You are on page 1of 18

Clean Technologies and Environmental Policy

https://doi.org/10.1007/s10098-019-01693-1

ORIGINAL PAPER

Eco‑efficiency and techno‑economic analysis for maleic anhydride


manufacturing processes
Patrick V. Mangili1 · Pedro G. Junqueira1 · Lizandro S. Santos1 · Diego M. Prata1

Received: 27 October 2018 / Accepted: 4 April 2019


© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Abstract
Maleic anhydride may be obtained from different technological routes, being the selective oxidation of benzene and oxida-
tion of butane the only ones that are currently in operation and, hence, represent competitive alternatives. In this paper, the
said technologies are compared with regard to their economics and ecological performances in order to assert which one
corresponds to the cleanest technology. The economics of each process was estimated on the basis of their respective cash
flows, while the environmental comparison was carried out through the Eco-efficiency Comparison Index method by estimat-
ing six different categories of eco-indicators and seven life cycle metrics. To the best of our knowledge, such technologies
have not been compared in terms of a joint evaluation of life cycle and eco-efficiency metrics, let alone considering the
design of their respective utility plants. Finally, a sensitivity analysis was performed in order to analyze how the heuristic
parameters for the utility plants considered in this work affect the estimation of the said indicators. The butane technology
was shown to be more sustainable than the benzene process, since it was approximately 72% more profitable and 38% more
eco-efficient than the latter.
Graphical abstract

Keywords  Cash flow · Eco-efficiency · Maleic anhydride · Process simulation · Utility plants · Waste reduction

Extended author information available on the last page of the article

13
Vol.:(0123456789)
P. V. Mangili et al.

List of symbols V3 Vessel 3


A Heat transfer area 𝜈̇ bfw Boiler feed water volumetric flow rate
C1 Column 1 𝜈̇ cw Cooling water volumetric flow rate
C2 Column 2 𝜈̇ hps High-pressure steam volumetric flow rate
Cbenzene Benzene composition 𝜈̇ lps Low-pressure steam volumetric flow rate
Cbutene n-Butane composition 𝜈̇ mps Medium-pressure steam volumetric flow rate
CMAN Maleic anhydride composition xbenzene Benzene molar fraction
Coxygen Oxygen composition xcumene Cumene molar fraction
Di Inner diameter xDIPB p-Diisopropyl benzene molar fraction
E1 Cooler 1
Greek symbols
E2 Cooler 2
α Cooling water loss factor (process)
E3 Cooler 3
ß Cooling water loss factor (cooling tower)
E4 Cooler 4
γ Cooling water loss factor (blowdown)
EC CO2 emissions eco-indicator
δ Cooling water make-up
Ecomb Total thermal energy from combustion
ε Steam loss factor (condensate losses)
consumption
ζ Steam loss factor (boiler blowdown)
EE Energy use eco-indicator
η Steam loss factor (feed water treatment)
EF Fuel consumption eco-indicator
θ Feed water make-up
Eind Total electricity consumption
Emcomb Total ­CO2 emissions due to thermal energy Abbreviations
consumption ATP Aquatic toxicity potential
Emfug Total ­CO2 emissions due burning off-gases in bfw Boiler feed water
the flare cw Cooling water
Emind Total ­CO2 emissions due to electricity ECI Eco-efficiency Comparison Index
consumption EPA US Environmental Protection Agency
ERM Raw material consumption eco-indicator GWP Global warming potential
EW Water consumption eco-indicator hps High-pressure steam
EWW Wastewater generation eco-indicator HTPI Human toxicity potential by ingestion
f Temperature difference factor HTPE Human toxicity potential by inhalation
F1 Fired heater IPCC Intergovernmental Panel on Climate Change
H Pump head LCA Lice cycle assessment
H1 Heater 1 lps Low-pressure steam
H2 Heater 2 MAN Maleic anhydride
H3 Heater 3 MOC Material of construction
K1 Compressor mps Medium-pressure steam
L Vessel length ng Natural gas
ṁMAN Maleic anhydride mass flow rate NPV Net present value
ṁNG Natural gas mass flow rate NRTL Non-random two liquid
ṁRM Raw material mass flow rate PCOP Photochemical oxidation potential
P1 Pump 1 PEI Potential environmental impact
P2 Pump 2 TTP Terrestrial toxicity potential
P3 Pump 3 UNCTAD United Nations Conference on Trade and
P4 Pump 4 Development
P5 Pump 5 WAR​ Waste reduction
P6 Pump 6
pB Partial pressure of butane
Pc Compressor power Introduction
pM Partial pressure of MAN
Q Energy requirement Maleic anhydride (MAN) is a chemical product associated
R1 Tubular reactor with a broad market due to its application as an intermediate
V̇ Volumetric flow rate for the production of unsaturated polyester resins, polymers,
V1 Vessel 1 vernices and paints, among several other commodities. The
V2 Vessel 2 global MAN market was estimated at approximately 1.8

