Bioresource Technology 124 (2012) 163–168
Contents lists available at SciVerse ScienceDirect
Bioresource Technology
journal homepage: www.elsevier .com/locate /bior tech
Feasibility of anaerobic co-digestion of pig waste and paper sludge
Prathap Parameswaran ⇑, Bruce E. RittmannSwette Center for Environmental Biotechnology, The Biodesign Institute at Arizona State University, P.O. Box 875701, Tempe, AZ 85287–5701, USA
h i g h l i g h t s
" Benefits of co-digestion of pig waste and paper sludge demonstrated by BMP assays." Hydrolysis constants for co-digestion 2–20 times higher than baseline wastes." Semi-continuous digester shows higher performance for co-digestion than pig waste.
a r t i c l e i n f o
Article history:Received 28 May 2012Received in revised form 24 July 2012Accepted 26 July 2012Available online 15 August 2012
Keywords:MethanogenesisPig wastePaper sludgeCo-digestionHydrolysis
0960-8524/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.biortech.2012.07.116
⇑ Corresponding author. Tel.: +1 480 727 0849; faxE-mail address: [email protected] (P. Parameswara
a b s t r a c t
Pig waste (PW) and paper sludge (PS) possess complementary properties that can be combined for successfulanaerobic digestion. Biochemical methane potential (BMP) tests revealed that a PW:PS 3:1 (v/v) ratio had thehighest normalized CH4–COD removal (54%), while PS had the lowest value (11%) and PW had 44%. BatchBMP tests revealed a significant decrease in lag times for methane production in the order of PW:PS 1:3(14 days) < PW:PS 1:1 (17 days) < PW:PS 3:1 (20 days) < PW (23 days). Hydrolysis constants (khyd) werehigher for all PW:PS combinations than for either of the individual waste streams: 0.004 d�1 (PS) < 0.02 d�1
(PW) < 0.024 d�1 (PW:PS 3:1) < 0.03 d�1 (PW:PS 1:1) < 0.05 d�1 (PW:PS 1:3). Semi-continuous reactors per-forming co-digestion of PW and PS at a 2:1 ratio showed 1.5 times higher methane production than baselinePW-only reactors, confirming the BMP results.
� 2012 Elsevier Ltd. All rights reserved.
1. Introduction
The carbon in organic wastes has high-energy electrons that canbe transformed into useful forms of energy for society. Anaerobicdigestion, a mature technology for capturing these electrons asmethane gas (CH4), is widely used worldwide. In the USA, forexample, over 1500 anaerobic digesters are currently in operation:approximately 135 treating livestock/agricultural wastes, 850 formunicipal solid waste removal, and 544 in wastewater treatmentplants (Alternative and Advanced Fuels – Biogas, 2009). Whilethe current application of anaerobic digestion is significant, muchmore anaerobic digestion is possible for animal waste and pulpand paper waste, which represent the two largest fractions ofwaste biomass generated in the USA; waste biomass amounts aresummarized in Table 1.
Pig waste (PW) represents a significant fraction of animal wastes,the largest waste stream in Table 1. Likewise, paper sludge (PS),which is mainly residues from various stages of paper mill opera-tion, is the largest pulp and paper waste. While containing a large
ll rights reserved.
: +1 480 727 0889.n).
potential for energy recovery, both of these large waste streamspose unique challenges when subjected to anaerobic digestion.
Due to the high protein content in the diet of young pigs, PWhas a very high organic-nitrogen (N) content that is converted tototal ammonia during hydrolysis and fermentation. Inhibition ofmethanogenesis due to high concentrations of total ammonia is awell-established fact (Van Velsen, 1979; Cheung et al., 2002; Sossaet al., 2004; Sawayama et al., 2004; Kayhanian, 1993; PoggiVeraldo et al., 1997; Koster and Lettinga, 1984; Hansen et al.,1998). The major inhibition is caused by unionized ammonia(NH3) at a concentration of 150 mg N/L or higher, but the ammo-nium ion NHþ4 also exhibits toxicity at very high concentrations,5 g N/L or above (Braun et al., 1981). Overcoming inhibition has ahigh payback, because 30–40% of the total COD in PW is solubleand immediately bioavailable (Jindal et al., 2006), a value muchhigher than in stabilized biomass, such as waste activated sludge(Jindal et al., 2006). In addition, the high N content of pig wastegives it a high alkalinity, because the organic N is hydrolyzed toNH3, a moderately strong base (pKb = 3.7, Snoeyink and Jenkins,1980): 1 mol of bicarbonate alkalinity is released for every moleof ammonia released, and this corresponds to 3.6 mg as CaCO3
per mg N. Alkalinity is essential for stable pH control.
