Scholarly article on topic 'Dynamic Modelling and Simulation of Pilot Scale Anaerobic Digestion Plant Treating Source Separated Food Waste and Effect of Recycling Sludge'

Dynamic Modelling and Simulation of Pilot Scale Anaerobic Digestion Plant Treating Source Separated Food Waste and Effect of Recycling Sludge Academic research paper on "Chemical engineering"

CC BY-NC-ND
0
0
Share paper
Academic journal
Procedia Environmental Sciences
OECD Field of science
Keywords
{ADM1 / "Anaerobic digestion" / "Food waste" / "Dynamic modeling"}

Abstract of research paper on Chemical engineering, author of scientific article — P.G. Rathnasiri

Abstract This study investigates the effect of recycling sludge and stability of pilot scale anaerobic digestion plant treating organic fraction of source separated food waste. Pilot plant comprises of pre-treatment, anaerobic digestion and post treatment. Anaerobic digestion is carried out under mesophilic conditions. At flow rate of 0.221 m3/d, slurry of food waste is introduced semi-continuously every day within 2 hr time span. Anaerobic digestion model No.1 (ADM1) was applied for modelling and simulation of continuous stirred tank anaerobic reactor including recycling and was implemented in AQUASIM 2.1f. Input food waste was characterized and parameters were determined as dictated by ADM1. Process parameters were obtained from pilot plant and kinetic parameters are standard parameters given in ADM1. Input Organic Loading Rate(OLR) to reactor is 4.81 kgCOD/m3.d and Hydraulic Retention Time (HRT) was 20 days. Pilot scale AD plant was simulated for 40 days. It was found that biogas production rate and gas composition vary according the intermittent feeding pattern and reactor head space contains average compositions of CH4 and CO2 of 56% and 30% (v/v) respectively. Average biogas production rate under this condition was 16.4 m3/d. Recycling of sludge in terms of biomass was implemented in model as an advective link. When model was simulated at 40 day solid retention time (SRT), average biogas production rate increased by 31%. By varying OLRs to reactor, stability was investigated. When input OLR was doubled in terms of hydraulic load, anaerobic reactor became unstable producing H2 (38% v/v) and CO2 (40%v/v) in reactor head space. Under this condition, bulk liquid phase pH was 4.78. Instability is further confirmed by accumulation of volatile fatty acids and inhibition of strict methanogens. ADM1 can be applied to model and simulate pilot scale anaerobic digestion plant and to screen different options before scaling up into large scale plants.

Academic research paper on topic "Dynamic Modelling and Simulation of Pilot Scale Anaerobic Digestion Plant Treating Source Separated Food Waste and Effect of Recycling Sludge"

CrossMark

Available online at www.sciencedirect.com

ScienceDirect

Procedía Environmental Sciences 35 (2016) 740 - 748

International Conference on Solid Waste Management, 5IconSWM 2015

Dynamic modelling and simulation of pilot scale anaerobic digestion plant treating source separated food waste and effect of

recycling sludge

P.G. Rathnasiri*

Senior Lecturer, Department of Chemical and Process Engineering, University ofMoratuwa, SriLanka

