Sustainable Energy Technologies and Assessments 47 (2021) 101475

Contents lists available at ScienceDirect

Sustainable Energy Technologies and Assessments

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

Developing efficient nutrient removal and resource recovery strategy

towards synergistic MLW treatment using macroalgae in a flat
panel photobioreactor
Akanksha Mhatre-Naik a, Gayatri Pillai a, Prashant Savvashe a, Mahesh Navale a, Juilee Palkar a,
Arvind M. Lali b, Reena Pandit a, *
a
DBT-ICT-Centre for Energy Biosciences, Institute of Chemical Technology, Matunga, Mumbai 400019, India
b
Department of Chemical Engineering, Institute of Chemical Technology, Matunga, Mumbai 400019, India

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

Keywords: The present study conceived a novel technology – ‘MARiNE photobioreactor (PBR): MacroAlgae Remediates
Bioremediation Nutrients for Energy in PBR’. The technology targets the simultaneous treatment of desalination effluent and
Desalination municipal liquid waste. The present study showcases the proof of concept wherein Ulva lactuca, a marine green
Macroalgae
macroalga, well-reported for its pollutant scavenging properties, has been explored for the synchronized treat­
Photobioreactor
ment of two effluents, namely desalination effluent and municipal liquid waste and making it safe for marine
Ulva lactuca
discharge. A total nutrient removal of 94.1%, 80.4%, and 100% for COD, TN, and TP was achieved at an op­
timum HRT of 12 h with the highest biomass productivity of 340 g DW m− 2 d-1. The study demonstrates MARiNE
PBR to be a biologically cohesive strategy for rapid MLW remediation. A presumptive proposal for biorefinery of
the generated biomass which enables recycling of the captured nutrients from the effluents is also proposed.

Introduction treatment efficiency reduces to 38% in upper-middle-income countries

and to 24% in lower-middle-income countries. The extent of treatment is
UN-Water report 2015 revealed that over 80% of municipal liquid critically reduced in low-income countries wherein only 8% of MLW
waste (MLW) is released untreated to the world’s waterways, leading to undergo treatment. Recovering the nutrients, water, and other precious
health hazards and environmental pollution [46]. Contamination of resources from the MLW is the opportunity to be seized. Currently, most
surface and groundwater indirectly aggravates the scarcity of potable technologies for MLW management are unviable for developing econ­
water. Untreated or partially treated MLW discharged to the marine omies. Moreover, challenges in MLW recycle have further exacerbated
environment accumulates on the ocean-surface owing to its positive the demand for freshwater owing to urbanization and the rising
buoyancy. This triggers eutrophication of ocean-surface leading to loss population.
of vegetation, deoxygenation, and subsequent death of under-dwelling The rising freshwater demand has probed many countries to adopt a
marine fauna. Currently, the wastewater outfall has caused dead-zones desalination approach [36]. Globally, 90 million m3 of water is desali­
inflating over an area about the size of the European Union nated every day by over 18,500 desalination plants that produce ~54
(~4,500,000 km2) [7]. The increasing dead-zones have severely affected million m3 of hypersaline effluent. The hypersaline effluent, when dis­
the yields of aquaculture, pushing the livelihood of over 500 million charged untreated into the ocean sinks to the ocean floor due to its
people in jeopardy. Thus, untreated MLW has indirectly impacted negative buoyancy. This results in acute salinity change causing osmotic
maricultural and related business, bringing about an annual loss of 1 shock to the sensitive marine flora-fauna. Nonetheless, many desalina­
billion USD, with the potential to upswing if the marine outfall continues tion facilities have adopted strategies of co-discharging the hypersaline
at the unabated rate. The management of MLW depends critically on the brine with untreated or partially treated MLW to meet the ocean’s
economic liberty of the countries to bear the treatment cost. On average, salinity. However, discharging a mixture of hypersaline brine and un­
most high-income countries treat almost 70% of their MLW, while the treated (or partially treated) MLW cannot be referred to as a full-proof

* Corresponding author.
E-mail address: [email protected] (R. Pandit).

https://doi.org/10.1016/j.seta.2021.101475
Received 27 August 2020; Received in revised form 8 June 2021; Accepted 1 July 2021
Available online 29 July 2021
2213-1388/© 2021 Elsevier Ltd. All rights reserved.
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

solution, as the mixed effluent still contains a reasonable amount of total and biorefinery-based bioproducts.
dissolved solids, organic pollutants, nitrogen, and other potential nu­ In the current study, a remediation technology titled ‘MARiNE PBR’
trients that trigger eutrophication and successive chain of environ­ which expands as ‘MacroAlgae Remediates Nutrients for Energy in
mental hazards [16]. Owing to the salinity from the hypersaline brine, Photobioreactor’ is proposed. Herein, the potential of macroalgae
the resultant mixture becomes incompatible for MLW treatment using U. lactuca for synergistic treatment of hypersaline brine and municipal
conventional remediation techniques such as activated sludge process liquid waste in a flat-panel PBR is evaluated. Preliminary experiments
(ASP) and up-flow anaerobic bioreactor (UASB) [16]. Moreover, these were undertaken to evaluate treatment efficiency by U. lactuca in
conventional methods are ineffective in removing the eutrophication comparison to other methods like activated sludge process (ASP),
triggering pollutants. Thus, the challenge is to develop a strategy that is anaerobic digestion (UASB), down flow hanging sponge (DHS) treat­
not only compatible but also efficient in transforming the co-effluent ment methods (ref Table 1). Thereafter, Response Surface Methodology
stream, making it safe for disposal. (RSM) was applied for optimisation of growth parameters such as
In recent years, green macroalgae have been frequently reported for inoculum density and irradiance to reduce the effective treatment
their high bioremediation potential in removing excess nitrogen, phos­ duration. Subsequently, nutrient dosing along with CO2 supplementa­
phorous, carbon, as well as heavy metals [11]. Seaweeds (marine mac­ tion was developed to maximize nutrient removal efficiency and
roalgae) are efficient in reducing nutrients and at the same time provide simultaneous biomass production.
extra income, when species of economic importance are used (Lab­
abpour, 2014). Removal of eutrophication pollutants N and P by using Materials and methods
macroalgae is clean and also yields biomass which can be used for
various applications. Among several marine macroalgae, Ulva lactuca is Procurement and maintenance of U. lactuca
acclaimed for its high growth rates, nutrient uptake, and CO2 seques­
tration [25]. It is also known to be an abundant and ubiquitous species U. lactuca biomass was collected from the intertidal zone in the
among coastal benthic communities. Many studies show that Ulva spp, Arabian Sea from Uran beach (18◦ 0.9′ N, Longitude 72◦ 0.9′ E), Mumbai,
grows uninhibited in polluted seawater, and sequesters pollutants in the India. The cultures were maintained in natural seawater (collected from
biomass [11]. These attributes make U. lactua, a choice of the candidate Dadar-Mahim coast, Mumbai, India, 19.03◦ N 72.825◦ E) with MP1
while developing any processes dealing with global scale remediation media. The composition of MP1 media includes NaNO3 (total nitrogen,
involving saline waters. However, being a photosynthetic organism, TN = 17 ppm), Na2HPO4⋅12H2O salt (total phosphate, TP = 2 ppm), and
U. lactuca requires irradiance and natural diurnal cycles to be functional. salts of trace metals and other micronutrients [41]. In the present study,
This makes conventional treatment infrastructure involving closed and MP1 media was used for optimisation studies and also as the control
opaque reactors impractical for macroalgae-based MLW treatment. media to compare the growth of U. lactuca while growing in effluents
MLW remediation by macroalgae can offer an added economic advan­ wherein additional nutrients were not provided.
tage as the biomass generated is rich in a plethora of commodity
chemicals and energy molecules [27,44]. Hence, it is required to Preparation of salinity amended water (SAW) treatment
develop a suitable infrastructure that facilitates the cultivation of
U. lactuca along with MLW treatment. Recently, we have demonstrated SAW was prepared by mixing municipal liquid waste (MLW) and
year-round cultivation of U. lactuca in pilot-scale flat panel photo­ synthetic hypersaline brine. The MLW (post sedimentation) was
bioreactor (F-PBR) with the estimation of potential biorefinery products collected from the residential complex at the Institute of Chemical
[25]. Thus, F-PBR becomes a convenient prototype to evaluate the Technology, Mumbai, India. The synthetic hypersaline brine was pre­
sustainability of the proposed MLW treatment. In previous reports, the pared by adding mineral sea-salt (common salt) to natural seawater to
application of the PBR system for biomass production has been debated make up the total salinity to 55 ppt (the salinity of hypersaline brine can
due to high Capital Expenditure (CapEx.) and Operational Expenditure range between 52 and 90 ppt depending on the source of the brine). The
(OpEx.) Further, most reports discourage biomass cultivation solely for MLW was mixed with synthetic hypersaline brine in 3:2 (volume: vol­
biofuel production for the additional cost involved in nutrient media and ume) ratio (hypersaline brine: MLW) (for every 1000 ml, 600 ml brine
freshwater. However, these economic drawbacks can be overcome by and 400 ml MLW was taken) such that the salinity of the resultant
striking a synergy between MLW remediation, high biomass production, mixture i.e. SAW was 35 ppt. Since SAW contains MLW that is bound to