Table 1Summary of various waste-biomass-to-energy sources is the USA. The values indicated are the total amount of waste generated by each source and the total quantity that isalready utilized for energy recovery.
Biomass waste category Amount of waste produced/year(million US dry tons)
Waste quantity currently withenergy recovery (million US dry tons)
Reference
Animal wastes 335 3.6 a 0.07b Manure & By-product Utilization Program (2006)Pulp and paper 149 142 McKeever (2004) and Perlack et al. (2005)Food processing 113 6 Milbrandt (2005) and Perlack et al. (2005)Municipal wastewater 7 1.3 Water Environment Federation (2002)Total 604 153
a Amount of biomass estimated to be recovered as energy for heating from landfills with livestock wastes.b The amount of biomass recovered as biogas in anaerobic digesters treating animal wastes.
164 P. Parameswaran, B.E. Rittmann / Bioresource Technology 124 (2012) 163–168
PS is generated in large quantities during several stages of apaper mill operation, such as chipping, paper machine rejects,and packaging (Mahmood and Elliott, 2006). The organic matterin these waste streams is lignocellulosic and has a very low N con-tent, poor buffering capacity, and low soluble COD; hence, it hasproven difficult to sustain a diverse anaerobic microbial commu-nity for good methanogenesis with paper sludge (Banks andHumphreys, 1998). PS from paper mill rejects is also known tohave a significant fraction of inorganics making anaerobic diges-tion difficult.
A potential solution for successful anaerobic digestion of bothlarge, but challenging waste streams is to mix pig and paper-sludgewastes, since they have complementary properties for methano-genesis. Mixing the two streams dilutes the high N content of thePW, which lowers inhibition, and adds readily biodegradableCOD and alkalinity from the PW to help establish a stable anaero-bic methanogenic community that can efficiently degrade the par-ticulate matter present in the PS.
Previous efforts to co-digest PW with other organic wastestreams have had success: e.g., with municipal solid wastes(Campos et al., 1999), food and vegetable wastes (Alvarez andLidén, 2007), wastewater from olive-oil bleaching and filtering(Ahring et al., 1992), grass silage (Xie et al., 2012), wasted sardineoil (Ferreira et al., 2012) and crude glycerol (Astals et al., 2011).Zhan et al (2012) already demonstrated the benefits of successfulco-digestion of grass silage and pig manure at 1:1 ratio. Anaerobicdigestion of pig manure with pure cellulose as a co-substrate wasresearched by Van Assche et al. (1983), who demonstrated threetimes higher gas production than with pig waste alone. However,no studies have evaluated co-digestion of the combination of PSand PW, which are the two largest biomass waste sources in theUSA and should have especially complementary characteristics.
The first goal of this study was to test whether or notco-digestion of PW and PS provided advantages for methanogene-sis. This goal was achieved using batch biochemical methane po-tential (BMP) tests for various ratios of PW:PS compared againstPW or PS by itself. The second goal was to interpret the data fromthe batch BMP tests and using a first-order model to estimate thehydrolysis constant (khyd). khyd values were used to explain theimproved efficiency for the co-digestion over baseline PW or PSanaerobic digestion and to fit the experimental data to a wellestablished empirical model, namely Gompertz equation. The thirdgoal was to confirm the benefits during long-term anaerobic diges-tion with a workable ratio of PW to PS. The semi-continuous reac-tor operation along with a control semi-continuous reactortreating PW alone helped to achieve the last goal.
2. Methods
2.1. Biochemical methane potential (BMP) tests
BMP tests (Owen et al., 1979; Angelidaki et al., 2009; Salerno etal., 2009) were performed to compare CH4 production from PW
and PS alone and with different volume ratios of co-digestion mix-tures. PW slurry was obtained from Hormel foods, Snowflake, AZand PS from Abitibi pulp and paper mill, Snowflake, AZ. Threeratios of PW:PS by volume were evaluated: 1:3, 1:1, and 3:1. Aninoculum control was additionally prepared that consisted of theanaerobic inoculum with added trace minerals. The anaerobicinoculum was obtained from the anaerobic digesters at Mesa(AZ) Northwest Wastewater Reclamation Plant (MNWWRP), whichis fed with mixed primary + waste activated sludge that is centri-fuged at about 2000 rpm for 10 min.
