Bioresource Technology 169 (2014) 428–438

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Effects of aeration rate on degradation process of oil palm empty fruit bunch with kinetic-dynamic modeling Ahmad Tarmezee Talib a, Mohd Noriznan Mokhtar a,⇑, Azhari Samsu Baharuddin a,b, Alawi Sulaiman c a

Department of Process and Food Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia Institute of Tropical Forestry and Forest Products, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia c Faculty of Plantation and Agrotechnology, Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia b

h i g h l i g h t s  Effect of different aeration rates on the EFB co-composting process was studied.  Lower aeration rates significantly effect on the EFB degradation.  A new kinetic model with mass and energy transfers and balances was introduced.  Mathematical modelling is implemented to describe the phenomena in OM degradation.

a r t i c l e

i n f o

Article history: Received 19 May 2014 Received in revised form 5 July 2014 Accepted 7 July 2014 Available online 14 July 2014 Keywords: Oil palm empty fruit bunch (EFB) Degradation Co-composting Kinetic modeling Aeration rate

a b s t r a c t The effect of different aeration rates on the organic matter (OM) degradation during the active phase of oil palm empty fruit bunch (EFB)-rabbit manure co-composting process under constant forced-aeration 1 1 1 1 system has been studied. Four different aeration rates, 0:13 L min kgDM ; 0:26 L min kgDM ; 0:49 1 1 1 1 1 1 L min kgDM and 0:74 L min kgDM were applied. 0:26 L min kgDM provided enough oxygen level (10%) for the rest of composting period, showing 40.5% of OM reduction that is better than other aeration rates. A dynamic mathematical model describing OM degradation, based on the ratio between OM content and initial OM content with correction functions of moisture content, free air space, oxygen and temperature has been proposed. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Composting is a controlled microbiological degradation process of organic matter (OM) which produces a useful stable material for plant and soil use (Gomes and Pereira, 2008; Kulcu and Yaldiz, 2004). The main products of this process are carbon dioxide, water and humified materials. The main factors affecting the composting process can be divided into two main categories, namely environmental parameters such as aeration rates, moisture content (MC), pH level or temperature; and the nature of the substrate parameters such as porosity, C/N ratio or nutrient content (Diaz et al., 2002). Among other parameters, aeration, MC and temperature are the major factors affecting the composting process, as these parameters are interdependent. An aeration rate which is too high will

⇑ Corresponding author. Tel.: +60 3 8946 6367; fax: +60 3 8946 4440. E-mail address: [email protected] (M.N. Mokhtar). http://dx.doi.org/10.1016/j.biortech.2014.07.033 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.

increase energy transfer, resulting in drop in temperature and MC, and when the aeration rate is too low, oxygen level will decrease which may lead to anaerobic condition, in addition of high moisture content. Although studies have been performed to examine the influences of aeration rate on the OM degradation process (Gao et al., 2010; Guo et al., 2012; Kulcu and Yaldiz, 2004), different raw materials with different composting systems result in different level of sufficient aeration rate, especially in the initial part of the process. This part is the active phase of the composting process involving the mesophilic and thermophilic phases. The later part of the process is the less active, cooling down phase toward ambient temperatures (Mason, 2007). This includes late mesophilic and also curing phase. Malaysia, as the second world largest palm oil producer processed 95 million tons of oil palm fresh fruit bunches (FFBs) in 2013 alone (MPOB, 2013). Since each FFB contains 22% oil palm empty fruit bunch (EFB) (Sulaiman et al., 2011), 21 million tons of EFB are produced annually and the trend is increasing. In conjunction with the ‘‘green technology’’ approaches, composting

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

429

Nomenclature Ac As b C material cpwet air cpash cpj cpOM DM F1 F in F out Gf Gs kFAS kleach KlO2 mash mH2 O _ vap m H O

Surface area of bioreactor, m2 Surface area of composting material, m2 Power constant for leachate run off, – Heat capacity of material, kJ K1 Specific heat capacity of wet air, kg kJ1 K1 Specific heat capacity of ash, kg kJ1 K1 Specific heat capacity gas j, kg kJ1 K1 Specific heat capacity of OM, kg kJ1 K1 Dry material, kg Moisture correction function, – Flow in, m3 h1 Flow out, m3 h1 Specific gravity of fixed fraction of solid material, – Specific gravity of solid material, – FAS correction function, – Leachate run off constant, kg h1 Oxygen transfer constant, % Mass of ash, kg Mass of water, kg Mass rate of water evaporation, kg h1

_ bio m H2 O

Mass rate of water generated by biological reaction, kg h1 Mass rate of water vapor intake, kg h1 Mass of water loss, kg Mass rate of gas j generated by biological reaction, kg h1 Mass rate of gas j intake, kg h1 Initial mass of OM, kg Mass of total composting material, kg Molecular weight of water, kg kmol1 Molecular weight of gas j, kg kmol1 Final mass fraction of OM, – Ratio of OM i with initial OM Partial pressure of water vapor, kPa Partial pressure of gas j, kPa Heat transfer rate to surrounding, kJ h1 Heat transfer rate to exit, kJ h1 Heat rate of intake air, kJ h1 Relative humidity, – Relative root mean squared error, % Ambient temperature, K Temperature of solid state, K Maximum temperature for OM i, K Overall heat transfer coefficient, kJ h1 m2 K1 Volume of bioreactor, m3 Volume of composting material, m3 Compost water holding capacity, % Yield of humified material, kghum kg1 OM Cross-section area of pipe, m2 Heat capacity of wet air, kJ K1 Discharge flow coefficient, – Concentration of gas j, % Specific heat capacity of air, kg kJ1 K1 Specific heat capacity of water vapor, kg kJ1 K1 Specific heat capacity of water, kg kJ1 K1 Initial dry material, kg Final dry material, kg Free air space, – Temperature correction function of OM i, – Mass fraction of gas j within intake air, –

