Bioresource Technology 341 (2021) 125759

Contents lists available at ScienceDirect

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

Food waste digestate composting: Feedstock optimization with sawdust and

mature compost
Bing Song a, 1, M.K. Manu a, 1, Dongyi Li a, Chen Wang a, b, Sunita Varjani c, Narsi Ladumor d,
Lui Michael e, Yunjie Xu f, Jonathan W.C. Wong a, f, *
a
Institute of Bioresource and Agriculture, Sino-Forest Applied Research Centre for Pearl River Delta Environment and Department of Biology, Hong Kong Baptist
University, Hong Kong
b
College of Land Science and Technology, China Agricultural University, Beijing 100193, China
c
Gujarat Pollution Control Board, Gandhinagar 382 010, Gujarat, India
d
OSCAR Bioenergy Joint Venture, Hong Kong
e
Environmental Protection Department, Hong Kong
f
School of Technology, Huzhou University, Huzhou 311800, China

H I G H L I G H T S

• Food waste digestate can be effectively co-composted with sawdust and mature compost.
• NH4+-N toxicity could be reduced to permissible limit within 2 weeks of composting.
• Good quality compost with SGI > 80% could be produced within 2 weeks of composting.
• Sawdust and mature compost amendments enhanced decomposition rate.
• Significant reduction in NH3 emission was observed during co-composting.

A R T I C L E I N F O A B S T R A C T

Keywords: Direct land application of food waste digestate (FWD) leads to 60–70% of nitrogen loss through NH3 volatili­
Food waste digestate zation due to its innate characteristics like high ammonium nitrogen (NH4+-N) (~6000 mg/kg dry matter) and
Composting high moisture content (~75%). Hence, bio stabilization of FWD through composting is a promising solution to
NH3 volatilization
curb the environmental and occupational hazards. Hence the aim of this study was to assess the feasibility of
Ammonium nitrogen
Compost quality
using sawdust and/or mature compost as a bulking agent to achieve effective composting. The results showed
that mixing of FWD with sawdust alone or together with mature compost could produce quality compost with
reduced NH4+-N (<700 mg/kg dry matter) and increased seed germination index (>80%) within 2 weeks of co-
composting. Composting FWD with both sawdust and mature compost effectively reduced ~ 83% of NH3
volatilization demonstrating that this approach can effectively produce mature nitrogen enriched FWD compost.

1. Introduction into bioenergy (Kaur et al., 2020). The residue produced after the AD
process is referred as digestate, which requires further wastewater
Rapid urban growth and change in lifestyle has increased the global treatment before disposal or reuse. The solid fraction obtained after
production of food waste which not just requires careful disposal strat­ dewatering is the food waste digestate (FWD) which comprises of
egy to reduce its environmental impact but at the same time to reutilize partially degraded organic matter, high NH4+-N and minerals (Manu
the residual energy and nutrient resource in food waste. Among the et al., 2021a). NH4+-N contents in the range 2540–7200 mg/kg dry
available technologies, anaerobic digestion (AD) has evolved as a sus­ matter and moisture content of 70–96% were reported for FWD from
tainable carbon negative technology for the conversion of food waste various countries irrespective of their food waste type (Manu et al.,

* Corresponding author at: Department of Biology, Hong Kong Baptist University, Hong Kong.
E-mail address: [email protected] (J.W.C. Wong).
1
Equal contribution.

https://doi.org/10.1016/j.biortech.2021.125759
Received 21 June 2021; Received in revised form 8 August 2021; Accepted 9 August 2021
Available online 12 August 2021
0960-8524/© 2021 Elsevier Ltd. All rights reserved.
B. Song et al. Bioresource Technology 341 (2021) 125759

