Start-up of dry anaerobic digestion system for processing solid poultry litter using adapted liquid inoculum

Start-up of dry anaerobic digestion system for processing solid poultry litter using adapted liquid inoculum

Accepted Manuscript Title: Start-up of dry anaerobic digestion system for processing solid poultry litter using adapted liquid inoculum Author: Rajini...

475KB Sizes 0 Downloads 16 Views

Accepted Manuscript Title: Start-up of dry anaerobic digestion system for processing solid poultry litter using adapted liquid inoculum Author: Rajinikanth Rajagopal Daniel I Mass´e PII: DOI: Reference:

S0957-5820(16)30050-7 http://dx.doi.org/doi:10.1016/j.psep.2016.05.003 PSEP 765

To appear in:

Process Safety and Environment Protection

Received date: Revised date: Accepted date:

14-1-2016 3-5-2016 5-5-2016

Please cite this article as: Rajagopal, R.,Start-up of dry anaerobic digestion system for processing solid poultry litter using adapted liquid inoculum, Process Safety and Environment Protection (2016), http://dx.doi.org/10.1016/j.psep.2016.05.003 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ip t cr

Highlights

Treatment of poultry litter (PL) at higher solid content of above 15% is limited

-

Low-temperature treatment of solid PL with high nitrogen has been scarcely reported

-

Tested the adapted liquid inoculum to start the solid dry anaerobic digester (DAD)

-

Higher total kjeldahl nitrogen concentrations (>30 g/L) in PL inhibited the DAD

Ac

ce pt

ed

M an

us

-

23 Page 1 of 28

ip t

cr

Start-up of dry anaerobic digestion system for processing solid poultry litter using adapted

Rajinikanth Rajagopal, Daniel I Massé*

M an

us

liquid inoculum

Corresponding author. Tel.: +1 819 780 7128; Fax: +1 819 564 5507

Ac



ce pt

Sherbrooke, Quebec, J1M 0C8

ed

Dairy and Swine Research and Development Center, Agriculture and Agri-Food Canada

E-mail addresses: [email protected] (D.I. Massé) ; [email protected] (R.Rajagopal)

Abstract

24 Page 2 of 28

ip t

cr

The objective is to obtain the basic design criteria for starting up of dry anaerobic-digestion (DAD) systems treating solid-poultry-litter (PL) with hay-

us

bedding (TSmixture: 68.6%) using adapted liquid inoculum. Effect of organic loading rates (OLR) and mode of operation (particularly liquid-inoculum recirculation-percolation mode) were evaluated in two phases; such that OLR of 5.4 and 21.6 gVS/kginoculumVS/d were maintained respectively for Phase-

M an

1 and-2; and top-down and down-up mode of liquid-inoculum recirculation into DAD-system were experimented. Digesters were operated at psychrophilic-temperature (@20oC) with cycle length of 26 and 38 d for Phase-1 and-2, respectively. Results show that specific methane yield of 0.147 0.162 L/gVSfed was obtained for Phase-1 with a methane content of 35-39%; whereas Phase-2 had 61-70% fewer yields compared to Phase-1. Though PLdigestion was possible at OLR <5.4 g VS/kginoculumVS/d, high nitrogen-content in PL inhibited the digestion process especially at higher OLR. However, adapted inoculum to TKN of >20g/L could minimize the inhibition. Top-down-recirculation is recommended for simpler operation.

ed

Key words: Ammonia; Poultry litter; Dry anaerobic digestion; Liquid inoculum; Percolation

ce pt

Introduction

Improperly managed poultry wastes can cause drastic damage to the environment, which can harmfully influence human and animal health. On the other hand, poultry litter (PL) is a potential organic substrate for biogas production through anaerobic digestion (AD),

Ac

which has not been fully utilized so far, due to the key issues linked with inhibition caused by high concentration of ammonia in chicken litter (Abouelenien, et al., 2014). In 2006, Canadian livestock produced over 180 million tons of manure over the year; of this total about 5 million tons produced by poultry (Statistics Canada, 2008). PL contains 20% or more dry matter, and the anaerobic decomposition of uric acid and undigested proteins in PL results in the production of high amounts of unionized ammonia and ammonium ions (Bujoczek et al., 2000; Abouelenien et al., 2009). In addition, PL is a poor substrate, due to presence of high total kjeldahl nitrogen (TKN), which led to imbalanced carbon to nitrogen ratio. The optimal carbon to nitrogen (C/N) ratio of 15-30 is 25 Page 3 of 28

ip t

cr

preferred for AD and hence external supplementation of carbon has to be regularly performed to dilute TKN concentration, in order to

us

achieve a stable and efficient process (Babaee et al., 2013). It is to be noted that dilution can be done by adding water; however, it is not practically feasible due to the huge amount of dilution necessities.

