within the di-

gester (Casey, 1986). Interestingly, in other literature

sources it has been reported that biogas recirculationis the most efficient mode of mixing for anaerobic digest-

ers (Morgan and Neuspiel, 1958; Kontandt and Roedi-

ger, 1977; Lee et al., 1995).

In the case of a digester mixed with biogas, various

parameters which can affect the flow pattern (mixing) in-

side the digester are: biogas recycling rate, bottom clear-

ance of the draft tube, slope of the hopper bottom, draft

tube to tank diameter ratio, position of the biogas injec-tion (sparger) and its design, solids loading rate, among

other factors. The current paper evaluates the effects of

biogas recycling rates and draft tube heights on the per-

formance of anaerobic digesters mixed with biogas

recirculation.

2. Methods

2.1. Experimental design

Six laboratory scale digesters (Digesters 1–6) having

a working volume of 3.73 L each were operated at a

controlled temperature of 35 ± 2 �C. Biogas generated

in each digester was collected in a Tedlar bag and was

recirculated from the top of the digester by an air pumpand draft tube arrangement, as shown in Fig. 1. The

digesters were inoculated with 373 mL anaerobic seed

sludge collected from a dairy farm operated by the

University of Tennessee. The seed sludge had total

suspended solids (TSS) of 66.13 g/L and volatile sus-

pended solids (VSS) of 35.63 g/L. The remaining 90%

of the working volume was filled with freshly prepared

manure slurry having a 5% dry solid concentration (thatis, 100 mL of slurry contains 5 g of dry solids). Hydrau-

lic retention time (HRT) was kept constant at 16.2 days,

resulting in a solid loading rate of 3.08 g TS/L-d and an

organic loading of 3.37 g TCOD/L-d. Approximately460 mL of effluent was taken from the bottom of the

digesters on alternate days and replaced with the same

amount of the prepared animal waste slurry. The digest-

ers were operated with four different biogas recirculation

rates and three different draft tube heights described in

Table 1. Digesters were operated under steady-state

conditions for about three to four weeks. Steady-state

conditions were assumed when the coefficient ofvariation for daily biogas production was less than

10% (Vartak et al., 1997).

2.2. Analytical methods

Feed and effluent samples were analyzed for total sol-

ids (TS), volatile solids (VS), TSS, VSS, volatile fatty

acids (VFA), total chemical oxygen demand (TCOD),dissolved chemical oxygen demand (DCOD), and total

nitrogen (TN). The total volume of the biogas generated

was measured, and the composition of the biogas was

analyzed three times each week. All the analysis were

performed per standard procedures (APHA, 1998) un-

less otherwise noted.

0

10

20

30

40

50

60

70

80

11-Oct 21-Oct 31-Oct 10-Nov 20-Nov 30-Nov 10-Dec 20-Dec

Date (2002)

Met

hane

Con

tent

(%

)

0.0

0.5

1.0

1.5

Bio

gas

Prod

uctio

n (L

/L-d

)

Gas Composition

Gas Production

Fig. 2. Performance characteristics of Digester 1.

0

10

20

30

40

50

60

70

80


Date (2002)

Met

hane

Con

tent

(%

)

0.0

0.5

1.0

1.5

Bio

gas

Prod

ucti

on (

L/L

-day

)

Gas Composition

Gas Production


K. Karim et al. / Bioresource Technology 96 (2005) 1607–1612 1609

Volatile fatty acids (formic, acetic, propionic, butyric,

and valeric acids) were determined by centrifuging a

small sample at 10,000 rpm for 5 min, filtering the liquid

through a 0.2 lm-pore-size filter, and injecting a 10 lLsample into a high pressure liquid chromatograph

(HPLC). In the HPLC, the mobile phase (filtered5 mM H2SO4) was pumped at 0.6 mL/min through a

300 mm · 7.8 mm (8 lm particle size) RHM Monosac-

charide column (Phenomenex, Torrance, CA) held at a

temperature of 65 �C to a refractive index detector

(Model 2410, Waters Corporation, Milford, MA) held

at a temperature of 40 �C.Total nitrogen was determined by diluting approxi-

mately 1 g of slurry to 50 mL with ultra pure water. Fivemilliliter of this dilute solution was further diluted with