2021a). Due to these innate characteristics of FWD, 60–70% of NH3 Table 1

emission and phytotoxic effects on plants may occur upon its direct land Physico-chemical properties of the different raw materials.
application (Rincón et al., 2019). Hence, further bio-stabilization Parameters FWD Sawdust Mature compost
through composting process is required to alleviate the raw FWD
Moisture content (%) 72.5 ± 0.4 11.5 ± 0.1 27.5 ± 0.3
properties into beneficial soil amendment in environmentally sustain­ pH 8.3 ± 0.2 5.3 ± 0.1 7.1 ± 0.1
able manner to meet the circular bio-economy criteria (Lu and Xu, 2021; EC (mS/cm) 11.8 ± 0.4 0.3 ± 0.1 3.1 ± 0.1
Manu et al., 2021a; Peng and Pivato, 2019). In view of the increasing Total organic carbon (TOC) (%) 34.4 ± 0.3 45.4 ± 0.1 29.8 ± 0.1
FWD generation around the world, it warrants more effort to develop Total Nitrogen (TKN) (%) 4.5 ± 0.2 0.3 ± 0.05 3.3 ± 0.1
C/N ratio 7.7 ± 0.1 151.3 ± 0.1 9.1 ± 0.2
effective composting process for FWD treatment (Manu et al., 2021a). NH4+-N (mg/kg DM) 6717 ± 217 – –
Hong Kong commenced its first biological waste treatment facility
‘Organic Resource Recovery Centre 1 (ORRC1)’ in 2018 for converting
200 tpd of food waste into biogas by AD followed by composting of compost is easily available and can replace part of the sawdust as a
FWD. ORRC1 composting facility has experienced lengthened com­ drying agent and also as a microbial inoculum. Use of sawdust and
posting process due to poor composting properties of FWD resulting in a mature compost individually for co-composting has been previously
long retention time and a larger footprint for the composting facility. reported, but their efficiency in combination has not been explored. The
This warrants detailed study to develop effective composting strategy experimental design consisted of four treatments to study the feasibility
for better management of FWD composting which could be adopted to of using sawdust and mature compost as bulking agents in FWD co-
any FWD composting facilities in any country irrespective of their food composting (Table 2). The compost mixtures were prepared as fol­
waste type since the FWD characteristics in many countries are similar. lows: FWD only with C/N 7.7 and moisture content 73% (R1 Control);
Despite of growing interest in AD technology, FWD composting has FWD and mature compost only with C/N 8.8 and moisture content 54%
received limited focus. In the study conducted by Zeng et al. (2012), NH3 (1:2.6 dry weight basis) (R2); FWD and sawdust only with C/N 29 and
emission was 10 times higher from FWD than fresh food waste within 10 moisture content 55% (1:8.3 dry weight basis); FWD, sawdust and
days of composting due to higher ammonia/ammonium, hence, co- mature compost with C/N 29 and moisture content 56% (1:5.3:2.6 dry
composting with bulking agents having high C/N ratios was recom­ weight basis).
mended for enhancing nitrification thereby mitigating NH3 emission. In Around 7 kg of properly mixed feedstock was added into each
a recent study, mixing of 62% FWD with 28% OFMSW and 10% sawdust reactor. Atmospheric air comprising 21% of oxygen and 78% of nitrogen
provided optimized composting conditions to produce quality compost was pumped at an air pressure of 0.1 MPa from the bottom of the re­
within two weeks (Espinosa-Salgado et al., 2020). Arab and McCartney actors achieving an aeration rate of 0.25 L/kg/hr which was maintained
(2017) co-composted FWD with organic fraction of MSW and found that throughout the process. Fig. 1 shows the schematic representation of
20–40% FWD content can result in enhanced degradation. Zeng et al. composting set up of this study.
(2016) used wood chips and plastic pall rings as bulking agents to un­
derstand the nitrogen dynamics during FWD co-composting process.
2.2. Sampling and analyses
Addition of woodchips as a bulking agent could increase the C/N ratio,
absorb water and adsorb NH4+/ NH3 to mitigate NH3 and N2O pro­
Around 200 g of sample was collected from each reactor after thor­
duction by 98% and 64%, respectively. These studies show the need of
ough mixing on days 0, 3, 7, 11, 15, 22 and 29 for physico-chemical and
suitable bulking agents for mitigating unwanted gaseous emissions
biological analyses. pH, electrical conductivity (EC), total organic car­
during FWD composting.
bon (TOC), total kjeldahl nitrogen (TN) and ammonium nitrogen (NH4+-
Conserving the nitrogen in FWD could be achieved through devel­
N) were measured by standard methods (Chan et al., 2016). The emitted
oping appropriate composting conditions with the addition of carbon
NH3 gas was trapped into H2SO4 solution (Lim et al., 2017) and
enriched bulking agents to produce nitrogen rich compost. Effective
composting of FWD could promote bio-circular economy and reduce the
Table 2
consumption of inorganic fertilizer thereby reducing the impact on
Feedstock composition and physico-chemical properties of the feedstock of the
climate change. Hence, in this study, a batch study was conducted in
various treatments.
bench scale reactors to develop a suitable compost mix consisted of a
combination of sawdust and mature compost for effective FWD com­ Parameters Treatment Treatment 2 Treatment 3 Treatment 4
1 (R1) (R2) (R3) (R4)
posting to produce mature compost in short duration by evaluating their
effect on carbon mineralization and nitrogen transformation. The out­ Feedstock Digestate Digestate Digestate Digestate,
only and compost and sawdust sawdust and
comes of the present study will provide an innovative approach to
compost
develop composting mix for FWD using sawdust and mature digestate Feedstock ratio: 1–0-0 1–0-1 1–2.6–0 1–1.7–1
compost. Digestate-
sawdust-
compost (fresh
2. Materials and methods
wt)
Feedstock ratio: 1–0-0 1–0-2.6 1–8.3–0 1–5.3–2.6
2.1. Feedstock and composting setup Digestate-
sawdust-
Dewatered FWD and mature digestate compost were collected from compost (dry
wt)
ORRC1, Hong Kong whereas sawdust was procured from Wong Yee Kee Moisture content 72.5 ± 0.4 53.5 ± 1.5 54.7 ± 1.2 56.2 ± 1.1
Co Ltd, Hong Kong. The characteristics of the feedstock materials are (%)
summarized in Table 1. Composting experiments were carried out in 20 pH 8.3 ± 0.2 8.1 ± 0.1 8.5 ± 0.2 7.9 ± 0.2
L bench scale stainless steel composting reactors (height: 55 cm, diam­ EC (mS/cm) 11.8 ± 0.4 8.5 ± 0.1 3 ± 0.3 2.9 ± 0.2
Total organic 34.4 ± 0.3 30.7 ± 0.2 43.4 ± 0.4 43.1 ± 0.3
eter: 25 cm) with thickness of 2 mm. The reactors were connected with a
carbon (%)
computer feedback control system for the online temperature mea­ Total Nitrogen 4.5 ± 0.2 3.5 ± 0.1 1.5 ± 0.2 1.5 ± 0.2
surement and ventilation control aeration (Chan et al., 2016). (%)
Sawdust and mature compost from ORRC1 were used as carbon C/N ratio 7.7 ± 0.1 8.8 ± 0.1 28.9 ± 0.2 28.7 ± 0.2
enriched bulking agents with the purpose to absorb moisture and to NH4+-N (mg/kg 6717 ± 217 4085 ± 231 2605 ± 134 1938 ± 117
DM)
provide seeding, respectively. Since sawdust is expensive, mature