2

_ intake m H2 O mloss H2 O _ bio m j _ intake m j mOM0 mtotal MWH2 O MWj OMT OMfi Pvap H2 O Pj Q_ ambient Q_ exhaust Q_ intake

RH rRMSE T ambient Ts T maxi U Vr Vc WHC Y hum Aout C wet air Cd cj cpair cpvap H2 O cpH2 O DM0 DMT FAS fT i fj

Gv ki k0i kw kO2 mgas _ cond m H O

Specific gravity of volatile fraction of solid material, – Degradation coefficient of OM i, h1 Reaction rate constant of OM i, h1 Heat transfer coefficient, kJ m2 h1 Oxygen correction function, – Mass of air inside bioreactor, kg Mass rate of water condensation, kg h1

_ ext m H2 O _ FHin O m

Mass rate of water vapor exit, kg h1

_ leach m H2 O

Mass rate of water leachate out, kg h1

2

2

mvap H2 O

mhum mj mOM mOMi MC ni OM0 P Patm Q_ bio Q_ feed H2 O q_ s Q_ trans

Mass rate of water addition, kg h1 Mass of water vapor, kg Mass of humified material, kg Mass of gas j, kg Mass of OM, kg Mass of OM i, kg Moisture content, % Substrate i limitation constant, – Initial mass fraction of OM, – Pressure inside bioreactor, kPa Atmospheric pressure, kPa Heat rate generated by biological reaction, kJ h1

r OMi R Tg

Heat rate of water addition, kJ h1 Mass flow rate of gas, kg h1 Heat transfer rate between compost material and air, kJ h1 Reaction rate of OM i, kg h1 Gas constant, kJ kmol1 K1 Temperature of gas state, K

T feed H2 O T mini T opti VM Vg Y cond Y O2

Temperature of feeding water, K Minimum temperature for OM i, K Optimal temperature for OM i, K Mass fraction of volatile matter, – Volume of gas inside bioreactor, m3 Condensate ratio, – Oxygen consumption ratio, kgO2 kg1 OM

Subscripts  Average of observed values A j Gas j (1: CO2, 2: O2, 3: N2) Or Observed value of profile r i OM i (1: ‘‘easy’’, 2: ‘‘moderate’’, 3: ‘‘hard’’) n Number of samples Pr Predicted value of profile r Greek letter DHbio Enthalpy of biological reaction, kJ kg1 DHvap Enthalpy of water vaporization, kJ kg1 qash Density of ash, kg m3 wet qair Density of wet air, kg m3 qH2 O Density of water, kg m3 qhum Density of humified material, kg m3 qOMi Density of OM i, kg m3 DHcond Enthalpy of water condensation, kJ kg1 c Isentropic expansion coefficient, – qair Density of air, kg m3 qDM Density of dry material, kg m3 vap qH2 O Density of water vapor, kg m3 qj Density of gas j, kg m3 W Outflow coefficient factor, –

430

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

treatment has been considered to be one of the options for overcoming this problem. Most are co-composted with other materials, such as EFB with sewage sludge, palm oil mill sludge with sawdust, EFB with chicken, goat or cattle manure, EFB with palm oil mill effluent (POME) sludge, or POME with sawdust (Mohammad et al., 2012). Many studies have been performed regarding the nature of substrate parameters i.e., different substrates as co-substrate or different microbes as inoculum (Baharuddin et al., 2010; Yahya et al., 2010). Surprisingly, less attention has been given to the effects of environmental parameters such as aeration rate on the EFB co-composting process. Mathematical modeling has been implemented and developed to improve understanding about the co-composting process. Different co-composting processes involve different composting systems, substrates and microenvironments. The complexity of interaction among all these factors makes it difficult for researchers to develop a general model for the co-composting processes. To our knowledge, though several models have been developed for lignocellulosic waste degradation (Kulcu and Yaldiz, 2004; Pommier et al., 2008; Ponsá et al., 2011), no such model has been developed for EFB co-composting process. Therefore the objectives of this study are to elucidate the effect of different aeration rates on the OM degradation during the active phase of EFB-rabbit manure co-composting process, as well as to propose its kinetic-dynamic modeling with mass and energy transfers.

of 67.5 L (0.30  0.45  0.50 m). Temperature sensor was inserted into the middle of the bioreactor which then connected to in-house data logger system with setting to measure the temperature for every 30 min interval. Aeration was provided using an air pump and then humidified by passing through the washing bottle containing distilled water. The oxygen level was monitored daily using an oxygen probe (Model OT-21, Demista Instruments, USA). 2.3. Experimental procedure and analysis Four different experiments with different aeration rates were performed in two batches. Forced aeration was continuously provided with an air pump at four different flow rates. The first batch (P1) consists of two experiments with low aeration rate set at 1 1 1 1 0:13L min kgDM (AR1) and 0:26L min kgDM (AR2), whereas for the second batch (P2), the aeration rate was set higher at 1 1 1 1 0:49L min kgDM (AR3) and 0:74L min kgDM (AR4). The aeration rate was measured by using a rotameter (Valve Acrylic Flowmeter, Cole-Parmer, USA). EFB was mixed with rabbit manure at 2:1 ratio (w.b). 5% (w/w) of matured compost was added as inoculum. The co-composting material was manually mixed each time before sample was taken from bioreactor. Initial MC was adjusted within 60–70% by water addition. Water was then added into the mixtures during co-composting process when the MC level was below 50%. MC was measured by drying of samples at 105 °C for 24 h. OM was determined by loss on ignition at 550 °C for 4 h using muffle furnace (KSL-1700X, MTI Corporation, USA).The percentage of OM loss and Total Carbon Content (TOC) were calculated by applying Eq. (1) and Eq. (2), respectively (Haug, 1993):

2. Methods 2.1. Materials

OMloss ð%Þ ¼ 100 

Pressed–shredded EFB and rabbit manure were used in this study. EFB that serves as carbon source was collected from Dengkil Palm Oil Mill Sdn. Bhd. (Selangor, Malaysia), whereas rabbit manure acting mainly as a nitrogen source was supplied by Animal Resource Unit, Faculty of Veterinary Medicine, Universiti Putra Malaysia.

