Professional Documents
Culture Documents
Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech
h i g h l i g h t s
" A methodology to recirculate medium in microalgae cultures is developed.
" The most adequate sterilization method is established.
" The recirculation reduces the demand of nutrients cutting costs in the process.
" N. gaditana continuous cultures were maintained using the recirculated medium.
" The biomass biochemical composition resulted of high interest for aquaculture.
a r t i c l e
i n f o
Article history:
Received 1 August 2012
Received in revised form 12 November 2012
Accepted 16 November 2012
Available online 28 November 2012
Keywords:
Medium recycling
Sterilization
Microalgae
Aquaculture
a b s t r a c t
Nannochloropsis gaditana is a good producer of proteins and valuable fatty acids for aquaculture. Recycling of culture medium is interesting for microalgae commercial production as it cuts costs and prevents
environmental contamination. The recycled medium must be sterilized to prevent the buildup of
unwanted metabolites and microorganisms. We tested several sterilization methods: ltration, ozonation, chlorination, addition of hydrogen peroxide and heating. Results showed that the most successful
method is ozonation lowering the bacterial load to 1.9 103 CFUs/mL, which is 1000-fold and 10-fold lower
than the supernatant obtained after harvesting and the initial ltered medium, respectively. Continuous
cultures of N. gaditana were grown using this recirculated supernatant. A maximum biomass productivity
of 0.8 g/L/d composed of 50% proteins and 40% lipids with more than 3% d.w. EPA was obtained making
this biomass very interesting for aquaculture.
2012 Elsevier Ltd. All rights reserved.
1. Introduction
The production of microalgae is essential for the commercial
cultivation of larvae of mollusks, crustaceans and sh. Microalgae
can be used directly as live food (for mollusks and crustaceans)
or indirectly as food for zooplankton species, such as the rotifer
Brachiomus plicatilis and nauplii of Artemia, used as prey for sh
larvae (Benemann, 1992). The microalgae used as food in the early
larval stages of sh should ensure the nutritional requirements of
the species, particularly in essential compounds like amino acids or
polyunsaturated fatty acids. The microalgal diet must contain high
levels of polyunsaturated fatty acids as docosahexanoic acid (DHA
22:6n3), eicosapentanoicacid (EPA 20:5n3) and araquidonic acid
(AA 20:4n6) to ensure a good larvae growth and high levels of survival (Navarro and Villanueva, 2003; Morais et al., 2005).
Corresponding author. Tel.: +34 950 015981; fax: +34 950 015484.
E-mail address: mcceron@ual.es (M.C. Cern-Garca).
0960-8524/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biortech.2012.11.061
pathogens associated with live prey, such as rotifers, in marine larval production systems or to disinfect the surface of sh eggs
(Grotmol and Totland, 2000). The ltration only eliminates particles of the size of the smallest lter and does not eliminate viruses.
A chemical sterilization method is the chlorination of water with
bleach (sodium hypochlorite) for several hours, followed by the removal of the excess of chlorine by the addition of sodium thiosulphate and aeration (Kawachi and Nel, 2005). Regarding thermal
disinfection, pasteurization has proved to be an effective method
for two species of algae: Rhodomonas lens (CCMP 739) and R. salina
(CCMP 1319) (Rhodes et al., 2008).
The use of ozone began in the 1990s, with the purication of
drinking water and ever since it has become widespread. Ozone
is used for disinfection of bottled water and in the oxidative
decomposition of harmful impurities in industrial and domestic
sewage (Martnez et al., 2011). It is a strong disinfectant with a
high oxidation potential and it is one of the most effective and frequently used for deactivating pathogens in drinking water treatments (Langlais et al., 1991). Ozone also acts against various
cellular substances including lipids, peptidoglycans, enzymes and
viruses. It is a powerful disinfectant against bacteria, fungi, etc.
and it can therefore be used to satisfy the highest disinfection standards (Tyrrell et al., 1995).
The production of microalgae at large-scale with medium recycling has not yet been commercialized. The life-cycle impacts of
large-scale microalgae productions are being debated, specially
the impact on water usage, i. e., the water consumption per hectare of land used for algal feedstock production. As far as we know,
the life-cycle of water when using seawater and recycling the
harvested water to produce microalgae in aquaculture has not
been established yet. The nearer attempts are those of Yang
et al. (2011) and Jung and Lovitt (2010) who recycled the harvested water for microalgae production with the objective to obtain biodiesel and PUFAs, respectively. Therefore, if the
supernatant obtained from the harvest of the biomass is not recirculated to the system, high amounts of nutrients are thrown out.