2
B. Song et al. Bioresource Technology 341 (2021) 125759

Fig. 1. Schematic representation of compost reactors used in the study. (R1 = digestate control, R2 = digestate + compost, R3 = digestate + sawdust, R4 = digestate
+ sawdust + compost).

determined using spectrometric method whereas CO2 emission was time ‘t’.
measured by a WMA-2 CO2 gas analyzer (PP systems, Herts, UK). The
seed germination index (SGI) was determined by germinating Cress 2.3.3. Levi-Minzi kinetic model
seeds in 1:5 ratio (w:v) extract of compost and distilled water for 72 h In this model, an exponential kinetic model was used to express
(HKORC, 2005). All the analyses were conducted in duplicate, and carbon mineralization as (Levi-Minzi et al., 1990):
means and standard deviation values were represented in results.
Ct = ktm (3)

2.3. Kinetic analysis where Ct is the CO2-C emission (%), k and m are constants which rep­
resents units of the variables (mg/kg day) and shape of the curve,
The kinetic models namely, first-first-order, Chen and Hashimoto, respectively.
and Levi-Minzi models were used for fitting cumulative CO2 emission The constants in all three models were estimated using non-linear
(CO2-C) data to determine the degradation rates of rapidly and slowly regression technique after minimizing the sum of the squares of errors
biodegradable carbon content during FWD composting (Tosun et al., by Levenberg-Marquardt algorithm.
2008). Three statistical parameters such as, root mean square error (RMSE),
chi-square (χ2) and model efficiency (EF) were used to assess the models
2.3.1. The first-first-order kinetic model (Külcu and Yaldiz, 2004):
In this model, the rapid and slow decomposition phases follow first √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
√N
order kinetics and hence it is the sum of two first order equations that √∑
√ (CO2pre,i − CO2exp,i )2
can be expressed as follows: √
RMSE = I=1 (4)
N
C = Cr (1 − exp( − kr t) ) + Cs (1 − exp( − ks t)) (1)
∑n
i=1 (CO2exp,i− CO2pre,i )2
where C is cumulative CO2-C emitted (%) at time ‘t’, Cr and Cs are χ2 = (5)
N− n
percentage of rapidly and slowly decomposable carbon fractions,
respectively, kr and ks are rapid and slow degradation rate constants (∑n ( ( )2 ∑n )2
(day− 1), respectively.
2
CO2exp,i − CO2exp.mean − i=1 (CO2pre,i − CO2exp,i ).
EF = i=1
∑n 2
(6)
i=1 (CO2exp,i − CO2exp,mean,i )
2.3.2. Chen and Hashimoto kinetic model
The percent CO2-C evolution in this model could be expressed using where CO2exp = experimental CO2-C evolution (%), CO2exp,mean = mean
the above equation as follows (Chen and Hashimoto, 1980): value of experimental CO2-C evolution (%), CO2pre = predicted CO2-C
( ) evolution (%), N = number of observations and n = number of
CO2 − C = 100 − 100 × R +
(1 − R)K
(2) constants.
μm t − 1 + K