M an

Anaerobic digesters can be operated under liquid (wet), semi-solid, or solid-state (dry) conditions, when the total solids (TS) of substrate are <10%, 10–15%, or >15%, respectively (Li et al., 2011). Largely, liquid AD is frequently applied in full scale operations, owing to reasons such as easy operation and maintenance, and increased methane yield (Fdez-Guelfo et al., 2010). Nevertheless, liquid AD is not suitable for high solid content wastes, for instance solids fraction of animal manure with bedding materials, due to the higher

ed

dilution requirements and hence large volume of digesters required to treat certain quantity of feed material. On the other hand, solid

ce pt

state digestion enables a higher volumetric organic loading rate, small reactor volume, lower energy requirements for heating, positive energy balance and limited environmental consequences such as it has no impact on phosphorous (Jha et al., 2011). However, solid state digesters need longer retention time and huge amount of inoculum over wet digestion systems and often resulted in poor start-up performance due to incomplete mixing and accumulation of volatile fatty acids (VFAs) (Jha et al., 2011; Li et al., 2011). Some

Ac

researchers have reported notable effects of temperature on the microbial community, process kinetics and stability, and methane yield. Lower temperatures during the AD process are known to decline the microbial growth, specific substrate utilization rates, and methane production (Babaee et al., 2013). On contrary, digesters operating at high temperatures resulted in lesser biogas yield due to the production of volatile gases such as ammonia which suppresses methanogenic activities (Khalid et al., 2011).

26 Page 4 of 28

ip t

cr

Co-digestion of manure with energy crops/crop residues (like straw or hay) or other carbonaceous wastes can increase the biogas

us

yield. In addition, it helps to maintain an optimal pH for methane producing bacteria, decreases free ammonia/ammonium inhibition, which may occur in AD of PL alone, and provides a better C/N ratio in the feedstock (Xie et al., 2011). Few researchers have explored

M an

the co-digestion of poultry manure with other organic wastes, as demonstrated in Table 1. However, (i) the study demonstrating the feasibility of treating chicken litter (with hay bedding) at higher TS content of above 15% is limited; (ii) low-temperature treatment of these co-substrates has not been reported (Table 1); (iii) in addition, starting up of solid dry anaerobic digester (DAD) using liquid inoculum from adapted inoculum storage has not been extensively studied, as huge amount of dry inoculum is hard to find for starting

ed

the digesters. This concept is expected to enrich the methane production and hence the performance in an economical way. Premixing

ce pt

of substrate with inoculum would provide a better waste-microbe interaction at the first hand. Liquid inoculum percolationrecirculation is expected to eliminate the need for mixing equipment within the digesters and thus, would enhance the waste-microbes interactions in the DAD reactor during the digestion process. Presence of hay or straw as a structural material in the waste matrix should improve the percolation rate.

Ac

Thus the objective of the present study was to determine the operational feasibility of solid dry anaerobic digester (DAD) treating PL with hay bedding using adapted liquid inoculum as a seeding source. Liquid inoculum recirculation-percolation method of operation was carried out at psychrophilic temperature (20 ±0.5oC). A separate liquid inoculum reservoir (reactor) was used to provide the aforementioned adapted inoculum (biomass), so that it can be used to speed up the digestion of DAD process and improve the energy production and biomass quantity compared to the static DAD operation without inoculum. The liquid inoculum reservoir also 27 Page 5 of 28

ip t

cr

produces energy from VFAs that are washed out of the DAD reactors by the inoculum recirculation. The transfer of VFAs from the

us

DAD reactor to the inoculum reservoir is expected to reduce the risk of VFAs accumulation and inhibition in the DAD reactors.

M an

Special attention was given to the mode of operation and its procedures to operate the reactors with minimal inhibition.

Methods and Materials Feedstock and Inoculum

ed

Fresh chicken litter was collected from a local commercial poultry operation located near Sherbrooke (Quebec), Canada. Collected PL contained hay bedding. For characterisation purpose, the manure samples (i.e. mixture of PL + hay) was then grinded to prepare

ce pt

homogenize feed samples; however, for experimentation purpose, raw manure with hay material (without grinding) was used as a feedstock and stored in a cold room at 4oC to prevent biological activity until needed for feeding. The liquid inoculum was obtained from a laboratory-scale wet-digestion bioreactor located at the Dairy and Swine Research

Ac

Development Centre (DSRDC), which was processing pig manure + chicken pellets at high ammonia concentrations (7-8 g TAN/L). The inoculum was diluted with water to bring down the TS content from 9% to 5%, so that it could percolate more easily into the void spaces of the porous solid substrate matrix of the DAD reactor. Liquid inoculum at 5% TS was stored in a gas tight reservoir (reactor), prior its intermittent recirculation into the solids DAD rector whenever needed.

