Chemical Engineering Journal 367 (2019) 76–85

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

Continuous bioreactor with cell recycle using tubular ceramic membrane for T
simultaneous wastewater treatment and bio-oil production by oleaginous
Rhodococcus opacus
⁎
Tanushree Paula, Divya Baskarana, Kannan Pakshirajana,b, , G. Pugazhenthia,c
a
Center for the Environment, Indian Institute of Technology Guwahati, Guwahati, Assam 781039, India
b
Department of Biosciences and Bioengineering, Indian Institute of Technology Guwahati, Guwahati, Assam 781039, India
c
Department of Chemical Engineering, Indian Institute of Technology Guwahati, Guwahati, Assam 781039, India

H I GH L IG H T S G R A P H I C A L A B S T R A C T

• Different bioreactors evaluated for re-

finery wastewater treatment by R.
opacus.
• Continuous operation mode with
membrane cell recycle proved for
better efficiency.
• Low cost ceramic based tubular
membrane was used for cell separa-
tion.
• Hydrothermal liquefaction of biomass
for suitable bio-fuel applications.

A R T I C LE I N FO A B S T R A C T

Keywords: Rapid consumption of fossil fuels has led to the search for alternative energy sources. Bio-fuels as an alternative
Bio-oil energy source require cheap and abundantly available substrates to keep the economics of the production
Continuous with cell-recycle process low. The present study was therefore focused on utilizing raw refinery wastewater by the oleaginous
Hydrothermal liquefaction bacterium Rhodococcus opacus for converting it into bio-oil via hydrothermal liquefaction of the lipid rich bio-
Rhodococcus opacus
mass produced during the treatment process. For treating the wastewater, different operating modes using a
Refinery wastewater treatment
bioreactor were evaluated including batch, fed-batch, sequential batch, continuous and continuous with cell
Tubular ceramic membrane
recycle using low cost tubular ceramic membrane. Among the different strategies, the continuous cell recycle
system proved efficient in terms of complete removal of chemical oxygen demand (COD) (99%) and high lipid
production (86%, w/w) at a hydraulic retention time (HRT) of 16 h (dilution rate of 0.06 h−1). Furthermore, the
residual bacterial biomass from the bioreactor was treated by HTL to produce bio-oil which showed excellent
bio-fuel properties. This study demonstrated the application of R. opacus for simultaneous wastewater treatment
and production of bio-oil for energy application.

⁎
Corresponding author at: Department of Biosciences and Bioengineering, Indian Institute of Technology Guwahati, Guwahati 781039, Assam, India.
E-mail address: pakshi@iitg.ac.in (K. Pakshirajan).