For all BMP tests, 70 mL of sample and 30 mL of inoculum wereadded to 160-mL serum bottles. The tests were performed in dupli-cate and the average values are presented here. To make theco-digestion mixtures, appropriate volumes were added to makeup 70 mL: e.g., 17.5 mL of PW and 52.5 mL PS added to 30 mL ofanaerobic inoculum for the 1:3 PW:PS condition. Before sealingthe serum bottles with rubber stoppers and aluminum caps, theserum bottles were purged with 100% N2 gas to create anaerobicconditions. During the BMP test, the serum bottles were shakenat 180 rpm and incubated at 37 ± 1 �C.
Gas production was periodically measured in the headspaceusing a frictionless glass syringe (Perfektum, NY) after allowingthe syringe to equilibrate with atmospheric pressure. The CH4
composition was analyzed by gas chromatography (GC-2010,Shimadzu, Japan) equipped with a thermal conductivity detector(TCD) and a packed column (Shincarbon ST 100/120, 2 m, Restek,Bellefonte, PA). The GC-TCD was operated at a 145 �C iso-thermalcondition (inlet 120 �C; detector 150 �C; current 45 mA) duringgas-composition analysis. Helium was used as carrier gas. Standardcurves were prepared using certified CH4, CO2, and H2 mixed gas(40%:30%:30%, Matheson Tri-Gas, Twinsburg, Ohio).
Chemical Oxygen Demand (COD) – total and soluble (after fil-tration through a 0.45-lm membrane filter) – for the initial andfinal samples of the BMP tests were also measured to establishCOD mass balances. COD, total nitrogen (TN), and ammonia nitro-gen (NH3-N) were measured using a HACH COD kit (concentrationrange 10–1500 mg/L) and measuring the absorbance at 620, 410,and 655 nm, respectively, using a spectrophotometer. Total sus-pended solids (TSS) and volatile suspended solids (VSS) were mea-sured per Standard Methods (APHA, 1998). Volatile organic acidswere analyzed using a Shimadzu HPLC equipped with an AMINEXHPX-87H column at 50 �C with 2.5 mM H2SO4 at a flow rate of0.6 mL/min.
2.2. Semi-continuous anaerobic reactors – pig wastes and co-digestion
A1-L semi-continuous methanogenic bioreactor treating PWalone was operated along with a 3-L semi-continuous methano-genic bioreactor performing anaerobic co-digestion of PW and PSat a 2:1 volume ratio. The reactor was semi-continuous withrespect to feeding/wasting material, since the reactors were oper-ated with hydraulic retention time (HRT) = solids retention time(SRT) = 35 days. This SRT was achieved by removing a defined
(a)
200
400
600
800
1000
1200
Cum
ulat
ive C
H4
prod
ucti
on (m
L)
PW PS PW:PS 1:3 PW:PS 1:1 PW:PS 3:1
P. Parameswaran, B.E. Rittmann / Bioresource Technology 124 (2012) 163–168 165
volume of liquid (29 mL for the 1-L pig-waste reactor and 86 mLfor the 3-L co-digestion reactor), at which time we also added freshfeed of the same type.
Anaerobic digested sludge from the MNWWRP was used asinoculum at a 1:1 volume ratio with the wastes. pH controllerscoupled with in situ pH meters were used to maintain a pH rangeof 6.8–7.6. A water jacket was also used with a temperature con-troller to maintain 35 ± 2 �C. Gas production was quantified in acumulative fashion using a wet-tip gas meter (Rebel Point wetgas meter company, TN). This equipment is sensitive enough todetect flow rates as low as 100 mL/day, but still able to measureup to 500 mL/min.
(b)
00 20 40 60 80 100
Duration of batch BMP tests (days)
0.1
0.2
0.3
0.4
0.5
0.6
CH
4-C
OD
nor
mal
ized
to fe
ed C
OD
PW PS PW:PS 1:3 PW:PS 1:1 PW:PS 3:1
3. Results and discussion
3.1. Characterization of waste streams
Characteristics of the PW and PS are summarized in Table 2.Very low and high availabilities of soluble nitrogen are evident inPS and PW, respectively. This leads to no alkalinity for PS, but sig-nificant alkalinity for PW. Soluble COD, one indicator of bioavail-able material for the anaerobic food web, is small in PScompared to PW. On the other hand, total solids and COD valuesare higher for PS, indicating the potential for electron capture asmethane. The values in the co-digestion mixture column indicatethe potential advantages of blending the complementary streamsfor anaerobic digestion, as no values are at high or low extremes.