It is possible to reuse those nutrients by using closed systems to
recirculate the water and take the most of the nutrients not
consumed.
The biggest challenge of this issue is to develop a method to
sterilize the supernatant obtained from the harvested biomass of
microalgae cultures and to recirculate it to the system as to work
in a closed circuit. The harvesting process involves the presence
of suspended solids in the supernatant. These solids are organic
matter that accumulates over time. So it is necessary to study
the inuence of this organic matter on the microalga growth. For
example, Chlorella sp. uses organic matter to promote its growth
and the subsequent reduction of chemical oxygen demand (COD)
(Li et al. 2011). COD is an important and easily reproducible
parameter in the analysis of domestic and industrial wastewaters
in terms of the evaluation of their pollution levels. Nevertheless,
the methodology for its determination is not well adapted to seawater so the results are not reliable. Saral and Goncalog (2008)
developed a method to determine COD of saline wastewater (up
to 14 gNaCl/L) but it involved the use of toxic chemicals. On the
other hand, Aziz and Tebbut (1980) assessed that the total organic
carbon (TOC) is an alternative to COD measurements. The TOC
measurement is not dependent on the salt content of the sample.
So to account the organic matter, the TOC content would be used
as the preferable parameter.
The aim of this work is to develop a methodology at laboratory
scale that allows recirculating the supernatant obtained after harvesting microalgae cultures. For that, it is necessary to establish the
best sterilization method to be used in order to avoid any inuence
on the microalgae production and/or on their biochemical composition. The experimentation has been designed bearing in mind
431
432
433
where S and S0 are the nal and initial substrate concentrations (g/
L), respectively and X is the average biomass concentration (g/L) at
the stationary state.
The substrate requirement (YS/X, g substrate consumed/g biomass) is easily obtained as the inverse of substrate yield.
The kinetic parameter used to control the uptake of the substrate was the biomass specic uptake rate (g substrate/day/g biomass), which was calculated with the following equation where D
is the dilution rate (0.3 1/d).
Y X=S
Y S=X
S S0
X
0.5
0.5
control
2 mg/L
1.5 mg/L
1 mg/L
0.86 mg/L
0.72 mg/L
0.44 mg/L
0.4
0.4
0.3
0.2
0.3
0.2
0.1
0.1
0.0
0.0
0
10
12
Time, d
Time, d
0.5
0.5
control
459 mg/L
396 mg/L
348 mg/L
300 mg/L
237 mg/L
174 mg/L
127 mg/L
95 mg/L
16 mg/L
0.4
0.3
control
10 %v/v
5 %v/v
3.8 %v/v
2.6 %v/v
Hydrogen peroxide
0.4
Absorbance, 680 nm
Ozone
Absorbance, 680 nm
control
200 mg/L
400 mg/L
Sodium Dichloroisocyanurate
Absorbance, 680 nm
Sodium hypochlorite
Absorbance, 680 nm
D S S0
X
qs
0.2
0.3
0.2
0.1
0.1
0.0
0.0
0
Time, d
Time, d
0.5
control
90 C
70 C
60 C
40 C
Pasteurization
Absorbance, 680 nm
0.4
0.3
0.2
0.1
0.0
0
10
Time, d
Fig. 1. Inuence of the sterilization method at different dosages and conditions in seawater culture medium (Algal medium). The rst number in abscises axis denotes
sterilization methods. 1: autoclave (control), 2: sodium hypochlorite, 3: sodium dichloroisocyanurate, 4: ozone, 5: hydrogen peroxide, 6: pasteurization.
434
2.0e+6
2.0e+6
(a)
Algal medium
1.0e+6
5.0e+5
4.0e+4
3.0e+4
1.5e+6
2.0e+4
1.0e+6
5.0e+5
4.0e+4
3.0e+4
2.0e+4
1.0e+4
1.0e+4
6.
90
C
5.
5
g/
L
g/
L
4.
95
g/
L
m
4.
95
6.
90
C
5.
5
g/
L
g/
L
m
g/
L
3.
20
0
2.