Experimental Set-up and Procedures 28 Page 6 of 28

ip t

cr

The operational feasibility of solid DAD treating PL +hay using liquid inoculum recirculation-percolation method of operation were

us

experimented in two phases. Fig. 1 presents the schematic representation of the DAD + liquid recirculation-percolation system. Phase 1 of the study consisted of two stage AD system: DAD reactor (BR1) +inoculum reservoir (BR2) [20-L active inoculum

M an

volume]. Two sets of digesters (in replicates) namely BR1a+BR2a and BR1b+BR2b were operated in parallel and were installed at a controlled-temperature room, which was maintained at 20±0.5o C. The solid substrate was fed in a batch feeding mode and a cycle length of 26 days was maintained. No external mixing was applied. About 4- L of inoculum from the inoculum reservoir was pumped and sprinkled at the top of the DAD system; such that it percolates in to the solid matrix of the DAD system. The inoculum leaching at

ed

the bottom of the DAD was collected every day and returned back to the inoculum reservoir for further digestion. By doing this,

ce pt

inoculum is expected to adapt with time to the new substrate type and also contribute to energy production of the inoculum reservoir. Daily biogas production was measured using wet tip gas meters. Phase 2 of the study also consisted of two stage AD system: DAD reactor + inoculum reservoir [20-L active inoculum volume]. The solid substrate was fed in a batch feeding mode and a cycle length of 38 days was maintained for this phase of study. The main

Ac

difference compared to Phase 1 was that two modes of recirculation-percolation was performed, such that (i) One set of reactors (DAD reactor (BR3) + inoculum reservoir (BR4)) was operated from top-down mode of liquid inoculum recirculation in to the DAD system. About 10-L of inoculum from the inoculum reservoir was pumped and sprinkled in the DAD, such that liquid inoculum percolates into the solid matrix in the DAD. The percolated inoculum was retained in the BR3 (DAD) for a night. The leachate at the bottom of the DAD was collected and returned to the inoculum reservoir in the day time. This process is likely to acclimatise the inoculum to the dry 29 Page 7 of 28

ip t

cr

PL with time and also contribute to the energy production in the inoculum reservoir. This operation was repeated every day; (ii)

us

another set of reactors (DAD reactor (BR5) + inoculum reservoir (BR6)) was operated from down-up mode of liquid recirculation in to the DAD system. About 10-L of inoculum from the inoculum reservoir were pumped in to the bottom of DAD system, such that liquid

M an

inoculum percolates in to the solid matrix in the DAD. The percolated inoculum was retained in the BR5 (DAD) for a night. The leachate at the bottom of the DAD was collected and returned to the inoculum reservoir in the day time. This operation was repeated every day.

ed

Analytical Methods

A mixed liquor samples of about 100 mL capacity was taken on weekly basis from the liquid inoculum reservoir; whereas samples

ce pt

from DAD system were collected at the beginning and at the end of each treatment cycle. These samples were analysed for total COD (TCOD), total solids (TS), volatile solids (VS), volatile fatty acids (VFAs) such as acetic, propionic, butyric, iso-butyric, valeric, isovaleric and caproic, pH, TKN and ammonia nitrogen.

Ac

TCOD were determined by the closed reflux colorimetric method (APHA, 1992). VFAs concentration was determined using a Perkin Elmer gas chromatograph model 8310 (Perkin Elmer, Waltham, MA), mounted with a DB-FFAP high resolution column. Before VFAs quantification, samples were conditioned according to the procedures described by Massé et al. [2011]. TS and VS were determined using standard methods (APHA, 1992). pH value was measured using PH meter (model, TIM840, France). TKN and NH4– N were analyzed using a Kjeltec auto-analyzer model TECATOR 1030 (Tecator AB, Hoganas, Sweden) according to the macro30 Page 8 of 28

ip t

cr

Kjeldahl method (APHA, 1992). Daily biogas production was measured using wet tip gas meters. Every week, biogas composition

us

(methane, carbon dioxide, and nitrogen) was determined with a HachCarle 400 AGC gas chromatograph (Hach, Loveland, CO). The GC column and thermal conductivity detector were operated at 80oC. The nitrogen content was subtracted from the biogas results,

Results and Discussion Characteristics of the feed material and inoculum

M an

because N2 gas was used as a filler gas in the digester during drawdown.

ed

Table 2 presents the characteristics of poultry litter (PL) with hay bedding and the liquid inoculum used. Although the PL+hay mixture

ce pt

contained higher TKN value, the initial COD/TKN ratio was in the range of 26.2, which is in the optimal C/N range of 15-30. TS content of the inoculum was maintained at around 5%, primarily to enhance the percolation of liquid seed in to the void spaces of the porous solid substrate matrix of the DAD reactor. The inoculum was already acclimatised to the higher ammonia concentration wastes

Ac

such as pig manure+chicken pellets (7-8 g TAN/L).