00 20 40 60 80 100
Duration of batch BMP tests (days)
Fig. 1. (a) Methane production from co-digestion experiments compared withcontrol experiments with PW or PS. The contribution from the inoculum controlwas too small to be plotted in the figure. (b) Normalized CH4-COD to the respectivefeed COD values for the BMP tests with different volumetric ratios of PW to PS.
3.2. Batch BMP tests
The results of the BMP tests for the different ratios of PW and PSare shown in Fig. 1a (CH4 production) and Fig. 1b (CH4 conversionnormalized to the feed total COD). At the end of 35 days, all threeco-digestion ratios reported higher CH4 production and normalizedTCOD conversion efficiency than either PW or PS control. This dem-onstrates benefits of co-digestion. Until about 60 days, the 1:1 PW/PS mixture had the greatest volumetric methane production. How-ever, CH4-COD removal normalized to the feed COD showed thatthe 1:1 and 3:1 ratios of PW:PS were similar at 60 days. At theend of 80 days, the normalized COD conversion efficiency washighest for the 3:1 ratio of PW and PS, even though PW led allthe samples in terms of volumetric CH4 production, since it hasthe highest total-COD concentration.
Table 3 summarizes the fate of COD at the end of the BMP tests.Initial total COD input was the highest for PW and lowest for PS. Atthe end of 80 days, PW:PS at a 3:1 ratio produced the highest TCODremoval efficiency of 49%, followed by PW at 44%. PS produced thelowest TCOD removal of 11%, indicating the limited availability ofits TCOD for methane production.Based on the BMP results in Fig. 1and Table 3, the following insights were gained:
Table 2Summary of feed characteristics for pig waste (PW) and paper sludge (PS) andestimated values for a 1:1 co-digestion mixture.
Paper sludge Pig waste 1:1 Mixture forco-digestion
Parameter (units) Value Value ValueTotal solids (g/L) 56 36 46Total COD (g/L) 68 53 60Soluble COD (g/L) 4.2 20 12SCOD/TCOD 0.06 0.39 0.20Total nitrogen (mg/L) 180 1350 760Soluble nitrogen (mg N/L) 16 860 440NHþ4 (mg N/L) 0 420 210
Alkalinity from NHþ4 0 1500 750
(i) Co-digestion of PW and PS had a strong mutual benefit onthe COD conversion efficiency of either waste stream. While thiseffect was most remarkable by comparison with PS by itself (11%versus any of three ratios of PW:PS shown in Table 2), co-digestionwith a lower volume ratio of PS (i.e., 1:1 and 3:1 PW:PS) increasedthe COD conversion efficiency of PW as well.
(ii) Co-digestion significantly decreased the lag time for achiev-ing rapid methane production. 1:1 PW:PS reported the shortest lagtime of 14 days (Fig. 1a), while PW alone reported the longest lagtime of 23 days. PS never entered rapid phase of methane produc-tion. All three ratios of PW and PS tested in the study had shorterlag times than either waste stream tested individually.
(iii) Higher SCOD in the initial feed mix correlated with higherCOD conversion to CH4 at the end of the test. This is supportedmost strongly by the PS control (lowest input SCOD), whichreported the lowest COD conversion efficiency to CH4 (11%).
(iv) Lower SCOD and NH4-N during start-up of the co-digestionstudies correlated with the shortest lag times for methanogenesis(Table 3). The effect of NH4 was probably two-fold: increasingthe lag time for CH4 production and reducing overall methanogenicactivity. In further support of the NH3 effect, the saturation regionof the methane-production curve for the PW control, which hadthe highest total and soluble COD input, was lower than or equalto other co-digestion mixtures in Fig. 1a.
3.3. Quantification of hydrolysis rate constants (khyd) and hydrolysisefficiencies with BMP data
Hydrolysis of particulate COD is often the rate-limiting step foranaerobic digestion and is included in anaerobic digestion models
Table 3Summary of COD values for the different batch BMP-test conditions. PW alone was quantified at the end of 97 days, as the CH4-production rate was still high (>100 mL/day ofbiogas) at the end of 80 days, when the other conditions were quantified.
Condition COD initial (mg) COD final (mg) COD to CH4 (mg)a Soluble COD final (mg) (% of feed TCOD) % TCOD removal as CH4
PS 3920 3540 326 27 (2.9%) 11PW:PS 1:3 3990 1980 1710 97 (6.3%) 43PW:PS 1:1 4910 2600 2310 144 (7.7%) 47PW:PS 3:1 4650 2160 2280 189 (8.7%) 49PW 6100 3330 2690 264 (7.9%) 44
ND – not detected.a COD equivalent of CH4 gas calculated by the following relationship: 1 mL CH4gas ¼ L
103 ml: 1 molCH4
22:4 L : 273 K313 K
8e�eqmolCH4
: 8CODe�eq :
103 mgg ¼ 2:52 mgCOD.