TOC ¼

A laboratory-scale co-composting process was carried out in a closed poly-styrene box reactor system (Fig. 1) with total volume

9

mOM  100 1:8

ð1Þ

ð2Þ

Total Kjeldahl Nitrogen (TKN) was determined according to the manufacturer’s manual using Kjeltec 2300 Analyzer (FOSS Analytical AB, Sweden). TOC and TKN were both used to calculate the carbon to nitrogen ratio (C/N). Thermogravimetric Analysis (TGA) and Derivative Thermogravimetry (DTG) were conducted using Thermogravimetric Analyzer (Pyris TGA 7, Perkin Elmer, USA) with heating

2.2. Experimental system

10

DM0  OM0  DMT  OMT DM0  OM0

8

7

6

1 3 5

2

4

Fig. 1. Schematic diagram for physical model of the composting system (1 – from air pump, 2 – valve, 3 – rotameter, 4 – air humidifer, 5 – leachate out, 6 – temperature sensor, 7 – to data logger, 8 – sampling port, 9 – oxygen probe port, 10 – exhaust pipe).

431

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

rate at 10 °C min1 under nitrogen gas purged at 20 mL min1 from 50 °C to 650 °C. Cellulose and lignin content were determined by using Acid Detergent Fibre (ADF) and Acid Detergent Lignin (ADL) method according manufacturer’s manual using Fibertec 2010 (FOSS Analytical AB, Sweden). Since ADF consists of cellulose and lignin whereas ADL consist of only lignin, cellulose content was determined by subtracting value of ADF with value of ADL. Electrical conductivity (EC) and pH level were determined by diluting sample with distilled water (1:10) and measured using pH/Conductivity meter (Hanna Instrument, USA). All chemicals and proximate analysis were measured in duplicate. 2.4. Model description

mOM

ð3Þ

ð4Þ

dmOMi ¼ r OMi dt

ð5Þ

With r OMi as the reaction kinetic for the co-composting process, which is introduced as:

r OMi ¼ mOM0  ki 

OMfi OMfi þ ni

ð6Þ

where

mOMi mOM0

ð7Þ

The production of humified materials is described as: 3 X dmhum ¼ Y hum  rOMi dt i¼1

ð8Þ

2.4.2. Substrate degradation coefficient and correction functions Substrate degradation coefficient, ki consists of four different correction functions as shown in Eq. (9) as follows:

ki ¼ k0i  F1  kFAS  fT i  kO2



1 þ ð1VMÞ G

ð12Þ

ð13Þ

f

mOM DM

VM ¼

ð14Þ

The model for oxygen correction function, kO2 was taken from Higgins and Walker (2001) as follows:

cj¼2 ðKlO2 þ cj¼2 Þ

ð9Þ

The model for correction functions for MC (F1) and FAS (kFAS ) are shown in Eq. (10) and Eq. (11), respectively (Haug, 1993).

ð15Þ

where

KlO2 ¼ 0:79  0:041  ðT s  273Þ þ 0:04  MC

ð16Þ

The temperature correction function, fT i was obtained from Rosso et al. (1993) as follows:

ðT s  T maxi ÞðT s  T mini Þ2    T mini Þ ðT opti  T mini ÞðT s  T opti Þ  ðT opti  T maxi ÞðT opti þ T mini  2T s Þ

transitional compound, and ultimately stable humified materials.The kinetics for OM degradation can be explained by the following equation:

OMfi ¼



qDM  1  mDM qDM  mDM total total FAS ¼ 1   Gs  qH2 O qH2 O

kO2 ¼

where mOMi is mass of degradable OM with three different components, such as easily degradable OM content, slower (‘‘moderate’’) degradable OM which is cellulosic constituent and hardly degradable which is lignin content within the co-composting material. The mhum is the OM component that produced by the OM degradation process for example dead microbial cells,

ðT opti

ð11Þ

where volatile matter fraction is:

i¼1

fT i ¼

ð10Þ

FAS and specific gravity ðGs Þ are defined according to Haug (1993) as in Eq. (12) and Eq. (13), respectively.

Gv

Total OM can be represented by: 3 X ¼ mOMi þ mhum

1 expð23:675  FAS þ 3:4945Þ þ 1

kFAS ¼

Gs ¼ VM

2.4.1. Substrate degradation kinetics The total composting material is defined as follows:

mtotal ¼ mOM þ mH2 O þ mash

1   MC exp 17:684  100 þ 7:0622 þ 1

F1 ¼

ð17Þ

There are three sets of temperature correction function, each for different OM degradation constituents. Each set has three temperature points based on the cardinal temperatures for microbial growth; minimum, optimum and maximum temperature. Three sets of temperature correction functions were used based on the assumption that different group of OM constituent is not degraded simultaneously, rather degraded one at a time. Microorganism within the composting material will consume easily degradable OM (OMi¼1 ) first, during the thermophilic phase. When easily degradable OM starts to deplete, the microorganism only then begins to utilize moderately degradable OM (OMi¼2 ) during the mesophilic phase, utilizing hardly degradable OM only after this moderately degradable OM becomes exhausted during the later stage of composting (curing phase). 2.4.3. Gas balance Air inside composting system is divided into four components which are CO2, O2, N2 and H2O vapor. The total mass of gas, mgas therefore constitute the mass for each of those four species (Eq. (18)), with mass of each component is represented by Eq. (19) and Eq. (20).

mgas ¼

3 X mj þ mvap H2 O

ð18Þ

j¼1

mj ¼ MWj 

Pj  V g R  Tg

ð19Þ

432

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

Pvap H2 O  V g

mvap H2 O ¼ MWH2 O 

ð20Þ

R  Tg

dmH2 O _ bio _ Fin _ vap _ cond _ leach ¼m H2 O þ mH2 O  mH2 O þ Y cond  mH2 O  mH2 O dt