5 mL of ultra pure water and then digested for 60 min at

121 �C, using 1.5 mL potassium persulfate solution (1 g

K2S2O8, 1 g NaOH, 17 mL ultra pure water) in Teflon-

capped vials. Digested samples were cooled and pH was

adjusted with H2SO4 (50%) to pH 2.7. This procedure

converted all organic and inorganic nitrogen to nitrate,

which was measured using a nitrate ion-sensitiveelectrode (Models 9307, 9300BN, and 900200; Thermo

Orion, Beverly, MA) (Smart et al., 1983; Ferree and

Shannon, 2001).

Biogas volume was measured using wet gas test meters

(GSA/Precision Scientific, Chicago, Ill), and 150 lL sam-

ples for biogas composition were collected using a gas-

tight syringe. The samples were injected in duplicate into

a Hewlett Packard gas chromatograph (Model 5890Series II, Avondale, PA) gas chromatograph (GC) with

a 0.53 mm · 30 m GS-Q phase capillary column (J &

W Scientific, Folsom, CA). The injector, oven, and ther-

mal conductivity detector (TCD) temperatures were

kept as 125, 50, and 250 �C, respectively. The carrier gas

(helium) flow rate through the column was maintained

at 4 mL/min. The samples were injected in a split mode

with approximately 10% of the sample going throughthe column. The column, make-up, and reference gas

in the GC was helium. Initially the gas chromatograph

was calibrated with separate runs of 99.9% pure methane

(CH4), carbon dioxide (CO2), and air injected in volumes

between 50 and 200 lL. Later, periodic calibration was

performed by injecting different amounts of air and using

the relationship between TCD response factors described

by Dietz (1967).

2.3. Data analysis

Steady-state digester performance data were collected

for four different biogas recirculation rates and three

different draft tube positions as described in Table 1.

Steady-state conditions were assumed when the coeffi-

cient of variation for daily biogas production was lessthan 10%. Average steady-state data and the standard

error presented in the paper were calculated as a mean

value over 22 days (i.e., 11 observations). Statistical

significance (P = 0.05) of the experimental data was

tested using one way ANOVA and t -test statistical

program (Microsoft Excel 2002).

3. Results and discussion

Laboratory scale anaerobic digesters were operated

to evaluate the effect of different biogas recirculation

rates and draft tube heights on the methane production

rate. As shown in Table 1, Digesters 1, 2, 4, and 5 had

the same experimental conditions except for the biogas

recirculation rate, while digesters 2, 3, and 6 had thesame biogas recirculation rate but different draft tube

heights. In Figs. 2–7 the performance characteristics of

the six digesters over approximately 60 days are shown.

Initially, the biogas production rate varied in all six

digesters. This corresponds to a period when a feeding

0

10

20

30

40

50

60

70

80


Date (2002)

Met

hane

Con

tent

(%

)

0.0

0.5

1.0

1.5

Bio

gas

Prod

uctio

n (L

/L-d

)

Gas Composition

Gas Production


0

10

20

30

40

50

60

70

80


Date (2002)

Met

hane

Con

tent

(%

)

0.0

0.5

1.0

1.5B

ioga

s Pr

oduc

tion

(L/L

-d)

Gas Composition

Gas Production


0

10

20

30

40

50

60

70

80


Date (2002)

Met

hane

Con

tent

(%

)

0.0

0.5

1.0

1.5

Bio

gas

Prod

uctio

n (L

/L-d

)

Gas Composition

Gas Production


0

10

20

30

40

50

60

70

80


Date (2002)

Met

hane

Con

tent

(%

)

0.0

0.5

1.0

1.5

Bio

gas

Prod

uctio

n (L

/L-d

)

Gas Composition

Gas Production



procedure was changed from three times a week (Mon-

day, Wednesday and Friday) to alternate day feeding.