PW y = -0.0225x
PS y = -0.0042x
PW: PS 1:3 y = -0.0496x
PW: PS 1:1 y = -0.0323x
PW: PS 3:1 y = -0.0242x
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
0 10 20 30 40 50 60 70 80
ln(1
-{Y
/Ym
ax})
Duration of test (days)
Fig. 2. Linear regression fits for estimation of hydrolysis rate constants (khyd) fromthe BMP assay data for the various samples.
Table 4Summary of hydrolysis constants, lag times, and time taken to achieve 90% of themaximum cumulative methane production for the various waste combinations testedby BMP tests.
Sampletype
Hydrolysisconstant(khyd, day�1)
Lag phase(days)
t90% CH4
(days)
PW 0.02 23 75PS 0.004 No exponential gas
production65
PW:PS 1:3 0.05 15 41PW:PS 1:1 0.03 17 60PW:PS 3:1 0.024 20 65
166 P. Parameswaran, B.E. Rittmann / Bioresource Technology 124 (2012) 163–168
(Batstone et al., 2002; Rittmann and McCarty, 2001). The positiveimpact of co-digestion is evident from the measurement of theultimate output of the anaerobic food web, methane. However,BMP tests do not yield direct experimental data to measure thehydrolysis rate. By assuming that hydrolysis was the rate-limitingstep, researchers have modeled batch BMP data using a first-orderhydrolysis model and obtained valuable interpretations abouthydrolysis kinetics (Bolzonella et al., 2005; Pavlostathis andGiraldo-Gomez, 1991). The first-order rate equation in terms ofmethane production can be expressed as:
Y ¼ Ymax½1� expð�khydtÞ� ð1Þ
where Y = cumulative methane production from the BMP assay attime t (mL). Ymax = ultimate methane yield from BMP assay at theend of the incubation time (mL). khyd = first-order hydrolysis rateconstant (day�1)
Eq. (1) provides an accurate representation of the BMP resultswhen (a) hydrolysis is the rate-limiting step, and (b) the maximummethane production at the end of the batch tests (Ymax) representsthe total concentration of hydrolyzable COD at the beginning of thetests.Rearranging Eq. (1) yields
�khydt ¼ ln½1� fY=Ymaxg� ð2Þ
which is an equation for a straight line with a slope whose magni-tude is the hydrolysis rate constant (khyd). The slope was obtainedby plotting ln½1� fY=Ymaxg� versus time and performing linearregression.
Fig. 2 shows the plot of Eq. (2) for all scenarios of PW and PScombinations evaluated with the BMP assays. The slopes of thelinear regression lines correspond to hydrolysis constants, andTable 4 summarizes the khyd values determined from the slopes.For comparison, hydrolysis constants reported from previous re-search include 0.22 day�1 for waste activated sludge, 0.06 day�1
for primary + waste activated sludge, and 0.07 day�1 for newsprint(Rittmann and McCarty, 2001; Lee et al, 2011). The hydrolysis con-stant for PW (0.03 day�1) was an order of magnitude higher thanPS (0.003 day�1). However, the khyd values for PW:PS 1:3(0.06 day�1) and PW:PS 1:1 (0.04 day�1) were 1.5 to 2 times higherthan PW by itself. An interesting trend is the increase in hydrolysisconstant as the fraction of PS increased, even though PS alone hadthe lowest hydrolysis constant. This suggests that co-digestion sig-nificantly increased the bioavailability of the organic content in pa-per sludge, since PS likely dominated the COD in the 1:3 mixture,since a 1:1 mixture of PW:PS has a higher COD than baseline PW(Table 2)
Other significant parameters that could be gleaned from thebatch BMP assays are the lag time and the time taken to achieve90% of the maximum methane production; these also are tabulatedin Table 4 based on the results in Fig. 1a. Lag times followed a trendsimilar to the hydrolysis constants. In particular, co-digestion mix-tures exhibited decreased lag time as the PS fraction in the mixincreased. The time taken to achieve 90% of maximum cumulativemethane production is another factor characterizing the bioavail-
ability of the organic matter. Similar to the lag time, the time to90% production decreased steadily for mixture that containedmore PS, even though PS itself did not have a low time to 90%production.