The model of gas balance within the co-composting process can be described as follows:

dmvap H2 O

F out  qj dmj _ intake _ bio ¼m þm  cj  j j dt 100

dmloss H2 O

ð21Þ

dt

Mass rate of each gas via intake air is as follows:

_ intake m j

¼ fj  F in  qj

dt

ð22Þ

ð33Þ

_ intake _ vap _ ext _ cond ¼m H2 O þ mH2 O  mH2 O  mH2 O

ð34Þ

_ cond _ leach ¼ ð1  Y cond Þ  m H2 O þ mH2 O

ð35Þ

When condensation process occurs, part of the condensate will return back into the composting material (represent as _ cond Y cond  m H O ) while the remaining part will goes into the bottom of 2

For evolution rate of gas species by microbial aerobic process, they can be explained per the following:

MWj¼1 _ bio m j¼2 MWj¼2

_ bio m j¼1 ¼ 

_ bio m j¼2 ¼ Y O2 

ð23Þ _ vap m H2 O ¼

3 X

r OMi

_ bio m j¼3 ¼ 0 mj

 100

ð25Þ



b WHC _ leach 1  ¼ k  ; MC > WHC m leach H2 O MC

q_ s

qwet air

ð37Þ

ð26Þ

Condensation occurs when temperature of air inside bioreactor (Tg) is higher than ambient temperature (T ambient ). Mass rate of condensate is measured as follows:

ð27Þ

_ cond m H2 O ¼

Exhaust flow rate is calculated based on the following:

F out ¼

ð36Þ

When MC is higher than compost water holding capacity (WHC), there will be leachate run off. Mass rate of leachate is defined as follows:

The concentration of gas in composting air is calculated as:

qj  V g

Q_ trans ; Ts > Tg DHvap

ð24Þ

i¼1

cj ¼

the composting system, leachate collection system (represent as _ cond ð1  Y cond Þ  m H2 O ).Water vaporization will happen when Ts is higher than Tg. Mass rate of water vaporization, can be defined as follows:

Q_ ambient ; T g > T ambient DHcond

ð38Þ

When pressure inside composting system, P is higher than surrounding (Patm ), there will be an outflow of air from the system. This can be modeled by applying a non-choked, subsonic gas flow equation (Van den Bosch and Weterings, 2005):

The enthalpy of water condensation, DHcond is equal but opposite to the value of enthalpy of water vaporization, DHvap :DHvap was correlated based on data that obtained from Çengel and Ghajar (2011) as in Eq. (39) as follows:

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u

ccþ1 u 1 2 t qs ¼ 3600  C d  Aout  W  1000  qwet  P  c  air cþ1

DHvap ¼ 2500:8  0:0016  ðT g  273Þ2  2:2789  ðT g  273Þ

_

Mass rate of water in supplied air is described using the following equation:

With the outflow coefficient factor, W2 is as follows:

2 W ¼  c1 2

"

ccþ1

2

c1 !# 1 Patm c Patm c   1 2 P P

cþ1

ð29Þ

With P is the sum of partial pressures of CO2, O2, N2 and also vapor pressure of water, which is described in Eq. (30) as follows: 3 X P¼ Pj þ Pvap H2 O

ð39Þ

ð28Þ

ð30Þ

_ intake m H2 O ¼ 0:03  RH  F in

ð40Þ

While water exit rate in the form of water vapor through the exhaust pipe is as follows:

_ ext m H2 O ¼ P3

mvap H2 O

j¼1 mj

þ mvap H2 O

 qwet air  F out

ð41Þ

j¼1

Total volume of gas within the bioreactor is defined as follows:

Vg ¼ Vr  Vc

ð31Þ

where

Vc ¼

3 X mOM

i

i¼1

qOMi

þ

mhum

qhum

þ

mash

qash

þ

mH2 O

qH2 O

ð32Þ

2.4.5. Thermal balance There will be two sets of energy balance, energy balance in gas phase and energy balance in solid phase of composting material as introduced in Eq. (42) and Eq. (43), respectively.

C wet air 

dT g ¼ Q_ intake  Q_ exhaust  Q_ ambient þ Q_ trans dt

C material  2.4.4. Water balance Water in the composting process is divided into three different species: the mass of water (mH2 O ) within the composting material that representing compost MC; mass of water in the form of water vapor (mvap H2 O ); and finally the mass of water lost in the form of both condensate and leachate (mloss H2 O ) as described in Eqs. 33–35, respectively.

dT s _ ¼ Q_ bio þ Q_ feed H2 O  Q trans dt

ð42Þ

ð43Þ

where wet C wet air ¼ cpair  mgas

ð44Þ

and

C material ¼ cpOM  mOM þ cpH2 O  mH2 O þ cpash  mash

ð45Þ

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

433

Fig. 2. Changes in (a) mass, (b) temperature, (c) oxygen level, (d) MC, (e) pH, (f) EC level and (g) TGA and DTG curves of the composting materials, during co-composting process. Arrows indicate water addition during co-composting process for AR1 and AR2.