https://doi.org/10.1016/j.cej.2019.02.050
Received 17 November 2018; Received in revised form 29 January 2019; Accepted 7 February 2019
Available online 08 February 2019
1385-8947/ © 2019 Elsevier B.V. All rights reserved.
T. Paul, et al.
1. Introduction
Global warming is influenced by climate change and several other
direct as well as indirect factors; one of the main reasons is due to the
greenhouse gas (GHG) emissions from fossil combustion. Hence, re-
searchers over the past decade have focused on alternative sources of
fuels from non-food crops which are GHG free [1]. Such fuels, com-
monly called as bio-fuels, are valuable in terms of meeting the energy
demands of developing countries like India. Bio-fuels can be produced
from different feedstock, e.g. lipid and lipid rich feedstock, for the
production of renewable hydrocarbon fuels [2]. Microbial lipids ob-
tained from oleaginous microorganisms containing more than 50% (w/
w) lipids are attractive for biodiesel production as such bacteria are
capable of lipid accumulation using cheaply accessible industrial was-
tewater [3,4]. However, production of bio-oil from lipid rich microbial
biomass by hydrothermal liquefaction (HTL) is less explored which is of
recent interest amongst the researchers.
On the other hand, the discharge of industrial wastewater con-
taining complex recalcitrant organics into water bodies causes serious
environmental issue. Moreover, the demand for pure water is ever in-
creasing in water scarcity regions which is also governed by population
growth. Petroleum refining industries use large amounts of water for
diverse processes and thus generate a huge volume of wastewater
which needs to be properly treated before it could be discharged. The
conventional chemical methods such as adsorption, gravity separation,
filtration, coagulation and coalescence that are frequently applied in
water treatment plants to treat refinery wastewater are dis-
advantageous due to high operation costs, low treatment efficiency,
corrosion, recontamination problems, etc. [5]. On the other hand,
biological treatment methods can overcome these drawbacks. Refinery
wastewater containing complex recalcitrant compounds are already
reported to be degraded by the hydrocarbonoclastic bacteria Rhodo-
coccus opacus. The major criteria for selecting the organism is based on
its ability to degrade and utilize complex compounds present in the
wastewater for its growth and metabolic activity. R. opacus has already
shown excellent potential for wastewater treatment and lipid accumu-
lation [6–8]. Also, the biodegradation potential of genus Rhodococcus
has been reported several years ago in the literature owing to its out-
standing properties to degrade a variety of compounds including phe-
nols, aromatics, nitriles etc. [9]. Hence, the oleaginous R. opacus bio-
mass can be employed to degrade recalcitrants in refinery wastewater
and subsequently for bio-oil production by hydrothermal liquefaction
(HTL) of the lipid-rich biomass which is not reported so far.
In a recent study by Saisriyoot et al. [10] production of fuel oils
from petroleum processing wastewater by R. opacus was reported but
the study was limited to bioreactor experiments without any sub-
sequent treatment of the residual biomass. Gargouri et al. [11] reported
95% COD removal from refinery wastewater using bacteria isolated
from contaminated soil but no lipid or bio-oil production was reported.
Janben et al. [12] focused on optimization of parameters for lipid
production by R. opacus using industrial wastewater; however, si-
multaneous treatment of the wastewater was not addressed. All these
reports suggest that for a successful application of the technique, si-
multaneous treatment of refinery wastewater and lipid accumulation by
R. opacus together with value addition of the method should be ex-
amined in detail.
Therefore, in the present study, R. opacus bacterial biomass was
used to treat refinery wastewater under different reactor operating
conditions: batch, fed-batch, sequential batch, continuous and con-
tinuous with cell-recycle using tubular ceramic membrane for evalu-
ating lipid accumulation and bio-oil production from the residual bio-
mass. For cell separation and recycle, low cost and tubular ceramic
membrane made from locally available inorganic precursors was em-
ployed. Separation process using inorganic membranes are becoming
popular and are found well suitable for large scale applications owing
to their excellent properties [13,14]. Hence, in order to keep the
77
Chemical Engineering Journal 367 (2019) 76–85
process cost low and upgrade efficiency of the wastewater treatment as
well as biomass separation and cell recycle, the tubular ceramic mem-
brane was used in this study. Bio-oil production by hydrothermal li-
quefaction is an efficient technique for converting biomass rich in lipid/
oil but it has been experimented only on micro-algal feedstock [15,16].
Therefore, this study focused on converting the lipid rich R. opacus
biomass produced following the refinery wastewater treatment to bio-
oil by HTL process which was further characterized to evaluate its
potential for biofuel applications. Thus, a novel strategy involving
continuous bioreactor operation with cell recycle for refinery waste-
water treatment coupled to bio-oil production is reported in this study.
2. Material and methods
2.1. Refinery wastewater collection and characterization
Raw refinery wastewater was collected from a petroleum refinery
plant located at Bongaigaon, Assam, India, and was stored at 4 °C until
required for further use. Characteristics of the raw wastewater are as
follows: pH 10.5, light brown color, petrol like odor, 3.5 mS/cm con-
ductivity, 5.19 mg L−1dissolved oxygen, 1253 mg L−1total dissolved
solids (TDS), 558 mg L−1 total suspended solids (TSS), 4500 mg L−1
chemical oxygen demand (COD), 0.1 mg L−1 sulphate content,
8.34 mg L−1 ammonia content and 0.582 mg L−1 total heavy metal
content. Additionally, the wastewater contained a high amount of toxic
aliphatic and aromatic hydrocarbons.
2.2. Rhodococcus opacus and seed culture cultivation
The oleaginous gram-positive bacterium Rhodococcus opacus PD630
was obtained from Microbial Type Culture Collection (MTCC),
Chandigarh, India. For maintenance, the strain was grown on 1.8% (w/
v) Nutrient Broth (NB) agar slants and was stored at 4 °C. The pure
culture was regularly sub-cultured in every four weeks by growing at
30 °C for 48 h. The seed culture of R. opacus was cultivated in a 250 ml
Erlenmeyer flask containing 50 ml of Luria Bertani (LB) broth. A full
loop of the bacteria was inoculated into the autoclaved medium fol-
lowed by incubation at 30 °C and 120 rpm until the absorbance of the