3.4. Gompertz-equation fit to the batch BMP data
Empirical equations often are used to interpret basic mecha-nisms underlying a complex system and to provide practical designand control strategies. The Gompertz equation (Lay et al., 1996) isone such expression often used to fit batch methanogenic data:
MP ¼ PM exp � expfRM=PMðxo � xÞeþ 1g½ � ð3Þ
where MP = observed cumulative methane production (mL).PM = ultimate methane production (mL). RM = observed methaneproduction rate (mL/day). xo = lag phase time (days). x = time ofobservations (days). e = exponential (2.718).
All the parameters in the Gompertz equation were evaluatedby performing regression with a Newtonian algorithm (Wen
0
100
200
300
400
500
600
700
800
900
1000
0 20 40 60 80 100
Cum
ulat
ive
CH
4pr
oduc
tion
(mL
)
Duration of batch BMP tests (days)
PS PW:PS 1:3 PW:PS 1:1PW:PS 3:1 Model PS Model PMModel PM:PS 1:3 Model PM:PS 1:1 Model PM:PS 3:1
Fig. 3. Gompertz equation to the batch BMP data for the various co-digestionscenarios under study.
Table 5Estimated parameters from the fit of Gompertz equation to the experimental BMPdata.
Sample type Pm (mL CH4) Rm (mL CH4/day) xo (days) R2
PW 1100 15.3 13 0.99PS 139 5.18 ND 0.98PW:PS 1:3 604 20 7.5 0.99PW:PS 1:1 865 24 12 0.98PW:PS 3:1 902 21 15 0.99
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 20 40 60 80 100 120 140 160 180 200
Met
hane
pro
duct
ion
rate
(L
CH
4/ L
reac
tor
d)
Duration of test (days)
PW:PS PW alone
1 HRT = 35 days
Fig. 4. Methane production rates of 3-L continuous co-digestion reactors (circles)compared with the baseline 1-L reactor treating pig wastes (diamonds). Methaneproduction is normalized to 1-L for direct comparison. Methane in the headspacewas 60% v/v in both cases. The much higher methane production by theco-digestion reactor is evident.
Table 6Summary of process metrics for co-digestion and baseline PW reactors at a steady-state operation with respect to methane production rate. Both semi-continuousdigesters had an HRT of 35 days and input TCOD of around 35 and 50 g/L for thebaseline PW and co-digestion reactors, respectively.
Parameter Co-digestion PW:PS 2:1(ratio by volume)
Baselinepig waste
Volatile solids removal (%) 32 25TCOD removal (%) 48 21CH4–COD conversion (%) 42 25Total nitrogen effluent (mg N/L) 2220 3330NH3-N effluent (mg NH3-N/L) 1450 930Effluent soluble COD (mg/L) 1820 4100
P. Parameswaran, B.E. Rittmann / Bioresource Technology 124 (2012) 163–168 167
et al., 1994) to minimize the sum of squares of errors (SSE)between the experiment and model estimation. Microsoft Excelspreadsheet was used for this purpose.
The model fits to the experimental data, plotted in Fig. 3, indi-cate that the model represented the BMP data well for each exper-iment. The Gompertz parameters, shown in Table 5, show the sametrends as the experimentally observed values in Table 4, In partic-ular, the highest RM values in Table 5 and khyd values in Table 4 arefor the PW:PS 1:3 and 1:1 mixtures. Likewise, the shortest lag-phase times are for the same two mixtures.
3.5. Semi-continuous reactors – co-digestion and baseline PW systems
The 1-L semi-continuous reactor treating PW (1 HRT = 35 days)and the 3-L 2:1 PW:PS co-digestion reactors (also at 35 day HRT)were operated for approximately 178 days. The availability of feed-stock from the suppliers dictated the choice of a 2:1 ratio of PW:PS.The 2:1 ratio struck a balance between the 1:1 and 3:1 PW:PSratios, which gave the best performance in the BMP tests.
In terms of CH4 production, shown in Fig. 4, the performance ofthe co-digestion reactor improved steadily after 40 days and ap-proached a steady-state CH4-production rate of 0.25 L CH4/L reac-tor-day. The digester treating PW alone stabilized at a lowerproduction rate, �0.14 L CH4/L-day in the same timeframe. Meth-ane in the headspace was about 60% v/v in both cases.