Abstract

This study investigates the effect of recycling sludge and stability of pilot scale anaerobic digestion plant treating organic fraction of source separated food waste. Pilot plant comprises of pre-treatment, anaerobic digestion and post treatment. Anaerobic digestion is carried out under mesophilic conditions. At flow rate of 0.221 m /d, slurry of food waste is introduced semi-continuously every day within 2 hr time span. Anaerobic digestion model No.1 (ADM1) was applied for modelling and simulation of continuous stirred tank anaerobic reactor including recycling and was implemented in AQUASIM 2.1f. Input food waste was characterized and parameters were determined as dictated by ADM1. Process parameters were obtained from pilot plant and kinetic parameters are standard parameters given in ADM1. Input Organic Loading Rate(OLR) to reactor is 4.81 kgCOD/m3.d and Hydraulic Retention Time (HRT) was 20 days. Pilot scale AD plant was simulated for 40 days. It was found that biogas production rate and gas composition vary according the intermittent feeding pattern and reactor head space contains average compositions of CH4 and CO2 of 56% and 30% (v/v) respectively. Average biogas production rate under this condition was 16.4 m3/d. Recycling of sludge in terms of biomass was implemented in model as an advective link. When model was simulated at 40 day solid retention time (SRT), average biogas production rate increased by 31%. By varying OLRs to reactor, stability was investigated. When input OLR was doubled in terms of hydraulic load, anaerobic reactor became unstable producing H2 (38% v/v) and CO2 (40%v/v) in reactor head space. Under this condition, bulk liquid phase pH was 4.78. Instability is further confirmed by accumulation of volatile fatty acids and inhibition of strict methanogens. ADM1 can be applied to model and simulate pilot scale anaerobic digestion plant and to screen different options before scaling up into large scale plants. © 2016 The Authors.Publishedby ElsevierB.V. Thisis an open access article under the CC BY-NC-ND license (http://creativecommons.Org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the organizing committee of 5IconSWM 2015

Keywords: ADM1, Anaerobic digestion, Food waste,Dynamic modeling;

* Corresponding author.

E-mail address: ratnasiri@cheng.mrt.ac.lk

1878-0296 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the organizing committee of 5IconSWM 2015

doi: 10.1016/j.proenv.2016.07.082

1. Introduction

Food waste is considered as a desirable input substrate for anaerobic digestion, but it is prone to over acidification and lower pH levels due to accumulation of volatile fatty acids (Curry and Pillay 2012). Anaerobic digestion of food waste through the operation of a mesophilic two stage and single stage pilot scale was conducted by Grimberg et al., 2015 with the objective of assessing variable loadings on system performance. When comparing two stage and single stage system, both produced similar quality effluent, but biomethane yield was higher in two stage system. Anaerobic Digestion Model No.1 (ADM1) developed by the International Water Association (IWA) task group for mathematical modelling and simulation (Batstone et al., 2002) has been widely used for modelling and on and validation of Anaerobic ADM1 simulations under different anaerobic reactor configurations, different operating conditions and different substrates have received great attention in the recent past (Atallah et al., 2014; Razarviarani et al., 2015; Shi et al., 2014). simulations of lab scale, pilot scale , large scale anaerobic reactors. Two stage pilot scale high solid anaerobic digester system developed by Yu et al., 2012, used food waste as substrate. The first high solids anaerobic digester was partially mixed and second reactor operated in the UASB mode and leachate recirculation was the sole purpose. This system was modelled by ADM1 and predictions indicated that recycled methenogenic bacteria increased methane concentration and decreased hydrogen concentration in the first reactor. However major limitation of latter study was the modelling of UASB reactor with the approximation of advective diffusive reactor model. Not only single substrate, ADM1 model has also been applied to simulate anaerobic co-digestion of organic fraction of municipal solid wastes along with activated sludge in mesophilic condition (Derbal et al.,2009). This simulation results showed a good agreement with measured, pH, methane, biogas volume in the reactor.

Effect of recycling of sludge towards stability of anaerobic digestion has not extensively studied in previous studies. Objectives of this study are to model and simulatepilot scale anaerobic digestion plant treating source separated canteen food waste and investigate stability of process via recycling of sludge and by varying hydraulic loading. Anaerobic digestion model No.1 (ADM1) developed by International Water Association ( Batstone et al., 2002) is built in a dynamic simulator called Aquasim 2.1f and simulation is performed to study dynamic behaviour of the process.

