Chemical Engineering and Processing 64 (2013) 19

Contents lists available at SciVerse ScienceDirect

Chemical Engineering and Processing:
Process Intensication
j our nal homepage: www. el sevi er . com/ l ocat e/ cep
Process intensication in lactic acid production by three stage membrane
integrated hybrid reactor system
Parimal Pal

, Pinaki Dey
Environment and Membrane Technology Laboratory, Department of Chemical Engineering, National Institute of Technology Durgapur, 713209, India
a r t i c l e i n f o
Article history:
Received 7 July 2012
Received in revised form
30 November 2012
Accepted 14 December 2012
Available online 22 December 2012
Keywords:
Lactic acid
Fermentation
Membrane separation
Hybrid process
Process intensication
a b s t r a c t
l(+) lactic acid was directly produced in a three-stage membrane-integrated hybrid reactor system start-
ingwith acheap, renewable and clean carbon source. Provision of at sheet cross ow membrane modules
for microltration andnanoltration permitted selective production of l(+) lactic acid under high cell den-
sity with recycling of cells and unconverted sugars. This modular design ensured high yield (0.96 g g
1
),
productivity (12.4 g L
1
h
1
), concentration (250 g L
1
) and purity (95%) in a very simple, environmen-
tally benign, compact and exible plant conguration reecting all the major characteristics of high
process intensication. Such process intensication has been analysed for the rst time in terms of space
intensication, application exibility, capacity exibility, energy reduction parameter and cost advantage
parameter. The value of space intensication parameter, energy reduction parameter and cost advantage
parameter that were computed to be 3.33, 0.022 and 0.35 respectively indicated high degree of process
intensication in such a novel and green production scheme.
2012 Elsevier B.V. All rights reserved.
1. Introduction
In the recent years, there has been a paradigm shift in the
business strategies of chemical and pharmaceutical industries
throughout the world towards greener and cleaner productions
embracing the sustainable development concept. The newstrategy
that envisages faster, more efcient and clean production through
green and more exible route involves less energy, less material
consumption and reduced manpower compared to a conventional
productionscheme. Suchaproductionregimewhichinother words
is termed process intensication (PI) is indeed what traditional
chemical process industry is drastically seeking for. PI not only
seeks to replace old, inefcient plant with new, intensied equip-
ment but alsothrows upnewbusiness model challenges byopening
up opportunities for new patentable products and processes with
scale up opportunities [1]. Smaller is safer! Hence, process inten-
sication dramatically increases the intrinsic safety of chemical
process. There may be basically two approaches in achieving pro-
cess intensication: equipment-based PI and method-based PI.
Equipment-based PI may or may not involve chemical reactions
but method-based process intensication involves chemical reac-
tions as is well illustrated by multifunctional devices like, reactive
distillation, reactive extraction or reactive crystallization units that

Corresponding author. Fax: +91 3432547375; mobile: +91 9434469750.

E-mail addresses: parimalpal2000@yahoo.com, parimal.pal@che.nitdgp.ac.in
(P. Pal).
full the overall objectives. Membrane based hybrid reactors and
use of solar energy as an alternative energy source, received con-
siderable attention fromchemical manufacturing industries in the
recent years. Membrane based technologies have already been rec-
ognized as the most exible tools which can compete with and
replace conventional energy-intensive techniques [2]. Membrane
engineering brings useful solutions to some of the major industrial
problems throughhighlyselectiveandefcient transport of specic
compounds and as a result hybrid processes like membrane inte-
grated separation systems have been successfully incorporated as
one of the efcient methods in PI [3]. Due to huge market demand
for biodegradable polymer, production of monomer grade lactic
acid by membrane-based process has been gaining importance in
the recent years. Lactic acid, an important carboxylic acid, can par-
ticipate in a wide variety of chemical reactions leading to a host of
products for use in the elds of textiles, pharmaceuticals, medical
and food processing industries. Chemical synthesis process for lac-
tic acid production from petroleum resources consistently yields
a racemic mixture of dl-lactic acid whereas demand is for pure
l(+) lactic acid. 50% of the total production cost in such case is
due to involvement of a large number of separation and purica-
tionsteps. Conventional processes not onlyinvolveharshchemicals
like acids and alkalis but also end up with huge solid waste in the
formof calciumsulphate [4]. Wasewar et al. [5] attempted reactive
extraction-based production of lactic acid for process intensi-
cation (PI). But such reactive extraction processes having certain
limitations such as direct exposure of microbes-bearing fermen-
tation broth to extraction reagents is often not very successful
0255-2701/$ see front matter 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cep.2012.12.006
2 P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19
due disturbed equilibrium conditions as well as lower separation
efciency resulting fromsuchexposure [6]. Amidst growing world-
wide energy insecurity, process industries are increasingly seeking
for energy-saving alternatives. A membrane having the potential
of effecting very high degree of fractionation of materials holds
the key to production of such high grade lactic acid [7] as sepa-
ration does not involve phase change and needs minimumenergy.
