Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 7

Abstract

Food processing wastewater (FPW) contains high levels of oil and grease (O&G), especially
from factories that produce fast food such as nuggets, sausages, and burger patties from beef,
poultry, and fish. Therefore, the feasibility of replacing conventional membranes for the
treatment of FPW with low-cost dynamic membranes (DM) as an alternative was investigated in
this study. An anaerobic DM bioreactor (DAnMBR) was operated for 90 days to evaluate the
treatment performance using real FPW under different organic loading rates (OLR) of 3.5, 5.0,
6.5, and 7.0 g COD (chemical oxygen demand)/L day. Once the reactor had reached a steady-
state of 90% COD removal, the feed to the reactor was supplemented incrementally with FPW
from 10% to 90% as COD to allow the methanogenic bacteria to acclimate any potential
inhibitory effects from its recalcitrant content. The bioreactor presented a stable performance at
OLR 5.0 g COD/L day with 97.5% removal of COD and reached 20 mg/L total suspended solids
(TSS) discharge. A significant correlation between COD fractions removed via acidogenesis and
methanogenesis with different OLR was found, indicating that the increase in treatment
performance is beneficial to the methanogenic archaea activity. The methane gas production
yield achieved a maximum of 0.40 L methane/g CODadded at OLR 3.5 and 5.0 g COD/L day.
The average permeate flux in these studies is around 60 L/m2 h. The DM fouled after 57 days
(at flux 27.16 L/m2 h immediate drop to 2.16 L/m2 h) when operated at 3.5 g COD/L day. After it
fouled, the membrane underwent physical cleaning, backwashed insitu for 5 min, and reused
again without any chemical cleaning. The improved filtration resistance is contributed by the
occurrence of DM fouling induced by the soluble microbial products (SMP) and extracellular
polymeric substances (EPS) release as well as the increased protein/carbohydrate (P/C) ratio in
the mixed liquor.

Introduction

Within the GCC, Oman is likely to witness the highest growth at a compound annual growth
(CAGR) rate of 4.6 per cent in terms of food consumption which is anticipated to grow from 3.2
million MT in 2018 to 4.0 million MT in 2023, a new report reveals. Revenue in the Food
segment is projected to reach US$391.10m in 2023.Revenue is expected to show an annual
growth rate (CAGR 2023-2027) of 16.52%, resulting in a projected market volume of
US$720.90m by 2027. The Sultanate has pledged investments to the tune of around $7 billion
in the further development of a countrywide wastewater treatment, recycling and reuse
infrastructure system over the next two decades as part of Oman’s commitment to ensuring
sustainable freshwater availability and supply over the long term.
Food processing industries consist of a variety of industries such as dairy, snacks, sweets,
beverages, and distillery. Wastewaters from these industries come from different plant
operations such as production, cleaning, sanitizing, cooling, and materials transport.[ Alvarez
PM, Pocostales JP, Beltrán FJ. Granular activated carbon promoted ozonation of a
food-processing secondary effluent. J Hazard Mater 2011;185:776-83] However, the
constituents of these wastewaters are biodegradable due to high organic substances and may
also be nontoxic. This eventually results in high concentrations of BOD, chemical oxygen
demand (COD), and suspended solids (SSs).[ Cristian O. Characteristics of untreated
wastewater produced by food industry. Analele Universilatii Oradeq Fascicula Protectia Mediului
2010;15:709-14.] The characteristics of wastewater play a major role in selecting the type of
treatment to be carried out on it.[ Debik E, Coskun T. Use of static Granular Bed reactor (SGBR)
with anaerobic sludge to treat poultry slaughter house wastewater and kineticmodelling. Biores
Technol 2009;100:2777-82.]
Typical characteristics of a food processing industry are shown in Table 1.[ Tariq M, Ali M, Shah
Z. Characteristics of industrial effluents and their possible impacts on quality of underground
water. Soil Environ 2006;25:64-9. Tikariha A, Sahu O. Study of characteristics and treatments of
dairy industry wastewater. J Appl Environ Microbiol 2014;2:16-22. Vanerkar AP, Sanjeev S,
Shanta S. Treatment of food processing industry wastewater by a coagulation/flocculation
process. Int J Chem Phys Sci 2013;2:2583-8.]
Table 1: Characteristics of a typical food processing industry