where R is the refractory coefficient, K is Chen and Hashimoto constant

and µm is maximum specific growth rate of microorganism (day− 1) at

3
B. Song et al. Bioresource Technology 341 (2021) 125759

2.4. Statistical analyses degree of decomposition was assessed based on CO2 evolution which
was higher in treatment R4 as compared to other treatments as discussed
The experimentally obtained parameter data were analyzed for sta­ in section 3.2. The thermophilic phase ensured pathogen destruction
tistical significance by analysis of variance (ANOVA) in SPSS 24.0 sta­ (Tchobanoglous et al., 1993; Manu et al., 2019). Similar temperature
tistical software to determine the significant differences using Turkey’s profile was observed during FWD co-composting with structuring ma­
significance difference test at the confidence level of 95% (P < 0.05). terial (i.e., inert fraction after OFMSW ballistic separation before AD)
(Seruga et al., 2020).
3. Results and discussion
3.1.2. pH
3.1. Changes in physico-chemical properties during FWD co-composting The pH of the initial composting matrix in all treatments varied be­
process tween 8 and 9. Addition of sawdust alone or with mature compost did
not significantly affect the alkalinity of FWD (Fig. 2b). As the com­
3.1.1. Temperature posting process progressed, the pH varied between 7.5 and 9 indicating
Temperature changes during composting indicates the microbial high buffering capacity of FWD. Consumption of easily degradable or­
decomposition of organic matter and helps in pathogen destruction ganics such as starch, fat etc. during AD process, and the presence of
(Manu et al., 2017; Wong et al., 2017). Temperature profile showed humic substances increases the buffering capacity of FWD (Adhikari
significant differences (P < 0.05) among the various treatments with et al., 2009; Manu et al., 2019; Haouas et al., 2021). However, pH still
different combinations of feedstock materials (Fig. 2a). Following the plays an important role in FWD composting as it alters the NH4+-N
initial increase of temperature to ~ 65 ◦ C in the control (R1) and oxidation and NH3 volatilization (Awasthi et al., 2017; Cai et al., 2021).
treatment with only mature compost (R2), temperature remain high till The pH stabilization near the process completion in R3 and R4 could be
the end of the composting period. In these treatments (R1 and R2), attributed to the decline in ammonification due to reduction in NH4+-N
presence of indigenous microorganisms in FWD and mature compost content.
initiated the biological activities and thus the sharp rise in temperature
was observed. Oversupply of nitrogen to the microbial needs caused the 3.1.3. Electrical conductivity
elevation in temperature in these treatments resulting in a longer com­ The salinity of compost provides the information of phytotoxic ef­
posting period (Zhou, 2017). In R3, FWD mixed with sawdust only, also fects on the plant growth due to high levels of salt (Voberkova et al.,
achieved a high initial temperature of 55 ◦ C and maintained at this 2020). Addition of sawdust and/or mature compost reduced EC of
temperature range until day 18 before entering the cooling phase. In composting matrix in R3 and R4 (Fig. 2c). As the composting process
treatment with both sawdust and mature compost amendments (R4), the progressed, no/ less changes in EC were observed in these treatments
temperature initially increased to 58 ◦ C, but after a short thermophilic due to the continuous production of humic substances in better com­
phase, temperature declined slowly till the end of the composting posting environment (Sun et al., 2021). In the control and treatment
period. The presence of sawdust in R3 and R4 has improved the com­ with only mature compost addition (R1 and R2), although the EC
posting conditions by increasing porosity for enhanced microbial ac­ decreased from ~ 12 mS/cm to ~ 5 mS/cm and ~ 9 mS/cm, respec­
tivity (Espinosa-Salgado et al., 2020). In R4, addition of mature compost tively, they were still higher (>4 mS/cm) than HKORC standards
significantly influenced the decomposition of FWD, hence the cooling (HKORC, 2005) indicating the possible phytotoxic effect on plants even
phase could be achieved earlier as compared to other treatments. The after one month of composting (Chan et al., 2016).

Fig. 2. Changes in (a) Temperature; (b) pH and (c) EC during the composting of food waste digestate with sawdust and /or compost. (R1 = digestate control, R2 =
digestate + compost, R3 = digestate + sawdust, R4 = digestate + sawdust + compost).

4
B. Song et al. Bioresource Technology 341 (2021) 125759

3.2. Carbon mineralization compost amended treatment (R4). The cumulative CO2 emission
(Fig. 3c) gives a better representation of carbon mineralization and the
Unlike food waste, FWD comprises less TOC content (~34%) due to profile suggested higher carbon loss in the control treatment R1 (40%)
the presence of partially degraded organic matter (Wong et al., 2017; Liu followed by R2 (28%), R3 (20%), and R4 (18%).
et al., 2020). Addition of mature compost in R2 reduced the TOC content Since FWD was the major components in the compost mixtures that
of the compost mixture to 32%. Amending carbonaceous sawdust alone would perform active decomposition, CO2 emission from FWD fraction
or with mature compost in R3 and R4, increased TOC content to ~ 44%. was calculated based on the amount of FWD present in each treatment to
As the composting process progressed, the microbial metabolism of give an assessment of FWD decomposition efficiency (Fig. 3d-3e). Based
organic matter reduced TOC content in first two treatments (R1 and R2) on this, the cumulative CO2 emission was found to be highest in treat­
to ~ 27%, whereas in sawdust amended reactors (R3 and R4), the final ment R4 amended with sawdust and mature compost (2147 g/kg FWD)
TOC was found to be ~ 42% (Fig. 3a). Changes in TOC reduction may followed by treatment R2 with only mature compost (999 g/kg FWD). In
not give accurate decomposition ability of composting matrix due to the contrast, cumulative CO2 emission in treatments with only sawdust (R3)
presence of sawdust and mature compost in R3 and R4. Hence, CO2 and control (R1) were 762 g/kg FWD and 367 g/kg FWD, respectively.
evolution was considered to assess the degree of decomposition during This shows that the increased C/N ratio of composting matrix to 29 by
the process (Yang et al., 2020). addition of sawdust and mature compost provided a suitable composting
Daily CO2 emission was higher in the initial stages of composting in condition for the microbial decomposition of FWD. Addition of mature
all treatments due to the provision of aeration that enhanced the mi­ compost also play a role in providing microbial inoculum to enhance
crobial decomposition (Fig. 3b) (Chan et al., 2016). The highest emis­ FWD biodegradation by 63% when added alone and by 83% when added
sion was observed from control R1 due to the higher FWD quantity (p > together with sawdust. However, addition of only mature compost
0.05), while the daily CO2 emission followed the order: only mature caused longer composting duration because of the low C/N ratio of
compost (R2), only sawdust (R3) and then both sawdust and mature feedstock. Although actual total CO2 emission from the control R1 was

Fig. 3. (a) TOC; (b) Daily CO2 evolution based on total composting dry mass; (c) Cumulative CO2 evolution based on total composting dry mass; (d) Daily CO2
evolution based on digestate dry mass; and (e) Cumulative CO2 evolution based on digestate dry mass. (R1 = digestate control, R2 = digestate + compost, R3 =
digestate + sawdust, R4 = digestate + sawdust + compost).