1
tio
n
Fi
ltr
a
1.
In
itia
l
0.0
0.0
3.
20
0
2.
1
1.
F
In
iti
al
Sterilization method
Sterilization method
2.0e+6
2.0e+6
(c)
Pasteurization temperature, C
bc
0.0
b
4.
19
0
6.
60
6.
40
In
itia
l
6.
90
6.
70
a
0.0
5.0e+5
4.
95
5.0e+5
1.0e+6
4.
47
1.0e+6
1.5e+6
In
iti
al
1.5e+6
(d)
Recirculated supernatant
Bacteria load, CFUs/mL
Recirculated supernatant
Bacteria load, CFUs/mL
(b)
Recirculated supernatant
iltr
at
io
n
1.5e+6
Fig. 2. Variation of bacterial load with the sterilization method used in Algal medium and recirculated supernatant (a, b). Inuence of different temperatures: 40, 60, 70 and
90 C for 5 min used in pasteurization (c). Ozonation: 2.5, 5, 10 min of exposition time (d). The rst number in abscises axis denotes sterilization methods. 1: ltration
(control), 2: sodium hypochlorite, 3: sodium dichloroisocyanurate, 4: ozone, 5: hydrogen peroxide, 6: pasteurization. Note, for example, that 5 min of ozone exposition
corresponds to 95 mg O3/L. All the following gures result from a Multifactorial ANOVA Statistical Study (mean values with different small letters denote signicant
difference, P < 0.05).
435
aa
a a
0.8
d d
0.6
ee
0.4
b b
bb
0.2
cc
c
c
1.
f
ilt
ra
t
2. ion
2
m
2.
g/
1.
L
5
m
2. g/L
1
m
2.
g
0.
86 /L
m
3.
20 g/L
0
m
g
4.
96 /L
m
4.
g
/L
16
0
m
4.
32 g/L
0
m
4.
48 g/L
0
m
g
5.
10 /L
%
5. v/v
5
%
v
6. /v
60
6. C
90
C
0.0
Used Method
Fig. 3. Inuence of the sterilization method on the biomass concentration of N.
gaditana. Error bars correspond to mean values of biomass concentration in
stationary state for each method. 1: ltration (control), 2: sodium hypochlorite, 3:
sodium dichloroisocyanurate, 4: ozone, 5: hydrogen peroxide, 6: pasteurization. All
the following gures result from a Multifactorial ANOVA Statistical Study (mean
values with different small letters denote signicant difference, P < 0.05).
between the group of methods 2, 3, 4, 5 and 6 (sodium hypochlorite, dichloroisocyanurate, ozone, hydrogen peroxide and pasteurization) and the ltration (1). Similarly Fig. 2b shows that all
methods of sterilization are effective for sterilizing the recirculated
supernatant with the exception of ltration. It was expected that
ltration would not cause a sharp decline in bacterial load because
the pore size employed was 1.2 lm, which is high. The results
show that with pasteurization (Fig. 2c) the best results were
achieved at 90 C (5 min). However, the most remarkable fact is
the high effectiveness of the ozonation (Fig. 2d) at a concentration
of 95 mg/L obtained with a short contact time (5 min) resulting in a
complete reduction of the bacterial load. Other chemical methods
such as the addition of hydrogen peroxide or sodium dichloroisocyanurate resulted in chemical alteration of the water, observed as
changes in color, especially when using dichloroisocyanurate.
To select the best method to reduce the bacterial load in the
recirculated supernatant it is also important to pay attention to
its inuence on the microalga growth. Therefore, laboratory scale
batch cultures of N. gaditana using the supernatant obtained after
harvesting the microalgae have been performed (see Section 2.4).
The aim is to check if the use of these sterilization methods involves a harmful effect on the growth of the microalga. For that,
the growth of Nannochloropsis is followed as shown in Fig. 3. It is
observed that, although several methods are able to sterilize in
any extent the supernatant (2, 3, 4, 5 and 6), only ltration and
ozonation are tolerated by the microalgae. It can be seen how only
the asks grown with the ozonized supernatant at low concentration (95 mg/L and 160 mg/L) and the ltered one are not affected
by the sterilization method attaining similar biomass concentration values (0.8 g/L). The statistical study reveals that these two
methods do not show signicant differences. Nevertheless,
Fig. 2a and b shows that ltration did not reduce the bacteria load
to suitable values. Kawachi and Nol (2005) reported that this procedure is inefcient killing the most resistant spores, although it
does kill most life in the water. Regarding the other sterilizing
methods it was observed that all of them affected negatively the
growth of the N. gaditana. The chemical methods probably alter
the composition of the medium and the pasteurization could oxidize compounds due to the high reached temperatures (usually
95 C). These changes in the medium composition could inuence
the microalga metabolism.