culture reached to 0.99, as measured at 660 nm using a UV–Vis
Spectrophotometer (Agilent Technologies, Singapore). For wastewater
treatment, biomass growth and lipid production by R. opacus, the
wastewater was supplemented with Mineral Salt Medium (MSM) in the
ratio 1:4 (v/v). The MSM contained the following ingredients (g L−1):
MgSO4·7H2O, 0.409; CaCl2·2H2O, 0.0265; KH2PO4, 1; Na2HPO4·12H2O,
6; FeCl3·6H2O, 0.0833 and 1% of trace metal solution, i.e., 50 µL of each
of the trace elements (g L-1): FeCl3, 1.7; CaCl2, 0.6; ZnSO4, 0.2;
CuSO4·7H2O, 0.2; MnSO4, 0.2; CoCl2, 0.8; H3BO3, 0.1 and
Na2MoO4·2H2O, 0.3. Composition of the R. opacus biomass was found to
be: 45.52% of carbon, 7.23% of hydrogen, 12.76% of nitrogen, 2.181%
of sulphur, 28 mg L−1 of carbohydrates, 0.7 mg L−1 of proteins and
52% (CDW) of lipids content.
2.3. Refinery wastewater treatment and lipid production
Simultaneous biodegradation of toxic hydrocarbons present in the
refinery wastewater and lipid production by R. opacus was investigated
using an indigenous lab-scale bioreactor (2.5 L total volume) under
different operating modes, viz, batch, fed-batch, sequential batch (SBR),
continuous and continuous with cell recycle using a low cost tubular
ceramic membrane. The wastewater was supplemented with MSM as
mentioned earlier in Section 2.2. Furthermore, supplementation of the
wastewater using MSM was found to be essential for supporting the
bacteria for COD utilization, biomass growth and lipid accumulation.
All the experiments were carried out with 1.5 L of working volume
under the controlled conditions of pH 7, temperature 30 °C and agita-
tion of 250 rpm. The influent pH was adjusted manually by addition of
T. Paul, et al.
1 M NaOH or 0.5 M H2SO4 as needed. The dissolved oxygen (DO)
concentration in the reactor was controlled at above 2 mg/L by con-
tinuous stirring at 250 rpm and aeration at 1 L/min. Initially the reactor
was operated under batch mode for 5 days using 10% (v/v) of R. opacus
as the inoculum and the refinery wastewater was supplemented with
MSM in the ratio of 4:1 (v/v). Based on the results of the batch ex-
periments, the reactor was operated under fed batch mode, and for
which the feed was started after 24 h of initial batch operation and
maintained for 18 h. The SBR operation was carried out for one cycle at
a HRT of 18 h following inoculation. Under continuous operating mode,
different HRTs of 8, 16 and 24 h were evaluated for its effect on the
reactor performance.
For biomass separation and recycle following wastewater treatment
and lipid accumulation by the bacteria, the bioreactor was connected to
an indigenously prepared tubular ceramic membrane. The membrane
was fabricated using locally available low-cost inorganic precursors; it
was found to have a nominal pore size of 0.339 µm and was char-
acterized as needed for corrosion resistance and flexural strength. Also,
permeate flux value of the membrane was 3.40 × 10−5 m3/m2s under a
pressure of 68 kPa by employing a cross flow filtration system [14]. For
biomass separation in this study, the microfiltration module was packed
into stainless steel pellicon holder fixed with diaphragm, pressure
gauges at the inlet and outlet ports. A peristaltic pump was used to
supply the driving force required for achieving the trans-membrane
flux. For cell-recycle following membrane separation using the tubular
membrane set up, effluent from the bioreactor was fed to the membrane
system and retentate containing the cells was fed directly to the bior-
eactor. The working volume of the reactor was maintained constant by
supplying the retentate i.e. cell suspension from the microfiltration
module into the bioreactor with HRT of 16 h. After each cycle of the
membrane operation, the membrane was washed with Milli Q water to
avoid reduction in the membrane flux and clog formation.
During the entire operation, the system was monitored closely for
maintaining the desired pressure. During the experiments, Samples
were taken at every 6 h interval and analyzed for biomass concentration
(g L−1), COD removal (%) and lipid concentration (g L−1). All the
analyses were carried out in triplicate and the results were obtained
within ± 4% standard deviation.
For the fed-batch mode of operation with the bioreactor, the feed
rate (F) was calculated using the following Eq. (1):
μX0 V0 e μt
F=
Y X S0 (1)
S average bio-oil yield obtained along with standard deviation was re-
−1
where F, X0, V0, S0, µ and YX/S denote feed flow rate (L h ) of the
wastewater, bacterial biomass concentration (g L−1) at the end of batch
operation, wastewater volume (L) in the bioreactor at the end of the
batch, initial wastewater COD (g L−1), biomass specific growth rate
(h−1) and biomass yield, respectively. The µ (specific growth rate) was
estimated from the results of a previously conducted batch experiment
and the biomass yield was calculated using the following Eq. (2):
X − X0
YX / S = m
S0 − Sm (2)
where Xm, X0, Sm and S0 represent maximum biomass concentration
(g L−1) at time (t), initial cell concentration (g L−1) at initial time
(t = 0), total substrate concentration (g L−1) at time (t) and total sub-
strate concentration (g L−1) at initial time (t) = 0, respectively.
Fig. 1 depicts a schematic of the continuous with cell recycle mode
of operation with bioreactor integrated with tubular ceramic mem-
brane. The cell recycle ratio in these experiments was calculated ac-
cording to the following Eq. (3):
YX / S (Sm − S0)
X1 =
(1 + α − αC ) (3)
where α, C, X1, S0 and Sm denote recycle ratio, concentration factor in
78
Chemical Engineering Journal 367 (2019) 76–85
the cell recycle stream, initial and final substrate concentrations
(g L−1), respectively.
2.4. Hydrocarbon degradation profile and eco-toxicity analysis
In order to determine the hydrocarbon degradation profile followed
by the bacterium in the continuous with cell recycle experiments with
the bioreactor, samples were taken at regular time interval and ana-
lyzed by GC–MS (Perkin Elmer Clarus 600).
For a better assessment of the wastewater treatment efficiency fol-
lowing the continuous with cell recycle mode of operation with the
bioreactor, eco toxicity of the permeate from the membrane was tested
for seed germination and the results were compared with distilled water
(control), tap water and the raw (untreated) refinery wastewater. For
this eco toxicity test, a required amount of chickpeas (Cicer arietinum L.)
were taken into Petri plates and added separately with 20 ml of distilled
water, tap water, raw refinery wastewater (untreated) and permeate
(treated water); the plates were then incubated at 25°Cand 72 h in the
dark. For determining the Germination Index (GI %) the following
Equation (4) was used.
Germination index (GI)%
Seed germination (%) × Root elongtaion (%)