Performance of Reactors (Phase 1) Two sets of bioreactors (in duplicates) BR1a+BR2a and BR1b+BR2b were operated in parallel under similar operating conditions as presented in Table 3, in which BR1a and BR1b represents substrate DAD systems and BR2a and BR2b represents inoculum reservoirs.

31 Page 9 of 28

ip t

cr

The results obtained for methane production and yield is also given in Table 3. An illustration of the profile for the cumulative biogas

us

and methane production, specific methane yield (SMY) and VFA accumulations is presented in Fig. 2 and Fig. 3, respectively. The results show that the SMY of 0.147 -0.162 L/g of VS fed was obtained, which are comparable to other studies with co-

M an

digestion as reported in Table 1. It is to be noted that the results reported in Table 1 were for (i) diluted wastes, which was indicated by the substantially lower value of ammonia concentrations and (ii) for mesophilic to thermophilic temperatures. However, for solid fractions of chicken litter with hay bedding at psychrophilic operating temperatures, the SMY obtained in this study need to be validated. For 26 days of operation cycle, the VFA`s especially acetic and propionic acid concentrations in the inoculum reactors

ed

(BR2a and BR2b) continuously increased during the treatment cycle. The average acetic acid concentration increased from about 323

ce pt

to 2381 mg/L and the average propionic acid concentration increased from 25 to 983 mg/L during the 26 days of operation. The inoculum reactor`s pH (BR2a and BR2b) went down from 8.05 to 7.32 due to VFAs accumulation. This indicates that (i) VFA`s from the substrate reactors i.e. DAD reactors (BR1a and BR1b) have been washing out to the inoculum reservoir (BR2a and BR2b) due to liquid percolation and (ii) more retention time would be required to digest the

Ac

accumulated acids and to extend the level of degradation of the chicken litter. On the other hand, methane content of substrate digesters (BR 1a and BR1b) was very low with values 14-16% compared to 42-43% in the liquid reservoirs (BR2a and BR2b). The probable reasons for the lower values recorded for substrate DAD digesters, could be due to hydrolysis process and this appeared to be caused by an inhibition of methanogens owing to high nitrogen concentration. The limited mass of micro-flora available for substrate digestion probably affected the rate of biogas production, as less than 3 to 4 L of inoculum was recirculated every day. An increment 32 Page 10 of 28

ip t

cr

of NH3-N and TKN levels was observed in the liquid reservoirs (BR2a and BR2b), such that average NH3–N and TKN concentrations

us

augmented from 3.2 and 4.1 g/L at the start of the cycle to 3.6 and 4.7 g/L at the end of the cycle, respectively. This increase was probably due to washout of ammonia from the substrate DAD systems. However, the residual high nitrogen content in the DAD

M an

systems inhibited the methanogens. It is to be noted that solids content present in the inoculum settled at the bottom of the inoculum reservoir with time. Due to this, when we pumped the inoculum from the bottom of inoculum reactor into the DAD system, solids were also pumped into the DAD system which accelerated the clogging of the solid matrix in the DAD system. The percolation process did not work properly.

ed

The reactor BR1b had another technical issue in the beginning of the cycle: as the thickness of the chicken litter (1.5”- 2”) used was

ce pt

minimal in the reactor to maintain the OLR of 5.4 g VS/kg VS/d, the sprinkling of the liquid from BR2b (liquid reservoir) to BR1b (DAD system) displaced the solids and created opening in the solid matrix. Due to this, an important short circuiting took place and the inoculum contact time with the solid substrate was probably reduced compared to BR1a (DAD system) for the initial 9-11 days of operation. However, after 11 days, this opening was slowly covered with solids transferred from the liquid digester by recirculation.

Ac

Towards the end of study, the biogas and CH4 production were quite similar in both sets of reactors, i.e. BR1a+BR2a and BR1b+2b. Therefore, the initial inoculum bypass did not affect the process considerably. However the CH4 content is quite low with values in the range of 35±1 and 39±4 %, respectively for BR1a+BR2a and BR1b+BR2b, which could make this DAD+liquid recirculation system practically less feasible.