Parameters Standard in mg/L except pH


pH 5.5-9.0
Oil and grease 10
BOD (5 days at 25°C) 50
COD 250
TSS 50
TN 10
TP 2

This high strength wastewater requires significant space, costly aeration, and additional toxic
coagulants/flocculants such as alum and polymer, which generates a considerable amount of
sludge that requires disposal. Therefore, food processing companies’ operating costs will
increase due to higher energy usage, purchase of chemicals, and disposal of scheduled waste.
Apart from that, most companies have limited space on their premises to install new treatment
facilities.
The practicality of using wastewater as a resource lies in the development of an efficient,
economical and environmental friendly treatment technology. Aerobic wastewater treatment
technologies have been in use since a century. However, these methods require high energy,
produce high amount of sludge, need of larger space for plant operation and higher
maintenance cost which offset the advantages of reusing wastewater (Krzeminski et al., 2017).
During the process, there is uncontrolled release of methane (CH4), carbon dioxide (CO2), and
nitrogen oxide (N2O) which are potent greenhouse gases (Foresti et al., 2006; Ghauri et al.,
2011). Hence, there is an immense need of developing alternate wastewater treatment
technology. Recently, membrane separation processes gained significant attention to recover
resources from waste streams, biofuels and especially wastewater to reduce potent greenhouse
gases (Khalid et al., 2019; McCarty et al., 2011). Anaerobic membrane bioreactors (AnMBR)
are emerged as a promising alternative to aerobic wastewater treatment technologies. AnMBR
have low energy input, and sludge production compared with aerobic processes. Furthermore,
employing AnMBRs reduce: the operational space, and number of unit operations compared
with conventional processes (as shown in Fig. 1). The also offer easy scale-up and selective
separation and recovery of nutrients and resources (Aslam et al., 2018b; Liao et al., 2006).
However, widespread adaption of technology is still hindered due to several issues including
membrane fouling, dissolved CH4 recovery, sulphide-induced low COD and alkalinity. Among
these, membrane fouling is a key bottleneck for their commercialization (Lin et al., 2013).
Novel AnMBR configurations are introduced to overcome the issues in scale-up of technology in
general, and 4 combat membrane fouling in particular. These reactor schemes have also
reduced the energy input and increase the CH4 production yield (Shin & Bae, 2018; Smith et al.,
2014). So far, a number of review papers have been published on AnMBRs for wastewater
treatments (Lin et al., 2013; Ozgun et al., 2013; Shin and Bae, 2018; Skouteris et al., 2012;
Smith et al., 2012; Stuckey, 2012).
this research aims to contribute to the development of AnMBR technology as well as raising
new concerns about two-stage DAnMBRs. Moreover, studies using DM for high strength real
wastewater treatment in anaerobic treatment is scarce. Thus, the present research aims at
assessing the performance of DAnMBR with phase separation treating FPW at mesophilic
temperature (35 ◦C). Although the submerged membrane offers a significant advantage over the
external membrane configuration, the system’s performance must be evaluated to cater to the
treatment of food processing waste (PFW). The system’s treatment performance was assessed
in three successive phases from fully synthetic to the highest organic loading rate of pure
wastewater that the system can treat. This research’s findings are expected to fill the knowledge
gap and provide new insights into combining two-stage anaerobic digestion process and DM for
treating high FOG wastewater. Local small and medium food processing plants can also benefit
by reducing their wastewater treatment cost while performing their social responsibility to the
country.