2. Materials and methods

2.1 Description on pilot plant

Schematic diagram of the pilot scale anaerobic digestion plant is shown in Fig. 1. This is a single stage mesophilic process. Source separated food waste daily collected from University student's canteen is shredded by adding dilution water and slurry of MSW is then transferred into buffer tank where submerged pump is installed. Then this slurry is semi continuously fed at a rate of 0.221 m3/d to completely mixed high rate anaerobic digesterthat has bulk liquid volume of 4.44 m3.' Biogas generated is upgraded using scrubbing and subsequently used for either flaring or production of heat. Digested slurry is transferred into sedimentation tank where sludge and water are separated. Water is recycled back to the pre-treatment section as dilution water.

Table 1: Process parameters from pilot plant

Total amount of source separated MSW

Moisture content of MSW

Total solid content in MSW slurry

Flow rate of slurry

Hydraulic retention time (HRT)

Bulk liquid volume of anaerobic reactor

Head space of reactor

Height of the reactor

Diameter of the reactor

60 kg/d 63% 10%

0.221 m3/d 20 d

4.438 m3 0.44 m3 2.93 1.46 m

Fig. 1.Schematic diagram of pilot scale anaerobic digestion plant.(1) Shredder; (2) Buffer tank; (3) Submerged pump; (4) Anaerobic digester; (5) Pressure gauge; (6) Pressure relief valve; (7)Moisture trap; (8) Sedimentation tank; (9) Compressor; (10).Scrubber; (11) Gas

collection; (12) Final digestate collection tank

Water contains in the sludge is further removed and following aerobic composting, digestate is used as fertilizer. Extracted flow sheet for simulations is shown in Figure 2.

Fig. 2. Extracted flow sheet for simulation

2.2 Input Waste characterization

The first step of ADM1 is the detailed characterization of input substrates into its constituents. Disintegration of composite food waste into particulate components is performed via shredding and biodegradable components are characterized into carbohydrates, Proteins and Lipids. Measured TVS (55.6 kg/m3) of food waste slurry (Rathnasiri et al., 2005) is multiplied by the flow rate and loading rates were calculated in TVS (kg/d). By distributing TVS load equally into its constituent'si.e.0.33% fat, 0.33% protein, and 0.33% carbohydrates, COD loads for each component were then calculated. Influent slurry flow to anaerobic reactor is shown in Fig. 3. Based on this flow pattern and input COD concentrations, OLR s were determined and shown in Fig.4. To evaluate the effect of varying input values of each constituent, sensitivity analysis was performed and discussed under results section.

Fig. 3. Input food waste slurry flow rate into anaerobic reactor

Fig. 4. Organic load into anaerobic reactor

2.3 Modelling and simulation

ADM1 is a mechanistic model comprising 19 biochemical reactions with physico chemical reactions. This ADM1 was built in Aquasim 2.1f(Reichert, 1998). All kinetics and physico-chemical parameters were taken from the original version of ADM1. Process parameters were obtained from pilot plant. The model was initialized and simulated by selecting step size (0.01) and no. of steps (4000) according to the desired time period for 40 days. Simulation was conducted at four different solid retention times.Recycling of sludge is modelled as an advective link. This recycle flow line contains biomass in terms of all microbial species involved.

3. Results and discussion

The influent at organic loading of 4.81kgCOD/m3.d was fed intermittently. During characterisation of this stream, equal fractions of particulate matters were assumed. But sensitivity analysis performed confirmed that the input carbohydrate concentration has the highest sensitivity towards the CO2 composition in the reactor head space. While operating at 20 day hydraulic retention time, three solid retention times were investigated and under each SRT volumetric biogas production rate is shown in Fig. 5. During the simulation period of 40days only two consecutive cycles are shown in Fig.5 and it can be observed that biogas production rate is increased by 31% and no significant difference among other different SRTs.

Fig. 5. Biogas production rate variation for two consecutive cycles of intermittent feeding at varying solid retention time

It can be seen that (Fig. 6), biogas compositions vary due to intermittent feeding, but there is no significant variation among different SRTs. Following daily feeding, CO2 composition increases up to peak and CH4 composition decreases during this time period. But average CH4 ( 55%) and CO2 (30%) are within stable operation conditions of anaerobic reactor.