Table 6 presents performance metrics observed when bothbaseline and PW:PS 2:1 co-digestion reactors were at steady state.The feed TCOD values for the co-digestion reactor were higher thanfor the baseline PW until about 40 days, after which the feed con-centrations for both reactors were similar. Regardless of the varia-tions of the feed TCOD concentrations, the effluent TCOD values forPW:PS 2:1 was significantly reduced compared to the correspond-ing feed, while PW reported only modest reductions in the effluentTCOD. This is further reflected in the average TCOD removal for the
PW (21%) compared to PW:PS 2:1 co-digestion mix (48%). The co-digestion reactor performed better in terms of COD and VSSremoval efficiencies. This is consistent with the roughly two timesgreater methane conversion (Fig. 4) in the co-digestion reactor,despite a lower effluent soluble COD. Thus, the concept thatco-digestion should convert more COD to methane was confirmedin the semi-continuous experiments.
The soluble COD concentrations from the reactor effluent aver-aged around 4000 mg/L for the PW reactor and 1800 mg/L of theco-digestion reactor. HPLC analysis indicated that acetate concen-trations were around 400 mg/L in both reactors, and no other peakswere detected at significant concentrations. Thus, it is likely that asignificant fraction of the effluent soluble COD was recalcitrantunder the operating conditions. This could possibly include humicacids that gave rise to the brown color of the reactor influent andeffluent filtrates. The lower soluble COD suggests that co-digestionresulted in a substantial decrease in recalcitrant soluble COD.
4. Conclusions
Co-digestion of PW and PS provided significant benefits formethane production and COD removal, compared to either PW orPS by itself. Batch BMP tests revealed a significant decrease in lagtimes for methane production in the order of PW:PS 1:3 (14 days) <PW:PS 1:1 (17 days) < PW:PS 3:1 (20 days) < PW (23 days), whichwas confirmed by the Gompertz-parameter estimation. Hydrolysisconstants for the co-digestion scenarios were 2–20 times higherthan baseline. Semi-continuous digester operation with a PW:PS
168 P. Parameswaran, B.E. Rittmann / Bioresource Technology 124 (2012) 163–168
2:1 co-digestion mixture produced 0:25 L CH4 L�1reactord
�1 comparedto a baseline PW-only digester that produced 0:14 L CH4 L�1
reactord�1.
References
Ahring, B., Angelidaki, I., Johansen, K., 1992. Anaerobic treatment of manuretogether with industrial waste. Water Science and Technology 25 (7), 311–318.
Alternative and Advanced Fuels – Biogas, 2009. U.S. Department of Energy-EnergyEfficiency and Renewable Energy. U.S. DOE, June 23.
Alvarez, R., Lidén, G., 2007. Semi-continuous co-digestion of solid slaughterhousewaste, manure and fruit and vegetable waste. Renewable Energy 33, 726–734.
Angelidaki, I., Alves, D., Bolonzella, L., Borzacconi, J.L., Campos, A.J., Guwy, S.,Kalyunzhnyi, P., Jenicek, P., Van Lier, J.B., 2009. Defining the biochemicalmethane potential (BMP) of solid organic wastes and energy crops: a proposedprotocol for batch assays. Water Science and Technology 59, 927–934.
American Public Health Association (APHA), 1998. Standard Methods for theExamination of Water and Wastewater, 19th ed. APHA-AWWA-WEF,Washington DC, USA..
Astals, S., Ariso, M., Gali, A., Mata-Alvarez, J., 2011. Co-digestion of pig manure andglycerine: experimental and modelling study. Journal of EnvironmentalManagement 92 (4), 1091–1096.
Banks, C.J., Humphreys, P.N., 1998. The anaerobic treatment of a lignocellulosicsubstrate offering little natural pH buffering capacity. Water Science andTechnology 38 (4–5), 29–35.
Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., Pavlostathis, S.G., Rozzi, A.,Sanders, W.T.M., Siegrist, H., Vavilin, V.A., 2002. Anaerobic Digestion Model No.1 (ADM1). Water Science and Technology 45 (10), 65–73.
Bolzonella, D., Fatone, F., Pavan, P., Secchi, F., 2005. Anaerobic fermentation oforganic municipal solid wastes for the production of soluble organiccompounds. Industrial and Engineering Chemistry Research 44, 3412–3418.
Braun, R., Huber, P., Meyrath, J., 1981. Ammonia toxicity in liquid piggery manuredigestion. Biotechnology Letters 3 (4), 159–164.
Campos, E., Palatsi, J., Flotats, X., 1999. Co-digestion of pig slurry and organic wastesfrom food industry. In: Proceedings of the Second International Symposium onAnaerobic Digestion of Solid Wastes, Barcelona, Spain, 192–195.
Cheung, W.S.P., Zhou, H., Hacker, R., 2002. Inhibitory effects on anaerobic digestionof swine manure. In: CSCS/EWRI of ASCE Environmental EngineeringConference, Niagara, NY, pp. 73–77.