33 Page 11 of 28

ip t

cr

As a result the percolation process was not taking place effectively. In addition, PL+hay contains TKN of in the range 30,000 mg/L,

us

which is quite high compared to the studies reported in Table 1. Due to this phenomenon along with lower methane content compared to the studies reported in Table 1, the operation strategy was modified in the Phase 2 of the study, such that the inoculum from the

M an

inoculum reservoir was pumped from the middle part of the digester to avoid pumping more solids. Also, an additional pump was used to mix the reactor content of the inoculum reactor before pumping in to the DAD system primarily to maintain homogeneity. In addition to this, one set of reactors was operated with down-up mode of recirculation, predominantly to avoid clogging and compare the performance with top-down recirculation mode. OLR was increased four times to that of Phase 1 of the study and such that the

ed

thickness of the chicken litter applied to the solid reactor was also augmented. To compensate this, longer cycle length was provided in

ce pt

the Phase 2 to digest the accumulated VFA`s and to have better performance.

Performance of Reactors (Phase 2)

Two sets of bioreactors BR (3-4) [top-down recirculation] and BR (5-6) [down-up recirculation] were operated in parallel under

Ac

similar operating conditions as presented in Table 4, in which BR4 and BR6 represents inoculum reservoir and BR3 and BR5 represents substrate DAD system. The results obtained for methane production and yield is also given in Table 4. An illustration of the profile for the cumulative methane, biogas production and specific methane yield (SMY), and VFA accumulation is presented in Fig. 4 and Fig. 5, respectively. The OLR during this phase of study was maintained in the range of 21.6 g VS/kg inoculum VS/d, i.e. 4 times compared to the Phase 1 of the study. 34 Page 12 of 28

ip t

cr

The results show that SMY obtained was much lower i.e. about 61-70% less yield compared to the Phase 1 of the study. For 38

us

days of operation cycle, the VFA`s especially acetic and propionic acids in the inoculum reactors (BR4 and BR6) were showed to have an increasing trend. That is to say that average acetic acid concentration increased from about 340 to 7130 and 6496 mg/L and average

M an

propionic acid concentration increased from 169 to 3318 and 2888 mg/L for BR4 and BR6 respectively, during the 38 days of operation. VFA concentration is increasing constantly. This fact, in addition to the short period studied, and with the low methane production seems to indicate that tests are not completed. Probably, methane has been generated by hydrogen based methanogens and the acetoclastic activities are underdeveloped. The hydrolysis and acetogenesis has been developed in the substrate reactors with high

ed

hydrogen production, which it has been consumed in order to produce methane but the acetoclastic activities are underdeveloped. The

ce pt

inoculum reservoir`s pH (BR4 and 6) decrease slightly from around 8 to 7.6. The alkalinity at the end of the cycle (BR4 and 6) was in the range of 13.5-15.5 gCaCO3/L. This indicates that VFA`s from the substrate (DAD) reactors (BR3 and BR5) have been washing out to the inoculum reservoir (BR4 and BR6) due to inoculum percolation and about 2.5 times higher VFA concentrations were recorded compared to Phase 1 due to its higher loading rates.

Ac

Methane production and SMY was still going strong at the end of the treatment cycle, which means that a longer cycle length would result in a higher methane production and SMY. The distressing factor is the methane content of biogas in the digesters BR3+4 and BR5+6, which were very low with values of about 18% and didn`t improve during the 38 days of operation. The methane content observed in this phase were 40-50% lower than that of Phase 1 of the study. The probable reasons for the lower values recorded, could be due higher loading rate and ammonia inhibition particularly higher TKN concentration. Average NH3–N and TKN concentrations in 35 Page 13 of 28

ip t

cr

the liquid reservoirs (BR4 and BR6) increased from 3.4 and 4.29 g/L at the start of the cycle to 3.8 and 4.95 g/L at the end of the cycle,

us

respectively. This increase was probably due to washout of ammonia from the substrate DAD systems. Nevertheless, the residual high nitrogen content in the DAD systems inhibited the methanogens. In this Phase of study, the technical problems faced in the Phase 1 of

M an

the study such as pumping solids from inoculum solids, short circuiting, clogging were not observed. This shows that the process did not work effectively at higher OLR (21.6 g VS/kg VS/d) and TKN (>33,000 mg/L) concentrations. Rather, well adapted inoculum to such high concentrations of TKN might reduce the inhibitions.