=
100 (4)
2.5. Hydrothermal liquefaction (HTL) of residual biomass
HTL of the lipid rich R. opacus biomass following wastewater
treatment was carried out using a 50 ml stainless steel vessel of 30 ml
working volume and at 250 °C temperature for a reaction time of 2 h.
The biomass (3–4 g) was added with 15 ml of deionized water and
mixed well for 5 min before transferring into the HTL vessel. After the
reaction time, the entire vessel was placed in an ice-water bath to
quench the reaction and liquid phases were collected following solvent
extraction with dichloromethane for further analysis. During collection,
the vessel was washed with chloroform and mixed well. The oil and
aqueous phases were separated and dichloromethane was evaporated
eventually. The bio-oil thus obtained was further weighed and char-
acterized for its properties. The other liquid phase containing mostly
water soluble products was dried in an oven at 105 °C and its final
weight was determined. The experiments were repeated thrice and the
ported. The following equation (Eq. (5)) was used to estimate the bio-
oil yield obtained in this study:
Mass of bio − oil
Bio − oil yield (wt%) =
Mass of R. opacus biomass
× 100
(5)
2.5.1. HTL product analysis
Gas Chromatography-Mass spectroscopy (GC–MS) was used to
analyze the bio-oil product derived from hydrothermal liquefaction
(HTL) of residual R. opacus biomass. The GC–MS (Perkin Elmer Clarus
600) was equipped with a column of dimensions (60 m × 250 µm) and
maintained at an initial oven temperature of 50 °C for 2 min and then
ramped to 300 °C with a heating value of 10 °C. The injector and de-
tector temperatures were set to 280 °C and 300 °C, respectively. For
analysis, samples were dissolved in methanol and injected at the split
ratio 10:1 with He as the carrier gas. Individual compounds present in
the bio-oil product were identified by comparing the GC–MS results
obtained with those available in the MS library of pure standard com-
pounds.
Functional groups present in the bio-oil product were identified by
Fourier Transform Infrared Spectroscopy (FTIR) using a FTIR system
(Spectrum series, Perkin Elmer, USA) with a range of 500–4000 cm−1.
T. Paul, et al.
Fig. 1. Schematic showing different reactor modes operated along with continuous with cell recycle mode of operation integrated with tubular ceramic membrane.
2.6. Analytical methods
2.6.1. Biomass growth, COD and total lipids
R. opacus biomass concentration (g/L) was determined by mea-
suring optical density of the culture at 660 nm wavelength using a
UV–Vis spectrophotometer (Agilent Technologies, Cary 100 series,
Singapore). Cell Dry Weight (CDW) was determined by lyophilizing
centrifuged biomass (10,000×g, 10 min) from 10 ml of its culture broth
and weighing.
For COD estimation, samples were centrifuged at 10,000×g for
10 min followed by digestion of the supernatant for 2 h at 120 °C (DRB
200, HACH, USA) and analysis as per the Standard Methods.
Lipid concentration (g L−1) in the R. opacus biomass was de-
termined by standard Folch method [17]; 1 g of dry biomass was
homogenized with 20 ml mixture of chloroform-methanol (2:1, v/v)
and the mixture kept for 20 min in an orbital shaking incubator under
ambient room temperature. The liquid phase was then recovered and
washed with 20% (v/v) of MilliQ water followed by vortexing and
centrifuging at 3500×g for separation into two distinct phases. The
upper aqueous phase was siphoned off whereas the lower chloroform
phase containing lipids was evaporated using a rotary evaporator
(Rotavapor R3, Buchi, Switzerland). The extracted lipids were finally
quantified by gravimetric analysis.
2.6.2. Microscopic analysis of R. opacus biomass after recycle
To examine any change in surface morphology of the R. opacus
biomass after the cell recycle operation, Field Emission Scanning
Electron Microscopy (FESEM) of the biomass was performed. For
sample analysis, 1 ml of a sample was taken before and after completion
of the continuous cell recycle experiment and was centrifuged at
10,000×g for 10 min followed by washing with sterile water (MilliQ).
Following centrifugation, the pellet obtained was suitably diluted and
mixed properly with MilliQ water and placed in the specimen stub and
coated with gold for analysis under FESEM (Zeiss, Sigma, Germany).
79
Chemical Engineering Journal 367 (2019) 76–85
3. Results and discussion
3.1. Wastewater treatment and lipid rich biomass production using R.
opacus
3.1.1. Batch, fed-batch and SBR bioreactor operation strategies
Refinery wastewater treatment and lipid rich biomass production by
R. opacus was first carried out using the bioreactor experiments oper-
ated under different modes as batch, fed-batch and sequential batch
(SBR). Table 1 presents the recently published reports on lipid pro-
duction and refinery wastewater treatment with bioreactor operated
under different modes. The wastewater was supplemented with MSM
(4:1) and inoculated with 10% (v/v) of R. opacus seed culture prior to
the batch operation. The results of these three modes of operation with
the bioreactor are shown in Fig. 2. Fig. 2a, shows the combined profile
of biomass growth (OD), lipid accumulation (g L−1) and COD removal
(%) in the batch operated bioreactor. Maximum biomass concentration
obtained was 1.5 g L−1 along with a lipid accumulation of 0.83 g L−1
(55%, w/w). Moreover, a maximum COD removal of 86% was achieved
in the batch experiments which confirms that the R. opacus could effi-
ciently degrade the toxic hydrocarbons present in refinery wastewater
for its metabolism and growth. Thus, 55% (w/w) of lipids were accu-
mulated in the biomass which indicates its potential for bio-oil pro-
duction from refinery wastewater. Saisriyoot et al. [10] studied bio-fuel
production by R. opacus grown on petroleum wastewater, and reported
52% (w/w) of lipid accumulation with wastewater supplemented with
molasses as a substrate. Interestingly, the lipid accumulation value
obtained in the present study is better in terms of the yield, in presence
of wastewater as the sole carbon source substrate [17].
Based on the results obtained using the batch operated bioreactor,
the biomass yield of the bacteria (Yx/s) was estimated to be 0.62 (Eq.
(2)). By using Eq. (1), feed flow rate required for fed-batch operation
was calculated and feeding was initiated at the end of 24 h batch
period. From Fig. 2b, which shows the results of the fed-batch experi-
ment in bioreactor, the maximum biomass growth and lipid accumu-
lation were 2.3 g L−1 and 1.8 g L−1, respectively. Approximately, 78%
(w/w) of lipids were accumulated in R. opacus biomass under the fed-
batch operation mode, which is higher than the value obtained in the
T. Paul, et al. Chemical Engineering Journal 367 (2019) 76–85