Literature Review

Materials and Methods

Food Industry Waste Water

The substrate, food-processing wastewater (FPW) was collected from a food processing
industry located at Batu Caves, Selangor, Malaysia. This factory produces varieties of fast food
products such as nuggets, burgers, and sausages from meat such as chicken, beef, lamb, and
fish. Processes include half-cooked food processing such as frying, boiling, mincing, and
seasoning. These activities produce a large amount of organic and inorganic materials such as
carbohydrates (starch for coating), proteins, and fat, oil, and grease (FOG) from the main
ingredient, which is animal meat. The raw sample was collected at the collection sump prior to
the in-situ plant wastewater treatment system. Before physical and chemical analyses were
done, the FOG solidified and floated on the upper layer of the substrate were manually skimmed
off and filtered using a kitchen strainer. This study only used this simple pretreated sample.
FPW was kept at 4 ◦C in a chiller to prevent the degradation of the organic substances. The
desired organic loading rates (OLR) for the experiments were achieved by dilution of the sample
with distilled water at the calculated proportion. The inoculum used in this study was an
anaerobically digested sewage sludge obtained from the anaerobic sludge tank of a municipal
sewage treatment plant located in Sohar, Sultanate of Oman. The high-rate anaerobic digester
system is used for sludge treatment from the aerobic lagoon in the same facilities. The digester
runs at mesophilic temperature with a flow rate of 300–400 m3 / day. The methane production
yield for the inoculum is analyzed during the preliminary study of biochemical methane potential
(BMP) for 54 days and achieved 54 mLCH4/g VS [59]. According to Md Huzir et al. [1-60], the
inoculum from anaerobic digester taken from MWTP, the specific methanogenic activity (SMA)
varied from 0.25 to 0.31 g COD/g VS day. Samples were kept at 4 ◦C to prevent the
degradation of organic/biomass substances. Table 2 tabulates the characteristics of the
substrate (FPW) and the inoculum and also methods and instruments used for the analysis. The
range of values for the substrate is because the collection of FPW is fully dependent in the
different production varies the range of the parameter concentration.
Table 2 Characteristics for FIWW (substrate) and anaerobically digested sewage sludge
(inoculum).

Parameter Unit Substrate


pH 3.20–3.70 7.00
BOD5 mg/ L 7800–9200 690
COD mg/L 15,500–17,500 1970
NH3-N mg/L 11.5–17.5 302
Total phosphorus mg/L 225–240 563
Total solid g/L 15.55–16.65 18.70
Volatile solid g/L 15.27–16.25 12.75
VS/TS 0.9760–0.9820 0.6818
Hydrogen % 8.722 4.8825
Sulfur % 2.1395 1.383
Carbon % 54.8 32.20
Nitrogen % 2.4685 4.51
C/N ratio 22.20 7.14
Oil & grease mg/L 2300–4400
Turbidity NTU 999
Heavy metal, , and Cu (0.506 mg/L)
Zn (5.024 mg/L)
Ni (0.498 mg/L)
Fe (5.094 mg/L)