ed

Final discussions

ce pt

Digester`s operating temperature have a critical effect on the course of methane production, as it could affected the length of the lag phase. Abouelenien et al (2014) observed slightly longer lag phase (13-14 d) under the mesophilic conditions (35 oC) compare to the thermophilic process (55oC) for treating fresh semi-solid chicken litter at 10% TS, which was caused by the rapid degradation of organic matter under the thermophilic conditions. In the present study, about 10 d was observed as a lag phase of methane production,

Ac

which was quite close to that of mesophilic operating conditions. The probable reason could be that use of liquid inoculum adapted to TKN and TAN concentrations of 8.5-9.7 g/L and 7-8 g/L, respectively in the present study. According to Wang et al. (2012) and Rajagopal et al. (2013), this scenario can probably be explained by the slower pace of uric degradation in the PL under low temperature conditions resulted in lesser free ammonia production, which is less toxic to the methanogens than under higher temperature conditions. Specifically, low temperature anaerobic digestion offers the some proven benefits such as (i) it operates at low 36 Page 14 of 28

ip t

cr

temperature (20°C), resulting in year-round positive energy balance, even in cold climates; (ii) ability to utilize a liquid and/or solid

us

feedstock(s); and (iii) minimal impact from loading/feeding ambient temperature manure, even in winter. The present preliminary study demonstrated that the PL with hay bedding digestion (TS=68.6%) could be possible at lower OLR

M an

less than 5.4 g VS/kg inoculum VS/d. Higher TKN concentrations probably inhibited the AD process especially at higher OLR. Although PL performed poorly, they may still have high potential as biomass for dry anaerobic digestion if appropriate designs are developed to prevent significant volatile fatty acid (VFA) accumulation and ammonia inhibition. As described in Table 1, co-digestion with other organic substrates like pig manure or cattle manure could probably improve the digestion process by diluting the TKN

ed

concentrations and also maintaining the optimal C/N ratio. In addition, adapted liquid inoculum to a TAN concentration of 7-8 g/L was

ce pt

used in the present study; but however, well acclimatised liquid inoculum especially to TKN in the range of 20-30g/L could probably minimize the ammonia inhibition and thus improve the performance of DAD. These techniques could also enhance the methane content of the biogas, so that biogas boiler efficiency can be improved. A further extensive research is required to obtain the inhibitive threshold level of TKN and its optimal operating conditions. Down-up mode of recirculation doesn’t show a significant improvement

Ac

compared to top-down recirculation. The top portion of the solid matrix was floating and was not in complete contact with the inoculum when down-up recirculation was done. Hence top-down recirculation is recommended for simpler operation. However, these two modes of operation have to be validated in pilot-scale reactors. In addition, a hybrid system, which uses both mode of operation, could have been useful, especially if clogging occurs at the top, then recirculation could be switched to the bottom. Nevertheless, this hypothesis needs to be validated in large pilot. 37 Page 15 of 28

ip t cr us

Conclusions

This study demonstrated the preliminary investigation of DAD of PL with hay bedding using liquid inoculum recirculation-percolation

M an

mode of operation at 20oC. The highest SMY of 0.147 -0.162 L/gVSfed was obtained for OLR of 5.4 gVS/kginoculumVS/d with a methane content of 35-39%. Due to liquid recirculation-percolation, an increase in VFA and ammonia concentrations was observed in liquid reservoirs. Higher TKN concentrations (>30 g/L) in PL inhibited the AD process especially at higher OLR. Nevertheless, well acclimatised liquid inoculum especially to TKN of >30g/L and/or co-digest with C rich substrates would improve the performance of

ce pt

Acknowledgements

ed

DAD. Further research is required to obtain the inhibitive threshold level of TKN.

This project has been financially supported by contributions of Agriculture and Agri-Food

Ac

Canada and Bio-Terre Systems Inc.

References

38 Page 16 of 28

ip t

cr

Abouelenien, F., Kitamura, Y., Nishio, N., Nakashimada, Y., 2009. Dry anaerobic ammonia–methane production from chicken

us

manure. Appl Microbiol Biotechnol 82, 757–764.

Abouelenien, F., Namba, Y., Kosseva, MR., Nishio, N., Nakashimada, Y., 2014. Enhancement of methane production from co-

M an

digestion of chicken manure with agricultural wastes. Bioresour Technol 159, 80–87. Ahn, HK., Smith, MC., Kondrad, SL., White, JW., 2010. Evaluation of biogas production potential by dry anaerobic digestion of switch grass-animal manure mixture. Appl Biochem Biotechnol 160, 965–975. APHA., 1992. Standard methods for the examination of water and waste water. 18th ed. Washington DC, USA: American Public

ed

Health Association.

ce pt

Babaee, A., Shayegan, J., Roshani, A., 2013. Anaerobic slurry co-digestion of poultry manure and straw: effect of organic loading and temperature. J Env Health Sci & Eng, 11-15.