The change in feeding procedure was made to achieve

a HRT of 16.2 days. The biogas production rate was af-

fected during this period, but the methane content of the

biogas always remained close to 65%. Gradually the

digesters stabilized, and steady-state conditions of lessthan 10% variation in daily biogas production were

achieved in approximately 28–30 days.

The average steady-state biogas production rate and

the methane content are presented in Table 2. Biogas

and methane production rates were calculated as the

volume of biogas/methane produced per unit volume

of digester per day. The digesters produced methane

at a rate of 0.40–0.45 L/L-day (Table 2). The averagesteady-state data of TS, VS, TSS, VSS, VFA, TCOD,

DCOD and TN in the feed and effluents from all six

reactors are given in Table 3. In all six digesters, TS

and VS reductions were about 30–38% and 40–48%,

respectively. Total COD in the feed was approximately

52.7 g/L, 18% of which was present in the form of dis-

solved COD. The observed reduction of TCOD was

44–50% for all six digesters. Volatile fatty acids didnot accumulate in any of the digesters. As shown in Ta-

ble 3, the only fatty acid detected in effluent was acetic

acid, and the concentration of VFA in effluents was very

low. The results obtained in the present study were in

accordance with the wide range of data reported in the

literature. For example, Quasim and Warren (1984)

operated a completely mixed mesophilic digester with

cattle manure at an organic loading rate of 3.2 kg/m3dand achieved a 52.9% volatile solids conversion to gas.

In an another study, Ghaly et al. (1992) observed 40%

VS conversion in a completely mixed dairy waste diges-

ter at an organic loading rate of 2 kg/m3d. On the other

hand, Robbins et al. (1983) observed 30% VS conversion

in a completely mixed mesophilic digester operated at an

organic loading rate of 2.6 kg/m3d.

In order to evaluate the effect of the biogas recircula-tion rates, the methane production rates of digesters 1,

2, 4, and 5 were plotted as a function of biogas circula-

tion rate, as shown in Fig. 8. The unmixed digester

Table 2

Performance data of the digesters

Biogas production rate (L/L-d) Methane content (% CH4) CH4 Production rate (L/L-d) H2S (ppm) O2 Infiltration (L/L-d)

Digester 1 0.68 ± 0.027 67 ± 0.1 0.45 ± 0.019 350 0.01

Digester 2 0.67 ± 0.013 66 ± 0.2 0.44 ± 0.008 <2.5 0.03

Digester 3 0.69 ± 0.025 65 ± 0.6 0.45 ± 0.019 <5 0.02

Digester 4 0.60 ± 0.023 66 ± 0.3 0.40 ± 0.016 <5 0.03

Digester 5 0.61 ± 0.022 67 ± 0.6 0.41 ± 0.016 10 0.02

Digester 6 0.65 ± 0.046 66 ± 0.3 0.44 ± 0.031 <2.5 0.03

Table 3

Average steady-state observation data for feed and effluents

TS (%) VS (%) TSS (%) VSS (%) VFA (g/L) TCOD (g/L) DCOD (g/L) TN (g/L)

Feed 5.0 2.5 2.9 1.5 3.0 52.7 9.45 1.33

Digester 1 3.1 1.3 2.6 1.2 0.06 26.5 4.22 1.37

Digester 2 3.5 1.5 2.6 1.3 0.01 28.2 3.75 1.65

Digester 3 3.2 1.4 2.0 0.9 0.01 28.6 3.83 1.37

Digester 4 3.5 1.6 3.2 1.5 0.03 29.3 3.96 NA

Digester 5 3.3 1.5 NA NA 0.03 28.8 4.09 NA

Digester 6 3.5 1.5 NA NA 0.02 27.2 3.95 NA

NA: Not analyzed.