In aquaculture it is usual to have large tanks containing seawater. In that case, ash sterilization is commonly carried out using a
titanium plate heat exchanger that could be quite expensive. On
the other hand, an ozone generator can also be costly but the ozone
is able to achieve complete bacteria elimination in contrast to ltration. Accordingly, ozonation seems to be the most adequate
method to be used for the sterilization of mediums to grow
microalgae.
As ozonation seems to be the best sterilizing method, the total
organic carbon concentration (TOC) of the samples treated by
ozonation and ltration was analyzed in order to compare them
(Table 1). The results obtained show that the organic matter in
the recirculated supernatant reaches a value of TOC of 73.7 mg/L.
This organic matter is mostly composed of cellular fragments and
to a lesser extent of other dissolved organic compounds. The dissolved organic compounds are those that are potentially rst oxidized and degraded in the presence of strong oxidizing agents
such as ozone. By this way, the application of 95 mg/L of ozone
to the sample reduces the TOC in 33.5%. Applying higher concentrations of ozone it is possible to achieve higher TOC depurations,
but it would affect the microalga growth, as previously described.
When the sample is just ltered at 1.2 lm the TOC depuration
reaches 50.4%, which is quite higher. If the ltered sample is afterwards ozonized it is possible to increase the TOC depuration up to
72.1% but this requires a very high ozone dosage (3133 mg/L), so
this process would show low effectiveness. It is possible to assess
that ltration is not enough to ensure a properly sterilization of
the medium in spite of these tests and taking into account the bacteria load measured in the previous experiments carried out. Once
the best sterilization method has been chosen it is necessary to
evaluate the behavior of the cultures of N. gaditana growing in
the ozonized medium. They were rst grown in batch mode in
100 mL asks using the ozonized recirculated supernatant as the
culture medium. No negative effect on the growth was observed
(Fig. 4a) as the obtained biomass productivities were over the
one attained in the control ask, which used Algal medium (0
100). The higher the proportion of supernatant used, the higher
the biomass concentration, reaching values from 25% to 31% higher
than in the control ask in the best case. That demonstrated that
Table 1
Results obtained from the recirculated supernatant treated with ozone and ltration to reduce the dissolved organic carbon.
Treatment
Initial
without
treatment
Ozone
16 mg/L
Ozone
95 mg/l
Ozone
197 mg/L
Filtration
1.2 lm
Filtration
1.2 lm + Ozone
16 mg/L
Filtration
1.2 lm + Ozone
95 mg/L
Filtration
1.2 lm + Ozone
197 mg/L
Filtration
1.2 lm + Ozone
3133 mg/L
TOC; mg/L
Depuration,
%
73.7 2.2
0.0
71.0 2.1
3.8
49.0 1.5
33.5
43.9 1.3
40.5
36.6 1.1
50.4
28.0 0.8
62.0
27.4 0.8
62.8
27.3 0.8
62.9
20.6 0.6
72.1
(a)
Batch cultivation
YX/S
3.0
1.5
YS/X
(a)
0.10
0.08
0.06
0.04
2.5
2.0
1.0
1.5
1.0
0.5
0.5
0.02
0.0
0.00
0.0
0-100
0-100
25-75
50-50
75-25
100-0
50-50
75-25
100-0
100*-0
100*-0
0-100
50-50
75-25
100-0
100*-0
0.06
1.0
(b)
(b)
b)
0.05
0.8
0.04
qS, g/d/g
436
0.6
0.03
0.02
0.4
0.01
0.2
0.00
Nitrates
Phosphates
K+
Mg+2
Sulphates
0.0
0-100
50-50
75-25
100-0
100*-0
Ash
Carbohidrate
Protein
Lipid
60
50
(c)
c)
40
30
20
10
0
0-100
50-50
75-25
100-0
100*-0
Table 2
Empirical formula of the biomass obtained in the steady state of the continuous cultures of N. gaditana at 0.3 1/d using different mixtures of recirculated supernatant and Algal
medium.