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

million tons in 2017 and is expected to reach 2.2 million calculating the environmental indicators, whereas the
tons by 2023, at an estimated compound annual growth rate assumptions to estimate the economics are provided in
of 3.69%, being the region of Asia–Pacific accountable for “Economic analysis” section. In “Results and discussion”
the largest market (Mordor Intelligence 2018). MAN can be section, we discuss the methodology required to establish
manufactured from different technological routes, namely the comparison between both technologies, while in “Sen-
the selective conversion of furfural (Alonso-Fagúndez et al. sitivity analysis” section a sensitivity analysis is carried out
2017), catalytic oxidation of butenes (Matar and Hacth to demonstrate the relevance of design parameters and utility
2000), as side product from phthalic anhydride production plant heuristics in estimating the economics and environ-
(Matar and Hacth 2000), oxidation of levulinic acid (Chatzi- mental impacts of both processes. Finally, in “Conclusion”
dimitriou and Bond 2015), selective oxidation of benzene section, we summarize the main results and provide some
(Turton et al. 2010) and oxidation of n-butane (Sharma et al. suggestions for future works.
1991). The furfural and levulinic acid routes are still under
development and have not yet been established as competi-
tive process alternatives. The phthalic anhydride-recovered Maleic anhydride production
MAN technology, in turn, has very low expressiveness in
the global production demand, whereas there is no butene- As previously mentioned, maleic anhydride may be obtained
based process currently in operation (Felthouse et al. 2001). from different technological routes, being the benzene-based
Although old-fashioned, the benzene route is still used by and butane-based designs the only ones that are currently
some companies in Asia and the Far East, while the butane in operation and, hence, represent competitive alternatives.
route corresponds to a most recent alternative since it not The selective oxidation of benzene induces the loss of two
only has lower costs but also is less aggressive to the envi- carbon atoms, thus decreasing mass yield and resulting in
ronment (Rase 2010). the formation of heavy by-products such as phthalic anhy-
This paper aims to compare the economics and environ- dride and benzoquinone. In the oxidation of n-butane, in
mental performances of both technologies for MAN manu- turn, lighter by-products such as acetic and formic acids are
facture. As far as we know, a detailed comparison of both obtained (Centi et al. 2001).
processes with regard to their eco-efficiency, combined with Both technologies also differ from each other in terms of
life cycle burdens, has not been made, let alone considering operating conditions that depend on several factors such as
their respective utility plants. Eco-efficiency relates the pro- feedstock conditions, process design. The reactor type may
cess’ environmental impacts to its economic performance. be of fixed, fluidized or moving bed. MAN recovery can be
Such an analysis is generally performed on the basis of the performed in either aqueous or organic phase, whereas its
quantitative evaluation of the process eco-indicators, which purification may be carried out through either azeotropic
are represented by a relationship between an environmen- distillation or thin-film evaporation (Baerns 2004).
tal variable (e.g., air emissions) and an economic variable The benzene and butane technologies for MAN manu-
(e.g., production rate or net profit) (Mangili et al. 2018a). In facture studied in this paper were originally proposed by
this regard, raw material consumption, fuel consumption, Turton et al. (2010) and Frank (1975), respectively. Both
energy use, C­ O2 emissions, water consumption and waste- processes were simulated in ­UniSim® Design Suite R390.1,
water generation were chosen as eco-indicators. Other envi- so we could obtain the required data to estimate the costs and
ronmental metrics considered in the ecological performance determine their eco-indicators. For the benzene technology,
assessment could be calculated through the waste reduction the same inlet feed conditions and equipment parameters
(WAR) algorithm, namely photochemical oxidation potential used by Turton et al. (2010) were considered. The butane
(PCOP), acidification potential (AP), ozone depletion poten- technology design provided by Frank (1975) was scaled up
tial (ODP), human toxicity potential by ingestion (HTPI), in order to match both plant capacities and thus allow a fair
human toxicity potential by exposure (HTPE), aquatic toxic- comparison regarding the processes’ economics. Both tech-
ity potential (ATP) and terrestrial toxicity potential (TTP). nologies were simulated under steady-state conditions by
The aforementioned indicators were then normalized and using the non-random two liquid (NRTL) thermodynamic
grouped through the Eco-efficiency Comparison Index (ECI) model (enthalpy, phase equilibrium, etc.).
method (Pereira et al. 2018). On the other hand, the eco-
nomic analysis was performed by evaluating cash flows of Benzene technology
the respective technologies in order to provide the scientific
community with a more complete comparison. According to Turton et  al. (2010), MAN production
In the next section, a brief revision regarding the maleic from oxidation of benzene is achieved by using a vana-
anhydride manufacturing processes is presented. In “Envi- dium–molybdenum catalyst and may not exceed 650 °C in
ronmental analysis” section, we describe the method for order to prevent sintering, as described by Eq. 1.

13
P. V. Mangili et al.

C6 H6 + 4.5O2 → C4 H2 O3 + 2CO2 + 2H2 O to 4 take place, whose kinetics are presented in Table 1.
(1) Assuming a catalyst bulk density of 900 kg/m3 (Westerink
benzene MAN
and Westerterp 1988), the catalyst load is 33,930 kg. One
Since the reaction is highly exothermic, an increase in the should note that R1 is cooled through a molten salt-cool-
temperature may result in the total oxidation of the feedstock ing system, which is integrated to distillation column C2’s
and further oxidation of MAN, as described by Eqs. 2 and reboiler and generates hps to be interpreted as credit in the
3, respectively. economic assessment. Such a system was not simulated due
C6 H6 + 7.5O2 → 6CO2 + 3H2 O (2) to the software limitations regarding the simulation of solids.
However, since Turton et al. (2010) provided the informa-
C4 H2 O3 + 3O2 → 4CO2 + H2 O (3) tion on Pump P2 and cooler E1, such equipment was taken
If lower temperatures are achieved, benzene may be oxi- into account in both economic and environmental analyses.
dized to quinone, as in Eq. 4. The reactor product (containing MAN, side products
and unreacted reagents), at 608.0 °C and 2825.3 kmol/h,
C6 H6 + 1.5O2 → C6 H4 O2 + 2H2 O
(4) is cooled down to 260.0 °C and fed to the bottom of an
quinone absorption column C1 having 14 sieve plates plus full-reflux
condenser and reboiler. A dibutyl phthalate make-up stream,
The process flow diagram for the MAN production tech-
nology via benzene oxidation is illustrated in Fig. 1, which
shows the results obtained by simulation. In the schematic,
cw, bfw, lps, hps and ng stand for cooling water, boiler feed Table 1  Reaction kinetics for the benzene technology
water, low-pressure steam, high-pressure steam and natural Reaction Rate expression (Turton et al. 2010)
gas, respectively.
Reaction 1—Eq. 1 r1= 7.70 × 106 exp(− 105,198/RT) Cbenzene
Fresh benzene at 42.3 kmol/h is pressurized and mixed
Reaction 2—Eq. 2 r2= 2.33 × 104 exp(− 89,659/RT) Cbenzene
with compressed air from compressor K1 at 2790.0 kmol/h.
Reaction 3—Eq. 3 r3= 7.20 × 105 exp(− 113,591RT) CMAN
The resulting mixture is heated to 460.0 °C prior to being fed
Reaction 4—Eq. 4 r4= 6.31 × 107 exp(− 124,892/RT) Cbenzene
to a tubular reactor R1, composed of 12,100 catalyst-filled
tubes with 6 m length and 2.5 cm diameter (reaction vol- Cbenzene: benzene composition. CMAN: MAN composition. Composi-
ume of 37.7 m3), in which the reactions described by Eqs. 1 tion units: kmol  m−3. Reaction rate units: kmol  m−3  s−1. Activation
energy units: kJ kmol−1. R: 8.314 kJ kmol−1 K