Through tireless efforts of several researchers [5,812] membrane-
integrated production system has gradually developed over the
years. However, process intensication that can be expected from
direct and continuous production of l(+) lactic acid in a fully
membrane-integrated systemneeds to be studied further. Detailed
analysis in terms of process intensication parameters of such a
fully membrane-integrated systemis of paramount importance to
indicate the nature and extent of process intensication resulting
frommembrane integrationso that scale upcondence canbe built
up for industrial production. To our knowledge, no such analysis
has yet been reported. The present study attempts to ll this gap
through development of a three stage membrane-integrated sys-
tem and through a detailed analysis of the process intensication
parameters.
2. Theoretical
2.1. Cross-ow microltration
Mass transport through microltration membranes at steady
state condition is dependent on the specic membrane and cake
resistances [13]. Separation through microltration membranes is
based on size exclusion mechanism. The solvent ux incorporating
those resistances can be expressed as:
j
m
=
^P

P
(R
m
+R
c
+R
f
)
(1)
where J
m
represents the solvent ux in permeate stream, ^P,
transmembrane pressure, R
m
, the membrane resistance, R
f
, the
membrane fouling resistance and R
c
stands for the cake resis-
tance. During cross ow microltration of fermentation broth for
cell separation with PVDF (Polyvinylidene Fluoride) membrane, R
f
was observed to exert insignicant inuence on solvent ux and
after chemical treatment membrane fouling was observed to be
totally reversible. Membrane resistance, R
m
, was calculated from
pure water ux experiment through HagenPoiseuille equation.
j
w
=
^P
R
m
(2)
where J
w
represents pure water ux andrepresents water viscos-
ity. Again this membrane resistance (R
m
) is dependent on porosity,
pore-size distribution and pore tortuosity of the membrane as:
R
m
=
32zl
m
d
2
(3)
wherez represents tortuosityof themembrane, l
m
is themembrane
thickness, is the membrane porosity and d is the pore diameter.
Thebacterial cakeresistance(R
c
) was computedusingthefollowing
relation:
R
c
=
m
/
TMP
n
(4)
where m, A, and n indicate bacterial mass, membrane surface
area, cake resistance coefcient and compressibility index respec-
tively [14]. The membrane module with at sheet membranes was
operated in cross ow mode when sweeping action of adequate
cross owover the membrane surface ensuredlong hours of largely
fouling-free separation. Reversible nature of fouling permittedsep-
aration of cells from fermentation broth for continuous recycling
without necessitating frequent replacement of membranes.
2.2. Cross-ow nanoltration
Separation of charged solutes by nanoltration membranes
depends on a transport mechanismthat involves diffusion, convec-
tion as well as electro migration phenomena. Retention of neutral
solutes like undissociated lactic acid largely follows size exclusion
mechanism which in turn is dependent on effective pore radius
(r
p
), thickness-porosity ratio (^x/A
k
) and also on stoke radius of
solute (r
s
). In nanoltration, solvent ow can be expressed using
HagenPoiseuille type relationship as:
dP
dX
=
8qv
r
2
p
(5)
where V represents solution velocity and q represents solvent
viscosity within membrane pores. Concentration gradient with
respect to membrane thickness can be expressed as:
dc
dx
=
v
D
p

K
c

D
p
R1
v
s
8q
r
2
p

C C
pn

(6)
where D
p
denotes the diffusivity of the solutes which is correlated
with solute bulk diffusion coefcient (D
a
), q denotes solvent vis-
cosity within the pore, C
pn
denotes uncharged solute permeate
concentration and C denotes average uncharged solute concen-
tration within the pore.
Uncharged solute ux based on extended NernstPlank equa-
tion considering diffusion, convection may be expressed as:
j
s
= K
c
cvD
p
dc
dx

cD
p
R1
v
s
dP
dX
(7)
Concentration gradient and pressure gradient along the membrane
thickness exert a negative impact on ultimate solute ux. Volumet-
ric ux (J
v
) and rejection (R
j
) of uncharged solutes that are directly
dependent on J
s
as well as concentrations of uncharged solute in
permeate and feed site were calculated following the relation:
j
v
=
j
s
C
pn
(8)
R
j
= 1
C
pn
C
fn
(9)
Details of capturing solute and solvent transport through nanol-
tration membranes during ltration by at sheet cross ow
modules can be found in Dey et al. [15].