Table 1
Comparison of current study to other effluent treatment methods spanning macroalgae based techniques and conventional method.
Wastewater Macroalgae Initial Treatment N P COD HRT BP Reference
Characteristic used density condition (d) g dw/
Initial RE Initial RE Initial RE
g/L m2/d
(mg/L) (%) (mg/L) (%) (mg/L) (%)

SAW Ulva lactuca 0.25 Flat panel PBR 20.5 96 5 83.1 237 94.08 4 30-36 Present
under natural light study
Primary settled Oedogonium sp. 0.25 Open ponds under 27.2 62 5.04 75 31 57 20 7-10 Park et al
sewage natural light [29]
Chaetomorpha 0.22 Bioreactor under 39.5 97.6 1.55 79.1 205 NA 12 10.7 Neveux
linum LED light et al.[14]

Wastewater Treatment method N P COD HRT Reference

(h)
Initial RE Initial RE Initial RE
(mg/L) (%) (mg/L) (%) (mg/L) (%)

Raw sewage Activated sludge process (ASP) 41.7 72.6 - - 576.5 96.1 12 S. Ge and P. Champagne
[42]
Anaerobic digestion (UASB) + Down flow 40.6 55.9 - - 599.6 89.7 8.5 S. Ge and P. Champagne
hanging sponge (DHS) [42]
UASB - - - - 540 88 12 M. Tandukar [35]

RE-Remediation efficiency. Sludge retention time (SRT) has not been considered during calculation of treatment capacity. Data shown as the average value of the
initial concentration of pollutant (or nutrient, viz., N, P and COD) in case where range of concentration value was available.

2
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

have inherent microorganisms it was imperative to check the influence spiking (HDC-S). It involved nutrient spiking (MP1 media nutrients)
of these microorganisms on the growth of U. lactuca. Hence a portion of along with the optimized high stocking density of 5 g L-1. The nutrients
SAW was autoclaved at 121 ◦ C and referred to as autoclaved-SAW (a- were spiked at an interval of 12 h. The amount of MP1 media required to
SAW), the remaining portion was termed as raw-SAW (r-SAW). reset the initial nutrient concentration (TN = 17 ppm, TP = 2 ppm) was
U. lactuca was grown in both a-SAW and r-SAW to compare the influence provided. Another strategy involved high density culture with nutrient
of micro-organisms in SAW on the growth of the biomass. All effluents spiking and biomass harvesting i.e., HDC-SH involved harvesting the
namely MLW, synthetic brine, a-SAW, and r-SAW were characterized for biomass after every 12 h. After harvesting the biomass, the culture
pH, salinity, TN, TP, COD, dissolved oxygen (DO), and presence of co­ density was restored to 5 g L-1 along with nutrient spiking. Yet another
liforms. The characteristics of MLW, synthetic brine, r-SAW, and a-SAW strategy was nutrient spiking, biomass harvesting, and CO2 supple­
are shown in supplementary material Table S1. mentation i.e. HDC-SHC. The media was enriched with CO2 by sparging
pure CO2 gas to achieve a concentration of 1000 ppm of CO2 by volume.
Experimental conditions These strategies were directed to reduce the effects of nutrient depletion,
light attenuation due to biomass crowding, and pH deviation. Growth
All of the studies were carried out in acrylic-based flat panel pho­ conditions were maintained as given in section 2.3. The media samples
tobioreactors (PBR) (0.32 m long × 0.12 m wide × 0.62 m in height) were collected in all the strategies for analysis of TN, TP to estimate
with a working volume of 20 L. Two air blowers were used to provide nutrient uptake rates.
aeration. Air was purged at the rate of 1 VVM through a sparger
(perforated pipe) installed longitudinally along the bottom centerline of Analysis of nutrient uptake efficiency in SAW
the PBR. These blowers also facilitated the suspension of U. lactuca
fronds in the PBR. The studies were carried out from September to June Based on the previous studies HDC-SHC strategy was found to be
under natural diurnal light in a state-of-the-art glasshouse facility, optimal for nutrient uptake and biomass productivity in MP1 media.
referred to as an environmental laboratory (EL). Since EL facilitates Hence further experiment was designed with HDC-SHC strategy using
penetration of natural light, the average natural irradiance on the sur­ SAW as the media. This experiment was carried out to analyze the
face of PBR was 1000 µmole photons m− 2 s− 1. EL was equipped to nutrient uptake efficiency of U. lactuca in SAW for efficient remediation
maintain the laboratory temperature at 25 ± 2 0C. Daily growth rate of SAW. U. lactuca was cultivated in SAW with the optimized parameters
(DGR, %), biomass productivity (BP, g DW m− 2 d-1), nutrient uptake (in of HDC-SHC strategy with a hydraulic retention time (HRT) of 24 h.
terms of TN, TP, COD) were the parameters evaluated during the ex­ Samples were collected at a frequency of 6 h to analyze TN, TP, and COD
periments throughout the study. for estimation of nutrient uptake rates at different time intervals.

Preliminary investigation for potential of U. lactuca for treatment of SAW Analytical estimations

This study was carried out to analyze the effect of innate composition Measurement of growth, biomass productivity, CO2 sequestration, and net
and micro-organisms in SAW on the growth and remediation potential of energy ratio
U. lactuca. To evaluate the potential of U. lactuca to treat SAW, it was Daily growth rate (DGR %) was calculated from the fresh weight
grown using a-SAW as well as r-SAW in the PBR. U. lactuca was also using the following standard equation (1),
grown in MP1 media (control media). During preliminary investigation,
[( )1t ]
U. lactuca was grown with a stocking density (SD) of 0.2 g L -1 for a wt
DGR (%) = − 1 × 100 (1)
period of 4 days. The reduction of COD, TN, and TP of a-SAW and r-SAW w0
and the control media was evaluated for nutrient removal rate (mg g− 1
DW L-1 d-1), removal efficiency (% d-1), and total nutrient removed (%) where W0 is the initial weight at time t = 0, Wt is the weight after time t
along with biomass productivity and daily growth rates. [15]. Biomass productivity, (g L-1 d-1 was calculated from the fresh
weight as follows;
Optimisation studies to improve nutrient uptake efficiency
(x2 − x1 )
Biomass productivity = , (2)
(t2 − t1 )
Since the preliminary investigation established the potential of
U. lactuca to treat SAW, optimisation studies were set up in order to where X1 and X2 were the biomass concentration (g FW L-1) on days t1
improve nutrient uptake. It has been reported that the nutrient uptake and t2, respectively.
efficiency of macroalgae is affected by initial SD [34] as well as by the Estimation of CO2 sequestration rate and net energy ratio (NER)
irradiance [26]. Hence, in the present study, two independent factors, i. The rate of CO2 sequestration (PCO2) was calculated as shown in Eq
e. SD and irradiance were modelled using Response Surface Methodol­ (3)
ogy (RSM) to achieve optimum response on biomass productivity, ( )
nutrient uptake, and pH of the media. Also, the RSM studies were carried PCO2 = Cc *P
MCO2
(3)
using MP1 media instead of SAW to avoid obscure variability from MLW Mc
composition. The results of these studies were then extended to establish
the improvement in treatment efficiency and reduction of HRT. where Cc is the carbon content in the dry biomass (gcarbon g-1
biomass), P is

Based on the leads received from RSM studies, high stocking density the biomass productivity in g m− 3 d-1, M CO2 is the molecular weight of
(HSD) of 5 g L-1 was considered to be optimal to improve nutrient uptake carbon dioxide and Mc is the molecular weight of carbon [2].
rate for efficient treatment of SAW. However, during RSM studies we The net energy ratio of the flat panel PBR system was calculated
encountered a lack of nutrients, biomass crowding, and an increase in according to Eq. (4)
pH while working with HSD which could be the potential hindrances in Wnet EB
reducing HRT of the treatment. Hence, strategies such as pulse feeding NER = (4)
Eaux
(nutrient spiking at intervals), daily biomass harvesting, and mainte­
nance of pH were explored with SD 5 g L-1 for further improvement of herein, Wnet is the net biomass produced over the total cultivation area,
treatment efficiency. Here again, MP1 media was used instead of SAW to EB is the higher heating value (HHV) of the dry biomass for U. lactuca
avoid obscure variability from MLW which is 11.7 MJ/kg, Eaux is the auxiliary energy input required in
The first strategy adopted was high density culture with nutrient pumping and cooling [38].