Bioreactor configuration and general operating procedure

A submerged DAnMBR was designed and fabricated, as shown in Fig. 1. The up-flow anaerobic
two-stage reactor using two cylindrical Plexiglass vessels has a total working volume of 30 L.
The working volume is separated into two stages. The first stage is an acidogenic reactor (AR)
with a dimension of 82 cm diameter x 245 cm height and a volume of 10 L. While the second
stage is a methanogenic reactor (MR) with a dimension of 160 cm diameter x 245 cm height
and a volume of 20 L. Both reactors had the same headspace height of 65 cm. A heating belt
and coil maintained both reactor temperatures between 35 and 40 ◦C. The transmembrane
pressure (TMP) through the filter medium was measured using a pressure gauge. The influent
is pumped into the bioreactor using the peristaltic pump (Longer Pump, BT600-2J). To adjust
the rate of flow passing through the membrane, a peristaltic pump was used, which vacuum the
effluent through the filter cloth filter into the permeate tank. The MR was equipped with two units
of woven fabric cloth as the supporting membrane, which was installed on the surface of the
tubular membrane casing as shown in Fig. 1. The tubular membrane casings have an external
diameter of 86 mm, a working height of 270 mm with a calculated total area of 0.0972 m2
outside surface area for two units of the membrane, and a nominal average pore size of 20 µm.
The assembly of the DAnMBR is similar to spiral-wound modules, which are similar to the flat
sheet type except that the whole assembly is wrapped up rather like a ’Swiss roll’ as shown in
Fig. 1. The support material for the dynamic membrane development is made of polypropylene
woven filter cloth with specifications as in Table 3. The influent was fed intermittently for 8 h/day
with a flow rate varied from 10 to 30 L/day depending on hydraulic retention time (HRT). Before
the substrate was pumped into the bioreactor, the pH was adjusted to between 6.9 and 7.2
using 3 N NaOH. The feed flows down the gravitationally and up-flow in each Tank 1 and Tank
2 to make sure the substrate has full contact with the biomass at the bottom of each tank.
Meanwhile, the permeate flows Table 1 Ingredients of meat and yeast extract (Bovril meat
extract and Marmite yeast extract). Ingredients Weight per 100 g Yeast extract Protein 38.4 g
Carbohydrates 19.2 g Fat 0.1 g Fiber 3.1 g Sodium 4.3 g Thiamine 5.8 mg Riboflavin 7.0 mg
Niacin 160.0 mg Folic acid 2500 µg Vitamin B12 15.0 µg Meat extract Protein 13.3 g
Carbohydrates 24.4 g Fat 0.1 g Potassium 1200 mg Sodium 3510 mg S.B. Mahat et al. Journal
of Environmental Chemical Engineering 9 (2021) 105245 4 tangentially through the membrane.
The pH was maintained between 6.8 and 7.2 in the acidogenic reactor and methanogenic
reactor, respectively using 3 N NaOH as needed. The supernatant samples from each operation
were collected every day. Sampling was done before the feeding process is started and by
using a special metal tube which was modified to collect the sample at the center (or any
location part) inside Tank 1 and Tank 2. The sample from Tank 1, a sample from Tank 2, raw
FPW, effluent, and permeate (samples filtered through the dynamic membrane) were evaluated
for standard wastewater analysis as explained in the following sections later. A peristaltic pump
(Longer Pump, BT600-2J) was used for the filtration during the experimental stage when the
membrane was fixed. Once the operation pressure reached a fouled TMP denoted by the
significant decrease in permeates flux (< 2 L/m2 h), the operation was ended. The in-situ
backwashing procedure commenced for 5 min using the same peristaltic pump for filtering
through the membrane by reversing the permeate flow backward for fouling mitigation purposes.
During the DAnMBR operations, biogas/N2 sparging and mechanical mixing was not used
because it was found to delay the DM formation either in AR and MR. There was no sludge
recycling between the two separated bioreactors because the mixed liquor is flowing
gravitationally from AR to MR with minimal sludge washout. Thus, the colloids and biomass
stayed at the bottom of the reactor. The two digestate outlets after MR were constructed
because one outlet (permeate lines) was used to determine the performance of the
development membrane and the effluent outlet was used to differentiate the performance
without DM treatment. The DAnMBR was fed intermittently with continuous operation for 280
days at an infinite sludge retention time (SRT) to identify possible relationships between the
filterability of mixed liquor and sludge properties such as SMP and EPS [61].