3. Experimental
3.1. Three-stage membrane-integrated reactor system
Flat sheet cross ow membrane modules in three stages were
integrated with a 30L fermenter fabricated with high quality stain-
less steel (SS316) as presented in Fig. 1. The rst stage was for
microltration of the fermentation broth using PVDF (Polyvinyli-
dene Fluoride) microltration membrane for cell separation and
recycling. The subsequent two stages consisted of nanoltration
membrane modules in the same ow mode. Characteristics of the
nanoltration membrane (NF-1) have been presented in Table 1.
3.2. Experimental procedure
Experimental investigations were carried out for continuous
mode fermentation and purication without any pH adjustment.
However, continuous fermentation and subsequent downstream
purication required an initial 15h of lag phase after which con-
tinuous membrane cell separation was carried out using cross ow
microltration membrane module with the fermenter. Recycling
of cells through the retentate of the microltration unit ensured
P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19 3
Fig. 1. Schematic diagramof membrane-integrated hybrid reactor systemfor continuous lactic acid production.
high cell concentration in the fermenter leading to high produc-
tivity. Lactobacillus delbrueckii (NCIM-2025), a homo-fermentative
l(+) lactic acid producing bacteriumwas used in the work and the
strainwas collectedfromNational Collectionof Industrial Microbes
(NCIM), National Chemical Laboratory, Pune, India. On reaching
steady state condition, the system was switched to second stage
nanoltrationwithpermeatefromrst stageof nanoltration. NF-2
membrane (Sepro Inc., USA) was used for this step of nanoltra-
tion and permeate was collected in holding vessel 2 as shown in
Fig. 1. Nanoltration was carried out at 15kgcm
2
transmembrane
pressure. In second stage nanoltration separation, 3 modules with
Table 1
Characteristics of NF-1 membrane.
Parameters Value
Water ux at 15kgcm
2
pressure 12.2L m
2
h
1
Membrane type Flat-sheet
Membrane surface area 0.012m
2
Membrane thickness 0.0165cm
Maximumprocess temperature 50

C
% MgSO
4
rejection at 150psi 99.5
% NaCl rejection at 150psi 90
% Lactose rejection at 150psi 99
NF-1 membranes were used at transmembrane pressure of 15bars.
This nanoltrationoperationcontinued for 24hat cross owveloc-
ity of 0.53ms
1
.
4. Results and discussion
4.1. Plant conguration: membrane-integrated systemvs.
conventional system
Despite presence of a host of chemical synthesis routes for lac-
tic acid production such as hydrolysis of acetonitrile, oxidation of
propylene glycol, reaction of acetaldehyde with carbon monox-
ide and water at elevated temperatures and pressure, hydrolysis
of chloropropionic acid and nitric acid oxidation of propylene,
hardly any of these processes has been commercialized [16] due
to involvement of high cost, energy and material. Typical conven-
tional fermentation based lactic acid production plant consists of
a number of downstreamtreatment units like chemical precipita-
tion, conventional ltration, acidication, carbon adsorption and
evaporation as it is presented in Fig. 2. In that production pro-
cess, addition of lime for controlling pH leads to the production of
calciumlactate. Calciumlactate is then separated fromthe micro-
bial cells by ltration and, further puried by activated carbon
4 P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19
Fig. 2. Typical conventional fermentation-based lactic acid production scheme.
adsorption. In next phase, calcium lactate is evaporated and acid-
ied by sulphuric acid to produce lactic acid. Gypsum (calcium
sulphate) is producedas a by-product inthe process. Further details
of this process may be found in Inskeep et al. [17]. To produce
high purity lactic acid, the liquor can be treated with few extra
processing steps like precipitation, extraction and crystallization
as shown in Fig. 2. That total number of unit operations involved in
conventional lactic acidproductionplant is muchhigher thanthose
involved in a membrane integrated lactic acid production system
is clear fromTable 2. Membrane integrated process consists of only
4unit operations compared to 12unit operations in a conventional
fermentation based plant. For a 1640tonnes/year capacity plant,
Table 2
Unit operations involved in conventional and membrane integrated plants for lactic acid production (capacity: 1640tonnes/year).
Conventional lactic acid production system Membrane integrated system
Sr. No Unit operations Sr. No Unit operations
1. Fermentation 1. Fermentation
2. pH adjustment through hydrated lime 2. Microltration
3. Rotary dramltration 3. First stage nanoltration
4. Carbon bleaching 4. Second stage nanoltration
5. Plate and frame ltration
6. Calciumlactate evaporation
7. Crystallization
8. Acid conversion
9. Extraction
10. Distillation unit
11. Colour removal by sodiumsulphide
12. Dilution
P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19 5
Table 3
Comparison of space requirement by conventional and membrane integrated systems (capacity: 1640tonnes/year).