3
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

Proximate and ultimate analysis of biomass Fv/Fm parameter was calculated as (Fm-Fo)/Fo after 20 mins of dark
The sulfo-phospho-vanillin method was used for lipid determination adaptation [37]. Non-photochemical quenching of fluorescence (NPQ),
as explained in the previous paper of the freeze-dried sample in was estimated according to Bilger and Björkman (Bilger et al., 1990).
concentrated sulphuric acid, using canola oil as standard (Barnes and
Blackstock, 1973).
CHNS/O Analyzer (Thermo Scientific TM FLASH 2000 CHNS/O Theoretical analysis
Analyzers) was used to determine protein content from freeze-dried
samples. The nitrogen to protein conversion used was 5 (Angell et al., Theoretical annual yields of biomass, protein, biocrude, bioethanol, and
2016). biomethane
The modified two-step acid hydrolysis procedure developed at the Analysis of seasonal variation in biomass was done by comparing the
National Renewable Energy Laboratory (NREL, Golden, CO) was used proximate and ultimate composition of biomass harvested each month
(Sluiter et al., 2011) to estimate the carbohydrate content of the algal from September to July.
biomass, and this acid-hydrolysed biomass was analysed by the phe­ Biocrude production involves a complex reaction cascade through
nol–sulphuric acid method (Dubois et al., 1956) to estimate the total hydrothermal liquefaction (HTL). The equation from the previous study
carbohydrate content (Agarwal et al., 1956). on hydrothermal liquefaction of Ulva sp. to biocrude [30] was used to
Moisture content was determined using halogen moisture analyzer calculate the theoretical yield of biocrude produced at 330–345 ◦ C for 5
(Mettler Toledo, HE73 115 V Model). The dry weight and ash were min. The biocrude yield (Ybiocrude, wt%) is according to the Eq (8)
calculated using Eq (5) and (6) respectively as explained in our previous Ybiocrude = 0.885 × Cbiomass − 7.455 (8)
paper [27].
The biomethane potential from biomass of U. lactuca was calculated
m3 − m2
DW = (5) using the Buswell equation (Eq. (9)) [9].
m1 − m2
( )
where, m1 (g) is the combined weight of the biomass dried at 40 ◦ C n
2
+ a8 − b4 − 3c8 × 22.4
and weight of the crucible, m2 (g) is the weight of the ceramic crucible Ymethane = (9)
12n + a + 16b + 14c
alone and m3 (g) is the combined weight of the biomass samples and
weight of ceramic crucible after 105 ◦ C treatment. wherein, the alphabets in the above equation correspond to the empir­
(
m4 − m2
) ical formula of the biomass i.e. CnHaObNc.
Ash(%) = × 100 (6) The theoretical yield of bioethanol was calculated from the mono­
m1 − m2
saccharide and reducing sugar concentration in Ulva biomass as per the
where, m4 (g) is the weight of the biomass after 550 ◦ C treatment. study optimised by Trivedi et al. [45].
HHV value was estimated using Eq (7)
( )
MJ Projected economic impact of MARiNE PBR system: sensitivity analysis
HHV = [(0.3536 × FC) + (0⋅1559 × VM) ] − (0.0078 × Ash) (7) To assess the potential value of macroalgal biomass at a commercial
kg
scale, it was converted from g m− 2 d-1 into t ha− 1 yr− 1. The projected
where VM is the volatile mass (%), FC is the fixed carbon content from values of bioproducts from the biomass per hectare per year was
elemental analysis (%) and A is the ash content (%) [31]. calculated from the current market value (two-year average values
observed in year 2018 and 2019) in US$. The value of WTI grade crude
Measurement of COD, TN, TP, and bacterial count analysis oil (58 US$ bbl− 1) was converted to tons− 1 using the specific gravity of
0.83 at 15 ◦ C. The barrel to tons− 1 conversion was considered assuming
COD, TN, and TP analysis. The COD analysis was performed using a that one barrel contains 158.987 L. The price of biocrude ($ ton-1) was
single lot of premixed digestion solution containing K2 Cr2 O7 as an then calculated from the current price of the WTI crude oil and from the
oxidation agent along with other reagents (HACH COD Digestion Kit). specific gravity of 0.97 at 15 ◦ C and adjustment for volume. The pro­
(The media samples for COD analysis were filtered through a 0.2 μm jected theoretical estimation of the value of protein extract was calcu­
filter). The analysis was conducted according to the manufacturer’s in­ lated from the value of soybean meal assuming that there are 50% (w/w)
structions and the readings were then measured using a spectropho­ amino acids in the soybean meal (390 US$ ton-1) whereas protein extract
tometer at 600 nm. with Shimadzu spectrophotometer, UV-2550, Japan. contains theoretically about 100% amino acids (w/w). The estimate
TN in media was estimated by the Total Nitrogen Measuring -Labo­ value of bioethanol (52.9 US$ bbl− 1) was converted to tons− 1 from the
ratory unit (TNM-L, Shimadzu, Japan) according to the standard pro­ specific gravity of 0.789 at 20 ◦ C. The barrel-to-ton conversion was
tocol defined by the manufacturer. based on the assumption that one barrel is equivalent to 158.98 L of
Total phosphate (Total-P) was determined colorimetrically at 610 ethanol. The value of biogas (2.78 US$ mmBTU-1) was converted to m− 3
nm (Shimadzu spectrophotometer, UV-2550, Japan) using the ammo­ to tons− 1 using the conversion factor of 1 MMBtu is equivalent to 28.26
nium heptamolybdate and malachite green [10]. m3 of natural gas. The projected price of biomass, protein, biocrude,
The TOC in the media was analysed using Total Organic Carbon- bioethanol, and biogas was calculated individually. Further, sequen­
Laboratory analyzer, (TOC-L Analyzer, Shimadzu, Japan). tially, the production of bioproducts obtained after extraction of protein
and conversion of protein-removed biomass to biofuel was determined.
Viable bacterial counts. Bacteriological examination of raw SAW was The sensitivity analysis tool was used to analyze the effect of seasonal
variation in biomass and composition on the yield and price of bio­
carried out on the samples taken at the start (0hr) and end of the
treatment. The estimation of total coliform and fecal coliform was done products. Firstly, the growth performance around the year was
analyzed. The lowest biomass was yielded in June and the highest
by determining Most Probable Number (MPN) per liter using the mul­
tiple tube fermentation technique (APHA, 2005). Further confirmation biomass productivity was observed in January. Yields for the biomass
productivity were thus defined with optimum being the center and the
was done using MacConkey’s agar plates.
annual average of the present study. Minimum is defined as the, left of
the center and corresponding to the average values from June, and
Chlorophyll-a fluorescence measurement and photosynthetic efficiency
Chlorophyll fluorescence was determined in vivo at the exponential maximum is defined as the right of the center corresponding to average
values during January. Similarly, yields were defined for the protein,
phase of growth using Dual pam 100 fluorometer from Walz Germany.
biocrude, bioethanol, and biogas. The price value of the biomass and

4
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

bioproducts were calculated from the market price of reference prod­

ucts: soybean meal, biocrude oil, bioethanol, and natural gas. The
average price of the product from last two years (obtained from the
Market Insiders [24] was defined as the optimum (centre), while the
lowest price was defined as the minimum (left of the centre) and the
highest price was defined as the maximum (right of the centre). The
projected value of the biomass (US$ ha− 1 yr− 1) from the sequential
extraction of protein followed by conversion of residue to biofuel was
calculated as in Eq (10).
( )
3.65 × P × Ybf − aa × Pricebf + (Waa × Priceaa )
Biomass value = (10)
100
Herein, biomass productivity in g m− 2 d-1 was converted to t ha− 1 y-1,
using the conversion factor 3.65. P is the biomass productivity in t ha− 1
y-1, dry weight. Ybf-aa is the biocrude or bioethanol or biogas yield
(weight %) after protein extraction, Pricebf is the average price of the
respective biofuel product during two years (US$ t− 1), Waa is the protein
content of U. lactuca (annual average) and Priceaa is the average price of
soybean meal (US$ t− 1) in last two years.