Fig. 1  MAN production process via oxidation of benzene

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

used as solvent for recovering MAN, is fed at 0.1 kmol/h and becomes richer in dibutyl phthalate, carrying MAN along
to the top of C1. The absorber’s top product consisting of with it.
unreacted components and combustion gases is obtained at
260.0 °C and sent to the flare at 2797.6 kmol/h. C1’s bot- Butane technology
tom product containing solvent and recovered MAN are fed
to the 27th plate of a 42 plate distillation column C2, from The n-butane route is carried out using a vanadium phosphorus
which MAN is obtained as distillate at 27.6 kmol/h with a oxide catalyst, according to Eq. 5, at temperatures kept below
purity of 94.7 mol%. C2’s bottom product, comprised of 500 °C in order to prevent catalyst deactivation (Ebner et al.
dibutyl phthalate, is recycled at 500.0 kmol/h to the top of 1999). The butane/oxygen concentration ratio must be kept
the absorption column. between 1.4 and 1.6% for safety purposes, since the butane’s
The NRTL model parameters for the pairs predicted by the flammability limit may increase rapidly when the temperature
software are given in Table 2, while Fig. 2 shows the ternary increases. Moreover, in order to avoid excessive loss of unre-
diagram and residue curve map of the MAN–water–dibutyl acted butane, a 99.5% purity oxygen feed is recommended
phthalate mixture at 0.8 bar. Dibutyl phthalate (DB-PHTH) (Baerns 2004).
is used as solvent to extract MAN from water, forming a two-
C4 H10 + 3.5O2 → C4 H2 O3 + 4H2 O
phase liquid region which, through distillation, separates water (5)
at the top section (raffinate) from MAN/solvent at the bottom butane MAN
section (extract). In fact, we note from Fig. 2a that, at 0.8 bar, Similarly to the benzene route, an increase in the tempera-
all mixtures of MAN and DB-PHTH are single phase and, ture may also result in further oxidation of MAN (as shown
therefore, only water–DB-PHTH mixtures present two-phase in Eq. 3) and total oxidation of the feedstock, as described by
behavior. From Fig. 2b, in turn, we see that water has the high- Eq. 6.
est boiling rate, whereas DH-PHTH has the lowest boiling
rate. Hence, during distillation, the liquid is depleted of water C4 H10 + 6.5O2 → 4CO2 + 5H2 O (6)

Table 2  NRTL model Benzene Oxygen MAN Water CO2 DB-PHTH


parameters for the benzene
route Benzene – − 12.163 762.556 4005.180 69.676 − 444.421
Oxygen 1480.016 – 2886.061 1863.905 0.000 1282.441
MAN 387.981 − 12.163 – 0.000 69.676 0.000
Water 3088.443 − 12.163 0.000 – 69.676 0.000
CO2 − 3918.325 0.000 − 2805.650 − 2941.997 – − 3986.726
DB-PHTH 1248.170 0.000 − 12.163 0.000 69.676 –

Fig. 2  Ternary equilibrium for the benzene process

13
P. V. Mangili et al.

The process flow diagram for the MAN production tech- Table 3  Reaction kinetics for the butane technology
nology via n-butane oxidation is illustrated in Fig. 3, which Reaction Rate expression (Sharma et al. 1991)
shows the results obtained by simulation. In the schematic,
mps stands for medium-pressure steam. Reaction 3—Eq. 3
( ( ))
−155000 1 1
2.9×10−6 exp R T
− 673 ×pM
r3 =
Fresh n-butane at 70.0 kmol/h is heated to 25.0 °C and
2
(1+310pM )
mixed with air at 820.2 kmol/h and a compressed gas recy- Reaction 4—Eq. 5
( ( ))
−93100 1 1
9.6×10−7 exp R T
− 673 ×p0.54
B
r4 =
cle stream from the absorption section, at 78.3 kmol/h. The 1+310pM

resulting mixture is heated to 420.0 °C prior to being fed Reaction 5—Eq. 6 r5 = 1.5 × 10−7 exp
(
−93100
(
1
− 1
))
× p0.54
to a tubular reactor R1 having 15,000 catalyst-filled tubes R T 673 B

of 2.5 cm diameter and 4.9 m length (reaction volume of pM: partial pressure of MAN. pB: partial pressure of butane. Partial
37.0 m3), in which the reactions described by Eqs. 3, 5 and pressure units: atm. Reaction rate units: kmol  m−3  s−1. Activation
6 take place, whose kinetics were retrieved from Sharma energy units: kJ kmol−1. R: 8.314 kJ kmol−1 K
et al. (1991)’s work and are presented in Table 3. Assum-
ing a catalyst bulk density of 1000 kg/m3 (Musa 2016), the
catalyst load is 37,000 kg. Here again we can observe that the temperature difference between the salt bath and the
R1 is cooled through a salt-cooling system, which was not reaction system may not exceed 80 °C in order to prevent
simulated. However, since the simulation provided the value catalyst degradation (Ebner et al. 1999).
of the required energy to cool down the reactor, we used the The reactor product (containing MAN and unreacted
said value to estimate the amount of bfw used to cool the reagents), at 470.0 °C and 998.7 kmol/h, is cooled down
molten salt and consequently generate hps which, similarly to 65.3 °C after passing through a series of three coolers
to the benzene technology, was considered as credit in the prior to being sent to a flash vessel V1. MAN is obtained
economic analysis. In a matter of operational importance, as V1’s bottom product at 98.6 kmol/h with a purity of

Fig. 3  MAN production process via oxidation of n-butane

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

70.0 mol% MAN and sent to a distillation column C2. In Utilities plant


this paper, we used UniSim’s shortcut method to perform
a preliminary design of the said column and obtain the The utilities plant was designed to allow a more realistic esti-
required number of stages, optimal feed section and reflux mation of the water consumption and wastewater generation
ratio. Such results were subsequently used in the simula- from both technologies. Its flow sheet is illustrated in Fig. 6,
tion of a rigorous distillation model, from which water is in which 𝜈̇ cw, 𝜈̇ bfw, 𝜈̇ lps, 𝜈̇ mps and 𝜈̇ hps correspond to the total
retrieved as distillate and recycled to the absorption col- volumetric flow rate of cooling water, boiler feed water, low-
umn at 68.6 kmol/h and MAN is obtained at 30.0 kmol/h at pressure steam, medium-pressure steam and high-pressure
the bottom with a purity of 99.9 mol%. The C2’s composi- steam, respectively, whereas α, ß and γ are the cooling water
tion profiles are shown in Fig. 4. loss factors due to the process, cooling tower and blowdown,
V1’s top product, containing unreacted components and respectively; and ε, ζ and η are the steam loss factors associ-
combustion gases, is fed at 900.1 kmol/h to the bottom of ated with the process (condensate losses), boiler blowdown
an absorption column C1, whose top section is fed by a and feed water treatment, respectively. One should note that
water make-up stream at 76.6 kmol/h. C1’s bottom product δ and θ represent the cooling water and steam generation
is then divided into a water recycle stream, which is sent systems’ make-up, respectively. Therefore, δ corresponds
back at 100.0 kmol/h to the seventh stage of the absorber, to α + ß + γ, while θ corresponds to ε + ζ + η.
and a maleic acid waste stream, which is obtained at
172.2  kmol/h containing about 99.5  mol% water. C1’s
top product, in turn, is divided into a gas recycle stream, Environmental analysis
which is pressurized in compressor K1 and sent back at
78.3 kmol/h to the feed section, and an off-gas stream, Numerous studies related to the evaluation of the environ-
obtained at 725.3 kmol/h and sent to the flare. mental impacts from maleic anhydride production processes
The NRTL model parameters for the pairs predicted by are available in the literature. The US Environmental Pro-
the software are given in Table 4. As shown in Fig. 5, all tection Agency (EPA 1980) analyzed the benzene emis-
separations may be deemed fairly easy to be performed sions from process vents of MAN plants and established a
by distillation since the curves are quite fat, which thus National Emission Standard for such an air pollutant. Chen
require low number of trays and reflux ratios. and Shonnard (2004), in turn, estimated the C­ O2 emission
rates for both processes on the basis of the EPA emission