Fig. 6. Biogas composition (CH4 and CO2) variation in reactor head space

In the ADM1 model, there are seven bacterial species defined as dynamic state variables and final conversion step of mehtenogenesis comprises of two bacterial species i.e. aceticlastic which is acetic acid degrading microorganisms and hydrogenotrphes which is hydrogen degrading microorganisms. Their total concentration under varying SRT is shown in Fig. 7. According to these results, biomass concentration increases by 78% due to recycling and also it increases with increasing SRT.

0 A-,-,-,-,

0 10 20 30 40

Time (d)

Fig. 7. Active biomass concentrations of methenogenic bacteria in reactor bulk liquid

ADM1 model biochemical reactions comprises of VFA as acetic, propionic, butyric, valeric, fatty acids and amino acids. Though these acids are present in the liquid phase, only acetic acid is shown in Fig.8. It can be clearly seen that acetic acid concentration reduces by 80% due to recycling of biomass. When SRT increases this acetic acid concentration further reduces improving anaerobic reactor performance. When there is no recycling, bulk liquid pH

varies between 7.1 and 7.5.and under recycling at SRT 25d, pH varies 7.3 -7.7. Thus improvement of acetic acid conversion and increase of active concentrations of methenogens are the reasons for increased biogas production rate.Two stage pilot scale anaerobic reactor treating highly degradable source separated food waste obtained from cafeteria was modelled by Yu et al., 2012. This comprises of high solids anaerobic digester followed by leachate recirculation reactor operating in UASB mode. The upper section of anaerobic reactor had more biodegradable solid wastes floating and the lower section contained more liquid leachate. When the recycled methenogenic bacteria were increased by two fold, the acetic acid concentration was reduced in the lower section of the anaerobic digester. These observations are consistent with the results obtained in this study. To investigate the effect of solids retention time on anaerobic digestion of swine manure, experiments were conducted using four bench scale anaerobic reactors at SRTs of 14, 21, 28 and 42 days (Kinyua et al., 2014). All reactors performed very well and VS removal was more than 60%. In addition, CH4 yield interms of m3 CH4/kg VS added also increased with increase SRT. Same trend was also observed in this study under varying SRTs.

Fig. 8. Acetic acid concentration inside reactor due to varying recycling rate (SRTs).

3.1 Stability of pilot plant due to overloading

Anaerobic reactors are frequently subjected to input load variations due to input waste compositions variations or hydraulic load variations such as dilution. In this study, input OLR was doubled in terms of hydraulic load and pilot scale anaerobic reactor stability was studied. It was noted that reactor was inhibited leading to produce high acetic acid concentration of 16000 mg/l (Figure 9). In addition higher concentrations of butyric and propionic acid also accumulated in liquid phase. This causes to decrease pH and inhibition has been incorporated into ADM1 model by multiplying substrate utilization rate of each species with inhibition functions which vary between 0 and 1.

Reactor failure is manifested in head space gas compositions and given in Figure 10. CH4 compositions decreases up to 3% (v/v) and H2and CO2 concentrations increases to 43%( v/v) and 38%v/ respectively. Under this condition, bulk liquid phase pH was 4.78. Accumulation of hydrogen is a dominant of acedogenic phase of anaerobic digestion process leading to accumulate more VFAs. According to results from inhibition functions, pH inhibition has directly affected to growth of acetic acid degrading and hydrogen degrading methanogens presence in the reactor.