Statistical analysis
Fig. 1. Fresh weight profile of U. lactuca in raw and autoclaved (r- and a-)
All data is represented as mean ± standard deviation. A one-way salinity amended water in comparison to control media.
analysis of variance (ANOVA) was performed to confirm significant
differences in results. Multiple comparison test by Tuckey’s honest sig­ when mixed with the hypersaline stream. Not all macroalgae are resis­
nificance difference (HSD) was carried out to find out significant dif­ tant to inhibitory factors in MLW. M.A. Doblin et al., reported the
ferences at P = 0.05 in response from controls. deleterious effect of secondary treated sewage effluent on the growth
and reproduction of two brown macroalgae [13]. In the present study,
Results and discussion the SAW composition didn’t interfere with the regular growth and
reproduction of the biological agent, thus strengthening our proposed
The hypersaline brine effluent from the desalination process is usu­ concept of treatment of SAW using macroalgae U. lactuca. Apart from
ally blended with untreated municipal liquid waste (MLW) to amend the compositional stress, multiple pathogenic microorganisms are reported
salinity and prevent potential damage to the ecosystem. However, this to cause disease to macroalgae leading to sporadic bleaching and loss of
method is an incomplete treatment, as the added MLW stream still biomass [22]. Thus, it was important to study the effect of the microbial
carries organic material which may trigger eutrophication. Thus, addi­ population in SAW on U. lactuca by comparing the growth profiles of r-
tional treatment is obligatory to quench the nutrients from the co- SAW and a-SAW. The fresh weight trendline of r-SAW closely resembled
effluent stream before it is being discharged. The characteristic of that of a-SAW (see Fig. 1). Also, the DGR and biomass productivity of
MLW and hypersaline brine is complex, undefined, and dynamic. Hence, U. lactuca in r-SAW (62.7 ± 2.1% d-1 and 30 ± 3 g DW m− 2 d-1, p > 0.05)
the extent of bioremediation of the co-effluent stream will depend on the was comparable to that in a-SAW (61.3 ± 1.3% d-1 and 29 ± 3.1 g DW
interaction of the remediating agent and the inherent organic, inorganic, m− 2 d-1, p > 0.05) (Refer to Supplementary Table S2). These results
and biological factors. In the present study, macroalgae U. lactuca is imply that microorganisms present in SAW do not interrupt the growth
proposed as an agent for synergistic remediation of salinity amended of U. lactuca. Reports on MLW treatment using algae suggest that algal
MLW (SAW) prepared from MLW and hypersaline brine via a potential density and growth rate increase pH and dissolved oxygen concentration
MARiNE PBR technology. Although the proposed concept seems of the MLW causing inactivation of microorganism in MLW [3]. This
straightforward, a comprehensive preliminary study is mandatory to could be the reason that the microorganisms did not interfere with
understand the response of U. lactuca towards SAW as a lack of knowl­ growth in r-SAW. The preliminary studies ascertained that the nutrient
edge may pose ambiguous challenges during the scale-up of MARiNE composition, as well as microflora of SAW, did not impede the growth
PBR technology. rate and biomass accumulation of U. lactuca. Having understood this,
the capacity of nutrient removal must be assessed to determine the
Preliminary investigation on potential of U. lactuca for treatment of SAW remediation potential of U. lactuca.
The nutrient removal rate (mg g− 1 DW L-1 d-1), removal efficiency (%
U. lactuca is reported to be an opportunistic macroalga resistant to -1
d ), and total nutrient removed (%) were evaluated to examine the
high nutrient concentration and heavy metals in MLW [40]. To evaluate reduction of chemical oxygen demand (COD), total nitrogen (TN), and
the potential of U. lactuca to treat SAW, it was grown simultaneously in total phosphate (TP) of SAW. The comparison of nutrient removal pa­
r-SAW, a-SAW, and MP1 media. Autoclaved SAW (a-SAW) was used as a rameters for r-SAW, a-SAW to that of control media is shown in sup­
microbial control to understand if the inherent microorganism’s con­ plementary Table S2. COD, an important indicator of effluent treatment,
sortium interferes with the growth and remediation potential of was reduced significantly by 94.1 % in both r-SAW and a-SAW (p value
U. lactuca. It is noteworthy that r-SAW, as well as a-SAW, are with > 0.01). Moreover, TN- and TP- removed in r-SAW (96 ± 2.1% of TN and
complex and undefined nutrients which may or may not be conducive 83.1 ± 2.1% of TP) was equivalent to that in a-SAW (shown in supple­
for U. lactuca. Therefore, growth of U. lactuca in r-SAW and a-SAW was mentary Table S2). Further, values of TN and TP removal in r-SAW
compared with defined culture media -MP1 media (nutrient control). closely corresponded to that of nutrient control (96.3 ± 2.1% of TN and
Fig. 1 depicts trendline for the fresh weight growth of U. lactuca in 83.6 ± 2.1% of TP), suggesting that U. lactuca is capable of assimilating
SAW and nutrient control media. The graph shows that the organic nutrients from the co-effluent (SAW) stream. The TN and TP removal
nutrients present in SAW are amenable to U. lactuca and thus support its efficiency of U. lactuca in the present study is comparable to that of algae
growth. This also indicates that the MLW stream in SAW is either devoid based reports on wastewater remediation [5]. The algal based treatment
of toxic growth inhibitory factors or its strength may have been diluted

5
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

processes offer merit over the conventional methods (e.g. activated from 0.1 to 10 g L− 1 were evaluated to study its effect on nutrient uptake
sludge process) which are inefficient in TN and TP removal and require efficiency. Being a photosynthetic organism, the nutrient uptake by
post-treatment strategies to meet the permissible limits for discharge. U. lactuca is also critically affected by irradiance. The range of irradiance
Preliminary studies demonstrated the potential of U. lactuca as a bio- from 50 to 1000 µE m− 2 s− 1 was considered for optimisation studies as
absorber which needs to be exploited to develop a novel and efficient this scale spans the annual average irradiance in sub-tropical regions.
MLW treatment method. Thus, the effect of two-independent variables viz., stocking density, and
The efficiency of a MLW treatment method can be determined by irradiance were studied to determine the response on biomass produc­
comparative analysis with the existing treatment technologies. A tivity, N-uptake rate, and on pH of the media. An additional response-
detailed correlation of macroalgae based and other parallel treatment pH was selected in the study as the growth and nutrient uptake by algae
methodologies are shown in Table 1. Among macroalgae based treat­ is reported to be severely limited at elevated pH [33]. With multiple
ments, the current study with U. lactuca stands superior in terms of N-, P- factors to be optimised across a wide range, Design Expert software v.11
and COD removal efficiency as the time required to remove the equiv­ was used to develop design of the experiment (DOE) and conduct
alent concentration of nutrient was significantly lower. For instance, screening via response surface methodology. All optimisation studies
Oedogonium species removed 16.9 ppm of TN in 20 days [29] while in were carried out in flat panel photobioreactor and in nutrient media-
the present study, U. lactuca remediated 19.7 ppm of TN within 4 days of MP1 to avoid obscure variability from MLW composition.
treatment. Similar efficiencies were recorded for TP and COD removal
by U. lactuca. In contrast to other macroalgae, the rate of nitrogen and Response surface methodology (RSM)
phosphorus removal by U. lactuca is reported to be high in domestic and The mathematical and statistical model of the interaction of stocking
aquaculture MLW [20]. However, none of these studies discuss the po­ density and irradiance was created through RSM. In this study, I-optimal
tential of macroalgae to deal with COD reduction. Extending the design was applied to achieve optimal response from the experiment.
remediation potential, the current study presents a pioneering work According to the I-optimal, 13 experimental trials were conducted for a
featuring the potential of macroalgae for treating MLW streams. range of irradiance level (50–1000 µE m− 2 s− 1) and stocking densities
The aforementioned results from the preliminary treatment desig­ (0.1–10 g L-1). The control for the experiment was cultures grown at a
nate U. lactuca as a potential candidate for macroalgae based MLW density of 0.25 g/L at 500 ± 50 µE m− 2 s− 1 of light. The translation of the
treatment with the lowest hydraulic retention time (HRT) i.e. 4 days. coded levels is represented in Table 2.
However, as indicated in Table 1., the conventional treatment methods The model derived in the current study was validated by thoroughly
such as activated sludge process (ASP) and up-flow anaerobic sludge reviewing the response of individual factors: stocking density and irra­
digestion (UASB) have low HRT ranging from 8.5 to 12 h. Low HRT diance on biomass productivity, nitrogen uptake rate, and pH of the
enables larger working volumes of MLW thus improving treatment ef­ media. The experimental data for the response of biomass productivity
ficiency. Hence, U. lactuca based treatment can qualify as the potential and pH was fitted to the quadratic model while the response data of N-
treatment method only if the HRT is reduced significantly. uptake was fitted to the linear regression model. Analysis of variance
While reducing the HRT, one of the major challenges in algae-based (ANOVA) for the response indicate that the assumed regression models
MLW treatment systems is the limited scope for altering the height: were significant and valid for each of the responses (p < 0.01). Thus, the
diameter ratio of the reactors. Algae being photosynthetic organisms are ANOVA results indicate that the tested model and equation can be used
heavily dependent on light (natural/artificial) driven machinery and for creating response surface plots to visualise the relationship between
thus require reactor systems supporting light penetration. This severely the factors and their response on U. lactuca.
limits the maximum volume to which the photobioreactor can be scaled The final coded regression equation for the response (Y1 = biomass
up by altering dimensions alone. Unlike conventional STP tanks/re­ productivity, Y2 = N-uptake and Y3 = pH) is represented in Eqs. (14),
actors, changes in dimensions of photobioreactors to accommodate (15) and (16) for factors A (Irradiance) and B (stocking density).
larger volumes, can turn out to be disproportionate and severely affect
Y1 = (+150) + 55.9*A + 36.9*B + 19.4*AB – 59.9*A2 – 65*B2 (14)
light penetration. Also, in MLW treatment using microalgae, the HRT of
the process has been reported to improve by increasing the initial Y2 = (+40.75) + 38.77*A + 8.45*B (15)
inoculum density of the culture and maintaining the pH of the medium
[23]. According to Pereira et al. stocking density of the culture and
2
Y3 = (+8.59) + 1.23*A + 0.4830*B + 0.4429*AB + 0.7516*A –
incident irradiance have been observed to significantly affect the 0.2707*B2 (16)
pollutant removal rate in the macroalgae Pyropia sp [34]. In the present
In these equations, the positive sign of the coefficient indicates that
study, with the stocking density, 0.2 g L− 1 the nutrient uptake efficiency
the response is synergistically influenced by the factors whereas the
of U. lactuca in r-SAW was 19.2% d-1 for TN, 16.6% d-1 for TP, and
18.8% d-1 for COD (see supplementary Table S2). With this stocking
density, it needed 4 days for complete treatment of SAW and may not Table 2
qualify as a suitable method during scale up. Thus, with an objective to Central composite design of two independent variables and the experimentally
derived responses.
reduce treatment time, broad spectrum optimisation studies were
designed with varying stocking densities (inoculum densities) of U lac­ Run A: Irradiance B: Stocking Productivity N- pH
tuca. Further, the interplay of irradiance and stocking density on the density uptake
2 − 1
µE m− s g L− 1 g DW m− 2
d− 1
% d− 1
growth and nutrient uptake efficiency in U. lactuca was also evaluated.
With strategies focussed on reducing HRT of pollutant removal, present 1 658 0.1 34.5 29.8 8.1
2 786.25 6.535 120 40.52 10.2
study can advance towards competent technology.
3 381.921 2.30483 67 20.3 8.1
4 857.5 10 147 68.23 10.23
Optimisation of growth parameters to improve nutrient uptake efficiency of 5 50 5.7925 0 0 8
U. lactuca 6 458.5 5 165 25.63 8.3
7 458.5 5.743 150 25.6 8.3
8 1000 3.7135 120 86.3 10.2
The nutrient uptake efficiency of macroalgae is affected by initial 9 50 10 0 0 8
stocking density [34]. Further, the optimum initial stocking density for 10 458.5 10 117 27.3 8.5
maximal nutrient uptake may vary among species, and hence, detailed 11 458.5 5.743 157 25.33 8.5
optimisation studies are mandatory before targeting intensive applica­ 12 1000 5 180 87.9 10.5
13 50 1.585 0 0 8
tion. In the current study, a wide range of stocking densities ranging