5
B. Song et al. Bioresource Technology 341 (2021) 125759

the highest, the cumulative CO2 emission per kg of FWD was the lowest and R4 due to the decomposition of readily available substrates into
indicating its poor composting properties. Hence, addition of sawdust as CO2, which resulted in total mass reduction, and concentration of N
a bulking agent along with mature digestate compost as an inoculum content (Zhou et al., 2018).
could significantly enhance the decomposition of FWD. The temperature and pH could shift the balance of the equilibrium of
NH4+-N and NH3 during composting process (Manu et al., 2021a).
3.3. Nitrogen transformations Under aerobic conditions, oxidation of FWD leads to increase in tem­
perature as a result of microbial activities and hence the NH3 volatili­
One of the beneficiary reasons to recycle FWD is its higher TKN zation is inevitable. Following an initial rise in NH4+-N due to
content (4–5%) (Peng and Pivato, 2019), but the presence of high ni­ ammonification process, NH4+-N showed a decreasing trend till the end
trogen content reduces C/N ratio making it unsuitable for effective of the composting process (Fig. 4b). Slight variations in NH4+-N con­
composting (Cerda et al., 2019). In addition, most of the nitrogen in centration could be observed on day 7 which could be due to the higher
digestate is present in the form of NH4+-N which inhibits the microbial nitrogen transformation reactions occurring at the initial stages of FWD
decomposition of organics (Ren et al., 2018). Therefore, addition of composting. In the present study, NH4+-N concentration reduced to the
sawdust alone or in combination with mature compost in treatments R3 permissible limit within 15 days only in treatments amended with
and R4 reduced TKN content to ~ 1.5% due to the mass dilution making sawdust alone or with mature compost (HKORC 2005) indicating the
it a more suitable C/N ratio for composting (Fig. 4a). As the composting possible phytotoxic nature in the control and treatment with compost
process progressed, TKN was reduced to ~ 2.8% from 3.5 to 4% in amendment only due to the poor composting properties of FWD even
treatments R1 and R2 due to N loss mostly in the form of NH3 emission after 30 days of composting. The maturation time is important for a
(Rincón et al., 2019). This reduction was prominent until first 10 days of compost plant operation as it directly affects the economic aspects as
composting which is in agreement with the study of Zeng et al. (2012). well as waste retention time.
In contrast, increasing trend up to 2% was observed in treatments R3 NH3 emission is an unavoidable environmental hazardous problem

Fig. 4. Changes in (a) TKN; (b) NH4+-N: (c) Daily NH3 emission based on total composting dry mass; (d) Cumulative NH3 emission based on total composting dry
mass; (e) Daily NH3 emission based on digestate dry mass; (f) Cumulative NH3 emission based on digestate dry mass. (R1 = digestate control, R2 = digestate +
compost, R3 = digestate + sawdust, R4 = digestate + sawdust + compost).

6
B. Song et al. Bioresource Technology 341 (2021) 125759

during composting process, but a good composting process should

reduce NH3 emission to produce N enriched end product (Chan et al.,
2016) (Fig. 4c-4f). Hence, development of mitigation strategies for NH3
emission through feedstock adjustment is one of the critical research
areas in composting (Zhao et al., 2020). In the present study, NH3
emission was higher in the initial period in the control R1 due to rapid
increase in temperature and alkaline pH which were the favorable
conditions for NH3 volatilization (Wang et al., 2016). Addition of
sawdust alone or with mature compost significantly reduced NH3
emission in R3 and R4 due to the lower total N contents and the suitable
composting environment which would enhance the nitrification process
resulting in conversion of the phytotoxic NH4+-N to NO3–-N (Zhou et al.,
2018). At the end of composting, cumulative NH3 emission was in the
order of R1 > R2 > R3 > R4.
The mass balance on nitrogen speciation showed TKN loss of 55%
and 39% in the control R1 and treatment amended only with mature
compost R2, respectively, whereas only 9% and 7% of TKN loss was
observed in sawdust amended treatments R3 and R4, respectively. The
majority of nitrogen loss in R1 and R2 were due to NH3 volatilization as
it accounted for 5.2% and 2.2% loss of initial N % after composting. In
contrast, only 1.3% and 0.9% loss of initial N % was observed from
sawdust amended treatments R3 and R4, respectively, indicating a more
effective nitrification process.