Fig. 4  Distillation column C2’s


mol composition profile

Table 4  NRTL model n-Butane Oxygen MAN Water CO2 Nitrogen


parameters for the butane route
n-Butane – − 12.163 2707.386 3525.34 69.676 21.508
Oxygen 1200.414 – 2886.061 1863.905 0.000 0.000
MAN 1862.748 − 12.163 – 0.000 69.676 20.032
Water 3796.107 − 12.163 0.000 – − 183.691 20.032
CO2 − 3986.46 0.000 − 2805.65 676.278 – 0.000
Nitrogen 3.468 0.000 2003.572 602.209 0.000 –

13
P. V. Mangili et al.

Fig. 5  Phase equilibrium for the butane process

Fig. 6  Utilities plant flow sheet

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

factors. Fermeglia et al. (2009) applied the waste reduction on the basis of the emission factors for Brazil (considering
(WAR) algorithm to compare the benzene and butane tech- utility plants, electricity, flare emissions, etc.), which repre-
nologies in terms of six life cycle assessment (LCA) indica- sents a more complete analysis.
tors, namely global warming potential (GWP), photochemi- The calculation of the fuel consumption was performed
cal oxidation potential (PCOP), acidification potential (AP), by assuming that only the hypothetical utility plant’s boiler
ozone depletion potential (ODP), human toxicity potential and the fired heaters are fuel (natural gas) consumers oper-
by ingestion (HTPI), human toxicity potential by exposure ating with 80% thermal efficiency (EPA 2008; Trinks et al.
(HTPE), aquatic toxicity potential (ATP) and terrestrial tox- 2004). The energy use and ­CO2 emission indicators were
icity potential (TTP). Mangili et al. (2018b) calculated the determined through the guidelines provided by Mangili
­CO2 emissions eco-indicators of both technologies on the et al. (2018a) by assuming that pumps and compressors
basis of the United Nations Conference on Trade and Devel- operate with electricity at 75% efficiency, whereas heaters
opment (UNCTAD 2004) and Intergovernmental Panel on and reboilers were interpreted to operate with steam gener-
Climate Change’s guidelines (IPCC 2006). ated in the utility plant’s boiler. The same conversion fac-
Nevertheless, to the best of our knowledge, MAN manu- tors for electricity (0.0227 tCO2 ∕GJ ) and natural gas thermal
facturing technologies have only been studied in terms of energy (0.0561 tCO2 ∕GJ ) to carbon dioxide emissions were
either single environmental indicators or LCA indicators. used. The fugitive emissions (flare) were estimated by con-
The determination of a single indicator is not sufficient to sidering total combustion of gases. The water consumption
assert which technology has the highest environmental per- eco-indicators were also determined according to Mangili
formance, since most indicators are calculated independently et al. (2018a)’s work. Cooling water supply and return tem-
of one another. In addition, LCA indicators do not take the peratures were assumed to be 30 °C and 45 °C, respectively
product value into consideration, which may represent a (Turton et al. 2010). Boiler feed water supply temperature
tricky evaluation when the process economics differ sig- was considered to be 130 °C (Tvedt and Holloway 1997). α
nificantly from each other. In this paper, we compared the and ß together were interpreted as 3% (Seider et al. 2008)
ecological efficiencies of the benzene and butane technolo- of v̇cw, whereas γ was assumed to be 0.3% (Smith 2005) of
gies not only in terms of the aforementioned LCA indices 𝜈̇ cw. ζ and η were together interpreted as 3% of 𝜈̇ bfw, while
(i.e., HTPI, HTPE, TTP, ATP, ODP, PCOP and AP) but also ε was assumed to be 20% (Seider et al. 2008) of the total
with regard to eco-efficiency indicators (i.e., raw materials steam flow rate ( 𝜈̇ lps + 𝜈̇ mps + 𝜈̇ ̇hps). The low-, medium- and
and fuel consumption, energy use, ­CO2 emissions, water high-pressure steams conditions used here were taken from
consumption and wastewater generation). The LCA metrics Seider et al. (2008)’s book and are shown in Table 6. Steam
are expressed in terms of potential environmental impact used as heating source was assumed to exit the heat exchang-
(PEI) per kg of MAN and were calculated with the aid of ers as saturated liquid.
the WAR tool, whereas the eco-indicators were determined The ecological performances of the MAN production
through their respective formulae presented in Table 5. One technologies were then quantitatively compared by means of
should note that the GWP index has not been considered in the Eco-efficiency Comparison Index (ECI) method, origi-
this study since we have already estimated the ­CO2 release nally proposed by Pereira et al. (2018). The ECI method is
intended to jointly evaluate process indicators in order to
define the most eco-efficient. Such a methodology was used
Table 5  Eco-indicators equations
by Mangili et al. (2018a) to compare three acetone–metha-
Indicator Formula Unit nol separation technologies in terms of their environmental
Raw material consumption ṁ RM tRM/tMAN
performance and proved to be functional for such a purpose.
ERM = ṁ MAN Junqueira et al. (2018) also applied the ECI approach for
Fuel consumption EF =
ṁ NG tNG/tMAN
ṁ MAN comparing six cumene manufacturing processes in terms
Energy use EE =
Eind +Ecomb GJ/tMAN of both environmental and economic aspects. Mangili and
ṁ MAN
CO2 emissions EC =
Emind +Emcomb +Emfug tCO2/tMAN Prata (2019), in turn, based their study on such a method in
ṁ MAN
Water consumption EW = 𝛿+𝜃 m3w/tMAN
ṁ MAN
Wastewater generation EWW = 𝛾+𝜁 +𝜀 m3ef/tMAN Table 6  Steam parameters
ṁ MAN