Test
0100
5050
7525
1000
1000
Empirical formula
C1H1.36N0.14O0.51
C1H1.05N0.14O0.51
C1H1.19N0.11O0.50
C1H1.22N0.09O0.42
C1H1.39N0.10O0.68
437
Table 3
Fatty acids prole expressed as percentage on a biomass dry weight basis for the continuous cultures of N. gaditana at 0.3 1/d dilution rate using different mixtures of recirculated
supernatant and Algal medium. Note: N.I.: not identied.
Fatty acid
14:0
16:0
16:1n7
18:0
18:1n9
18:2n6
20:4n6
20:5n3
N.I.
Total
Test
0100
5050
7525
1000
100-0
1.2310.047
3.9020.150
3.4420.132
0.1400.005
0.8610.033
0.4110.016
0.7920.030
2.7020.104
2.3330.089
15.8650.608
0.6650.033
4.7120.236
4.8160.241
0.1050.005
0.6450.032
0.2280.011
0.7440.037
3.1400.157
1.7590.088
18.1890.909
1.0970.055
6.9220.346
6.7720.309
0.1060.005
0.8780.044
0.5440.027
1.1220.056
3.3730.199
2.3310.117
23.1451.157
1.0610.053
8.0510.403
6.9600.348
0.2110.011
1.3280.066
0.3780.019
0.9560.048
3.4400.172
2.0180.101
25.6221.281
1.5430.027
5.7550.102
4.4300.078
0.1950.003
1.2190.022
0.4620.008
0.9290.016
2.7070.048
2.4660.044
19.7070.348
medium reaching a biomass productivity near 0.8 g/L/d. Nevertheless, when the ratio of supernatant recirculation is increased (75
25 and 1000) the biomass productivity decreases to values of
0.4 g/L/d mainly because of the lack of nutrients in the medium.
There are hardly any difference between the tests 7525 and
1000 probably because of the lack of any essential nutrient to
the growth in both cases. On the other hand, when using only
the recirculated supernatant after the replenishment of the consumed nutrients (1000) it was possible to achieve a biomass productivity near 0.8 g/L/d, as the one obtained with Algal medium
(0100) with no statistically signicant differences between them.
The use of mixtures of recirculated supernatant and Algal medium
50:50 v/v or the use of only recirculated supernatant after the
replenishment of the lacking nutrients lead to a reduction of 50%
or 100%, respectively, in water usage (water evaporation and water
loss in harvesting are not considered). It also leads to a reduction in
the addition of nutrients up to 50% or 15%, respectively.
The elemental composition of the microalgal biomass harvested
in the steady state of the continuous cultures was analyzed in order to establish the empirical formula (Table 2). This formula is
quite similar whatever the composition of the culture medium
used. When using pure Algal medium the obtained formula is
C1H1.36N0.14O0.51. From the elemental analysis reported by Pan
et al. (2010) it is possible to determine an empirical formula for
Nannochloropsis sp. of C1H1.63N0.08O0.40, which is analogous to the
one obtained in this work. Having established the empirical formula it is possible to calculate the nitrate yield (Y X=S ) and the nitrate requirement (Y S=X ). It was calculated for the different ratios
of recirculated supernatant and Algal medium (0100,
5050, 7525, 1000 and 1000). It was shown (Fig. 5a) that the
higher the ratio of recirculated supernatant to Algal medium the
higher the yield coefcient for nitrate (YX/S) and the lower the
requirement of nitrate (YS/X). This would lead to an increase in
the content in lipids and a decrease in proteins. However, with
the recirculated supernatant after replenishing the nutrients
(1000) is more than 30% lower than the obtained when using Algal medium (0100).
The biomass-specic uptake rates of key inorganic nutrients,
3
2
+
+2
such as NO
3 , PO4 , K , Mg , and SO4 , was determined in order
to compare them (Fig. 5b). As shown, the recirculation of the
supernatant can signicantly reduce the uptake rate of nutrients.