3.4. Compost maturity evaluation

3.4.1. C/N ratio

C/N ratio is one of the conventional process parameters used to
assess the composting process (Zhou et al., 2018). Reduction in C/N
ratio due to the simultaneous consumption of carbonaceous and
nitrogenous organic substrates indicates an effective composting process
(Wang et al., 2016) and compost with a C/N ratio of < 20 is considered
as matured (HKORC, 2005). The initial composting feedstock in treat­
ments R1 and R2 were<10 indicating a poor composting condition and
hence the C/N ratio of these two treatments remained at around the
same low C/N ratio (Fig. 5a). R3 and R4 showed a decreasing trend in
C/N ratio from an initial C/N of 29 down to < 20 at around day 15
showing a good microbial decomposition. Fig. 5. Changes in (a) C/N ratio and (b) SGI during the composting of food
waste digestate with sawdust and/or compost. (R1 = digestate control, R2 =
3.4.2. Seed germination index (SGI) digestate + compost, R3 = digestate + sawdust, R4 = digestate + sawdust
Seed germination index (SGI) of >50% (dry weight basis) is + compost).
considered as the standard limit for matured compost to use it as a soil
amendment (Zucconi et al., 1981). In the present study, the initial SGI suitable bulking agent to enhance FWD degradation rate. First-order
values (i.e., day 0) of the composting mass in all four treatments were kinetics have been widely used to predict the biodegradation of
found to be 0% indicating the highly phytotoxic nature of FWD (Fig. 5b). organic waste (Baptista et al., 2010; Manu et al., 2016). In this study, the
As the composting process progressed, SGI values from treatments R1 curve fitting of different models was based on cumulative CO2-C evo­
and R2 with high FWD showed increasing trend only after 11 days due to lution data as it directly represents the microbial decomposition process
the presence of high NH4+-N that inhibited the seed germination in the (Fig. 6) (Table 3).
beginning days. In contrast, the sawdust amended treatment R3 showed The first-first-order kinetic model showed that the rate of degrada­
increasing SGI trend from the beginning and reached > 50% on day 11 tion of rapidly (kr) and slowly (ks) degradable organic matter increased
due to the reduction in phytotoxicity as a result of improved aeration with addition of sawdust as a co-composting feedstock. The kr value was
and enhanced nitrification. Similarly, addition of sawdust and mature 0.5384 day− 1 when FWD was treated alone in control R1 whereas
compost in treatment R4 had a significantly higher reduction in addition of mature compost reduces the degradation rate to 0.2923
phytotoxicity of FWD than R3 as indicated by the significant increase in day− 1 as the mature compost does not help in increasing C/N ratio of
SGI values after one week of composting. The slight variations on day 7 FWD. In contrast, addition of sawdust alone and with mature compost in
could be due to the changes in NH4+-N concentration as seen in Fig. 4b. R3 and R4 significantly improved kr value to 07837 day− 1 0.69 day− 1,
This indicates that FWD could be bio stabilized into good quality respectively, by creating suitable composting environment. Besides, the
compost by co-composting with sawdust and mature compost within a degradation rate of slowly degradable organic matter in these treat­
week duration. ments was in the range 0.3865–0.3893 day− 1 that was almost 2.5 times
higher than control treatment R1. This model was efficient (EF =
3.5. Composting kinetics 0.97–0.99) in fitting into the experimental data from all reactors,
however, using sawdust alone or sawdust with mature compost as co-
The determination of rate of biodegradation during composting is composting feedstock was found to be fitting better based on χ2 and
vital as it alters the design parameters, hence enhancing the decompo­ RMSE values.
sition rate is one of the main objectives in composting research (Tosun In Chen and Hashimoto model, the refractory coefficient (R) and
et al., 2008; Manu et al., 2021b). The present study aims at determining

7
B. Song et al. Bioresource Technology 341 (2021) 125759

Fig. 6. Comparison of different kinetic models fitting into cumulative CO2-C evolution during composting in (a) R1; (b) R2; (c) R3 and (d) R4.

increasing, respectively with sawdust and mature compost addition in

Table 3
treatments R3 and R4 as compared to the control reactor R1. Compared
Kinetic parameters estimated from different models for the various treatments.
to other two models, Levi-Minzi model showed larger errors between the
R1 R2 R3 R4 experimental and predicted CO2-C evolution during FWD composting.
Cumulative CO2-Cexp (%) 49.83 27.43 17.11 15.95
First-first-order model 4. Conclusions
Cumulative CO2-Cpred (%) 49.32 27.42 17.11 15.72
Cr (%) 77.87 80.17 68.39 65.88
Cs (%) 22.13 19.83 31.61 34.12 Co-composting of FWD with sawdust and mature compost signifi­
kr (day− 1) 0.5384 0.2923 0.7837 0.69 cantly enhanced the decomposition rate to produce quality compost
ks (day− 1) 0.1633 0.092 0.3865 0.3893 within a week. The NH4+-N and SGI values of compost from these
χ2 7.63 5.76 0.18 0.29 treatments were within the prescribed limits to use it as soil conditioner.
RMSE 1.78 1.34 0.11 0.31
Besides, nitrogen loss could be reduced by enhancing nitrification to
EF 0.99 0.97 0.99 0.99
Chen and Hashimoto model avoid NH3 volatilization. The first-first-order and, Chen and Hashimoto
Cumulative CO2-Cpred (%) 49.63 28.28 17.27 15.93 kinetic models demonstrated to be efficient in predicting cumulative
R 0.3 0.59 0.65 0.71 CO2 emission. The outcome of this study could potentially benefit to
K 80 88.15 298.1 345.6
develop effective composting strategies for FWD at ORRC1, Hong Kong
µm (day− 1) 7 7.23 10.87 14.19
χ2 1.85 0.36 0.24 0.22 as well as any FWD composting facility in any country.
RMSE 1.18 0.52 0.23 0.21
EF 0.99 0.99 0.99 0.99 CRediT authorship contribution statement
Levi-Minzi model
Cumulative CO2-Cpred (%) 52.06 31.04 18.59 17.03
Bing Song: Conceptualization, Investigation, Formal analysis,
k (mg/kg day) 10.24 4.47 1.54 1.55
m 0.49 0.58 0.74 0.71 Writing - original draft, Writing - review & editing. M.K. Manu:
χ2 2.39 3.11 0.56 0.41 Conceptualization, Investigation, Formal analysis, Writing - original
RMSE 1.34 1.53 0.65 0.42 draft, Writing - review & editing. Dongyi Li: Writing - review & editing.
EF 0.99 0.99 0.99 0.99
Chen Wang: Writing - review & editing. Sunita Varjani: Writing - re­
view & editing. Narsi Ladumor: Resources. Lui Michael: Resources.
kinetic constant (K) values varied significantly among the treatments Yunjie Xu: Writing - review & editing. Jonathan W.C. Wong:
during composting of FWD. The R and K values increased with addition Conceptualization, Writing - original draft, Writing - review & editing,
of sawdust alone or with mature compost in R3 and R4. The maximum Supervision, Project administration, Funding acquisition.
specific growth rate of microorganisms (µm) value increased by 55% and
101% in treatments amended with only sawdust (R3) and with both Declaration of Competing Interest
sawdust and mature compost (R4), respectively.
Levi-Minzi model demonstrated better fitting for first rapid decom­ The authors declare that they have no known competing financial
position phase. The exponential increase in the second phase of slow interests or personal relationships that could have appeared to influence
decomposition process does not fit the CO2 emission data during FWD the work reported in this paper.
composting. The k and m values were found to be decreasing and