Malakahmad et al. (2011) [33]

Wiszniowski et al. [26]

Saisriyoot et al. [10]

Gargouri et al. [11]

Thakur et al. [19]

This study
Reference
COD removal (%)

∼99
93
95
80

83
–
Lipid (%, w/
w)

54

86
–
–
–

–
Recently published reports on lipid production and petroleum refinery wastewater treatment using bioreactor operated under different modes.

Continuous cell-recycle with

Plug flow membrane

membrane system
Bioreactor mode

Fed-batch
CSTR
SBR

SBR
Mixed Culture (Rhodospirilium-like bacteria, Gomphonema-like algae,
and sulfate reducing-like bacteria)

Fig. 2. Time profiles of biomass production, lipid accumulation and COD re-
Municipal activated sludge

Rhodococcus opacus PD630

Rhodococcus opacus PD630

Sewage activated sludge

moval by R. opacus in the bioreactor operated under (a) batch, (b) fed-batch and
(c) sequential batch mode.
Contaminated soil
Microorganism

batch bioreactor experiments. Moreover, the COD removal efficiency is

found to be nearly 95% after 54 h of treatment which is relatively
higher than the removal efficiency obtained in the batch operated re-
actor. The increased biomass production is mainly due to an increase in
the volume of the wastewater treated under fed-batch mode than under
batch mode. However, a discontinuous trend in lipid production is
observed, which could be attributed to lipid utilization as a secondary
Raw petroleum wastewater supplemented with

carbon source by the bacteria [18]. Disintegration of the lipid rich R.

opacus biomass could be another reason for sudden decrease in the lipid
concentration at the end.
Synthetic petrochemical wastewater
Real petroleum refinery wastewater

Raw petroleum refinery wastewater

Real petroleum refinery wastewater

SBR is based on fill-and-draw batch system which is conventionally

Synthetic refinery wastewater

applied for treating wastewater containing high amount of recalcitrant

compounds [19,20]. In this study, sequencing batch mode was carried
out for a total cycle period of 18 h (HRT) consisting of 2 h of filling, 15 h
molasses (20 g/L)

reaction time, 45 min settling and 15 min of decanting phase with SRT
of 3 days. During the reaction phase, the organic load was maintained
Carbon source

at 3.5 g L−1. Fig. 2c shows poor performance of SBR mode than the
batch/fed-batch mode in terms of lipid accumulation. Maximum bio-
Table 1

mass concentration, lipid production and COD removal efficiency in

these experiments were 1.87 g L−1, 1.23 g L−1 and 87%, respectively.