Conventional lactic acid production system Membrane integrated system
Sr. No Processing stages Area required Sr. No Processing stages Area required
1. Fermentation 67.02m
2
1. Fermentation 12.06m
2
2. Culture tanks 12.67m
2
2. Microltration 104m
2
3. Mixing tanks 169m
2
3. First stage nanoltration 39m
2
4. Storage tanks 144m
2
4. Second stage nanoltration 79m
2
5. Rotary lters 48m
2
5. Storage tank 36m
2
6. Plate lters 76m
2
7. Evaporators 56.52m
2
8. Crystallizer 156m
2
9. Extraction unit 70m
2
10. Distillation unit 150m
2
Total area required (including pipe line) 1000m
2
(approximate) 300m
2
(approximate)
the space requirement for hybrid process comes down drastically
comparedto conventional plant as presentedinTable 3. Total space
required for microltration and nanoltration modules (including
excess modules) in the membrane integrated system came down
to 222m
2
due to the stack and parallel adjustment of the modules.
The efciency in space reduction can be represented in terms of
space intensication parameter which may be dened as:
space intensication =
total area required for conventional system
total area required for membrane hybrid system
The parameter value was evaluated as 3.33 and it signies that
space requirement for lactic acid production in the new scheme is
less than one-third the space required for conventional plant for
the same capacity.
4.2. Flexibility in the membrane based plant design over
conventional process
Three stage membrane integrated reactor system as presented
in Fig. 1 offers great exibility in operation compared to a conven-
tional process by virtue of modular design. Provision for using any
number of membrane modules depending on the desired quantum
of product ensures hassle-free continuous production. In compari-
son, conventional lactic acid production plant consists of a series of
unit operations with little exibility. This modular design in hybrid
systemcanbe usedfor productionof not only lactic acidbut also for
similar organic acids. A large number of other fermentation based
production processes and other membrane-based separation oper-
ations can be carried out with this plant conguration. The ratio
of number of possible applications in the hybrid process to the
number of possible applications in conventional process may be
expressed in terms of application exibility parameter:
application exibility =
possible applications in hybrid process
possible applications in conventional process
Applicationexibility parameter inthis case will be greater than
1. A comparative description of systemexibility for both the pro-
cesses has been presented in Table 4. By virtue of modular design,
the total production in the hybrid plant can be changed based on
the market demand for the product. A large number of membrane
modules can be arranged in parallel or in series and any number of
modules at any instant can be switched on or off. In a conventional
production plant, total production cannot be changed in this way.
Capacity exibility in terms of total throughput may be expressed
as:
capacity exibility =
variations in production capacity by hybrid process
variations in production capacity by conventional system
Again the ratio of variations in total production process in a hybrid
process to the total production in a conventional process is greater
than 1.
Thus membrane-integrated hybrid treatment scheme can offer
exibility bothinterms of applications as well as productioncapac-
ity. Typical congurationof the developedhybridprocess canallow
aproductionof anumber of similar products inthesameplant. Such
exibility is extremely difcult to attain in conventional plant due
to rigid process conguration involving several energy-intensive
steps wherephasechangetakes place. Verymodular designpermits
connection or disconnection of any number of membrane mod-
ules to the plant depending on the requirement of total production
quantity. Such exibility cannot expected in a conventional plant
conguration due non-modular design.
4.3. Eco-friendly process design of the membrane integrated
system
The developedmembrane integratedprocess of this study offers
an entirely green process of production vis-a-vis a conventional
process. It does not incorporate any phase change and thus mini-
mizes energy requirement. Conventional production process uses
hydratedlimeinfermentationandproduces calciumlactateinstead
of lactic acid. Direct disposal of such material into environment
leads to health hazards as it is susceptible to produce H
2
S gas.
Huge heat generation due to the exothermic reactions fromdiffer-
ent stages of conventional systemaffect surrounding environment
substantially. Amount of activatedcarbonrequiredfor decolouriza-
tion and bleaching in conventional lactic acid production process
for the same production capacity stands at 4530tonnes and it as
also harmful for open environment as ne pulverized carbon is
vulnerable to dust explosions although activated carbon may be
reused to some extent with due chemical or physical treatment.
On the other hand, in membrane integrated reactor system, liquid
phase is always maintained throughout the system. The membrane
integrated process with non-neutralizing approach ensures pro-
duction in a clean environment with minimum consumption of
resources and minimum generation of waste. Table 5 presents a
comparative description of the newscheme of production and the
conventional process in this context of environment-friendliness.
Thenewplant withmembrane-integrationis expectedtoovercome
the adverse environmental impacts of a conventional plant where
disposal of hazardous wastes anddepletionof natural resources are
quite common.
4.4. Reduction of energy consumption through membrane
bio-reactor system
Total annual energy consumption of a conventional plant of
1640tonnes/year production capacity stands at 3.3910
6
kWh.
Energy required for fermentation and total mixing procedures
were comparatively higher than the other individual processing
steps in conventional processing stage. Power consumption of each
6 P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19
Table 4
Comparison of systemexibility.