K. Karim et al. / Bioresource Technology 96 (2005) 1607–1612 1611

(Digester 1) produced more methane than the others,

and increasing the biogas recirculation rate reduced

the methane production rate. The steady-state methane

production rate data were subjected to analysis of vari-

ance (ANOVA), and a significant difference in methane

production rates was observed at the 5% level

(F = 28.86, df = 3, 40) for the four digesters with differ-

ent biogas recirculation rates. Out of the four digesters,Digester 1 (unmixed) had the highest methane produc-

tion rate, 0.45 L/L-d. Based on a t-test at the 5% level

it is apparent that the Digester 2 (mixed at 1 L/min bio-

gas recirculation rate) produced methane at a rate simi-

lar to that of Digester 1 (t = 1.26, df = 40). Based on

(0.02)

(0.01)(0.03)

(0.03)

0.3

0.4

0.5

0.6

0 1 2 3

Gas Circulation Rate (L/min)

CH

4 Pr

oduc

tion

(L/L

-d)

Fig. 8. Methane production rate for different biogas recirculation

rates. (Value in parenthesis corresponds to estimated O2 infiltration

rates in L O2/L-day.)

these results, it can be concluded that increased gas cir-

culation rates reduced the methane production rate. It is

important to note that biogas circulation in laboratory

digesters increases the chances for infiltration of air into

the system (due to air permeable tubing, leakage, and

other factors) as shown in Table 2. Since the presence

of oxygen is known to have a detrimental effect on meth-

ane production (Patel et al., 1984), the difference in theperformance of the digesters could be attributed to oxy-

gen infiltration. The oxygen infiltration was estimated

from the nitrogen content in the biogas, which ranged

from 3% in digester 1 to 15% in Digester 4, assuming

all nitrogen in the biogas was due to air infiltration.

However, no clear relationship between O2 infiltration

and methane productivity could be established. One

important observation during this study was that slightair infiltration into the digesters resulted in the reduction

of hydrogen sulfide (H2S) in the biogas, as shown in

Table 2. Higher methane production rates in unmixed

digesters have also been reported by Ghaly and Ben-

Hassan (1989). On the other hand Bello-Mendoza and

Sharratt (1998) observed that unmixed reactors perform

poorly, especially when they were large.

The effect of the draft tube heights on the methaneproduction rate of the digesters (Digesters 2, 3 and 6)

mixed by biogas recirculation at a rate of 1 L/min is

shown in Fig. 9. The digesters performed quite similarly.

The steady-state methane production rate data were

subjected to analysis of variance (ANOVA) and found

to vary insignificantly at the 5% level (F = 1.64, df = 2,

30). Thus, there was no effect of different draft tube

heights on the methane production rate of the digesters.

(0.03)

(0.02)(0.03)

0.3

0.4

0.5

0.6

0 5 10 15 20 25 30 35 40 45

Bottom Position of Draft Tube (mm)

CH

4 P

rodu

ctio

n (L

/L-d

)

Fig. 9. Methane production for digesters having different draft tube

heights. (Value in parenthesis corresponds to estimated O2 infiltration

rates in L O2/L-day.)


4. Conclusions

There was no difference in the performance of all six

digesters with different mixing conditions. This could be

because of the low solids concentration in the fed animal

slurry or because of the long HRT of 16.2 days, during

which mixing created by the naturally produced gas wassufficient for substrate distribution. However, it would

be interesting to see if the mixing patterns inside the

digesters changed with the applied physical changes in

the mixing conditions or not.

Acknowledgements

The authors would like to thank the United States

Department of Energy for sponsoring the research pro-

ject (Identification Number: DE-FC36-01GO11054).

References

APHA, 1998. Standard methods for the examination of water and

wastewater. 20th ed., American Public Health Association, Wash-

ington DC.

Bello-Mendoza, R., Sharratt, P.N., 1998. Modeling the effects of

imperfect mixing on the performance of anaerobic reactors for

sewage sludge treatment. J. Chem. Technol. Biotechnol. 71, 121–130.

Brade, C.E., Noone, G.P., 1981. Anaerobic digestion—need it be

expensive? Wat. Pollut. Control 80, 70–76.

Casey, T.J., 1986. Requirements and methods for mixing in anaerobic

digesters. Anaerobic Digestion of Sewage Sludge and Organic

Agricultural Wastes. Elsevier App. Sci. Pub., 90–103.