80
T. Paul, et al.
Fig. 3. Time profiles of biomass production, lipid accumulation and COD re-
moval by R. opacus during the continuous bioreactor experiments at different
HRTs: (a) 8 h (b) 16 h and (c) 24 h.
Approximately, 66% (w/w) of lipids were produced under the SBR
mode which is less than that obtained under fed-batch mode of op-
eration. Moreover, biomass settleability was poor due to disintegration
of the biomass. Also, inhibition of microbial activity due to hydro-
carbons present in such wastewater may result in poor settling of the
biomass. Thakur et al. [19] performed aerobic degradation of petro-
leum wastewater by activated sludge in SBR, and observed maximum
83% of COD removal efficiency at 0.83 d HRT (19 h) with 2 h of filling
time; a further increase in the HRT decreased the COD removal effi-
ciency. In the present study conducted at 18 h HRT and 2 h filling time,
the COD removal efficiency obtained is 87% which is higher than the
literature reported value. These results reveal exceptional efficiency of
the bacteria R. opacus for wastewater treatment under SBR mode, but
not in terms of lipid production.
In the above mentioned bioreactor experiments, the lipid
81
Chemical Engineering Journal 367 (2019) 76–85
Fig. 4. Time profiles of biomass production, lipid accumulation and COD re-
moval by R. opacus during the continuous with cell recycle experiment at 16 h
HRT.
accumulation rate ultimately decreased which contributes to the fact
that bacteria may utilize lipids as a substrate in the absence of primary
carbon source [6,21]. Moreover, no significant growth in the biomass
was observed after 110 h due to inhibition in the growth by the residual
COD, which however needs to be confirmed further. A similar ob-
servation was made by Thakur et al. [19] where COD removal effi-
ciency decreased or was maintained on increasing the HRT from 19 to
24 h during the treatment of petroleum refinery wastewater in SBR
mode. Gargouri et al. [11] operated an aerobic CSTR with petroleum
wastewater by using activated sludge for 225 days and reported max-
imum 95% of COD which is relatively lower than the value obtained in
this experimental study (98%).
3.1.2. Continuous mode of operation
For improving the bioreactor performance, biodegradation of hy-
drocarbons and lipid production was studied under continuous (CSTR)
operating mode of the bioreactor. The fermenter was initially operated
for 24 h under batch mode and then shifted to continuous mode; in this
continuous operation mode, the effect of different HRT (8, 16 and 24 h)
was investigated which corresponded to dilution rate of 0.12, 0.06 and
0.04 h−1, respectively. Fig. 3 shows the combined profile of biomass
growth (OD), lipid production (g L−1) and COD removal (%) efficiency
at different HRTs with continuously operated bioreactor. Fig. 3a–c
shows an increasing trend in the biomass concentration from 1.7 to
2.9 g L−1 along with an increase in the HRT, beyond which the biomass
concentration slightly decreased to 2.7 g L−1. More than 95% of COD
was removed under the continuous mode for an optimum HRT of 16 h
(dilution rate 0.06 h−1). Maximum accumulated lipid content in the
biomass of 1.1 g L−1 (64%, w/w), 2.35 g L−1 (81%, w/w) and
2.16 g L−1 (80%, w/w) were obtained at 8, 16 and 24 h HRT, respec-
tively. Thus, 16 h (dilution rate 0.06 h−1) HRT prove to be optimum for
wastewater treatment, lipid production and biomass growth. However,
biomass with lipid accumulation exhibited poor settling characteristics
similar to that observed in the previous bioreactor experiments. Re-
sidual COD in the treated wastewater was 180 mg L−1 which was suc-
cessfully achieved following treatment. Hence, a novel bioreactor
strategy based on continuous operation with cell recycle using in-
digenous tubular ceramic membrane was effectively examined to im-
prove the process efficiency and biomass separation.
3.1.3. Continuous mode with cell recycle using low cost tubular ceramic
membrane
For enhancing the process efficiency by biomass recycle in the
bioreactor following continuous mode of treatment, the bioreactor was
operated at the same HRT of 16 h and by integrating it with a low cost
tubular ceramic membrane. Prior to the experiments, the membrane
flux was calculated which was found to be 2.39 × 10−5 m3/m2s under
T. Paul, et al. Chemical Engineering Journal 367 (2019) 76–85

Fig. 5. FESEM images of R. opacus biomass collected from (a) recycle stream and (b) effluent stream of the bioreactor operated in continuous with cell recycle mode.

Fig. 6. GC–MS analysis results showing hydrocarbon profile of the wastewater at (a) 0 h and (b) 168 h of treatment under continuous with cell recycle membrane
process.

Fig. 7. Germinated Cicer arietinum L. seeds in-

cubated with (a) distilled water (b) tap water (c) raw
refinery wastewater (untreated) and (d) permeate
(treated water) for their eco-toxicity analysis.

a pressure of 68 kPa. For better assessment, continuous feeding of re- production along with COD removal from the wastewater [6]. More-
finery wastewater and simultaneous recycling of the residual biomass in over, membrane process (organic/inorganic) have been widely used as
the effluent to the fermenter was carried out following the micro-fil- reported in the literature [22,23,26,32]. The results of the continuous
tration step. Cell-recycling helps in achieving improved lipid mode with cell recycle are shown in Fig. 4, which depicts the biomass

82
T. Paul, et al. Chemical Engineering Journal 367 (2019) 76–85

Fig. 8. GC–MS analysis results showing different compounds present in the bio- oil obtained by HTL of the residual R. opacus biomass.