Membrane integrated system Conventional lactic acid production system
Flexibility parameters
Modules Flexible in size and numbers Fixed conguration
Production capacity Flexible Fixed
Steps of operations Fewsteps Large no of steps with little exibility
Table 5
Comparison of eco-friendliness.
Membrane integrated plant Conventional lactic acid production plant
1. Does not incorporate any phase change. 1. Multiphase process consists of liquid phage, gas phase and solid phase.
2. Does not produce any harmful
by-products.
2. Generation of 2676.833tonnes of calciumsulphate (gypsum)/year due to the addition of
1928.287tonnes of sulphuric acid and its direct disposal into open environment can create
environmental problem.
3. Free fromadditional pre-treatment or
post-treatment processes.
3. Direct disposal of 4530tonnes of activated carbon into open environment without any
pre-treatment can also create problemalthough it is reusable up to certain extent with chemical
treatment.
4. No exothermic reactions. 4. Heat generation due to exothermic reactions.
5. Lowenergy consumption. 5. Highly energy-intensive.
6. Use of no harsh chemicals. 6. Uses of harsh chemicals like acids and alkalis.
stage in conventional lactic acid production plant as well as mem-
brane integrated system has been shown in Table 6. Total annual
power consumption by a membrane integrated lactic acid produc-
tion plant of similar capacity has been computed to be around
7.2010
4
kWh only. Membrane integrated lactic acid systemthus
turns out to be quite energy efcient system as power consump-
tion is only 2.13% of total power consumption by a conventional
system. All the power calculations for scaled up system were
carried out with the help of six-tenths-factor rule [18], nal capac-
ity requirement for the instrument and power number equations
[19]. For conventional lactic acid production system, energy con-
sumption is 210
3
kWh/tonne of lactic acid production against
only 44kWh for membrane integrated system. The efciency in
energy consumption can also be expressed through energy reduc-
tion parameter as:
energy reduction parameter =
energy required by hybrid systemfor per unit production
energy required by conventional systemfor per unit production
The value of energy reduction parameter calculated for our case
was 0.022 which indicate towards high energy intensication.
Eco-friendliness of a membrane-integrated process may be
reected in lower chemical consumption, reduced waste genera-
tion, reduced energy consumption, reduced waste discharge and
reduced emission into the atmosphere. Low energy consumption
means low generation of energy which often comes fromburning
of fossil fuel. Thus low energy consumption implies less environ-
mental pollution. These are only some of the easily obtainable
parameters which may serve as indicators of Eco friendliness.
Overall impact of operation of a plant on the environment may
be reected in several other parameters and some of which
may not be easily measurable. Table 7 shows how a conven-
tional plant involves huge consumption of material and energy
vis-a-vis the membrane-integrated plant for the comparable
capacity. In the present investigation, it may thus be concluded
that expressed eco friendliness is partly qualitative and partly
quantitative.
4.5. Economics of production
The capital investment cost for yearlyproductionof 1640tonnes
of l(+) lactic acid including equipment installation cost and
Table 6
Comparison of total energy consumption by conventional and proposed plant of lactic acid production (capacity: 1640tonnes/year).
Conventional lactic acid production system Membrane integrated system
Equipment Energy (kWh) Equipment Energy (kWh)
Fermenter 2.8810
6
Fermenter (including water circulating bath) 6.5710
4
Heat supplying water circulating bath 66,083 Power required by 3 pumps 6340.83
Culture tanks 9019.44
Total mixing tank 251,641.66
Rotary dramlter 135
All pumps 133,491.66
Evaporator 10,397
Acid conversion tank 19,613
Rotary dramlter 85
Evaporator 2361.11
Total energy requirement 3.3710
6
Calciumlactate crystallizer 2166.67
Liquidliquid extraction 11,836.11
Distillation of lactate esters 2283.33
Total energy requirement 3.3910
6
7.2010
4
Energy requirement 210
3
kWh/tonne of LA 44kWh/tonne of LA
P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19 7
Table 7
Operational cost comparison for the conventional and membrane integrated systems (capacity: 1640tonnes/year).