Experimental procedure for treatment performance of DAnMBR

The experimental procedures to evaluate the treatment performance Table 2 Characteristics for
FPW (substrate) and anaerobically digested sewage sludge (inoculum). Parameter Unit
Substrate Inoculum Methods/Instrument pH – 3.20–3.70 7.00 pH meter (Brand: Trans) BOD5
mg/L 7800–9200 690 APHA method (5210B) COD mg/L 15,500–17,500 1970 APHA method
(5220C) NH3-N mg/L 11.5–17.5 302 APHA method (4500C) Total phosphorus mg/L 225–240
563 HACH DR890 Spectrophotometer Total solid g/L 15.55–16.65 18.70 APHA method (2540B)
Volatile solid g/L 15.27–16.25 12.75 APHA method (2540E) VS/TS – 0.9760–0.9820 0.6818
Hydrogen % 8.722 4.8825 Elemental analyzer (LECO) Sulfur % 2.1395 1.383 Elemental
analyzer (LECO) Carbon % 54.8 32.20 Elemental analyzer (LECO) Nitrogen % 2.4685 4.51
Elemental analyzer (LECO) C/N ratio – 22.20 7.14 Oil & grease (O&G) mg/L 2300–4400 –
APHA method (5520B) Turbidity NTU 999 – HACH 2100N Turbidimeter Heavy metal detection
for FPW: Cu (0.506 mg/L), Zn (5.024 mg/L), Ni (0.498 mg/L) and Fe (5.094 mg/L) ICP-MS
(Inductively Coupled Plasma Mass Spectroscopy) Fig. 1. Schematic diagram of DAnMBR. S.B.
Mahat et al. Journal of Environmental Chemical Engineering 9 (2021) 105245 5 of the two-
phase DAnMBR reactor commence in four phases. Firstly, the preliminary biochemical methane
potential (BMP) test is introduced to find the best ratio for the FPW substrate and anaerobic
sludge from municipal wastewater treatment plant (MWWTP). The inoculum to substrate (S/I)
ratio of 1.0 was proven the best ratio for performance treatment and methane production
throughout the analysis [59]. After the BMP test is finished within 50 operating days, the
DAnMBR is filled with 50% of anaerobic sludge in both tanks and filled with tap water until
reaches 100% working volume. The reactor is left for 1 week to make sure all the supernatant in
the initially collected inoculum is fully digested before introduced with synthetic wastewater.
Secondly, after the reactor is stable, the start-up phase is commenced where the respective
OLR were calculated to vary between 0.07 and 0.75 g COD/L day. The hydraulic retention time
(HRT) was reduced from 1.3 to 0.5 during the start-up phase. The application of varied HRT
was to ensure the archaea is fully stabilized with high OLR conditions without inhibition factor
before introduced with real FPW for further acclimatization. The DAnMBR was started-up using
synthetic food wastewater (SFW). In this study, meat extract (Bovril) and yeast extract (Marmite)
was used as the carbon sources. Table 4 shows the parameters during the start-up and
acclimatization phases. It also shows reactor operating conditions during the treatment process.
The start-up period lasted for 120 days. Once the COD removal in the start-up phase reached
around 90%, the acclimatization procedure commenced in the third stage.
The FPW was introduced into DAnMBR and gradually mixed with synthetic wastewater (SW)
until the feed reached 100% FPW. This phase is called the acclimatization phase. Since the pH
value of the raw FPW was about 3.0–3.5, 3.0 N NaOH was used to adjust the pH of the feed
solution to around 6.8–7.2 before introducing it into DAnMBR. The mixture of SW and FPW for
the influent had an average COD concentration of 3000 mg/L, where OLR was calculated to
vary between 2.0 and 2.5 g COD/L.d using HRT of 0.5 days during the first half of the
acclimatization phase. Increasing average influent COD of 4500 mg/L resulted in OLR of 2.50–
3.50 g COD/L.d which was proceeded in the second half of the acclimatization phase. The
acclimatization period lasted for 15 days. After the mixed liquor suspended solids (MLSS)
reached 3000 mg/L in both tanks and reached stable condition with 90% COD removal, the
treatment phase was started immediately. The treatment phase denotes the starting time of the
support membranes were installed in the methanogenic reactor. The same OLR of 3.5 g COD/
L.d. was used at the beginning of the treatment phase. During the initial phase of filtration, the
filter itself cannot reject fine particles in the mixed liquor due to relatively large pores. This
resulted in the presence of particles in the effluent. Thus, the permeate containing high
suspended solids concentration was returned into the MR until it reaches high permeate quality
with total suspended solids (TSS) reduction to 20 mg/L and low turbidity. Once COD removal
achieved 85–90% and flux starting to decrease (fouled membrane), the in-situ backwash
procedure commenced. After that, the OLR was increased to 5.0, 6.0, and 7.5 g COD/L day
using the step-wise approach.
2.5. Analytical methods and calculations
The typical parameters for the characterization of wastewater were determined to assess the
effectiveness of the treatment process. Supernatant in Tank 1, Tank 2, raw wastewater,
synthetic wastewater, liquid effluent and permeate analyses were performed according to
Standard Methods for Examination of Water and Wastewater [62]. The reactor feedstock and
digestate contained high fats, oils, and grease (FOG), and small particles, which pose problems
for sample extraction. Because the analysis of raw wastewater was quite difficult, the layer of
FOG on the liquid surface is filtered using a kitchen strainer first before proceed with filtration
and dilution of the fresh samples. Sample pre-treatment involves centrifugation at 2000 rpm for
15 min to separate the supernatant. Centrifugation was effective at separating roughly 80% of
supernatant.

You might also like