Table 2
Major compounds present in the bio-oil identified by GC–MS analysis of the bio-
oil product.
Retention Compound Area% Formula
Time (min)

30.81 5-Methyl-Z-5-Docosene 0.06 C23H46

30.82 1-Dodecanol 0.29 C13H28O
31.48 2-isopropyl-5-Methyl-1-Heptanol 0.04 C11H24O
32.93 1-Octanol 0.26 C12H26O
33.83 Decane 0.73 C20H42O
34.74 IsoHexylOctadecyl Ester 0.20 C26H50O4
35.33 DotriacontylPentafluoropropionate 1.93 C35H65O2F5
36.64 OctatriacontylTrifluoroacetate 1.82 C40H77O2F3
37.14 OctacosylHeptafluorobutyrate 2.30 C32H57O2F7
37.70 Pentatriacontane 3.25 C35H72
38.15 Trihexadecyl Borate 3.31 C48H99O3B
38.42 Tetrapentacontane 2.13 C54H110
38.74 Trihexadecyl Borate 4.68 C48H99O3B
Fig. 9. FTIR spectra of the bio-oil produced by HTL of the residual R. opacus
38.91 2-Methyl-Z-4-Tetradecene 3.39 C15H30
biomass.
39.33 2-Ethylhexyl Octadecyl Ester 4.66 C26H54O3S
39.66 10-Methylnonadecane 4.56 C20H42
39.67 Octadecyl 2,2,2-Trichloroethyl Ester 4.52 C21H39O3Cl3
treated petrochemical wastewater with a membrane bioreactor and
39.96 TetratriacontylHeptafluorobutyrate 2.78 C38H69O2F7
40.19 5-Methyl-Z-5-Docosene 3.84 C23H46
obtained a high COD removal efficiency of 78–98%, which is lower
40.62 1-Pentacontanol 12.16 C50H102O than the value obtained in this study. Wiszniowski et al. [26] evaluated
41.40 HexatriacontylTrifluoroacetate 3.84 C38H73O2F3 the performance of a plug-flow membrane bioreactor for treatment of
41.62 [bi-1,4-cyclohexadien-1-yl]-3,3′,6,6′- 12.51 C16H14O6 petroleum wastewater and observed 93% of COD removal which is
Tetrone
distinctly less than that obtained in the present work. Moreover, the
41.69 Methyl 2-Hydroxy-Pentacosanoate 9.81 C26H52O3
42.70 Cyclohexane 2.15 C14H28 lipid production value of 89% (w/w) obtained in this study was without
42.73 Methyl 2-Hydroxy-16-Methyl- 4.17 C19H38O3 any addition of extra carbon source and the value is quite high com-
Heptadecanoate pared to that reported in the literature by Saisroyoot et al. [10].
42.98 5-Methyl-Z-5-Docosene 4.68 C23H46
Fig. 5a and b reveals no major change in surface morphology of the
43.29 Methyl 2-Hydroxy-Octadecanoate 1.98 C19H38O3
43.58 Cyclohexane 2.47 C28H54
R. opacus biomass taken from the cell recycle streams and effluent
43.48 9-Octadecene 2.67 C39H78O2 sample port of the membrane integrated bioreactor system; the biomass
43.97 Dodecyl Cis-9 4.62 C30H58O3 was intact (∼2 µm) without any visible damage to it. The figure also
44.62 I-Propyl 10-Methyl-Dodecanoate 3.44 C16H32O2 indicates agglomeration of the bacterial biomass following its passage
through the membrane for cell recycle; the cells were, however, rela-
tively less disrupted prior to its recycle into the reactor.
concentration (OD), lipid accumulation (g L−1) and COD removal (%)
GC–MS analysis was performed to determine the hydrocarbon pro-
efficiency profile with time (h). As shown in Fig. 4, maximum biomass
file in the influent and effluent streams of the continuous with cell re-
concentration and lipid production of 3.6 g L−1 and 3.2 g L−1 (89%, w/
cycle system. The analysis was conducted using a sample of the un-
w), respectively, were obtained. Moreover, complete removal of COD
treated wastewater (0 h) and treated wastewater (168 h). Fig. 6 shows
was achieved in the permeate obtained from the membrane system.
the abundant presence of petroleum hydrocarbon compounds in the
Membrane bioreactor has been already reported for toxicity removal
influent (0 h) whereas the amount was greatly reduced in the effluent
from wastewater due to its potential to be used in low sludge load and
stream (168 h) which reveals that the compounds were efficiently de-
high sludge age which supports the survival of the bacteria even in the
graded by R. opacus in the bioreactor and completely eliminated in the
presence of toxic compounds in the wastewater [24,25]. Hence, these
effluent through the membrane treatment step. These results are found
results distinctly demonstrate that the continuous with cell recycle is
to be in agreement with those reported in the literature [11].
best suited for achieving maximum COD removal from wastewater
Germination Index (GI) of Cicer arietinum L. was used in this study to
along with high lipid production using R. opacus.
test the toxicity of the untreated (raw) and treated wastewater obtained
Gupta et al. [6] studied real time lipid production from dairy was-
from the continuous with cell recycle experiment. The selection of the
tewater using R. opacus and reported highest lipid accumulation of
seed is owing to its high sensitivity towards toxic pollutants [27] and its
nearly 3.4 g L−1 (78% CDW) in continuous with cell recycle experi-
potential for protein digestibility following germination. The GI index is
ments, but without membrane integrated system. Yaopo et al. [25] first
a sensitive parameter for evaluating toxicity assessment and is used to