Membrane integrated plant Conventional lactic acid production plant
Item Quantity/year Cost (US$) Item Quantity/year Cost (US$)
1. Sugar cane (including transport cost) 28.2410
6
kg 3.1210
6
15% cane sugar 1050tonnes 5.1510
6
2. Yeast extract and other nutrients 96,000kg 2.1110
6
10% calciumcarbonates 700tonnes 1410
4
0.37% malt sprouts 25.9tonnes 1.1010
6
3. Nitrogen gas 18,900kg 1890 74.37% water 5180tonnes 1.0310
3
4. Electricity (total) 7.2010
4
kWh 5.0410
3
0.25% ammoniumhypophosphate 17.5tonnes 1.5710
4
5. Membranes replacement and cleaning 300m
2
310
5
Calciumhydroxide 11tonnes 16.8910
4
84.86m
2
1.0110
5
H
2
SO
4
1930tonnes 44.3510
4
Carbon for bleach 4530tonnes 1.3510
6
Sodiumsulphite 6.8tonnes 3740
Electricity (total) 3.2710
6
kWh 2.2810
5
6. Labour 127 persons 5.0810
5
Labour 410 persons 1.9610
6
7. Maintenance 7.7410
5
Maintenance 3.1310
6
8. Operating supplies 3.8710
5
Operating supplies 1.5610
6
9. Depreciation 1.8610
6
Depreciation 10.6510
6
10. Insurance 1.2410
4
Insurance 5.0110
4
11. General and administrative expenses 5.4210
5
General and administrative expenses 2.1910
6
Total production cost/year 9.7210
6
28.1410
6
instrumentation cost for both the systems were calculated taking
into account the experimental design shown in Figs. 1, 2 and the
scale up factors. The total annual capital investment cost involved
for conventional lactic acid production as has been evaluated here
stands at 71.5410
6
US$ for 1640tonnes/year capacity whereas
it drops to 15.4810
6
US$ for a membrane integrated lactic acid
production plant of same capacity. Direct capital investments costs
were calculated considering equipment purchase cost, equipment
installation cost, piping cost, building cost, electrical instrumen-
tation cost, yard improvements cost, service facilities and land
purchase cost. Total equipment purchase cost for conventional sys-
tem was calculated as 7.4010
6
US$ which is 3.78 times higher
than the membrane integrated lactic acid systemfor same produc-
tion capacity of lactic acid. Similarly, other associated costs and
indirect costs like engineering supervision, contractors fee and
contingency are in the same way on the higher side for a conven-
tional system. For membrane integrated system, costs for pumps
and membrane modules contributed about 36.81% and 19% respec-
tively to the total equipment cost. Cost of every equipment and
process has been calculated using the standard equation below
[18]:
cost of equip.a (C
1
) = cost of equip.b (C
0
)

capac.equip.a(Cap
1
)
capac.equip.b(Cap
0
)

n
Here n represents the scale-up factor and for different equipment,
n values were taken fromthe list mentioned in [18].
Table 7 presents a comparative economic evaluation. The ef-
ciency of product formation from initial as well as unit mass of
raw material is more in the case of membrane integrated reactor
system as it consumes 4040tonnes of sugarcane juice to produce
1640tonnes of lactic acid whereas for the same amount of lactic
acid production media requirement for conventional system was
foundtobe7000tonnes. It was assumedthat theconventional plant
workedfor 210days/year withtotal 70batches. Eachbatchfermen-
tation contains 3 days with 100m
3
of initial mediumrequirement
including 15% cane sugar, 10% calcium carbonate, 0.375% Matt
sprouts, 74.38% water and 0.25% ammonium hypophosphate. The
cost involvement in such system due to the addition of harsh
chemicals like sulphuric acid and calcium hydroxide for acidica-
tion and alkalinization was estimated at 62.2410
4
US$, whereas
membrane-integrated continuous process is totally free fromthose
stages due to adoption of non-neutralizing approach (without pH
adjustment). The overall cost decreases with decreasing pH [20]
and it also signies that the continuous production scheme in
non-neutralizing approach is much more economical compared to
that of aconventional plant. Simultaneously, labour cost andannual
maintenance cost are muchlower for membrane integratedreactor
system due to involvement of a few operating units and simplic-
ity of the overall processes. Amount of activated carbon, sodium
sulphite (Na
2
SO
3
) for bleaching and decolourization of lactic acid
in conventional process also very signicantly adds to the annual
productioncost of a conventional plant. For conventional lactic acid
production process, yearly total production cost was estimated as
28.1410
6
US$ or US$ 17.18/kg of l(+) lactic acid.
The continuous membrane-based l(+) lactic acid production
process with sugarcane juice as initial raw material was designed
in such a way that the process could be free fromthe initial invest-
ment cost associated with pre-treatment stages like liquefaction
and saccharication. Operating cost for such membrane integrated
system including all nutrients cost, membrane replacement cost,
electricity and labour cost was evaluated as US$ 9.7210
6
for
production of 1640tonnes of l(+) lactic acid per year. One of the
major cost contributors in yearly production process of l(+) lactic
acid was yeast extract with combination of other nutrients (total
21.70%) althoughtotal productioncost canbesubstantiallyreduced
with the use of cheap nitrogen source like corn steep liquor (CSL),
waste fromcorn wet milling industry. Membrane replacement and
cleaning charge was calculated considering the reusability of the
membranes (micro and nano) in a year through chemical treat-
ments. According to manufacturer, life time of the membranes was
assumed to be 5 years with a plant pay-off period of 15 years.