Bujoczek, G., Oleszkiewicz, J., Sparling, R., Cenkowski, S., 2000. High solid anaerobic digestion of chicken manure. J Agr Eng Res 76, 51–60.

Ac

Callaghan, FJ., Wase, DAJ., Thayanithy, K., Forster, CF., 2002. Continuous codigestion of cattle slurry with fruit and vegetable wastes and chicken manure. Biomass Bioenergy 27, 71–77. Fdez-Guelfo, LA., Alvarez-Gallego, C., Sales Marquez, D., Romero Garcia, LI., 2010. Start-up of thermophilic-dry anaerobic digestion of OFMSW using adapted modified SEBAC inoculum. Bioresour Technol 101, 9031–9039.

39 Page 17 of 28

ip t

cr

Gelegenis, J., Georgakakis, D., Angelidaki, I., Mavris, V., 2007. Optimization of biogas production by co-digesting whey with diluted

us

poultry manure. Renew Energy 32, 2147–2160.

Jha, AK., Li, J., Nies, L., Zhang, L., 2011. Research advances in dry anaerobic digestion process of solid organic wastes. African J

M an

Biotechnol 10 (65), 14242-14253.

Khalid, A., Arshad, M., Anjum, M., Mahmood, T., Dawson, L., 2011. The anaerobic digestion of solid organic waste. Waste Manag 31, 1737–1744.

Li, YB., Park, SY., Zhu, JY., 2011. Solid-state anaerobic digestion for methane production from organic waste. Renew Sust Energy

ed

Rev 15, 821–826.

ce pt

Magbanua, BS., Adams, TT., Johnston, P., 2001. Anaerobic co-digestion of hog and poultry waste. Bioresour Technol 76, 165–168. Massé, D., Gilbert,Y., Topp, E., 2011. Pathogen removal in farm-scale psychrophilic anaerobic digesters processing swine manure. Bioresour Technol 102, 641–646.

Rajagopal, R., Massé, DI., Singh, G., 2013. A critical review on inhibition of anaerobic digestion process by excess ammonia.

Ac

Bioresour Technol 143, 632–641.

Statistics Canada., 2008, EnviroStats. Catalogue no. 16-002-X. http://www.statcan.gc.ca/pub/16-002-x/16-002-x2008004-eng.pdf (accessed on April 20, 2016) Wang, X., Yang, G., Feng, Y., Ren, G., Han, X., 2012. Optimizing feeding composition and carbon–nitrogen ratios for improved methane yield during anaerobic codigestion of dairy, chicken manure and wheat straw. Bioresour Technol 120, 78–83. 40 Page 18 of 28

ip t

cr

Xie, S., Lawlor, PG., Frost, JP., Hu, Z., Zhan, X., 2011. Effect of pig manure to grass silage ratio on methane production in batch

M an

us

anaerobic co-digestion of pig manure and grass silage. Bioresour Technol 102, 5728–5733.

Figure captions

ed

Fig. 1 Schematic representation of DAD + liquid inoculum recirculation system

ce pt

Fig. 2 (A) Cumulative biogas production; (B) Cumulative methane production; and (C) Specific Methane Yield (SMY)–Phase 1 Fig. 3 VFA profile for the inoculum reservoirs (A) BR1b and (B) BR2b- Phase 1 Fig. 4 (A) Cumulative biogas production, and (B) Cumulative methane production and (C) Specific methane yield (SMY) – for Phase 2

Ac

Fig. 5 VFA and pH profiles for the inoculum reservoirs (A) BR4 and (B) BR6- for Phase 2

41 Page 19 of 28

ip t

CM or DM or SM with switch grass

CM: switch grass (%) = 0.95: 2

CM, DM, and WS

DM:CM; 100:0; 0:100; 50:50 and WS added to DMCM mixture to adjust C/N ratio to 25/1

C/N ratio

pH

us 35

35

35

55

Ac

Diluted CM and Whey

Whey in CM (15%, 25%,35% and 50% v/v)

TS (%)

Initial (CM) = 6 Initial (HW) = 5 Remaining =7-7.5

M an

CS:CM; 100:0; 70:30; 50:50; 25:75, 10:90

Retention time (d)

13

-

21

-

10-15 (CM was diluted to 15% TS before mix with CS) 6 (CM was diluted to Manure = 3.9 6% TS before mix with whey)