Heat transfer rate to the environment is:

  Q_ ambient ¼ U  Ac  T g  T ambient

ð46Þ

wet Q_ exhaust ¼ cpwet air  T g  qair  F out

While heat rate from supplied air is simplified as:

Q_ intake ¼ cpair  T ambient  qair  F in

Heat rate that released from the composting system via the exhaust pipe is defined with the following equation:

ð47Þ

The density of wet air is calculated as:

ð48Þ

434

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

P3

qwet air ¼

j¼1

vap qj  mj þ qvap H2 O  mH2 O

P3

j¼1 mj

þ

ð49Þ

mvap H2 O

where

qvap H2 O ¼ MWH2 O 

Pvap H2 O

ð50Þ

R  Tg feed

Heat rate due to water addition, dotQ H2 O is defined as: feed _ Fin Q_ feed H2 O ¼ cpH2 O  mH2 O  ðT H2 O  T s Þ

ð51Þ

Heat rate generated by biological reaction, dotQ bio is described as follows:

Q_ bio ¼ DHbio 

3 X r OMi

ð52Þ

i¼1

Heat rate transfer between composting solid material and air inside the system, Q_ trans is defined as follows:

As Q_ trans ¼ kw   ðT s  T g Þ FAS

ð53Þ

The specific heat capacity of wet air is calculated as:

P3 cpwet air ¼

vap  mj þ cpvap H2 O  mH2 O P3 vap j¼1 mj þ mH2 O

j¼1 cpj

ð54Þ

where the specific heat capacity for water vapor was correlated from data that obtained from Çengel and Ghajar (2011): 7 cpvap  ðT g  273Þ3 þ 2:2  106  ðT g  273Þ2 H2 O ¼ 1  10

þ 5:359  104  ðT g  273Þ þ 1:8545

ð55Þ

2.4.6. Modeling performance and sensitivity analysis Relative root mean squared error (rRMSE) was used to evaluate modeling performance by calculating sum of differences between  predicted and measured data for seven profiles cj¼2 ; mtotal ; mOM ; mOMi¼2 ; mOMi¼3 ; MCandTsÞ. The rRMSE is calculated using the following formula:

1 rRMSE ¼ 100   A

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 Xn ðOr  Pr Þ2 r¼1 n

separate stages, i.e., early mesophilic stage where temperature rose from 30 °C to 50 °C, then thermophilic stage where temperature continues to build up until 70 °C before cooling down into late mesophilic i.e., cooling stage with a temperature around 30–40 °C. All experiments, except for AR1 had a very short thermophilic stage, which is around two days. The results indicated that at high aeration rate contributes to a short thermophilic phase. Lower aeration rate (AR1) retained a longer thermophilic stage, although the temperature achieved was not high like others (58 °C as compared to around 67 °C, average for other aeration rates). This was mainly due to the lower oxygen concentration reaching levels lower than 5% (Fig. 2c) which was not adequate enough for microbial consumption, thus limiting the aerobic process and generating less heat. The oxygen concentration as well as temperature profile for batch P2, i.e., AR3 and AR4 show very similar trend. It also demonstrates that by providing higher aeration rate in this particular composting mixture is not necessary as oxygen concentration already achieved maximum level (equal to ambient oxygen level) as early as fifth days in both AR3 and AR4 aeration rates. Throughout the composting process, adequate level of oxygen concentration i.e., in the range of 10–18% (Petric et al., 2012) was achieved for other aeration rates. Oxygen level increases over the time due

ð56Þ

Sensitivity analysis was performed to study the impact of different aeration rates on the model based on generated temperature and OM degradation profiles. The estimated parameters in Table 3 were used as default parameters using five different aeration rates, which were 0.09 L min1 kg1 DM (25% lower than AR1), 0.18, 0.36, 0.64 and 0.90 L min1 kg1 DM (25% higher than AR4). Those five different aeration rates were labeled from SA1 to SA5, respectively. Parameter estimation and dynamic simulation were performed using gPROMs software (Process Systems Enterprise Ltd, UK).

Table 1 Values for TOC, TKN and C/N ratio. Aeration rate

AR1 AR2 AR3 AR4

TOC (g/kg)

TKN (g/kg)

C/N ratio

Initial

Final

Initial

Final

Initial

Final

533 533 523 523

509 501 494 495

10.8 10.8 14.9 14.9

18.8 20.1 18.8 18.5

49.4 49.4 35.1 35.1

27.1 25.0 26.3 26.8

OM loss (%)

29.3 40.5 30.1 35.6

Table 2 Values for coefficient and parameters used in the model. Parameter

Value

Reference

qOM c

450 1.4 1.32 0.848 1 2 293 283 277 3.96 70 0.95 0.00039 0.20 0.77

This study Çengel and Ghajar (2011) Petric and Selimbasic´ (2008) Petric and Selimbasic´ (2008) Haug (1993) Haug (1993) This study This study This study This study This study This study This study This study This study

cpOM cpash Gv Gf T mini¼1 T mini¼2 T mini¼3 U WHC RH fi¼1 fi¼2 fi¼3

3. Results and discussion 3.1. Changes in compost chemical and physical characteristic Fig. 2 shows changes in total mass, temperature, oxygen level, MC, pH and EC level throughout the composting process. There is significant decrease in total amount of mass (Fig. 2a) and this was not due to the decrease in MC since moisture content did not decrease towards the end of the process (Fig. 2d). The reduction of total mass was due to the degradation of simple substrates and nutrients (‘‘easy’’ OM fraction) within the compost materials, leaving more complex and recalcitrant OM (‘‘moderate’’ and ‘‘hard’’ OM constituents). The temperature profile (Fig. 2b) shows a typical compost temperature pattern, which can be divided into three

Table 3 Result of parameter estimation. Parameter

Value

Parameter

Value

DH b Cd k0i¼1 k0i¼2 k0i¼3 kleach kw ni¼1 ni¼2

2634.7 0.26 0.0016 16.89 1.96 0.18 0.84 2.81 64.67 489.52

ni¼3 T max i¼1 T max i¼2 T max i¼3 T opti¼1 T opti¼2 T opti¼3 Ycond Yhum Y O2

545.10 342 324 314 331 318 309 0.13 0.74 0.60

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

435

Fig. 3. Experiment (dots) and predicted (lines) data of (i) OM degradation (mOM and mOMi) of composting materials at different aeration rates (a) AR1, (b) AR2, (c) AR3 and (d) AR4, and (ii) other examples such as mass (mtotal), temperature (Ts), MC and O2 (cj=2) (only for AR2).