ṁ RM: total mass flow rate of raw materials. ṁ MAN: total mass flow rate Steam Supply temperature Pressure (bar)
of MAN product. ṁ NG: total mass flow rate of natural gas consumed (°C)
by fuel consumers. Eind: total electricity consumed by pumps and
compressors. Ecomb: total thermal energy consumed by heaters and Low pressure 135.0 3.10
reboilers. ­Emind: total ­CO2 emissions due to electricity consumption. Medium pressure 185.5 11.35
­Emcomb: total ­CO2 emissions due to thermal energy consumption. High pressure 254.0 42.38
­Emfug: total ­CO2 emissions due to burning gases in the flare

13
P. V. Mangili et al.

order to improve different butyl acetate production plants Table 8. A Plant Cost Index of 616.4 was assumed (Access
with regard to ecological, economic and safety indices. The Intelligence LLC 2019).
same procedure applied by the said references was used in It is worth mentioning that both reactors were interpreted
this study. as heat exchangers and that flash/reflux drums were consid-
ered to be horizontal pressure vessels (Seider et al. 2008).
Furthermore, we assumed that the catalyst purchase is a
Economic analysis one-time investment (capital cost) having a price of $30/kg
(Godefroy et al. 2009). The depreciable capital costs were
The economics of both technologies was determined by then estimated by taking into account the general site expen-
estimating their fixed capital investment (FCI) and produc- ditures disclosed in Table 9.
tion expenditures and subsequently carrying out a cash flow Regarding the production costs, benzene and n-butane
analysis. The capital costs for each process were calculated were priced at 3.45 $/gal (ICIS—Independent Chemical
on the basis of the equipment cost equations given by Seider Information System 2018) and 0.65 $/kg (The Chemical
et al. (2008)’s book and are presented in Table 7 by taking Company 2016), respectively, while MAN was priced at
into consideration the equipment specifications provided 0.75 $/lb (ICIS 2018) and maleic acid was priced at 2.165
by the reference authors and simulation results, as given in $/ton (ICIS 2013). Labor costs were assumed to be 50,000
$/operator/year, with 3 shifts and 5 operators per shift. High-
pressure steam credit was 12.33 $/GJ. Electricity and cool-
Table 7  Equipment cost basis ing water were assumed to be 16.80 $/GJ and 1.04 $/GJ,
Equipment Cost equation respectively. Medium- and low-pressure steams were priced
at 14.83 $/GJ and 7.78 $/GJ, respectively, while natural gas
Pumps exp{9.7171 − 0.6019[ln( was assumed to be priced at 4.24 $/GJ. All utility costs were
V̇(H)0.5)] + 0.0519[ln(V̇
(H)0.5)]2} retrieved from Turton et al. (2010), with the exception of the
Compressors exp{7.5800 + 0.8000[ln(Pc)]} natural gas price, which was given by The Chemical Com-
Heat exchangers exp{11.667 − 0.8709[ln(A)]  pany (2016). Other escalation parameters used to estimate
+ 0.09005[ln(A)]2} the production costs are shown in Table 10.
Columns/vessels 361.8(Di)0.7396(L)0.70684 The cash flow analysis was performed by considering that
Fired heaters exp{7.5800 + 0.8000[ln(Q)]} both processes were redesigned, since MAN production pro-
Cooling tower 164fQ0.61 cesses have already been implemented, with a digital control
Boiler 0.36Q0.77 structure and installed on grass-root sites in Brazil. Their
economic life was assumed to be of 13 years with 1 year
V̇  ̇: volumetric flow rate (gal/min). H: pump head (ft). Pc: power con- period of design and 2 years of construction. A 10% straight
sumed (HP). A: heat transfer area (­ft2). Di: inner diameter (in). L:
length (in). Q: energy requirement (Btu/h), f: temperature difference line depreciation method (PwC—PricewaterhouseCoop-
factor ers 2018) was considered along with a 6.90% interest rate

Table 8  Equipment design parameters


Equipment Parameters

Pumps Centrifugal, MOC stainless steel, 75% efficiency


Compressors Centrifugal, MOC carbon steel, 75% efficiency
Coolers/condensers Shell and tube, floating head, MOC stainless steel
Heaters/reboilers Kettle type, MOC stainless steel
Reactor R1 (benzene) Tubular, 12,100 tubes with 2.5 cm diameter and 6 m length, 900 kg/m3 catalyst bulk density, MOC stainless steel
Reactor R1 (butane) Tubular, 15,000 tubes with 2.5- cm diameter and 4.9 m length, 1000 kg/m3 catalyst bulk density, MOC stainless steel
Absorber C1 (benzene) 14 sieve trays with 4.1 m diameter, 0.61-m tray spacing, 0.21 reflux ratio, MOC nickel-based alloy
Column C2 (benzene) 42 sieve trays with 1.0 m diameter, 0.40-m tray spacing, 1.2 reflux ratio, MOC nickel-based alloy
Absorber C1 (butane) 43 sieve trays with 1.83 m diameter, 0.5-m tray spacing, 0.67 reflux ratio, MOC nickel-based alloy
Column C2 (butane) 8 sieve trays with 0.61 m diameter, 0.55-m tray spacing, 0.07 reflux ratio, MOC nickel-based alloy
Fired heaters 80% thermal efficiency, MOC low alloy
Cooling towers 15 °C temperature range, MOC carbon steel
Boilers 80% thermal efficiency, MOC carbon steel

MOC: material of construction

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

Table 9  Site expenditures
Basis Parameter Value References

Inside battery limits (ISBL) Process equipment costs 30% of FCI Peters et al. (2003)
Bulk items (piping, instruments, etc.) 25% of FCI
Outside battery limits (OSBL) Utility plant equipment costs 20% of FCI Towler and Sinnott (2008)
Warehouse, lighting, waste handling, etc. 40% of ISBL
Other Land development costs 1% of FCI Silla (2003)
Site development Civil works (roads, buildings, ditches, etc.) 5% of FCI Sinnott (2005)
Field expenses (canteens, overtime, etc.)
Miscellaneous Agent fees, import duties, etc. 5% of FCI Sinnott (2005)
Design and engineering Administration, inspection, etc. 7% of FCI Peters et al. (2003)
General Contingency 7% of FCI Peters et al. (2003)
Working capital 5% of FCI Towler and Sinnott (2008)
Location factor Brazil 1.14 Towler and Sinnott (2008)

Table 10  General production expenses


Basis Parameter Value References

Fixed costs (FC) Maintenance 5% of FCI Sinnott (2005)