Considering all the factors including maintenance, operating sup-
plies, depreciation and insurance in membrane integrated plant,
cost of 95% pure 250gL
1
of l(+) lactic acid was evaluated as US$
5.93/kg. The cost efciency for lactic acid production in membrane
integrated hybrid reactor systemin comparison with conventional
systemcan be evaluated by estimating cost advantage parameter.
Cost advantage =
cost involved by hybrid systemfor per unit production
cost involved by conventional systemfor per unit production
This particular parameter is estimated as 0.35 which indicates
that membrane hybrid reactor system is economically sounder
than conventional system. Economic advantage of the developed
membrane-integrated process lies in involvement of less number
of unit operations, less equipment cost, less material, and energy
consumption for the production of the same quantity of lactic
acid vis-a-vis a conventional process. Moreover, high selectivity
of membranes and microbes ensures higher purity of the desired
product. Continuous production and fast separation in a com-
pact plant leads to higher yield, productivity and purity attracting
8 P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19
Fig. 3. Separation of l(+) lactic acid by NF-1 membrane: (a) rejection of l(+) lactic acid with increasing transmembrane pressure and (b) permeate ux at 15kgcm
2
pressure
and lactic acid concentration with increasing time.
Fig. 4. Determination of sample purity by peak purity software tool by Agilent HPLC (Series 1200): (a) sample fromfermenter, (b) sample after microltration and (c) sample
after nanoltration.
better process for the product. All these add to the value of the nal
product.
4.6. Improvement in production process and product quality
After rst stagenanoltrationwithNF
2
membraneat 13kgcm
2
pressure, lactic acid concentration reached to 66.97gL
1
. In sec-
ond stage nanoltration, NF-1 membrane turned out to be very
much effective to concentrate l(+) lactic acid in holding vessel 3,
as the membrane operation was carried out with 80% rejection
efciency of lactic acid into the same vessel. At 15kgcm
1
pres-
sure, a steady permeate ow rate was observed from 3 modules
around 14L m
2
h
1
. Effects of transmembrane pressure on rejec-
tionand ux of l(+) lactic acid by NF-1 membrane have beenshown
in Fig. 3. The second part of this gure also indicates howin second
stage of nanoltration, concentration of nal lactic acid solution
reaches 250gL
1
after 23h of continues operation. The l(+) lac-
tic acid, produced in membrane-integrated systemcan directly be
used inthe productionof biodegradable polylactic acid (PLA). Prod-
uct yield (96.5%) achieved in this membrane integrated continuous
production scheme is better than the yield (85%) achieved by con-
ventional system[17]. Final product concentration (250gL
1
) and
density (1.17gcm
3
) are almost same for both the systems. The
continuous membrane-based separation scheme conjugated with
homo-lactic acid fermentation leads to high purity and in terms
of l(+) lactic acid, it is 95% pure as measured by HPLC peak purity
software tool. Fig. 4 indicates howsample purity increases with l-
tration. InFig. 4, maximumnumber of diamonds inthe greenregion
Table 8
Production parameters: comparison of conventional lactic acid with newmembrane based lactic acid.
Membrane integrated plant Conventional lactic acid production plant
Product yield 96.5% 85% [17]
Product concentration 250gL
1
250gL
1
(assumed, as it is nally diluted 5080%)
Productivity of the process 12.40gL
1
h
1
Must be lower than (12.40gL
1
h
1
), as it is weekly batch fermentation
process where initial sugar concentration as substrate is 15%
Purity 95% for l(+) lactic acid Solution contains l(+) lactic acid or mixture of two lactic acid isomers
P. Pal, P. Dey / Chemical Engineering and Processing 64 (2013) 19 9
indicates the high purity of the sample. Purity described for techni-
cal or food grade lactic acid produced fromconventional systemis
interms of onlylactic acidas awholeandnot interms of anyspecic
isomer. Selected microorganismL. delbrueckii, for the fermentation
of sugar cane juice largelyproducedpure varieties of l(+) lactic acid.
The overall production parameters and product characteristics for
both conventional and membrane integrated systemfor lactic acid
productionhas beentabulatedinTable 8. The productionprocess of
such pure l(+) lactic acid with selected strain (L. delbrueckii, NCIM-
2025) in membrane integrated system is highly desirable in the
production of linear homopolymer (molecular size >70) which has
wider range of medical application in respect of other polymers.
5. Conclusion
It transpires fromthe study that integration of membrane sepa-
rationwithconventional fermentationprocess canbringabout very
extensive process intensication in lactic acid production process.
A simple, eco-friendly, energy saving and exible plant congu-
ration that results from membrane integration not only promises
better quality of l(+) lactic acid at reduced cost but also opens up
possibility of similar process intensication for many other chem-
ical process industries through adoption of membrane technology.