to constant supply of air and also because of degradation process at later stage is not as intense as the thermophilic stage in which biological oxygen demand is not as high as the first part of the composting process. MC for batch P2 was relatively stable during the experiment whereas batch P1 shows increase over the time (Fig. 2d). Batch

P1 was added with water due to the fact that the initial MC was low (57%) unlike batch P2 which was around the optimal level (within 70%). The design of the bioreactor which has small exhaust pipe (15 mm in diameter) helps to minimize water loss through exhaust pipe in the form of water vapor particularly after thermophilic phase ended where condensate was returned back into the

436

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

compost mixtures. Humidified inlet air also contributed to a stable moisture level. Although MC remains stable throughout composting process and remains higher than WHC, large portion of leachate run-off occurred only during the early composting period which is within thermophilic stage and less leachate was observed on the later stage. This is particularly due to the fact that condensation process of water vapor since temperature of air inside composting system is higher than ambient temperature during the thermophilic stage of composting process. The initial pH (Fig. 2e) for both batches were slightly higher than neutral (7.10 and 7.60 for P1 and P2, respectively) and increased during the thermophilic phase, indicating that organic acids (e.g., uric acid) were decomposed and nitrogenous compounds such as proteins and amino acids were degraded and converted into inorganic nitrogen in the form of ammonia (Guo et al., 2012; Liu et al., 2011; Petric et al., 2012). Electrical conductivity (EC) indicates the level of soluble salts within the compost mixture. Fig. 2f shows that the level of EC for all different aeration rates are stable in the range of 0.99–1.66 dS m1. This is because of high moisture level reduced the salinity of the compost materials, therefore reducing the EC value. Fig. 2g shows TGA and DTG curves of the composting material for initial (P1 Day 0 and P2 Day 0) and Day 20 of the composting period. There are three major weight losses due to thermal decomposition, which can be seen clearly on the DTG graph (I, II and III). The first stage occurred from 50 °C to 150 °C which constitutes up to 10% weight loss. This is due to the water evolution and also other volatile substances (Pasangulapati et al., 2012; Yang et al., 2007). The second, which is the major decomposition happened from 250 °C to 400 °C. This decomposition stage resulted in a major weight loss around 45–50% for P1 and P2 Day 0, and to lesser extent, around 35–45% for composting materials at Day 20. These major losses constitute primarily due to the degradation of hemicellulose and cellulose content within the biomass (Carrier et al., 2011; Omar et al., 2011; Pasangulapati et al., 2012; Yang et al., 2007). This result was proven with the data obtained from proximate analysis of cellulose content, indicating that cellulosic materials on Day 20 is lower than at Day 0, due to degradation of OM material. The last stage of thermal degradation occurs from 450 °C to 600 °C. This constitutes about 20% losses in biomass weight, which mainly due to the degradation of lignin composition (Omar et al., 2011) and structural hemicellulose (Lyons et al., 2006). The degradation curves for Day 0 has different slope than at Day 20 for all aeration rates. This may due to degradation of structural hemicellulose and the result does not show any significant lignin degradation since proximate analysis shows that lignin content does not change throughout composting period and also because lignin is degraded slowly on a broad temperature range (Barneto et al., 2010). Lower aeration rates in batch P1 (AR1 and AR2) show higher reduction in C/N ratio as compared to higher aeration rates in batch P2 (AR3 and AR4) as tabulated in Table 1. The highest C/N ratio reduction was observed at AR2 aeration rate, from 49.4 to 24.9 decreases by 49.6%, while the lowest reduction was observed at AR4 aeration rate with 23.6% decrease, from 35.1 to 26.8. Lower aeration rate, coupled with lower temperature profile could almost maintain the content of inorganic N. A higher aeration rate, coupled with higher temperature profile enhanced the loss of inorganic N due to volatilization, contributing to lower TKN levels throughout the process. The rate of dry matter loss as CO2 and water were faster than the rate of N loss in the form of NH3 in lower aeration rates (Guo et al., 2012), resulting in greater C/N reduction as compared to higher one. One of the important aspects to look when assessing OM degradation process is the loss of OM. Table 1 shows that the lowest OM loss is achieved by AR1 aeration rate due to the insufficient oxygen

as discussed above. However, the AR2 aeration rate had the highest OM loss (40.5%). This result generally agrees with other findings such as by Gao et al. (2010) and Guo et al. (2012) that conclude lower (with adequate oxygen level) aeration rates proved to be better than higher one. Both AR3 and AR4 aeration rates had 30.1% and 35.6% in term of OM losses, respectively. At higher aeration rates, thus higher energy transfer to the exit of composting system, resulting lower OM degradation due to heat in composting material that is necessary for microbes was lower especially during the mesophilic or curing stage. However, during thermophilic stage, higher aeration rates seem better due to high degradation rate of simple OM that required higher oxygen supply. 3.2. Modeling and parameter estimation Due to the complexity of nitrogen cycle dynamic, the biological consumption and transformation (nitrification, denitrification) of nitrogen were not considered into this study. It has been simplified by considering only one species of nitrogen gas and only taking into account the intake and outflow of it in the composting system. Among four correction functions applied, kFAS had the least impact on the rate of the total OM degradation (data not shown). FAS was not a limiting factor because of the nature of the composting material itself, with a high lignocellulosic content that contributed to the low bulk density of total material hence high FAS value. Table 2 shows the parameter values that had been used for model development and estimating other unknown parameters. Unknown parameter values were estimated (Table 3) by fitting the models with experimental results using gPROMS. Fig. 3 shows a comparison between experimental data and simulation profile for OM content, mass, temperature, MC and oxygen level within the composting material throughout co-composting period. The estimated parameters for optimum and maximum temperature are generally higher than those found in literature as expected, due to the higher average temperature of a tropical climate that has higher average temperature than temperate climate, which provides a higher microbial temperature range. 3.2.1. Modeling performance As shown in Table 4, the moderately OM constituent ðmOMi¼2 Þ was poorly predicted with a rRMSE value higher than 15%. The hardly degradable OM constituent ðmOMi¼3 Þ also less successfully predicted even though the rRMSE value was not as high as OMi¼2 . The complexity of lignin-cellulose-hemicellulose superstructure leads to different rates of substrate degradation by microorganisms (Bertrand et al., 2006; Eklind and Kirchmann, 2000; Zhang et al., 2012). This makes it difficult to model and predict the degradation rates of that particular OM constituents especially for mOMi¼2 . This overestimation of OM constituents affecting the prediction of total OM (mOM ) although AR3 showed a good fit between measured and predicted data with rRMSE value equals 6.23%. Other profiles (cj¼2 ; mtotal , MC and Ts) show good fit, with exception to the oxygen profile for AR1 which was poorly predicted (rRMSE equals 49.85%). Table 4 Relative root mean squared error (rRMSE) analysis for accuracy of the model used. Profile