Overheads 5% of labor
Insurance, license fees and royalties 2% of FCI
Variable costs (VC) Personal protective equipment, cleaning materials, 1% of maintenance Sinnott (2005)
charts and accessories, etc.
Other Research and development 1% of revenue Towler and Sinnott (2008)
Sales and marketing 2% of FC + VC
Human resources, accounting, finance, etc. 35% of labor

(Banco Central do Brasil 2018) and 34% tax rate (Deloitte With regard to C ­ O 2 emissions, Chen and Shonnard
2017). (2004) found that, in fact, the benzene alternative has
a higher emission rate (2.06 tCO2/t MAN) than the butane
technology’s (1.21 tCO2/tMAN). One should note, however,
Results and discussion that the results found by the said reference differ from our
results since the authors considered different parameters
Environmental analysis for estimating the C ­ O2 emissions (e.g., emission factors,
waste management, etc.). In spite of this, we note that the
Table 11 shows the results for the LCA indices obtained benzene process emits almost twice as much ­CO2 as the
through the WAR tool and the eco-indicators calculated with n-butane route.
the formulae from Table 5. We can verify that the butane The contributions of each equipment to the respective
technology has the lowest eco-indicators for all categories eco-indicators of both processes are shown in Fig. 7. We
evaluated, except for the raw material consumption, water can note that the fired heaters F1 and the utility boilers have
consumption and PCOP indices. The higher water consump- the higher energy requirements when compared to other
tion is mainly due to the feed water requirement in the first equipment in both technologies. Also, it is evident that the
column, whereas, as in accordance with Fermeglia et al. ­CO2 emissions are mainly due to the off-gases burned in the
(2009)’s work, the highest photochemical oxidation potential flare, which corresponds to approximately 65% and 49% of
of the butane route may be associated with the PCOP of such the total emissions in the benzene and butane processes,
a feedstock (0.8425), which is higher than the benzene’s respectively. In terms of water consumption, we see that
(0.3884). On the other hand, disregarding the presence of column C1’s condenser has the highest water demand in
the acrylic acid considered by the said reference, the sum both scenarios.
of all other LCA impact categories of the former is lower The indicators shown in Table 11 were normalized by
than the latter’s. dividing them by the highest value present in the same

13
P. V. Mangili et al.

Table 11  Environmental Process Benzene technology Butane technology


indicator results
Production rate, t­MAN/h 2.44 2.20
Raw materials eco-indicator, ­tRM/tMAN 1.36 1.41
Fuel consumption eco-indicator, t­fuel/tMAN 0.53 0.32
Energy use eco-indicator, GJ/tMAN 34.78 29.69
CO2 emissions eco-indicator, tCO2/tMAN 4.12 2.68
Water use eco-indicator, m3W/tMAN 11.13 11.34
Wastewater eco-indicator, m3ef /tMAN 19.12 12.42
Human toxicity by ingestion, PEI/kgMAN 0.54 0.43
Human toxicity by exposure, PEI/kgMAN 0.25 0.24
Terrestrial toxicity potential, PEI/kgMAN 0.54 0.43
Aquatic toxicity potential, PEI/kgMAN 0.52 0.01
Ozone depletion potential, PEI/kgMAN 9.74E−09 5.56E−09
Photochemical oxidation potential, PEI/kgMAN 7.06E−04 0.11
Acidification potential, PEI/kgMAN 0.03 0.02

Fig. 7  Equipment contributions to eco-indicators

category, as suggested by Pereira et al. (2018), and are pre- plotting the normalized values in an irregular tridecagon-
sented in Table 12. shaped radar-type chart, as shown in Fig. 8.
We should note that the results are between 0 and 1, The quantitative evaluation was carried out by sum-
thus corresponding to a quantitative range that varies from ming up the area of each minor triangle calculated with
the most eco-efficient to the least eco-efficient, respec- the law of sines, as shown in Table 13, from which we
tively. The qualitative analysis was then performed by

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

Table 12  Normalized indicators Table 13  Quantitative ECI


Process Benzene Butane Eco-indicator × eco- Benzene technol- Butane technology
technology tech- indicator ogy
nology
ERM × EF 0.96 0.60
Raw materials eco-indicator, t­RM/tMAN 0.96 1.00 EF × EE 1.00 0.52
Fuel consumption eco-indicator, t­fuel/tMAN 1.00 0.60 EE × EC 1.00 0.56
Energy use eco-indicator, GJ/tMAN 1.00 0.85 EC × EW 0.98 0.65
CO2 emissions eco-indicator, tCO2/tMAN 1.00 0.65 EW × EWW 0.98 0.65
Water use eco-indicator, m3W/tMAN 0.98 1.00 EWW × HTPI 1.00 0.52
Wastewater eco-indicator, m3ef /tMAN 1.00 0.65 HTPI × HTPE 1.00 0.75
Human toxicity by ingestion, PEI/kgMAN 1.00 0.80 HTPE × TTP 1.00 0.75
Human toxicity by exposure, PEI/kgMAN 1.00 0.94 TTP × ATP 1.00 0.01
Terrestrial toxicity potential, PEI/kgMAN 1.00 0.80 ATP × ODP 1.00 0.01
Aquatic toxicity potential, PEI/kgMAN 1.00 0.01 ODP × PCOP 0.01 0.57
Ozone depletion potential, PEI/kgMAN 1.00 0.57 PCOP × AP 0.01 0.57
Photochemical oxidation potential, PEI/kgMAN 0.01 1.00 AP × ERM 0.96 0.57
Acidification potential, PEI/kgMAN 1.00 0.57 Sum 10.90 6.73
Tridecagon area 2.53 1.56
Highest values by indicator group
ECI – 38.29%

Table 14  Economic analysis results


Parameters Benzene Butane
technology technol-
ogy

ISBL investment, M$ 12.20 8.01


OSBL investment, M$ 5.00 3.28
Indirect costs (site development, design 3.98 2.67
and engineering, general), M$
Fixed operating costs, M$/year 4.28 4.06
Variable operating costs, M$/year 29.32 28.51
Other operating costs, M$/year 1.46 1.44
Depreciable capital cost, M$ 26.41 23.60
Total production cost, M$/year 35.05 34.00
Revenue (product + credits), M$/year 41.00 40.70
Gross profit, M$/year 5.95 6.70
Net present value (NPV), M$ 8.40 14.42
Internal rate of return, % 13.54 19.23
Return on investment, % 13.18 16.11
Payback period, years 6.82 5.27
Fig. 8  Qualitative ECI
Minimum MAN selling price, $/kg 0.71 0.69