Selection of renewable raw material, membranes and operating
modules has also signicantly contributed to the novelty of the
process. In this era of emaciated prot margin in a highly com-
petitive global market, such process intensication is absolutely
essential to development of sustainable technology. Future studies
should explore further development with synthesis and use of bet-
ter membranes offering still higher selectivity, ux and durability
at reduced cost.
Acknowledgement
The authors are thankful to the Department of Science and
Technology (DST), Government of India for nancial support under
Green Technology/Chemistry Program.
References
[1] F.J. Keil, Modelling of Process Intensication An Introduction and Overview,
WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, 2007, ISBN 978-3-527-
31143-9 (Copyright @ 2007).
[2] E. Drioli, A.I. Stankiewicz, F. Macedonio, Membrane engineering in pro-
cess intensication an overview, Journal of Membrane Science 380 (2011)
18.
[3] B.V. Bruggen, E. Curcio, E. Drioli, Process intensication in the textile industry:
the role of membrane technology, Journal of Environment Management 73
(2004) 267274.
[4] J.B. Chaudhuri, D.L. Pyle, Emulation liquid membrane extraction of organic
acids. 1. A theoretical model for lactic acid extraction with emulsion swelling,
Chemical Engineering Science 47 (1992) 4148.
[5] K.L. Wasewar, V.G. Pangarkar, A.B.M. Heesink, G.F. Versteeg, Intensication of
enzymatic conversion of glucose to lactic acid by reactive extraction, Chemical
Engineering Science 58 (2003) 33853393.
[6] H.G. Joglekar, I. Rahman, S. Babu, B.D. Kulkarni, A. Joshi, Comparative assess-
ment of downstream processing options for lactic acid, Separation and
Purication Technology 52 (2006) 117.
[7] P. Pal, J. Sikder, S. Roy, L. Giorno, Process intensicationinlactic acidproduction:
a review of membrane based processes, Chemical Engineering and Processing
48 (2009) 15491559.
[8] G.Q. Xu, J. Chu, Y.-h. Wang, Y.-p. Zhuang, Development of a continuous
cellrecycle fermentation systemfor production of lactic acid by Lactobacillus
paracasei, Process Biochemistry 41 (2006) 24582463.
[9] Y.J. Wee, H.W. Ryu, Lactic acid production by Lactobacillus sp. RKY2 in
a cellrecycle continuous fermentation using lignocellulosic hydrolysate
as inexpensive raw materials, Bioresource Technology 100 (2009)
42624270.
[10] L. Giorno, K. Chojnacka, L. Donato, E. Drioli, Study of a cellrecycle membrane
fermenter for the productionof lactic acid by Lactobacillus bulgaricus, Industrial
and Engineering Chemistry Research 41 (2002) 433440.
[11] A.M.R.B. Xavier, L.M.D. Goncalves, J.L. Moreira, M.J.T. Carrondo, Operational pat-
terns affecting lactic acid production in ultraltration cell recycle bioreactor,
Biotechnology and Bioengineering 45 (1995) 320327.
[12] P. Dey, P. Pal, Direct production of l(+) lactic acid in a continuous and fully
membrane integrated hybrid reactor system under non-neutralizing condi-
tions, Journal of Membrane Science 389 (2012) 355362.
[13] F.Y. Yamashita, T. Kokugan, An improved model for cross-ow microltration
properties of lactic acid fermentation broth, Journal of Chemical Engineering
of Japan 43 (2010) 993997.
[14] H. Carrere, F. Blaszkowa, H.R. deBalmann, Modellingthemicroltrationof lactic
acidfermentationbroths andcomparisonof operatingmodes, Desalination145
(2002) 201206.
[15] P. Dey, L. Linnanen, P. Pal, Separation of lactic acid fromfermentation broth by
cross ownanoltration: membrane characterizationandtransport modelling,
Desalination 288 (2012) 4757.
[16] R. Datta, S.P. Tsai, P. Bonsignore, S.H. Moon, J.R. Frank, Technological and
economic potential of poly (lactic acid) and lactic acid derivatives, FEMS Micro-
biology Review16 (1995) 221231.
[17] G.C. Inskeep, G.G. Taylor, W.C. Breitzke, Lactic acid fromcorn sugar, Industrial
and Engineering Chemistry 44 (1952) 19551966.
[18] M.S. Peters, K.D. Timmerhaus, Plant Design and Economics for Chemical Engi-
neers, fourth ed., McGraw-Hill, NewYork, 1991.
[19] P.M. Doran, Bioprocess Engineering Principles, second ed., Elsevier Ltd., New
York, 2008.
[20] C. Akerberg, G. Zacchi, An economic evaluation of the fermentative pro-
duction of lactic acid from wheat our, Bioresource Technology 75 (2000)
119126.