6
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

Biomass productivity (g DW m− 2 d-1) = (-71.17) + (0.355*Irradiance) +

negative sign indicates antagonist effect. Hence, from Eq. (14), it was
(30.47 * Stocking density) + (8.23 × 10-03 * Irradiance * Stocking density) –
observed that biomass productivity was greatly influenced by individual
(2.6 × 10-04 * Irradiance2) – (2.65 * Stocking density2) (17)
factors stocking density and irradiance and also by the interaction be­
tween them. In Eq. (15), the N-uptake is influenced by the individual N-uptake (% d-1) = (-10.72) + (0.082*Irradiance) + (1.7 * Stocking density)
factors. Being a linear equation, it appears that there was no interaction (18)
between the factors. In Eq. (16), the individual factors- irradiance and
stocking density along with the interaction between them greatly pH = (+7.87) - (1.85 × 10-03 *Irradiance) + (0.11 * Stocking density) + (1.88
influenced the pH of the nutrient media. The polynomial regression × 10-04 * Irradiance * Stocking density) + (3.3 × 10-06 * Irradiance2) – (0.01 *
equations (17), (18), (19) are presented below with the actual values of Stocking density) (19)
the coefficients. The response surface plots derived from the above regression equa­
tions are shown in Fig. 2. This allows graphical visual observations of the

Fig. 2. The contour (left panel) and the 3D graphs (right panel) for studying interaction of initial stocking density (SD) and irradiance and their response on biomass
productivity (a, b), nitrogen uptake rate (c, d) and pH (e, f). The graphs are generated through I-optimal analysis by the Design Expert software.

7
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

interaction between the factors and the resultant response. The contour shading effect’. Moreover, the pH of the media spiked from 8.3 to 10.5
plots and 3D graphs in Fig. 2, reveal that, the trend of observed re­ which is detrimental to the growth. The above observations suggest that
sponses increased significantly with increase in irradiance. In our pre­ the productivity obtained at 5 g L− 1 (180 g DW m− 2 d− 1) may not be the
vious study, we have reported 10–20% increase in biomass productivity maximum achievable productivity. Therefore, the possible maximum
under summer intensities (200 to 2500 µ mole photons m− 2 s− 1) for biomass productivity at 5 g L− 1 density must have been masked by the
U. lactuca adapted to high irradiance [26]. In yet another study, Graci­ limitation of irradiance, nitrogen depletion, and deviation in pH. The
laria vermiculophylla exhibited maximum nutrient removal rates and polynomial regression equations (Eq., 14, 15, 16) and the model graph
biomass productivity at irradiance of 1470 ± 18 µmol m− 2 s− 1 and high (Fig. 2) also indicate the critical role of irradiance to achieve maximum
stocking density of 5 g L− 1 [1]. biomass productivity. Thus, N-uptake and biomass productivity of
The trend of response to stocking density was similar to the response U. lactuca could be improved further by preventing the shelf-shading
with irradiance wherein higher values of nutrient uptake and produc­ effect, increasing N-dosage, and maintaining pH.
tivity were observed at higher stocking densities. Higher stocking den­
sities are known to result in higher biomass productivities and nitrogen Improvement in treatment efficiencies by various strategies
uptake in macroalgae [34]. However, the optimum stocking density With an aim to prevent pre-mature termination of the treatment
varies among macroalgal species. In a study by PL Nagler et al., red process, strategies were implemented with high density culture (HDC) i.
macroalgae Gracilaria parvispora showed, biomass productivity up to 57 e. at a stocking density of 5 g L− 1. Fig. 3 shows results achieved using
g DW m− 2 d− 1 was attained at a stocking density of 2 g L− 1 [28]. In a these strategies and their performance in comparison with low density
similar study, macroalgae Chondrus crispus and Palmanaria palmata culture (LDC) i.e. 0.2 g L− 1. Fig. 3 shows that the strategies employed
exhibited the highest productivity and nutrient uptake rates at a stock­ had a significant effect on the nutrient uptake rate by U. lactuca. The TN-
ing density of 2 g L− 1 and 4 g L− 1 respectively [21]. Thus, from the uptake efficiency increased significantly (p < 0.001) from 62.08% in
reported studies as well as the current study it is evident that irradiance, HDC to 76.4% d− 1 in nutrient spiking strategy (HDC-S). Harvesting
as well as stocking density, interact closely to influence the growth strategy (HDC-SH) did not increase N-uptake efficiency when compared
response of macroalgae. to HDC-S. The highest N-uptake efficiency of 92.5% d− 1 was achieved
The point prediction tool of Design Expert software was used to with the HDC-SHC strategy attributing to the effect of carbon supple­
statistically select the best stocking density. Table 2, shows results ob­ mentation and pH maintenance. A similar trend was observed for TP
tained from combinations of the studied factors that yielded the highest (total phosphate) uptake efficiency where uptake efficiency of 92% d-1
nutrient uptake and biomass productivity. The combination of variables was achieved with HDC-SHC culture as compared to 63.7% d− 1 in HDC.
with an irradiance of 1000 µE m− 2 s− 1 and stocking density of 5 g L-1 (see Thus, with initial nutrient concentration doubled and subsequently with
Table 3, sr. no.2) resulted in nitrogen uptake efficiency of 87.9% d- the strategies adopted, the nutrient uptake efficiency of over 90% d− 1
1
which is ~2.7 times higher than that observed with preliminary was achieved. Further, the impact of increased nutrient uptake effi­
studies. These results also provide confidence that high cell densities can ciency on biomass productivity was evaluated. The biomass productivity
result in high nitrogen uptake efficiency and thus can reduce HRT increased significantly (p < 0.001, Tuckey’s HSD test) from 187 g DW
significantly. High nitrogen uptake efficiency although favourable in m− 2 d− 1 HDC to 322.2 g DW m− 2 d− 1 in HDC-SHC cultures. It has been
reducing the HRT, also indicates that the culture media was N-depleted reported that supplementation of CO2 enhances nutrient uptake effi­
within a day. This aggressive rate of N-uptake makes the model sus­ ciency and biomass productivity in macroalgae such as Gracilaria sp.,
ceptible to sporadic crashes as the cultures starve of nutrients on sub­ Hizia fusiforme, Porphyra leucostica and Ulva rigida [17]. In studies with
sequent days. In the absence of critical nutrient concentration, biomass MLW, the microalgae Chlorella vulgaris showed increased biomass
productivity of U. lactuca may be limited leading to untimely treatment accumulation and nutrient removal with the incorporation of CO2 [19].
termination. These possibilities and limitations were considered while Thus, with strategy optimised in nutrient medium, it is apparent that
designing the strategies for SAW treatment. HDC-SHC strategy may enable nutrient removal within a day. The effi­
The biomass productivity at the best combination of variables was cacy of the HDC-SHC strategy was further evaluated for remediation of
observed to be significantly high, reaching 180 g DW m− 2 d− 1 (see synergistic effluent and decrease HRT using SAW.
Table 3). This productivity is 6 times higher than the preliminary studies
wherein the stocking density was 0.2 g L− 1. As apparent from the pro­
ductivity with a stocking density of 5 g L− 1, the biomass density Conceptualizing MARiNE PBR technology for remediation of SAW
increased by almost 3 times (~15 g L− 1), resulting in biomass crowding (salinity amended water)
which in turn led to attenuation in light intensity due to the ‘shelf-
Among photobioreactor systems, a flat panel PBR (F-PBR)

Table 3
Best combination of studied parameters yielding maximum productivity and N-uptake.
Sr.no Irradiance Stocking density Response variables Observed response Predicted response SD Mean
µE m− 2 s− 1 g L-1 SE

1 458 5 Productivity 165 140.7 30.6 14.9

Nitrogen uptake 45.6 35.2 12.6 3.6
pH 8.3 8.4 0.234 0.114

2 1000 5 Productivity 180 145.4 30.6 19.9

Nitrogen uptake 87.9 88.66 2.79 2.65
pH 10.5 10.56 0.23 0.15

3 1000 3 Productivity 120 125 30.62 21.4

Nitrogen uptake 86.3 85.46 12.5 6.5
pH 10.2 10.3 0.234 0.16

4 857 10 Productivity 147 148 30.6 26.4

Nitrogen uptake 76.3 68.2 12.5 7.3
pH 8.5 8.5 0.234 0.2

8
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

Fig. 3. The box and whisker plots of biomass productivity, nitrogen and phosphate removal efficiency for various strategies. Herein we compare results obtained
from 5 experimental replicates. The comparative analysis of LDC (low density culture, 0.2 gL-1) to HDC (high density culture, 5 gL-1), HDC-S (nutrient spiking), HDC-
SH (nutrient spiking plus biomass harvesting) and HDC-SHC (nutrient spiking, biomass harvesting and CO2 supplementation). The intensity of irradiance during the
experiment was around 1000 ± 200 μE m− 2 s− 1. The error bars indicate standard deviation. Box with the same alphabets are not significantly different (p value <
0.05, Tukey’s HSD test).