Ferreira, L., Duarte, E., Figueiredo, D., 2012. Utilization of wasted sardine oil as co-substrate with pig slurry for biogas production – A pilot experience ofdecentralized industrial organic waste management in a Portuguese pig farm.Bioresource Technology 116, 285–289.
Hansen, K.H., Angelidaki, I., Ahring, B.K., 1998. Anaerobic digestion of swinemanure: Inhibition by ammonia. Water Research 32 (1), 5–12.
Jindal, S., Kocherginskaya, A., Mehboob, M., Robert, M., Mackie, R.I., Raskin, L., Zilles,J.L., 2006. Antimicrobial use and resistance in swine waste treatment systems.Applied and Environmental Microbiology 72 (12), 7813–7820.
Koster, I.W., Lettinga, G., 1984. Anaerobic digestion at extreme ammoniaconcentrations. Biological Wastes 25, 51–59.
Lay, J.J., Li, Y.Y., Noike, T., 1996. Effect of moisture content and chemical nature onmethane fermentation characteristics of municipal solid wastes. Journal ofEnvironment System and Engineering, 101–108.
Mahmood, T., Elliott, A., 2006. A review of secondary sludge reduction technologiesfor the pulp and paper industry. Water Research 40, 2093–2112.
McKeever, D.B., 2004. Inventories of Woody Residues and Solid Wood Waste in theUnited States, 2002 (Forest Products Laboratory, USDA Forest Service, Madison,WI).
Milbrandt, A., 2005. A Geographic Perspective on the Current Biomass Resourceavailability in the United States. (NREL/TP-560-39181, National RenewableEnergy Laboratory, Golden, CO).
Owen, W.F., Stuckey, D.C., Healy, J.B., Young, L.Y., McCarty, P.L., 1979. Bioassay formonitoring biochemical methane potential and anaerobic toxicity. WaterResearch 13, 485–492.
Pavlostathis, S., Giraldo-Gomez, E., 1991. Kinetics of anaerobic treatment: a criticalreview. Critical Reviews in Environmental Control 21 (5–6), 411–490.
Perlack, R.L. et al., 2005. Biomass as Feedstock for a Bioenergy and BioproductsIndustry: The Technical Feasibility of a Billion-Ton Annual Supply (ORNL/TM-2005/66, Oak Ridge National Laboratory, Oak Ridge, TN).
Poggi Veraldo, H.M., Rodriguez-Vasquez, R., Villagomez, A., Garcia, E., 1997.Inhibition of mesophilic solids substrate anaerobic digestion by ammonianitrogen. Applied and Environmental Microbiology 71 (12), 8191–8200.
Rittmann, B.E., McCarty, P.L., 2001. Environmental Biotechnology: Principles andApplications. McGraw-Hill, NY (Chapter 13).
Salerno, M.B., Lee, H.S., Parameswaran, P., Rittmann, B.E., 2009. Using a pulsedelectric field as a pretreatment for improved biosolids digestion andmethanogenesis. Water Environment Research 81 (8), 831–839.
Sawayama, S., Tada, C., Tsukahara, K., Yagishita, T., 2004. Effect of ammoniumaddition on methanogenic community in a fluidized bed anaerobic digestion.Journal of Bioscience and Bioengineering 97 (1), 65–70.
Snoeyink, V.L., Jenkins, D., 1980. Water Chemistry. Wiley, NY.Sossa, K., Alarcón, M., Aspé, E., Urrutia, H., 2004. Effect of ammonia on the
methanogenic activity of methylaminotrophic methane producing Archaeaenriched biofilm. Anaerobe 10 (1), 13–18.
Van Assche, P., Poels, J., Verstraete, W., 1983. Anaerobic co-digestion of pig manurewith cellulose as co-substrate. Biotechnology Letters 5 (11), 749–754.
Van Velsen, A.F.M., 1979. Adaptation of methanogenic sludge to ammonia toxicity.Water Research 13, 995–999.
Water Environment Federation, 2002. A Guide to Understanding Biosolids Issues.Wen, T.C., Cheng, S.S., Lay, J.J., 1994. A kinetic model of a recirculated upflow
anaerobic sludge blanket treating phenolic wastewater. Water EnvironmentResearch 66, 794–799.
Xie, S., Lawlor, P.G., Frost, J.P., Hu, Z., Zhan, X., 2012. Effect of pig manure to grasssilage ratio on methane production in batch anaerobic co-digestion ofconcentrated pig manure and grass silage. Bioresource Technology 102,5728–5733.