83
T. Paul, et al.
monitor changes in phyto-toxicity of the wastewater following its
treatment. Using Eq. (4), GI Index of the tap water, permeate (treated)
and untreated (raw refinery wastewater) was calculated with distilled
water as the control. Fig. 7 shows the germinated seeds for GI calcu-
lation. A GI of 73.63% with the permeate water proves that the treated
wastewater is free from any harmful phyto-toxic agents, whereas the
untreated (raw) refinery wastewater showed a G1 % of 34.63. However,
tap water showed a high GI of 91.30%.
3.2. Assessment of bio-oil yield and characterization
R. opacus biomass obtained in the experiments was subjected to
hydrothermal liquefaction for bio-oil production; in this process a high
temperature is used to degrade R. opacus and the carbonaceous matter
is converted into single monomer units. Addition of solvent at the end
of the process leads to the separation of two phases (aqueous and oil
phase) with different solubility in water. The yield of bio oil obtained
was thus calculated based on Eq. (5) and was found to be 18% (w/w)
which is relatively lower than yield obtained from microalgae by hy-
drothermal liquefaction [28].
The bio-oil derived in this study contained a mixture of complex
hydrocarbons as shown by GC–MS analysis of the product (Fig. 8). The
chief constituents present in the bio-oil are presented in Table 2 which
clearly reveals a number of complex organic hydrocarbons with dif-
ferent structure and molecular weights were present in the oil phase.
These compounds can be classified into different groups such as mono
aromatics, aliphatics, polyaromatics and saturated/unsaturated fatty
acids [29].
FTIR analysis of the biomass revealed an identical spectra of the
produced bio-oil from R. opacus biomass using refinery wastewater
(Fig. 9). The OeH stretching vibrations at 3435 cm−1 shows the pre-
sence of alcohols and phenols. Most importantly the vibrations at
2853.86 and 2924 cm−1 is related to the presence of methylene groups
from alkanes or fatty acids such as n-hexadecaric acid and docosene
[30]. Furthermore, the carbonyl signal of ketones, aldehydes are ob-
tained at 1631.84 cm−1, which denotes the presence of fatty acids
arising from the degradation of lipids. The absorbance peak at
1384.43 cm−1 reveals the presence of alcohols, ester and phenolic
ethers. The presence of all these functional groups match with the re-
sults reported in the literature on bio-oil from micro-algae produced by
using HTL process [28]. The chemical functional groups identified by
FTIR analysis also matched with the compounds detected by GC–MS
analysis. Also, the techno-economic feasibility analysis of the process
for bio-oil production can be carried out in future for evaluating its bio-
fuel potential, as reported in the literature [31].
4. Conclusions
This study demonstrated the robust nature of the R. opacus for si-
multaneous wastewater treatment and lipid production in bioreactor
operated under different modes. Highest lipid accumulation and com-
plete COD removal was achieved in continuous with cell recycle mode
using a tubular ceramic membrane suggesting that the integrated ap-
proach was best suitable for achieving maximum lipid production from
refinery wastewater for subsequent bio-oil production by HTL. In ad-
dition, GC–MS analysis and eco-toxicity analysis of the wastewater re-
vealed almost complete removal of toxic compounds from the waste-
water suggesting this process to be highly feasible.
Acknowledgements
The authors thank the Centre for the Environment, IIT Guwahati, for
providing the necessary facilities to carry out this research work. The
authors also acknowledge Central Instrument Facility (CIF), IIT
Guwahati for FESEM analysis. The financial support by the Department
of Science and Technology, Government of India (DST/TM/WTI/2K16/
84
Chemical Engineering Journal 367 (2019) 76–85
11(G)), for this research work is gratefully acknowledged.
References
[1] J. Fargione, J. Hill, D. Tilman, S. Polasky, P. Hawthorne, Land clearing and the
biofuel carbon debt, Science 319 (2008) 1235–1238.
[2] S.C. Davis, R.M. Boddey, B.J. Alves, A.L. Cowie, B.H. George, S.M. Ogle, P. Smith,
M. van Noordwijk, M.T. van Wijk, Management swing potential for bioenergy
crops, GCB Bioenergy 5 (2013) 623–638.
[3] A. Daverey, K. Pakshirajan, S. Sumalatha, Sophorolipids production by Candida
bombicola using dairy industry wastewater, Clean Technol. Environ. Policy 13
(2011) 481–488.
[4] L. Goswami, M.T. Namboodiri, R.V. Kumar, K. Pakshirajan, G. Pugazhenthi,
Biodiesel production potential of oleaginous Rhodococcus opacus grown on biomass
gasification wastewater, Renewable Energy 105 (2017) 400–406.
[5] A. Rezvanpour, R. Roostaazad, M. Hesampour, M. Nyström, C. Ghotbi, Effective
factors in the treatment of kerosene–water emulsion by using UF membranes, J.
Hazard. Mater. 161 (2009) 1216–1224.
[6] N. Gupta, A. Manikandan, K. Pakshirajan, Real-time lipid production and dairy
wastewater treatment using Rhodococcus opacus in a bioreactor under fed-batch,
continuous and continuous cell recycling modes for potential biodiesel application,
Biofuels 9 (2018) 239–245.
[7] L. Goswami, R.V. Kumar, N. Arul Manikandan, K. Pakshirajan, G. Pugazhenthi,
Anthracene biodegradation by oleaginous Rhodococcus opacus for biodiesel pro-
duction and its characterization, Polycyclic Aromat. Compd. (2017) 1–13.
[8] L. Goswami, R.V. Kumar, N.A. Manikandan, K. Pakshirajan, G. Pugazhenthi,
Simultaneous polycyclic aromatic hydrocarbon degradation and lipid accumulation
by Rhodococcus opacus for potential biodiesel production, J. Water Process Eng. 17
(2017) 1–10.
[9] L. Martinkova, B. Uhnakova, M. Patek, J. Nesvera, V. Kren, Biodegradation po-
tential of the genus Rhodococcus, Environ. Int. 35 (2009) 162–177.
[10] M. Saisriyoot, T. Sahaya, A. Thanapimmetha, Y. Chisti, P. Srinophakun, Production
of potential fuel oils by Rhodococcus opacus grown on petroleum processing was-
tewaters, J. Renewable Sustainable Energy 8 (2016) 063106.
[11] B. Gargouri, F. Karray, N. Mhiri, F. Aloui, S. Sayadi, Application of a continuously
stirred tank bioreactor (CSTR) for bioremediation of hydrocarbon-rich industrial
wastewater effluents, J. Hazard. Mater. 189 (2011) 427–434.
[12] H.J. Janben, M.H. Ibrahim, D. Broker, A. Steinbuchel, Optimization of macro-ele-
ment concentrations, pH and osmolarity for triacylglycerol accumulation in
Rhodococcus opacus strain PD630, AMB Express 3 (2013) 38.
[13] B.K. Nandi, A. Moparthi, R. Uppaluri, M.K. Purkait, Treatment of oily wastewater
using low cost ceramic membrane: comparative assessment of pore blocking and
artificial neural network models, Chem. Eng. Res. Des. 88 (2010) 881–892.
[14] R.V. Kumar, A.K. Ghoshal, G. Pugazhenthi, Elaboration of novel tubular ceramic
membrane from inexpensive raw materials by extrusion method and its perfor-
mance in microfiltration of synthetic oily wastewater treatment, J. Membr. Sci. 490
(2015) 92–102.
[15] L.G. Alba, C. Torri, D. Fabbri, S.R. Kersten, D.W.W. Brilman, Microalgae growth on
the aqueous phase from hydrothermal liquefaction of the same microalgae, Chem.
Eng. J. 228 (2013) 214–223.
[16] T.K. Vo, O.K. Lee, E.Y. Lee, C.H. Kim, J.W. Seo, J. Kim, S.S. Kim, Kinetics study of
the hydrothermal liquefaction of the microalga Aurantiochytrium sp. KRS101, Chem.
Eng. J. 306 (2016) 763–771.
[17] H.M. Alvarez, R. Kalscheuer, A. Steinbüchel, Accumulation and mobilization of
storage lipids by Rhodococcus opacus PD630 and Rhodococcus ruber NCIMB 40126,
Appl. Microbiol. Biotechnol. 54 (2000) 218–223.
[18] A. Mahavi, Sequencing Batch Reactor: A Promising Technology in Wastewater
Treatment, University of Theran, 2008, pp. 79–90.
[19] C. Thakur, V.C. Srivastava, I.D. Mall, Aerobic degradation of petroleum refinery
wastewater in sequential batch reactor, J. Environ. Sci. Health 49 (2014)
1436–1444.
[20] A.R. Rajab, M.R. Salim, J. Sohaili, A.N. Anuar, S.K. Lakkaboyana, Performance of
integrated anaerobic/aerobic sequencing batch reactor treating poultry slaughter-
house wastewater, Chem. Eng. J. 313 (2017) 967–974.
[21] H. Xu, X.L. Miao, Q.Y. Wu, High quality biodiesel production from a microalga
Chlorella protothecoides by heterotrophic growth in fermenters, J. Biotechnol. 126
(2006) 499–507.
[22] J. Zhong, X. Sun, C. Wang, Treatment of oily wastewater produced from refinery
processes using flocculation and ceramic membrane filtration, Sep. Purif. Technol.
32 (2003) 93–98.
[23] S. Gonzalez, M. Petrovic, D. Barcelo, Removal of a broad range of surfactants from
municipal wastewater–comparison between membrane bioreactor and conven-
tional activated sludge treatment, Chemosphere 67 (2007) 335–343.
[24] S. Ishak, A. Malakahmed, M.H. Isa, Refinery wastewater biological treatment – a
short review, J. Sci. Ind. Res. 71 (2012) 251–256.
[25] F. Yaopo, W. Jusi, J. Zhaochun, Treatment of petrochemical wastewater with a
membrane bioreactor, Acta Sci. Circum. 1 (1997).
[26] J. Wiszniowski, A. Ziembińska, S. Ciesielski, Removal of petroleum pollutants and
monitoring of bacterial community structure in a membrane bioreactor,
Chemosphere 83 (2011) 49–56.
[27] S. Divyapriya, D. Dimi, K.P. Deepthi, Biochemical effect of industrial effluence on
germinating seeds of Cicer arientum, Int. J. Pharm. Pharm. Sci. 6 (2014) 538–542.
[28] D.L. Barreiro, B.R. Gómez, U. Hornung, A. Kruse, W. Prins, Hydrothermal lique-
faction of microalgae in a continuous stirred-tank reactor, Energy Fuels 29 (2015)
6422–6432.
T. Paul, et al.
[29] D.R. Vardon, B.K. Sharma, G.V. Blazina, K. Rajagopalan, T.J. Strathmann,
Thermochemical conversion of raw and defatted algal biomass via hydrothermal
liquefaction and slow pyrolysis, Bioresour. Technol. 109 (2012) 178–187.
[30] S. Zou, Y. Wu, M. Yang, C. Li, J. Tong, Bio-oil production from sub-and supercritical
water liquefaction of microalgae Dunaliella tertiolecta and related properties, Energy
Environ. Sci. 3 (2010) 1073–1078.
[31] Y. Zhang, T.R. Brown, G. Hu, R.C. Brown, Techno-economic analysis of two bio-oil
85
Chemical Engineering Journal 367 (2019) 76–85
upgrading pathways, Chem. Eng. J. 225 (2013) 895–904.
[32] M. Waszak, M. Gryta, The ultrafiltration ceramic membrane used for broth se-
paration in membrane bioreactor, Chem. Eng. J. 305 (2016) 129–135.
[33] A. Malakahmad, A. Hasani, M. Eisakhani, M.H. Isa, Sequencing Batch Reactor (SBR)
for the removal of Hg2+ and Cd2+ from synthetic petrochemical factory waste-
water, J. Hazard. Mater. 191 (2011) 118–125.