demonstrates a low land footprint and higher illumination surface area nutrient uptake, the present study offers merits over the conventional
(Mhatre et al 2017). As light plays a critical role in nutrient removal by methods as along with reduction in COD, significant reduction in N- and
macroalgae, F- PBR was selected to test the efficacy of the proposed P- can also be achieved simultaneously.
MARiNE (MacroAlgae Remediates Nutrients for Energy in PBR) PBR Having analysed the major indicators of MLW treatment, the treated
technology (Fig. 4). Further, macroalgae being a photosynthetic or­ SAW was also be evaluated for certain auxiliary yet important param­
ganism, light significantly contributes to biomass growth. High illumi­ eters to qualify for discharge to the sea.
nation surface area of F-PBR averts light attenuation at high biomass Table 5 indicates characteristics of SAW before and post treatment.
density, thus, preventing system collapse. Along with major nutrient removal, the treatment also achieved the
As observed in Table 4, the residual concentrations of COD, TN, and inactivation of faecal coliform. The reason for this may have been the
TP were significantly reduced in the samples at 12 h compared to the rise in dissolved oxygen levels during r-SAW treatment. The increased
samples at 6 h (p value < 0.0001, Tuckey’s ± SD test) signifying HRT of concentration of DO in treated SAW is a desired attribute for the marine
12 h to be better compared to the six hours. No significant change (p ecosystem. Further, CO2 enrichment supplementation, in turn, led to the
value > 0.01, Tuckey’s HSD test) in residual nutrient concentration was maintenance of pH of the treated SAW within the neutral range (7.5 ±
observed at HRT of 18 h and 24 h. The COD removal efficiency (see 0.2). Thus, the overall attributes of the treated SAW make it safe for
Table 4, under NRE) was highest at 12 h HRT, while TN and TP removal marine discharge (EPA guidelines for effluent discharge).
efficiency were maximum at HRT of 6 h. The highest value for ‘total The biomass productivity of U. lactuca with SAW treatment was 340
nutrient removed (%) for COD, TN and TP was 94.1%, 80.4%, and 100% g DW m− 2 d-1 which corresponded close to that obtained in nutrient
respectively at HRT of 12 h. There was no significant reduction in total control media. The monumental high biomass productivity indicates the
nutrient removed (%) at HRT of 18 h and 24 h (p value > 0.01, Tuckey’s critical role of CO2 enrichment in accelerating the overall growth rate to
HSD test). Thus, with HRT of 12 h, the current strategy is now at par with 48.3% d-1 compared to unoptimized conditions. Moreover, the high
conventional methods in terms of treatment time. With regards to total growth rate may have been the reason for the observed increase in total

Fig. 4. The schematic diagram of SAW treatment process with U. lactuca in flat panel MARiNE PBR.

9
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

Table 4
Treatment efficiency of U. lactuca at various hydraulic retention time (HRT).
Residual concentration (ppm) COD TN TP

HRT COD TN TP RR NRE TNR RR NRE TNR RR NRE TNR

(hours)

6 143.8 ± 10.0 ± 7 2.2 ± 1.9 9.5 ± 0.9 154.2 ± 35.8 ± 3.2 1.0 ± 0.1 198.1 ± 49.4 ± 3.7 0.2 ± 0.01 201.6 ± 49.3 ±
0.05 16 14 9.6 2.6
12 13.6 ± 0.04 3.9 ± 0.0 ± 6.6 ± 190.2 ± 94.1 ± 0.5 ± 160.5 ± 80.4 ± 0.7 0.1 ± 198.5 ± 100 ± 0
1.2 0.35 0.09 2.6 0.32 0.01 1.8 0.004 2.5
18 13.4 ± 0.02 3.9 ± 0.0 ± 0 3.8 ± 122.4 ± 94.1 ± 0.3 ± 0 104.0 ± 79.9 ± 0.1 ± 129.9 ± 0 100 ± 0
0.5 0.05 0.5 0.48 0.8 0.75 0.002
24 13.9 ± 0.01 4.6 ± 0.0 ± 0 2.3 ± 94.1 ± 0.3 93.9 ± 0.2 ± 0 76.9 ± 0.7 76.7 ± 0.0 ± 100.0 ± 0 100 ± 0
0.4 0.03 0.34 0.32 0.001

O, S content and HHV for biomass from SAW treatment corresponded to

Table 5 that of nutrient control media. Thus, the MARiNE PBR model can be
Analysis of SAW treatment by U. lactuca under optimized HDC-SHC strategy and
successfully applied for SAW treatment along with biomass application
comparison with control.
in biorefinery.
Wastewater Units Untreated Treated Permissible
characteristics SAW SAW limit
SHC
Sustainability analysis of macro algae in MARiNE PBR
The sustainability of a biological model depends on the physiological
COD mg/L 237 ± 11.3 14 ± 1.4 125
state of the culture and the environmental impact of the process. The
Nitrogen 20 ± 0.6 4 ± 0.3 15
Phosphorous 5 ± 0.04 0±0 2 physiology of U. lactuca during SAW treatment was evaluated to deter­
TOC 43 ± 3.6 1.8 ± 0.3 – mine the efficacy of photosynthesis. The photosynthetic performance is
pH 6.5 ± 0.2 7.5 ± 0.2 6–8.5 determined by Fv/Fm ratio, which is a proxy for maximum photosyn­
Faecal coliforms CFU/ 50.E103 200 ± 2 1000 thetic efficiency [18]. The value of photosynthetic efficiency for a
L
Dissolved Oxygen ppm 4 ± 0.3 13 ± 1.2 >10
healthy culture is ~0.8 [32]. From Table 6, it is evident that photo­
synthetic efficiency was close to 0.8 in U. lactuca under SAW treatment.
Apart from photosynthetic efficiency, non-photochemical quenching
nutrient uptake. Macroalgae being a photosynthetic organism requires (NPQ) is a physiological parameter studied to evaluate the fate of photon
CO2, which acts as a substrate for the most important biochemical re­ energy captured by the photosynthetic organism. A high value of NPQ is
actions i.e. Calvin-Benson cycle. However, seawater has a very low level indicative of wasteful loss of photon energy as heat, reducing the flux of
of CO2 (10–15 µM) in equilibrium with the atmosphere. In the case of photon towards photosynthesis and growth [12]. In the present study,
high culture density, CO2 from the seawater CO2 is depleted at rates the NPQ value was significantly low compared to that in unoptimized
higher than the rate of CO2 dissolution from the atmosphere. Depletion conditions (see Table 6), thus supporting higher growth performance.
of inorganic carbon results increase in pH of the media leading to a
reduced growth rate [39]. In the present study, the major advantage of Environmental sustainability analysis of MARiNE PBR
the SHC strategy is the extraordinary biomass productivity of 340 g DW The environmental sustainability of MARiNE PBR was gauged by
m− 2 d-1 which translates to 1230 dry tons h− 1 y-1. As per our knowledge, calculating the CO2 capture potential and net energy ratio. CO2
this is the first report to experimentally demonstrate biomass produc­ sequestration and mitigation are under the spotlight ever since the
tivity in U. lactuca which is 10-fold higher than published reports [8]. negative effect of excess CO2 in the environment became evident. Mass
High biomass production is advantageous for biorefinery applications cultivation of algae is being promoted by the global scientific and in­
provided the proximate and ultimate composition of biomass did not dustrial community for bio sequestration of CO2. Having said that, in the
alter during the effluent treatment. As observed in Table 6, the proxi­ present study we analysed the CO2 capture potential by U. lactuca during
mate composition (carbohydrate, protein, and lipid) of U. lactuca SAW treatment. The CO2 sequestration was >2000 g m− 3 d-1 with SHC
biomass produced from SAW treatment was similar to that of nutrient strategy in both r-SAW and control media. Furthermore, we analysed the
control media. Moreover, the ultimate composition in terms of C, H, N, net energy ratio (NER, energy return on energy invested) for the current
process as we target the application of biomass for bioenergy produc­
Table 6 tion. For sustainability, a high value of NER is desired. From Table 6, it is
Biomass composition and physiological analysis of U. lactuca post-SHC treat­ observed that a significantly high NER value of 24.3 was obtained using
ment in r-SAW and control media. the HDC-SHC strategy of SAW treatment in contrast to unoptimized
Parameter r-SAW-SHC Control-SHC conditions. The NER for the HDC-SHC strategy can be justified by the
− 2 -1
fact that the net high biomass production (335 g m− 2 d-1) compensates
Productivity (g m d ) 340 ± 5 336 ± 10
Carbohydrate (wt %) 58 ± 1.4 56 ± 1.1
for the energy invested for pumping and cooling. Having developed the
Protein (wt %) 16 ± 1.2 15 ± 0.5 SHC strategy and evaluated the performance of the MARiNE PBR tech­
Lipid (wt %) 2.1 ± 0.16 2 ± 0.1 nology, it will be noteworthy to project the benefits of this scheme to
Ash (wt %) 15.78 ± 1.3 16.1 ± 0.8 large-scale desalination systems.
Moisture (wt %) 8.12 ± 1.8 10.9 ± 2.3
C (wt %) 26.5 ± 2.2 24.8 ± 1.8
H (wt %) 4.8 ± 0.2 5.2 ± 0.4 Presumptive economic and sustainability analysis of MARiNE PBR for
O (wt %) 42.45 ± 2.1 44.1 ± 2.5 biorefinery application
N (wt %) 3.3 ± 0.3 3.5 ± 0.2
S (wt %) 5.8 ± 0.1 5.4 ± 0.1
HHV (MJ) 11.7 ± 1.3 11.7 ± 1.1 In order to predict the economic viability of MARiNE PBR system at
NER 24.28 ± 0.2 24.3 ± 0.12 working volume of 2000 MLD, we analysed what could be the key cost
Photosynthetic efficiency (Rel units) 0.79 ± 0.07 0.8 ± 0.01 contributing factors required in the SAW treatment and biomass pro­
Non-photochemical quenching (Rel units) 0.25 ± 0.01 0.21 ± 0.02 duction. Further, the biorefinery application of the feedstock from
CO2 sequestration (g CO2 m− 3 d− 1) 2137.6 ± 210 2186.3 ± 233
MARiNE PBR was evaluated from the annual biomass productivity and