cj¼2 mtotal mOM mOMi¼2 mOMi¼3 MC Ts

rRMSE (%) AR1

AR2

AR3

AR4

49.9 0.86 17.7 17.8 13.9 9.32 5.00

15.1 1.26 14.6 18.8 11.2 7.89 13.4

8.28 2.17 6.23 29.6 20.3 4.15 13.2

2.80 2.40 11.1 26.3 12.2 4.95 12.3

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

437

Fig. 4. Result for sensitivity analysis of different aeration rates on (a) Ts and (b) mOM .

3.2.2. Sensitivity analysis Fig. 4 depicts the result of sensitivity analysis of different aeration rates against temperature and total OM degradation. SA1 had the lowest temperature peak (50 °C) achieved after a 140 h composting period, followed by SA2 (63 °C) after 70 h. SA3, SA4 and SA5 aeration rates had similar temperature peaks at 68 °C achieved within 15 to 30 h of composting. The same trends were found for OM degradation, in which high aeration rates (SA3–SA5) showed no significant difference in OM reduction. For example at 45 h, where temperature drop rate began to decelerate slowly, SA3, SA4 and SA5 showed 11.8%, 12.4% and 13.2% OM reduction, respectively. During the same period, SA1 had 5.26% while SA2 showed 8.42% of OM reduction. This implies that temperature and OM degradation were affected more at lower aeration rates (SA1 and SA2) than at higher rates (SA3–SA5). This is because of the rate of biological oxygen demand is higher than the rate of oxygen supplied that negatively affected on the total OM degradation process, resulting in a lower and delayed thermophilic phase, unlike higher aeration rates. 4. Conclusion The study shows that at the early stage of composting process, lower aeration rates have a greater effect on the OM degradation than at higher levels due to insufficient oxygen supply, although higher aeration rates demonstrate high energy transfer and increase the loss of heat from the composting system. OM degradation during the active phase of co-composting process with aera1 1 tion rate at 0:26L min kgDM , was superior to other aeration rates and resulted in greater OM loss (40.5% reduction). Sensitivity analysis demonstrated that lower aeration rates have a greater effect on the rate of OM degradation than higher rates. Acknowledgements This work was supported by Fundamental Research Grant Scheme (03-04-10-792FR) and Long-Term Research Grant Scheme (600-RMI/LRGS 5/3) by Ministry of Higher Education, Malaysia. References Baharuddin, A.S., Hock, L.S., Yusof, M.Z., Abdul, N.A., Shah, U., Hassan, M.A., Wakisaka, M., Sakai, K., Shirai, Y., 2010. Effects of palm oil mill effluent (POME) anaerobic sludge from 500 m3 of closed anaerobic methane digested tank on pressed-shredded empty fruit bunch (EFB) composting process. Afr. J. Biotechnol. 9, 2427–2436. Barneto, A.G., Ariza Carmona, J., Díaz Blanco, M.J., 2010. Effect of the previous composting on volatiles production during biomass pyrolysis. J. Phys. Chem. A 114, 3756–3763. http://dx.doi.org/10.1021/jp903994p.