Economic analysis
note that the butane process is approximately 38% more
ecologically efficient than the benzene process. Table 14 presents a summary of the economic evaluation
In light of such results, future studies can be performed results of both technologies, whereas Fig. 9 illustrates the
with regard to the use of high-pressure steam generated net present value (NPV) variation during the project lives.
in the benzene technology. For instance, carbon-free hps It can be verified that the butane route proved to be the most
generated in coolers E1 and E2 could be used to heat the economically attractive since it is approximately 72% more
column C1’s reboiler in order to reduce the process oper- profitable than the benzene process and achieves positive
ating costs and ­CO2 emissions. NPV values earlier than the latter. This is in accordance with

13
P. V. Mangili et al.

Fig. 9  NPV variation

Rase (2010)’s statement regarding that the butane alterna- Sensitivity analysis
tive is not only more environmentally friendly but also more
economically attractive. These results also corroborate Chen The results obtained in this work are based on the original
and Shonnard (2004)’s study, which showed that the butane designs of both technologies described in “Maleic anhydride
process has a higher NPV. production” section. However, one should note that different
The minimum MAN selling prices for both technologies, equipment sizes and/or process configurations may be alter-
in turn, were estimated by means of the cash flow model natively considered. Although the flow sheet optimization of
itself by varying the product revenue in order to set the the processes under evaluation lies beyond the scope of this
respective NPVs equal to zero, given the specified discount paper, a sensitivity assessment can be performed in order to
rate, and were found to be 0.711 $/lb and 0.685 $/lb for the understand how specific design parameters affect the envi-
benzene-based and butane-based routes, respectively. This ronmental and economic results. Since several design con-
clearly evidences the competitiveness of the latter over the straints could be evaluated, only the main parameters were
former. considered in the analysis. For instance, Fig. 10 shows the

Fig. 10  Effects of design parameters on the sustainability: a benzene route. b Butane route

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

effects of varying the number of C1 plates and reactor tubes The energy consumption and carbon emissions were esti-
in both technological routes with regard to their respective mated by considering that the fired heaters and utility plant’s
energy use indicator, water consumption indicator and capi- boiler operate with a thermal efficiency of 80%, which corre-
tal cost. We note that although not particularly expressive, sponds to the worst-case scenario according to EPA (2008).
increasing the number of stages in the distillation columns The water consumption of both technologies, in turn, was
renders the separation easier and, consequently, reduces estimated by considering the worst-case scenario; that is,
the reboiler duties, thus resulting in a decrease in the steam the utility plant’s water blowdown from the boiler and cool-
requirements. On the other hand, increasing the number of ing tower was 0.3% each and the condensate loss from the
reactor tubes results in an increase in both the capital costs process to the steam plant was 20%. Therefore, a sensitivity
and boiler feed water demand. analysis was performed by varying the said parameters in
Moreover, the energy use, ­CO2 emissions and water order to analyze how they affect such indicators, as shown
consumption indicators were calculated in this work on the in Table 15. The results for the energy and C ­ O2 indicators
basis of the heuristic parameters present in the literature. are shown in Fig. 11 and Table 16, while the results for the
water and wastewater indicators are presented in Fig. 12 and
Table 17.
Table 15  Parameters evaluated in the sensitivity analysis From Fig. 11, we verify that, for the Benzene technol-
Parameters Range (%) ogy, the energy and C ­ O2 indicators are more sensitive to the
fired heater efficiency than to the boiler efficiency, which is
Fired heater efficiency 80–90
due to the former’s higher-energy requirements (accounting
Boiler efficiency 80–90
for 57.3% of the total energy use, as shown in Fig. 7). For
ζ (boiler blowdown factor) 0.1–0.3
the Butane technology, however, since the boiler thermal
γ (cooling tower blowdown factor) 0.1–0.3
requirement is the highest (53.8% of the total) as given in
ε (steam condensate loss) 10–20
Fig. 7, the energy and C ­ O2 indicators are more sensitive to

Fig. 11  Influence of thermal efficiencies on the energy and ­CO2 indicators: a benzene process. b Butane process

13
P. V. Mangili et al.

Table 16  Sensitivity analysis—EE and EC eco-indicators of the total in the benzene technology) and E4 (71.1% of the
Indicators Benzene technology Butane technology
total in the butane technology), since the blowdown heu-
ristics are based on the total water circulating flow rate of
Worst case Best case Worst case Best case
both production plants. Varying the cooling tower and boiler
EE 34.78 31.2 29.69 26.33 blowdowns from 0.3 to 0.1% and the condensate loss from
EC 4.12 3.66 2.68 2.34 20 to 10% resulted in decreases of higher than 60% in the
water and wastewater indicators for both processes.
The worst-case scenario: efficiencies = 80%. The best-scale scenario: The results obtained in the sensitivity analysis could be
efficiencies = 90%
used to recalculate the processes’ eco-efficiencies. Figure 13

its efficiency. In fact, varying the boiler and fired heater effi-
ciencies from 80 to 90% in the benzene technology resulted
Table 17  Sensitivity analysis—EW and EWW eco-indicators
in a decrease of 10.3% in the energy indicator and 11.2% in
the ­CO2 indicator. For the butane process, such values were Indicator Benzene technology Butane technology
of 11.3% and 12.7%, respectively. Worst case Best case Worst case Best case
From Fig. 12, we can note that, for both processes, the
water consumption and wastewater generation indicators EW 11.13 4.01 11.34 4.50
depend much more on the cooling tower and boiler’s blow- EWW 19.12 7.53 12.42 4.82
down than on the percentage of condensate loss. This is due The worst-case scenario: ζ = 0.3%, γ = 0.3%, ε = 20%. The best-case
to the higher water requirements on condensers E3 (87.3% scenario: ζ = 0.1%, γ = 0.1%, ε = 10%

Fig. 12  Influence of blowdown and condensate loss on the water consumption and wastewater generation indicators: a benzene process. b
Butane process

13
Eco‑efficiency and techno‑economic analysis for maleic anhydride manufacturing processes

integration and design new MAN purification methodologies


for reducing costs and mitigating ecological impacts.

Acknowledgements  This study was financed in part by the Coorde-


nação de Aperfeiçoamento de Pessoal de Nível Superior —Brasil
(CAPES)—Finance Code 001.