10
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

was used to calculate the projected yield and price of products in terms individual yields and commodity prices of protein and biofuel. Fig. 5a,
of per unit hectare by scaling production. represents value of feedstock as the sum of protein and biocrude yield.
Sensitivity analysis was used in the present study to predict the Here, the feedstock value is influenced by biomass productivity with the
relative influence of individual products of biorefinery (including potential to increase the value by 59.4%. The second most influential
biomass yields and commodity prices) on the overall value of the parameter is the biocrude yield that increases the feedstock value by
biomass or feedstock (USD h− 1 y-1). The maximum (MAX) biomass 46.2% respectively. The protein yield and commodity price of crude oil
productivity was obtained in January while the minimum (MIN) value and protein meal sparsely affected the feedstock value. The feedstock
was observed in June. The average value (AVG) of biomass productivity value in relation to biogas refinery is represented in Fig. 5b, where the
was calculated as the mean of the annual biomass productivities. The feedstock value is greatly influenced by biomass productivity (59.4%)
MAX, MIN, and AVG values for protein were deduced experimentally followed by protein yield (53.2%). Finally, we analysed the influence of
whereas, for biocrude, biogas and bioethanol was calculated from the bioethanol refinery on the feedstock value as shown in Fig. 5c. The in­
proximate and ultimate composition of the biomass (supplementary fluence of biomass productivity on feedstock value remains unchanged
Table S5). The MAX, MIN, and AVG feedstock values used for sensitivity (59.4%) and is followed by protein yield that displays the potential to
analysis in relation to biomass and biorefinery products are presented in increase the feedstock value by 50.8%. Amongst the three approaches
Supplementary Data. analysed, the highest value of feedstock obtained was ~41 Million US$
Tornado plots as seen in Fig. 5, were derived from the results of y-1 with biocrude biorefinery. Assuming that, the downstream process­
sensitivity analysis. In Fig. 5, the value of feedstock (US$ ton-1) on ing facility for protein extraction and biocrude production through hy­
sequential extraction of protein and biofuel has been plotted against drothermal liquefaction is already available, the only expenditure
required is the capital and operational expenditure for establishing PBR
system for SAW treatment and the operational cost on biomass pro­
cessing (protein and biocrude). The operational cost for protein sepa­
ration and hydrothermal liquefaction from seaweed is estimated to be
433 USD [4] per ton dry biomass and 93 USD per ton dry biomass
respectively [6].
The presumptive economic distribution for MARiNE PBR systems of
total capacity 2000 MLD is presented in table supplementary Table S5.
In our previous report, we have already reported the capital and oper­
ational expenditure for 2 KLD PBR system [25]. Hence, for calculating
the CapEx and OpEx cost for 2000 MLD system scale-up factor (φ) of 0.7
was used [43]. It is evident that the cost of treatment is 120,000 US
$/MLD which although is substantially high, the overall revenue on
biorefinery products is over 108 Million US$ (180,500 US$/MLD) thus,
making the treatment process free of cost (refer supplementary
Table S5). Thus, with overall economic and environmental sustainabil­
ity, the MARiNE PBR system can be a rewarding system to bridge the gap
between the volume of MLW produced and being treated.

Conclusion

The present study demonstrates a novel sustainable MARiNE PBR

system for synergistic remediation of hypersaline brine and municipal
liquid waste integrated with macroalgal biomass production. Response
surface analysis and broad-spectrum experiments were conducted to
accelerate treatment to an HRT of 12 h with total removal of 94.1%,
80.4%, and 100% for COD, TN, and TP along with a maximized biomass
productivity of 340 g DW m− 2 d1. Overall, MARiNE PBR can emerge to
be amputating the bottleneck of the economic feasibility of MLW
treatment and macroalgae cultivation for biofuels, without secondary
input on freshwater, land, and energy.

CRediT authorship contribution statement

Akanksha Mhatre-Naik: Conceptualization, Data curation, Formal

analysis, Investigation, Methodology, Software, Validation, Visualiza­
tion, Writing – original draft, Writing – review & editing. Gayatri Pillai:
Conceptualization, Data curation, Formal analysis, Investigation,
Methodology, Validation, Visualization, Writing – original draft,
Writing – review & editing. Prashant Savvashe: Conceptualization,
Data curation, Formal analysis, Investigation, Methodology. Mahesh
Navale: Conceptualization, Data curation, Formal analysis, Investiga­
tion, Methodology. Juilee Palkar: Formal analysis, Investigation,
Validation, Methodology. Arvind M. Lali: Funding acquisition, Project
Fig. 5. Sensitivity analysis. Cascading biorefinery with sequential extraction of
protein followed by conversion of residue to a) biocrude, b) biogas, c) bio­ administration, Resources, Supervision. Reena Pandit: Investigation,
ethanol. Variation in the value of feedstock (US$ yr− 1) for 2000 MLD system Formal analysis, Validation, Visualization, Funding acquisition, Project
associated with individual parameters while the other parameters remain same. administration, Resources, Supervision.
(The error bars in graph represent the standard deviation).

11
A. Mhatre-Naik et al. Sustainable Energy Technologies and Assessments 47 (2021) 101475

Declaration of Competing Interest recirculating aquaculture system. Aquaculture 2011;322–323:201–9. https://doi.