Dietz, W.A., 1967. Response factors for gas chromatographic Ana-

lyses. J. Gas Chromatogr. 5, 68–71.

Ferree, M.A., Shannon, R.D., 2001. Evaluation of a second derivative

UV/Visible spectroscopy technique for nitrate and total nitrogen

analysis of wastewater samples. Wat. Res. 35 (1), 327–332.

Ghaly, A.E., Ben-Hassan, R.M., 1989. Continuous production of

biogas from dairy manure using an innovative no-mix reactor.

Appl. Biochem. Biotechnol. 20/21, 541–559.

Ghaly, A.E., Echiegu, E.A., Ben-Hassan, R.M., 1992. Performance of

a continuous mix anaerobic reactor operating under diurnally

cyclic temperature. In: Presented at the ASAE International

Summer Meeting, Charlotte, North Carolina. ASAE Paper No.

92-6025.

Ghosh, S., 1997. Anaerobic digestion for renewable energy and

environmental restoration. The 8th International Conference on

Anaerobic Digestion, Sendai International Center, Sendai, Japan,

Ministry of Education, Japan.

Kontandt, H.G., Roediger, A.G., 1977. Engineering operation and

economics of methane gas production. In: Schlegel, H.G., Barnea,

J. (Eds.), Microbial Energy Conversion. Pergamon Press, pp. 379–

392.

Lee, S.R., Cho, N.K., Maeng, W.J., 1995. Using the pressure of biogas

created during anaerobic digestion as the source of mixing power.

J. Ferment. Bioengng. 80 (4), 415–417.

Lusk, P., 1998. Methane recovery from animal manures: A current

opportunities casebook. 3rd ed. NREL/SR-580-25145. Golden,

CO: National Renewable Energy Laboratory. Work performed by

Resource Development Associates, Washington, DC.

Meynell, P.-J., 1976. Methane: Planning a Digester. Prism Press,

London, 55–57.

Morgan, P.F., Neuspiel, P.J., 1958. Environmental control of

anaerobic digestion with gas diffusion. In: McCabe, J., Eckenfelder,

W.W. (Eds.), Biological Treatment for Sewage and Industrial

Wastes, vol. 2. Reinhold, New York.

Patel, G.B., Roth, L.A., Agnew, B.J., 1984. Death rates of obligate

anaerobes exposed to oxygen and the effect of media prereduction

on cell viability. Can. J. Microbiol. 30, 228–235.

Quasim, S.R., Warren, K., 1984. Methane gas production from

anaerobic digestion of cattle manure. Energy Sources 7 (4), 319–

341.

Robbins, J.E., Armold, M.T., Weiel, J.E., 1983. Anaerobic digestion of

cellulose dairy cattle manure mixture. Agric. Wastes 8, 105–118.

Sawyer, C.N., Grumbling, A.M., 1960. Fundamental consideration in

high-rate digestion. Inc. Sew. Engng. Div. ASCE, 86–92.

Sheffield, J., 2002. Financial approaches to animal manure manage-

ment. In: Animal Residuals 2002 Conference and Workshop

Report (No: JIEE 2002-04).

Smart, M.M., Rada, R.G., Donnermeyer, G.N., 1983. Determination

of total nitrogen in sediments and plants using persulfate digestion.

Wat. Res. 17 (9), 1207–1211.

Smith, L.C., Elliot, D.J., James, A., 1996. Mixing in upflow anaerobic

filters and its influence on performance and scale-up. Wat. Res. 30

(12), 3061.

Speece, R.E., 1996. Anaerobic Biotechnology for Industrial Waste-

waters. Vanderbilt University, Arachae Press.

Vartak, D.R., Engler, C.R., McFarland, M.J., Ricke, S.C., 1997.

Attached-film media performance in psychrophilic anaerobic

treatment of dairy cattle wastewater. Biores. Technol. 62 (3), 79–

84.

anaerobic digestion of animal waste: effect of mixing

Documents