8
B. Song et al. Bioresource Technology 341 (2021) 125759

Acknowledgement Manu, M.K., Kumar, R., Garg, A., 2017. Performance assessment of improved composting
system for food waste with varying aeration and use of microbial inoculum.
Bioresour. Technol. 234, 167–177.
The financial support from the OSCAR Bioenergy Joint Venture, Manu, M.K., Kumar, R., Garg, A., 2019. Decentralized composting of household wet
Hong Kong is greatly acknowledged. biodegradable waste in plastic drums: effect of waste turning, microbial inoculum
and bulking agent on product quality. J. Clean. Prod. 226, 233–241.
Manu, M.K., Li, D., Liwen, L., Jun, Z., Varjani, S., Wong, J.W.C., 2021a. A review on
References nitrogen dynamics and mitigation strategies of food waste digestate composting.
Bioresource Technology 125032.
Adhikari, B.K., Barrington, S., Martinez, J., King, S., 2009. Effective-ness of three bulking Manu, M.K., Kumar, R., Garg, A., 2021b. Effect of microbial inoculum and leachate
agents for food waste composting. Waste Manage. 29, 197–203. circulation on the performance of rotary drum composter used for household wet
Arab, G., McCartney, D., 2017. Benefits to decomposition rates when using digestate as biodegradable waste. Waste Biomass Valorization.
compost co-feedstock: part I - Focus on physicochemical parameters. Waste Manage. Peng, W., Pivato, A., 2019. Sustainable management of digestate from the organic
68, 74–84. fraction of municipal solid waste and food waste under the concepts of back to earth
Awasthi, M.K., Selvam, A., Lai, K.M., Wong, J.W.C., 2017. Critical evaluation of post- alternatives and circular economy. Waste Biomass Valorization 10, 465–481.
consumption food waste composting employing thermophilic bacterial consortium. Ren, L., Cai, C., Zhang, J., Yang, Y., Wu, G., Luo, L., Huang, H., Zhou, Y., Qin, P., Yu, M.,
Bioresour. Technol. 245, 665–672. 2018. Key environmental factors to variation of ammonia-oxidizing archaea
Baptista, M., Antunes, F., Gonçalves, M.S., Morvan, B., Silveira, A., 2010. Composting community and potential ammonia oxidation rate during agricultural waste
kinetics in full-scale mechanical–biological treatment plants. Waste Manage. 30 (10), composting. Bioresour. Technol. 270, 278–285.
1908–1921. Rincón, C.A., DeGuardia, A., Couvert, A., LeRoux, S., Soutrel, I., Daumoin, M., Benoist, J.
Cai, L., Cao, M., Chen, T., Guo, H., Zheng, G., 2021. Microbial degradation in the co- C., 2019. Chemical and odor characterization of gas emissions released during
composting of pig manure and biogas residue using a recyclable cement-based composting of solid wastes and digestates. J. Environ. Manage. 233, 39–53.
synthetic amendment. Waste Manage. 126, 30–40. Seruga, P., Krzywonos, M., Wilk, M., 2020. Treatment of by-products generated from
Cerda, A., Mejias, L., Rodríguez, P., Rodríguez, A., Artola, A., Font, X., Gea, T., anaerobic digestion of municipal solid waste. Waste Biomass Valorization 11,
Sánchez, A., 2019. Valorisation of digestate from biowaste through solid-state 4933–4940.
fermentation to obtain value added bioproducts: a first approach. Bioresour. Sun, C., Wei, Y., Kou, J., Han, Z., Shi, Q., Liu, L., Sun, Z., 2021. Improve spent mushroom
Technol. 271, 409–416. substrate decomposition, bacterial community and mature compost quality by
Chan, M.T., Selvam, A., Wong, J.W.C., 2016. Reducing nitrogen loss and salinity during adding cellulase during composting. J. Clean. Prod. 299, 126928.
‘struvite’ food waste composting by zeolite amendment. Bioresour. Technol. 200, Tosun, L., Gönüllü, M.T., Arslankaya, E., Günay, A., 2008. Co-composting kinetics of rose
838–844. processing waste with OFMSW. Bioresour. Technol. 99 (14), 6143–6149.
Chen, Y.R., Hashimoto, A.G., 1980. Substrate utilization kinetic model for biological Tchobanoglous, G., Theisen, H., Samuel, A.V., 1993. Integrated solid waste management