org/10.1016/j.aquaculture.2011.09.030.
[21] Kim JK, Duston J, Corey P, Garbary DJ. Marine finfish effluent bioremediation:
The authors declare that they have no known competing financial effects of stocking density and temperature on nitrogen removal capacity of
interests or personal relationships that could have appeared to influence Chondrus crispus and Palmaria palmata (Rhodophyta). Aquaculture 2013;414-415:
the work reported in this paper. 210–6.
[22] Kumar V, Zozaya-Valdes E, Kjelleberg S, Thomas T, Egan S. Multiple opportunistic
pathogens can cause a bleaching disease in the red seaweed Delisea pulchra.
Acknowledgement Environ Microbiol 2016;18(11):3962–75.
[23] Lau PS, Tam NFY, Wong YS. Effect of algal density on nutrient removal from
primary settled wastewater. Environ Pollut 1995;89(1):59–66. https://doi.org/
The authors acknowledge the Department of Scientific and Industrial 10.1016/0269-7491(94)00044-E.
Research (DSIR), New Delhi for the support of this project under Grant [24] Market insiders, n.d. Market Insiders [WWW Document]. URL https://markets.
Agreement No. DSIR/PACE/TDD, APPL/17/2013-2014, and also Third businessinsider.com/.
[25] Mhatre A, Navale M, Trivedi N, Pandit R, Lali AMAM. Pilot scale flat panel
phase of Technical Education Quality Improvement Programme (TEQIP- photobioreactor system for mass production of Ulva lactuca (Chlorophyta).
III) for the support. Bioresour Technol 2018;249:582–91. https://doi.org/10.1016/j.
biortech.2017.10.058.
[26] Mhatre A, Patil S, Agarwal A, Pandit R, Lali AM. Influence of nitrogen source on
Appendix A. Supplementary data photochemistry and antenna size of the photosystems in marine green macroalgae,
Ulva lactuca. Photosynth Res 2019;139(1-3):539–51. https://doi.org/10.1007/
Supplementary data to this article can be found online at https://doi. s11120-018-0554-4.
[27] Mhatre A, Gore S, Mhatre A, Trivedi N, Sharma M, Pandit R, et al. Effect of multiple
org/10.1016/j.seta.2021.101475.
product extractions on bio-methane potential of marine macrophytic green alga
Ulva lactuca. Renew Energy 2019;132:742–51. https://doi.org/10.1016/j.
References renene.2018.08.012.
[28] Nagler PL, Glenn EP, Nelson SG, Napolean S, 2003. Effects of fertilization
[1] Abreu MH, Pereira R, Buschmann AH, Sousa-Pinto I, Yarish C. Nitrogen uptake treatment and stocking density on the growth and production of the economic
responses of Gracilaria vermiculophylla (Ohmi) Papenfuss under combined and seaweed Gracilaria parvispora (Rhodophyta) in cage culture at Molokai , Hawaii
single addition of nitrate and ammonium. J Exp Mar Bio Ecol 2011;407(2):190–9. 219, 379–391.
https://doi.org/10.1016/j.jembe.2011.06.034. [29] Neveux N, Magnusson M, Mata L, Whelan A, de Nys R, Paul NA. The treatment of
[2] Adamczyk M, Lasek J, Skawińska A. CO 2 biofixation and growth kinetics of municipal wastewater by the macroalga Oedogonium sp. and its potential for the
Chlorella vulgaris and Nannochloropsis gaditana. Appl Biochem Biotechnol 2016; production of biocrude. Algal Res 2016;13:284–92. https://doi.org/10.1016/j.
179(7):1248–61. algal.2015.12.010.
[3] Ansa EDO, Lubberding HJ, Gijzen HJ. The effect of algal biomass on the removal of [30] Neveux N, Yuen AKL, Jazrawi C, Magnusson M, Haynes BS, Masters AF, et al.
faecal coliform from domestic wastewater. Appl Water Sci 2012;2(2):87–94. Biocrude yield and productivity from the hydrothermal liquefaction of marine and
[4] Asiedu A, Ben S, Resurreccion E, Kumar S. Techno-economic analysis of protein freshwater green macroalgae. Bioresour Technol 2014;155:334–41. https://doi.
concentrate produced by flash hydrolysis of microalgae. Environ Prog Sustain org/10.1016/j.biortech.2013.12.083.
Energy 2018;37(2):881–90. https://doi.org/10.1002/ep.v37.210.1002/ep.12722. [31] Nhuchhen D, Afzal M. HHV predicting correlations for torrefied biomass using
[5] Aslan S, Kapdan IK. Batch kinetics of nitrogen and phosphorus removal from proximate and ultimate analyses. Bioengineering 2017;4:7. https://doi.org/
synthetic wastewater by algae. Ecol Eng 2006;28(1):64–70. 10.3390/bioengineering4010007.
[6] Billing, J., Hallen, R., Schmidt, A., Snowden-Swan, L., 2017. Hydrothermal [32] Oquist G, Wass R. A portable, microprocessor operated instrument for measuring
Processing of Biomass Waste to Energy. chlorophyll fluorescence kinetics in stress physiology. Physiol Plant 1988;73(2):
[7] Breitburg D, Levin LA, Oschlies A, Grégoire M, Chavez FP, Conley DJ, et al. 211–7.
Declining oxygen in the global ocean and coastal waters. Science (80-.) 2018;359 [33] Park JBK, Craggs RJ, Shilton AN. Wastewater treatment high rate algal ponds for
(6371):eaam7240. https://doi.org/10.1126/science:aam7240. biofuel production. Bioresour Technol 2011;102(1):35–42.
[8] Bruhn A, Dahl J, Nielsen HB, Nikolaisen L, Rasmussen MB, Markager S, et al. [34] Pereira R, Yarish C, Sousa-Pinto I. The influence of stocking density, light and
Bioenergy potential of Ulva lactuca: biomass yield, methane production and temperature on the growth, production and nutrient removal capacity of Porphyra
combustion. Bioresour Technol 2011;102(3):2595–604. https://doi.org/10.1016/j. dioica (Bangiales, Rhodophyta). Aquaculture 2006;252(1):66–78.
biortech.2010.10.010. [35] Rizvi H, Ahmad N, Abbas F, Bukhari IH, Yasar A, Ali S, et al. Start-up of UASB
[9] Buswell AM, Mueller HF. Mechanism of methane fermentation. Ind Eng Chem reactors treating municipal wastewater and effect of temperature/sludge age and
1952;44(3):550–2. hydraulic retention time (HRT) on its performance. Arab J Chem 2015;8(6):780–6.
[10] Cogan EB, Birrell GB, Griffith OH. A robotics-based automated assay for inorganic https://doi.org/10.1016/j.arabjc.2013.12.016.
and organic phosphates. Anal Biochem 1999;271(1):29–35. [36] Roberts DA, Johnston EL, Knott NA. Impacts of desalination plant discharges on the
[11] Deniz F, Ersanli ET. An ecofriendly approach for bioremediation of contaminated marine environment: a critical review of published studies. Water Res 2010;44(18):
water environment: Potential contribution of a coastal seaweed community to 5117–28.
environmental improvement. Int J Phytoremediation 2018;20(3):256–63. https:// [37] Schreiber U, Hormann H, Neubauer C, Klughammer C. Assessment of photosystem
doi.org/10.1080/15226514.2017.1374335. II photochemical quantum yield by chlorophyll fluorescence quenching analysis.
[12] Derks A, Schaven K, Bruce D. Diverse mechanisms for photoprotection in Funct Plant Biol 1995;22(2):209. https://doi.org/10.1071/PP9950209.
photosynthesis. Dynamic regulation of photosystem II excitation in response to [38] Shen Y, Yang T, Zhu W, Zhao Y. Wastewater treatment and biofuel production
rapid environmental change. Biochim Biophys Acta – Bioenerg 2015;1847(4-5): through attached culture of Chlorella vulgaris in a porous substratum biofilm
468–85. reactor. J Appl Phycol 2017;29(2):833–41.
[13] Doblin MA, Clayton MN. Effects of secondarily-treated sewage effluent on the early [39] Shene C, Chisti Y, Bustamante M, Rubilar M. Effect of CO2 in the aeration gas on
life-history stages of two species of brown macroalgae: Hormosira banksii and cultivation of the microalga Nannochloropsis oculata: experimental study and
Durvillaea potatorum. Mar Biol 1995;122(4):689–98. mathematical modeling of CO2 assimilation. Algal Res. 2016;13:16–29.
[14] Ge S, Champagne P. Cultivation of the marine macroalgae chaetomorpha linum in [40] Sode, Sidsel, Annette Bruhn, Thorsten JS Balsby, Martin Mørk Larsen, Annemarie
municipal wastewater for nutrient recovery and biomass production. Environ Sci Gotfredsen, and Michael Bo Rasmussen. “Bioremediation of reject water from
Technol 2017;51(6):3558–66. https://doi.org/10.1021/acs.est.6b0603910.1021/ anaerobically digested waste water sludge with macroalgae (Ulva lactuca,
acs.est.6b06039.s001. Chlorophyta).” Bioresource technology 146 (2013): 426-435.
[15] Gerung GS, Ohno M. Growth rates of Eucheuma denticulatum (Burman) Collins et [41] Suto, S., 1959. Skeletonema no tame no jinkou baiyoueki. Suisan Zoushoku 7, 17-
Harvey and Kappaphycus striatum (Schmitz) Doty under different conditions in 19 (Japanese).
warm waters of Southern Japan. J Appl Phycol 1997;9(5):413–5. [42] Tandukar M, Ohashi A, Harada H. Performance comparison of a pilot-scale UASB
[16] Giwa A, Dufour V, Al Marzooqi F, Al Kaabi M, Hasan SW. Brine management and DHS system and activated sludge process for the treatment of municipal
methods: Recent innovations and current status. Desalination 2017;407:1–23. wastewater. Water Res 2007;41(12):2697–705. https://doi.org/10.1016/j.
https://doi.org/10.1016/j.desal.2016.12.008. watres.2007.02.027.
[17] Gordillo FJL, Niell FX, Figueroa FL. Non-photosynthetic enhancement of growth by [43] Taylor B, Xiao N, Sikorski J, Yong M, Harris T, Helme T, et al. Techno-economic
high CO2 level in the nitrophilic seaweed Ulva rigida C. Agardh (Chlorophyta). assessment of carbon-negative algal biodiesel for transport solutions. Appl Energy
Planta 2001;213(1):64–70. 2013;106:262–74.
[18] Govindjee G, 2004. Chlorophyll a fluorescence: a bit of basics and history. [44] Trivedi N, Baghel RS, Bothwell J, Gupta V, Reddy CRK, Lali AM, et al. An
Chlorophyll a Fluoresc. a Signat. Photosynth. Springer, Dordr. 1–42. integrated process for the extraction of fuel and chemicals from marine macroalgal
[19] Ji M-K, Abou-Shanab RAI, Kim S-H, Salama E-S, Lee S-H, Kabra AN, et al. biomass. Sci Rep 2016;6(1). https://doi.org/10.1038/srep30728.
Cultivation of microalgae species in tertiary municipal wastewater supplemented [45] Trivedi N, Gupta V, Reddy CRK, Jha B. Enzymatic hydrolysis and production of
with CO2for nutrient removal and biomass production. Ecol Eng 2013;58:142–8. bioethanol from common macrophytic green alga Ulva fasciata Delile. Bioresour
https://doi.org/10.1016/j.ecoleng.2013.06.020. Technol 2013;150:106–12. https://doi.org/10.1016/j.biortech.2013.09.103.
[20] Khoi LV, Fotedar R. Integration of western king prawn (Penaeus latisulcatus [46] Water, U.N., 2015. Water for a sustainable world. United Nations World Water
Kishinouye, 1896) and green seaweed (Ulva lactuca Linnaeus, 1753) in a closed Dev. Rep.

12