3
+
The major usages correspond to NO
3 , PO4 and K with values of
0.049, 0.017 and 0.014 g/d/g, respectively. In the test with the
replenished supernatant (1000) the uptake rate of nitrate decreases in 50% with regard to the test with Algal medium. Nevertheless, the rate for the uptake of the other inorganic nutrients
remains constant. The reduction of nutrients consumption (up to
50%) is not noteworthy in the context of aquafeeds as biomass production for this purpose is not large. Otherwise, this reduction
would have an important impact with regard to the sustainability
of the processes related to the production of commodities.
4. Conclusions
If mass production of microalgae is ever to come about it is necessary that the culture medium be reused. This work has shown
that the supernatant obtained from harvested N. gaditana cultures
can be reused after ozonation at 95 mgO3/L as this ensures a proper
sterilization. It will allow reducing the nutrient costs, in a extent
we expect to be able to quantify in future contributions, once the
method is scaled-up. It has also been shown that N. gaditana tolerates the recirculation of the medium replenishing the depleted
nutrients, and the obtained microalgal biomass is of high quality,
suitable for aquaculture because of its high content in proteins
(50%) and lipids (40%) with more than 3% EPA (d.w.).
Acknowledgements
This research was supported by the General Secretariat of Universities, Research and Technology of Andalusian Government
438
Lpez, C.V., Cern, M.C., Acin, F.G., Segovia, C., Chisti, Y., Fernndez, J.M., 2010.
Protein measurements of microalgal and cyanobacterial biomass. Bioresour.
Technol. 101, 75877591.
Martnez, S.B., Prez-Parra, J., Suay, R., 2011. Use of ozone in wastewater treatment
to produce water suitable for irrigation. Water Resour. Manage. 25, 21092124.
Morais, S., Koven, W., Ronnertad, I., Conceicao, L.E.C., 2005. Lipid enrichment for
Senegal sole (Soleasenegalensis) larvae: effect on growth survival and fatty acid
prole. Aquac. Int. 12, 281298.
Navarro, J.C., Villanueva, R., 2003. The fatty acid composition of Octopus vulgaris par
larvae reared with live and inert food: deviation from their natural fatty acid
prole. Aquaculture 14, 613631.
Pan, P., Hu, C., Yang, W., Li, Y., Dong, L., Zhu, L., Tong, D., Qing, R., Fan, Y., 2010. The
direct pyrolysis and catalytic pyrolysis of Nannochloropsis sp. Residue for
renewable bio-oils. Bioresour. Technol. 101, 45934599.
Rathore, A.S., Shirke, A., 2011. Recent developments in membrane-based
separations in biotechnology processes: review. Prep. Biochem. Biotechnol.
41, 398421.
Rhodes, M.A., Sedlacek, C., Marcus, N.H., 2008. Large-scale batch algae cultivation:
comparison of two seawater sterilization techniques. Aquac. Res. 39, 1128
1130.
Rodol, L., Chini Zitelli, G., Barsanti, L., Rosati, G., Tredici, M.R., 2003. Growth
medium recycling in Nannochloropsis sp. Mass cultivation. Biomolec. Eng. 20,
243248.
Rodrguez-Ruiz, J., Belarbi, E., Snchez, J.L.G., Alonso, D.L., 1998. Rapid simultaneous
lipid extraction and transesterication for fatty acid analyses. Biotechnol. Tech.
12, 689691.
Roessler, P.G., 1990. Environmental control of glycolipid metabolism in microalgae:
commercial implications and future research directions. J. Phycol. 26, 393
399.
Saral, A.B., Goncalog, B.I., 2008. Determination of real COD in highly chlorinated
wastewaters. Clean 36, 9961000.
Tan, C.K., Johns, M.R., 1996. Screening of diatoms for heterotrophic
eicosapentaenoic acid production. J. Appl. Phycol. 8, 5964.
Tyrrell, S.A., Ryppey, S.R., Watkins, W.D., 1995. Inactivation of bacterial and viral
indicators in secondary sewage efuents using chlorine and ozone. Water Res.
29, 24832490.
Wen, Z.Y., Chen, F., 2000. Production potential of eicosapentaenoic acid by the
diatom Nitzschia. Biotechnol. Lett. 22, 727733.
Yang, J., Xu, M., Zhang, X., Hu, Q., Sommerfeld, M., Chen, Y., 2011. Life-cycle analysis
on biodiesel production from microalgae: Water footprint and nutrients
balance. Bioresour. Technol. 102, 159165.