Bertrand, I., Chabbert, B., Kurek, B., Recous, S., 2006. Can the biochemical features and histology of wheat residues explain their decomposition in soil? Plant Soil 281, 291–307. http://dx.doi.org/10.1007/s11104-005-4628-7. Carrier, M., Loppinet-Serani, A., Denux, D., Lasnier, J.-M., Ham-Pichavant, F., Cansell, F., Aymonier, C., 2011. Thermogravimetric analysis as a new method to determine the lignocellulosic composition of biomass. Biomass Bioenergy 35, 298–307. http://dx.doi.org/10.1016/j.biombioe.2010.08.067. Çengel, Y.A., Ghajar, A.J., 2011. Heat and Mass Transfer: Fundamentals and Applications, Fourth ed. McGraw Hill Higher Education, New York. Diaz, M.J., Madejon, E., Lopez, F., Lopez, R., Cabrera, F., 2002. Optimization of the rate vinasse/grape marc for co-composting process. Process Biochem. 37, 1143– 1150. Eklind, Y., Kirchmann, H., 2000. Composting and storage of organic household waste with different litter amendments. I: carbon turnover. Bioresour. Technol. 74, 115–124. http://dx.doi.org/10.1016/S0960-8524(00)00004-3. Gao, M., Li, B., Yu, A., Liang, F., Yang, L., Sun, Y., 2010. The effect of aeration rate on forced-aeration composting of chicken manure and sawdust. Bioresour. Technol. 101, 1899–1903. http://dx.doi.org/10.1016/j.biortech.2009.10.027. Gomes, A.P., Pereira, F.A., 2008. Mathematical modelling of a composting process, and validation with experimental data. Waste Manage. Res. 26, 276–287. http:// dx.doi.org/10.1177/0734242X07086514. Guo, R., Li, G., Jiang, T., Schuchardt, F., Chen, T., Zhao, Y., Shen, Y., 2012. Effect of aeration rate, C/N ratio and moisture content on the stability and maturity of compost. Bioresour. Technol. 112, 171–178. http://dx.doi.org/10.1016/ j.biortech.2012.02.099. Haug, R., 1993. The Practical Handbook of Compost Engineering. Lewis Publishers, Florida. Higgins, C.W., Walker, L.P., 2001. Validation of a new model for aerobic organic solids decomposition: simulations with substrate specific kinetics. Process Biochem. 36, 875–884. http://dx.doi.org/10.1016/S0032-9592(00)00285-5. Kulcu, R., Yaldiz, O., 2004. Determination of aeration rate and kinetics of composting some agricultural wastes. Bioresour. Technol. 93, 49–57. http://dx.doi.org/ 10.1016/j.biortech.2003.10.007. Liu, D., Zhang, R., Wu, H., Xu, D., Tang, Z., Yu, G., Xu, Z., Shen, Q., 2011. Changes in biochemical and microbiological parameters during the period of rapid composting of dairy manure with rice chaff. Bioresour. Technol. 102, 9040– 9049. http://dx.doi.org/10.1016/j.biortech.2011.07.052. Lyons, G.A., Sharma, H.S.S., Kilpatrick, M., Cheung, L., Moore, S., 2006. Monitoring of changes in substrate characteristics during mushroom compost production. J. Agric. Food Chem. 54, 4658–4667. http://dx.doi.org/10.1021/jf052934i. Mason, I.G., 2007. A Study of Power, Kinetics, and Modelling in the Composting Process (Ph.D. Thesis). University of Canterbury, Christchurch, New Zealand. Mohammad, N., Alam, M.Z., Kabbashi, N.A., Ahsan, A., 2012. Effective composting of oil palm industrial waste by filamentous fungi: a review. Resour. Conserv. Recycl. 58, 69–78. http://dx.doi.org/10.1016/j.resconrec.2011.10.009. MPOB, 2013. FFB Processed by Mill 2013 [WWW Document]. URL (accessed 3.12.14). Omar, R., Idris, A., Yunus, R., Khalid, K., Aida Isma, M.I., 2011. Characterization of empty fruit bunch for microwave-assisted pyrolysis. Fuel 90, 1536–1544. http://dx.doi.org/10.1016/j.fuel.2011.01.023. Pasangulapati, V., Ramachandriya, K.D., Kumar, A., Wilkins, M.R., Jones, C.L., Huhnke, R.L., 2012. Effects of cellulose, hemicellulose and lignin on thermochemical conversion characteristics of the selected biomass. Bioresour. Technol. 114, 663–669. http://dx.doi.org/10.1016/j.biortech.2012.03.036. Petric, I., Avdihodz, E., Helic, A., Helic´, A., Avdic´, E.A., 2012. Evolution of process parameters and determination of kinetics for co-composting of organic fraction of municipal solid waste with poultry manure. Bioresour. Technol. 117, 107– 116. http://dx.doi.org/10.1016/j.biortech.2012.04.046. Petric, I., Selimbasic´, V., 2008. Composting of poultry manure and wheat straw in a closed reactor: optimum mixture ratio and evolution of parameters. Biodegradation 19, 53–63. http://dx.doi.org/10.1007/s10532-007-9114-x. Pommier, S., Chenu, D., Quintard, M., Lefebvre, X., 2008. Modelling of moisturedependent aerobic degradation of solid waste. Waste Manage. 28, 1188–1200. http://dx.doi.org/10.1016/j.wasman.2007.05.002.

438

A.T. Talib et al. / Bioresource Technology 169 (2014) 428–438

Ponsá, S., Puyuelo, B., Gea, T., Sánchez, A., 2011. Modelling the aerobic degradation of organic wastes based on slowly and rapidly degradable fractions. Waste Manage. 31, 1472–1479. http://dx.doi.org/10.1016/j.wasman.2011.02.013. Rosso, L., Lobry, J.R., Flandrois, J.P., 1993. An unexpected correlation between cardinal temperatures of microbial growth highlighted by a new model. J. Theor. Biol. 162, 447–463. http://dx.doi.org/10.1006/jtbi.1993.1099. Sulaiman, F., Abdullah, N., Gerhauser, H., Shariff, A., 2011. An outlook of Malaysian energy, oil palm industry and its utilization of wastes as useful resources. Biomass Bioenergy 35, 3775–3786. http://dx.doi.org/10.1016/j.biombioe.2011. 06.018. Van den Bosch, C., Weterings, R. (Eds.), 2005. Methods for the Calculation of Physical Effects: Due to Releases of Hazardous Materials (liquids and gases), third ed. Committee for the Prevention of Disasters, The Hague, NL.

Yahya, A., Sye, C.P., Ishola, T.A., Suryanto, H., 2010. Effect of adding palm oil mill decanter cake slurry with regular turning operation on the composting process and quality of compost from oil palm empty fruit bunches. Bioresour. Technol. 101, 8736–8741. http://dx.doi.org/10.1016/j.biortech. 2010.05.073. Yang, H., Yan, R., Chen, H., Lee, D.H., Zheng, C., 2007. Characteristics of hemicellulose, cellulose and lignin pyrolysis. Fuel 86, 1781–1788. http:// dx.doi.org/10.1016/j.fuel.2006.12.013. Zhang, Y., Lashermes, G., Houot, S., Doublet, J., Steyer, J.P., Zhu, Y.G., Barriuso, E., Garnier, P., 2012. Modelling of organic matter dynamics during the composting process. Waste Manage. 32, 19–30. http://dx.doi.org/10.1016/j.wasman.2011. 09.008.

Effects of aeration rate on degradation process of oil palm empty fruit bunch with kinetic-dynamic modeling.

The effect of different aeration rates on the organic matter (OM) degradation during the active phase of oil palm empty fruit bunch (EFB)-rabbit manur...
1MB Sizes 1 Downloads 4 Views