Fig. 9. Bulk liquid phase VFA concentrations when organic loading rate is doubled

g. 50 E o O

......H2

**X / *%v f

40 -L>< JX J>' Ji 4.70 Ï

31 31.5 32 32.5 33 33.5 34 34.5 35 Time (d)

Fig. 10. Head space gas composition when input organic loading rate is doubled

4. Conclusions

This study has demonstrated that pilot scale anaerobic plant treating source separated canteen food waste can be successfully dynamically modelled and simulated using ADM1. The model predictions indicated that under operating organic loading rate of 4.8 kgCOD/m3.d, anaerobic process was stable and produced high quality biogas. Effect of recycling biomass into anaerobic reactor was studied under varying solid retention times. It was found that increase of biomass recycling causes to increase biogas production rate. This is due to the increase of active methanogens and enhancement of conversion of acetic acid presence in the reactor. Recycling of biomass not greatly affected for variations of pH inside the reactor. Reactor stability was studied by increasing the OLR and found that reactor was completely inhibited due to accumulation of VFAs. Screen of different options at pilot scale greatly enhances the large scale anaerobic process developments

5. Acknowledgements.

Author thanks the Ministry of science and Technology in Srilanka providing funding to develop pilot scale biogas plant under Biomethane as Transport fuel project.

References

1) Atallah, N.M., El-Fadel, M., Ghanimeh, S., Saikaly, P. and Abou-Najm, M. (2014). Performance optimization and validation of ADM1 simulations under anaerobic thermophilic conditions, Bioresource Technology, 174, 243-255.

2) Batstone, D.J., Keller, J., Angelidaki, I., Kalyuzhnyi, S.V., Pavlostathis, S.G., Rozzi, A., Sanders, W.T.M., Siegrist, H. and Vavilin, V.A. (2002). Anaerobic Digestion Model No.1 (ADM1), IWA Task Group for Mathematical Modelling of Anaerobic Digestion Processes. London:IWA publishing.

3) Curry, N. and Pillay, P. (2012). Biogas prediction and design of a food waste to energy system for the urban environment, Renewable Energy, 41,200-209.

4) Derbal, K., Bencheikh-lehocine, M., Cecchi, F., Meniai, A.-H. and Pavan, P. (2009). Application of the IWA ADM1 model to simulate anaerobic co-digestion of organic waste with waste activated sludge in mesophilic condition, Bioresource Technology, 100, 1539-1543.

5) Grimberg, S.J., Hilderbrandt, D., Kinnunen, M. and Rogers, S. (2015). Anaerobic digestion of food waste through the operation of a mesophilic two-phase pilot scale digester-Assessment of variable loadings on system performance, Bioresource Technology, 178, 226-229.

6) Maureen, N. and Kinyua, J.C. and Sarina, J. E. (2014). Effect of solids retention time on the bioavailability of organic carbon in anaerobically digested swine waste,Bioresource Technology, 162, 14-20.

7) Rathnasiri,P.G., Johansen, J.E., Drangeid, C.B.and Bakke R. (2005). Transient co-digestion plant behavior. Paper presented at the Proceedings of the 4th International Symposium on Anaerobic Digestion of Solid Waste (Denmark),Copenhagen.

8) Razaviarani,V. andBuchanan,I.D. ( 2015). Calibration of the Anaerobic Digestion Model No. 1 (ADM1) for steady-state anaerobic co-digestion of municipal wastewater sludge with restaurant grease trap waste,Chemical Engineering Journal, 266, 91-99.

9) Reichert P. (1998).User manual of AQUASIM 2.0 for the identification and simulation of aquatic Systems. Swiss Federal Institute for Environmental Science and Technology, Dubendorf, Switzerland

10) Shi,X.S., Yuan,X.Z., Wang,Y.P., Zeng,S. J.,Qiu,Y.L.,Guo,R.B. andWanga,L.S. (2014). Modeling of the methane production and pH value during the anaerobic co-digestion of dairy manure and spent mushroom substrate, Chemical Engineering Journal, 244, 258-263.

11) Yu,L., Zhao,Q., Ma,Jingwei., Frear,Craig. andChen,S. (2009). Experimental and modeling study of a two-stage pilot scale high solid anaerobic digester system,Bioresource Technology, 124, 8-17.