treatment processes. Biotechnol. Bioeng. 22, 2081–2095. engineering principles and management issues. McGraw-Hill International Edition
Espinosa-Salgado, R., Saucedo-Castañeda, G., Monroy-Hermosillo, O., 2020. Composting (1993), New York.
a digestate from the organic fraction of urban solid wastes. Rev. Mex. Ing. Quim. 19 Voberkova, S., Maxianova, A., Schlosserova, N., Adamcova, D., Vrsanska, M.,
(1), 1–8. Richtera, L., Gagic, M., Zloch, J., Vaverkova, M.D., 2020. Food waste composting - is
HKORC. 2005. Compost and Soil Conditioner Quality Standards. Hong Kong Organic it really so simple as stated in scientific literature? - a case study. Sci. Total Environ.
Resource Centre. http://www.hkorc-cert.org/download/COMPOST-SD-080124-A- 723, 138202.
Eng.pdf (accessed on 15th Sep 2019). Wang, X, Selvam, A, Wong, J.W.C., 2016. Influence of lime on struvite formation and
Haouas, A., Modafar, C.E., Douira, A., Ibnsouda-Koraichi, S., Filali-Maltouf, A., nitrogen conservation during food waste composting. Bioresour. Technol. 217,
Moukhli, A., Amir, S., 2021. Evaluation of the nutrients cycle, humification process, 227–232.
and agronomic efficiency of organic wastes composting enriched with phosphate Wong, J.W.C., Karthikeyan, O.P., Selvam, A., 2017. Biological nutrient transformation
sludge. J. Clean. Prod. 302, 127051. during composting of pig manure and paper waste. Environ. Technol. 38, 754–761.
Kaur, G., Johnravindar, D., Wong, J.W.C., 2020. Enhanced volatile fatty acid degradation Yang, Y., Awasthi, M.K., Du, W., Ren, X., Lei, T., Lv, J., 2020. Compost supplementation
and methane production efficiency by biochar addition in food waste-sludge co- with nitrogen loss and greenhouse gas emissions during pig manure composting.
digestion: a step towards increased organic loading efficiency in co-digestion. Bioresour. Technol. 297, 122435.
Bioresour. Technol. 308, 123250. Zeng, Y., deGuardia, A., Dabert, P., 2016. Improving composting as a post-treatment of
Külcu, R., Yaldiz, O., 2004. Determination of aeration rate and kinetics of composting anaerobic digestate. Bioresour. Technol. 201, 293–303.
some agricultural wastes. Bioresour. Technol. 93 (1), 49–57. Zeng, Y., deGuardia, A., Daumoin, M., Benoist, J.C., 2012. Characterizing the
Levi-Minzi, R., Riffaldi, R., Saviozzi, A., 1990. Carbon mineralization in soil amended transformation and transfer of nitrogen during the aerobic treatment of organic
with different organic materials. Agric. Ecosyst. Environ. 31, 325–335. wastes and digestates. Waste Manage. 32, 2239–2247.
Lim, S., Park, H., Hao, X., Lee, S., Jeon, B., Kwak, J., Choi, W., 2017. Nitrogen, carbon, Zhao, S., Schmidt, S., Qin, W., Li, J., Li, G., Zhang, W., 2020. Towards the circular
and dry matter losses during composting of livestock manure with two bulking nitrogen economy – a global meta-analysis of composting technologies reveals much
agents as affected by co-amendments of phosphogypsum and zeolite. Ecol. Eng. 102, potential for mitigating nitrogen losses. Sci. Total Environ. 704, 135401.
280–290. Zhou, J., 2017. The effect of different c/n ratios on the composting of pig manure and
Liu, J., Huang, S., Chen, K., Wang, T., Mei, M., Li, J., 2020. Preparation of biochar from edible fungus residue with rice bran. Compost Sci. Util. 25 (2), 120–129.
food waste digestate: pyrolysis behaviour and product properties. Bioresour. Zhou, Y., Selvam, A., Wong, J.W.C., 2018. Chinese medicinal herbal residues as a bulking
Technol. 302, 122841. agent for food waste composting. Bioresour. Technol. 249, 182–188.
Lu, J., Xu, S., 2021. Post-treatment of food waste digestate towards land application: a Zucconi, F., Pera, A., Forte, M., De Bertoldi, M., 1981. Evaluating toxicity in immature
review. J. Clean. Prod. 127033. compost. Biocycle 22, 54–57.
Manu, M.K., Kumar, R., Garg, A., 2016. Drum composting of food waste: a kinetic study.
Procedia Environ. Sci. 35, 456–463.