7.8-8

ed

CM with CS

HW:CM100:0; 80:20; 60:40;

Temperature (oC)

ce pt

CM and HW

Ratios of substrates

cr

Table 1. Summaries of results for co-digestion of chicken manure (CM) with various organic wastes Cosubstrates*

35

18

Manure = 7.4

62

15

32.9

6.9 ± 0.1

30

14 (DM) 29 (CM) 86 (WS)

27.2

7.03-7.34

Methane/ Biogas yields

Biogas yield (H80) = 0.20 ± 0.03 m3/kgVS CH4 yield = 0.13 ± 0.03 m3/kgVS CH4 yield = 0.12 m3/kg VS

Biogas production (1.5– 2.2 L/L of reactor/d) CH4 yield = 0.02 m3/kgVS

CH4 = 0.235 m3/kgVS

Ammonia/ ammonium values

[NH3-N] = 2.6–7.9 g/L

[NH3] > 0.10 g/L

References

Magbanua et al. (2001)

Callaghan et al. (2002)

Total ammonia = 3.2 gN/L

Gelegenis et al. (2007)

[NH3-N] = 1.5 g/L

Ahn et al. (2010)

Total ammonia = 1.8 gN/L

Wang et al. (2012)

* CM: Chicken Manure; HW: Hog Waste; CS: Cattle slurry; DM: Dairy manure; SM: Swine Manure; WS: Wheat Straw

42 Page 20 of 28

Inoculum

Total COD g/kg

888

--

NH3-N mg/kg

7,000

3619

TKN mg/kg

33,800

4606

TS %

68.6

5.03

VS%

57.7

3.7

VS/TS

0.84

0.74

pH

--

8.3

M

2 3

8 9 10 11

te

7

Ac ce p

6

d

4 5

cr

PL + Hay

us

Parameters

ip t

Table 2 Characteristics of the feed material and inoculum

an

1

12 13 14 15

43 Page 21 of 28

Table 3: Operating conditions and the methane production (1st cycle of operation) Units

VS Content %

OLR

3.7 (BR1b)

57.70 (BR2a)

57.70 (BR2b)

0.577 (BR2a)

g VS/kg 5.4

Cycle length

days

26

Quantity of PL + Hay

kg

Total VS fed

g

Total cumulative CH4

L

an

inoculum VS/d

0.037 (BR1b) 0.577 (BR2b)

M

0.216

5.4 26 0.216 125

18.33

20.24

- Inoculum reservoir %

43.4 (BR2a)

42.2 (BR2b)

- DAD reactor %

14.7 (BR1a)

15.4 (BR1b)

35 (BR1a+BR2a)

39 (BR1b+BR2b)

0.147

0.162

d

125

Ac ce p

te

CH4 content:

- DAD+Inoculum reservoir % Specific CH4 yield (SMY)

18

3.7 (BR1a)

0.037 (BR1a) kg VS/ kg

17

BR1b+BR2b

cr

VS

BR1a+BR2a

ip t

Parameter

us

16

L/g of VS fed

19 20 21 22

44 Page 22 of 28

Table 4: Operating conditions and the methane production (2nd cycle of operation) Units

BR (3+4)

BR (5+6)

VS Content

%

57.70 (BR3)

57.70 (BR5)

VS

kg VS/kg

0.577 (BR3)

0.577 (BR3)

OLR

g VS/kg 21.6

Cycle length

days

38

Quantity of PL + Hay

kg

0.865

0.865

Total VS fed

g

500

500

M

us

28

%

Specific CH4 yield (SMY)

L/g of VS fed

24

18±3 (BR3+4)

18±5 (BR5+6)

0.056

0.048

d

CH4 content

38

te

26

21.6

Ac ce p

25

cr

inoculum VS/d

Total cumulative CH4 (BR3+4) L

24

ip t

Parameter

an

23

45 Page 23 of 28

26

Fig. 1

30 31 32 33 34 35

d te

29

Ac ce p

28

M

an

us

cr

ip t

27

36 37 38 39 40 46 Page 24 of 28

Fig. 2

us

cr

ip t

41

Ac ce p

43

te

d

M

an

42

44 45 46 47 Page 25 of 28

Fig. 3

an

us

cr

ip t

47

Ac ce p

te

d

M

48

49 50 51 52 48 Page 26 of 28

Fig. 4

us

cr

ip t

53

55

Ac ce p

te

d

M

an

54

56 49 Page 27 of 28

Fig. 5

M

an

us

cr

ip t

57

Ac ce p

te

d

58 59

60 61 62 63 50 Page 28 of 28