References
Access Intelligence LLC (2019) Economic indicators: CECPI february
2019. Chemical engineering. https​://www.cheme​ngonl​ine.com/
pci. Accessed 10 Mar 2019
Alonso-Fagúndez N, Ojeda M, Mariscal R, Fierro JLG, Granados ML
(2017) Gas phase oxidation of furfural to maleic anhydride on
­V2O5/γ-Al2O3 catalysts: reaction conditions to slow down the
deactivation. J Catal 348:265–275
Baerns M (2004) Basic principles in applied catalysis. Springer series
in chemical physics 75. Springer, Berlin
Banco Central do Brasil (2018) Interest rates. Banco Central do Bra-
Fig. 13  Qualitative ECI considering the best-case scenarios sil. https​: //www.bcb.gov.br/Pec/Copom​/ Ingl/taxaS​e lic-i.asp.
Accessed 08 May 2018
Centi G, Cavani F, Trifirò F (2001) Selective oxidation by heterogene-
illustrates an ECI result considering the best-case scenario ous catalysis, fundamental and applied catalysis. Springer, New
for both technologies, from which we can observe a slight York
Chatzidimitriou A, Bond JQ (2015) Oxidation of levulinic acid for the
difference on the tridecagon areas when compared to Fig. 8.
production of maleic anhydride: breathing new life into biochemi-
In fact, if we considered the best-case scenario, the butane cals. Green Chem 17:4367–4376
route would be approximately 35% more eco-efficient than Chen H, Shonnard DR (2004) Systematic framework for environmen-
the benzene technology, instead of 38% as previously calcu- tally conscious chemical process design: early and detailed design
stages. Ind Eng Chem Res 43:535–552
lated for the worst scenario.
Deloitte (2017) International tax: Brazil highlights 2017. Deloitte. https​
://www2.deloi​tte.com/conte​nt/dam/Deloi​tte/cn/Docum​ents/inter​
natio​nal-busin​ess-suppo​rt/deloi​tte-cn-ibs-brazi​l-int-tax-en-2017.
pdf. Accessed 08 May 2018
Ebner JR, Keppel RA, Mummey MJ (1999) High productivity process
Conclusion for the production of maleic anhydride, US Pat. 6005121. Hunts-
man Petrochemical Corporation
This study compared two maleic anhydride production pro- EPA – U.S. Environmental Protection Agency (1980) Benzene emis-
cesses with regard to their respective economics and ecologi- sions from maleic anhydride industry—background information
for proposed standards. EPA, Research Triangle Park
cal performance. The environmental analysis was performed
EPA – U.S Environmental Protection Agency (2008) Project type:
by means of the Eco-efficiency Comparison Index method industrial boiler efficiency. NSCEP, https​://www.epa.gov/nscep​.
based on six categories of eco-indicators and seven LCA Accessed 08 May 2018
metrics. The processes’ economics, in turn, was evaluated on Felthouse TR, Burnett JC, Horrell B, Mummey MJ, Kuo YJ (2001)
Maleic anhydride, maleic acid and fumaric acid. In: Kirk–Othmer
the basis of the equipment specifications provided by the ref-
encyclopedia of chemical technology. Wiley, New York
erence authors and utility requirements provided by simula- Fermeglia M, Longo G, Toma L (2009) Computer aided design for
tion. The comparison demonstrated that the butane technol- sustainable industrial processes: specific tools and applications.
ogy has higher environmental performance (approximately AIChE J 55:1065–1078
Frank ME (1975) Recycle process for oxidation of n-butane to maleic
38% more eco-efficient) and profitability (approximately
anhydride, US Pat. 3904652. Standard Oil Company
72%) when compared to the benzene technology, which is Godefroy A, Patience GS, Tzakova T, Garrait D, Dubois JL (2009)
mainly due to the former’s lower-energy requirements and Reactor technologies for propane partial oxidation to acrylic acid.
raw material costs. Therefore, the butane oxidation technol- Chem Eng Technol 32:373–379
ICIS – Independent Chemical Information System (2013) Price
ogy was shown to be the best alternative for MAN produc-
and market trends: US maleic acid contracts dip and buy-
tion in terms of both economics and sustainability. ers seek further decreases. https​: //www.icis.com/resou​r ces/
In this regard, process simulation proved to be a conveni- news/2013/04/21/96606​68/price​-and-marke​t-trend​s-us-malei​
ent technique for estimating the economics and environmen- c-acid-contr ​ a cts-dip-and-buyer ​ s -seek-furth ​ e r-decre ​ a ses/.
Accessed 08 May 2018
tal footprint of industrial processes. In addition, based on
ICIS – Independent Chemical Information System (2018) Indicative
the results provided in this paper, future studies on MAN chemical prices A-Z. ICIS. https:​ //www.icis.com/chemic​ als/chann​
production may be developed in order to perform thermal el-info-chemi​cals-a-z/. Accessed 08 May 2018

13
P. V. Mangili et al.

IPCC – Intergovernmental Panel on Climate Change (2006) Guidelines Sharma RK, Cresswell DL, Newson EJ (1991) Kinetics and fixed-bed
for national greenhouse gas inventories. IGES, Hayama reactor modeling of butane oxidation to maleic anhydride. AIChE
Junqueira PG, Mangili PV, Santos RO, Santos LS, Prata DM (2018) J 37:39–47
Economic and environmental analysis of the cumene produc- Silla H (2003) Chemical process engineering: design and economics.
tion process using computational simulation. Chem Eng Process Marcel Dekker Inc., New York
130:309–325 Sinnott RK (2005) Chemical engineering design, vol. 6. Coulson and
Mangili PV, Prata DM (2019) Improvement of the butyl acetate process Richardson’s chemical engineering series, vol 6. Elsevier Butter-
through heat integration: a sustainability-based assessment. Chem woth-Heinemann, Oxford
Eng Process 135:93–107 Smith R (2005) Chemical process design and integration. Wiley, West
Mangili PV, Dias RF, Santos LS, Prata DM (2018a) Comparison of the Sessex
­CO2 emissions from two maleic anhydride production processes The Chemical Company (2016) EPCA special edition: critical raw
through computational simulation. Latin Am Appl Res. (accepted) materials and chemical markets. The Chemical Company. https​
Mangili PV, Souza YPDM, Menezes DQF, Santos LS, Prata DM ://thech​emco.com/epca-speci​al-editi​on/. Accessed 08 May 2018
(2018b) Eco-efficiency evaluation of acetone-methanol separa- Towler G, Sinnott R (2008) Chemical engineering design: principles,
tion processes using computational simulation. Chem Eng Process practice and economics of plant and process design. Elsevier
123:100–110 Butterwoth-Heinemann, Oxford
Matar S, Hacth LF (2000) Chemistry of petrochemical processes, 2nd Trinks W, Mawhinney MH, Shannon RA, Reed RJ, Garvey JR (2004)
edn. Gulf Publishing Company, Houston Industrial furnaces, 6th edn. Wiley, Hoboken
Mordor Intelligence (2018) Maleic anhydride market - segmented by Turton R, Bailie RC, Whiting WB, Shaelwitz JA (2010) Analysis, syn-
type, end-user industry, and geography - growth, trends, and fore- thesis and design of chemical processes, 4th edn. Prentice Hall,
cast (2019–2024). Mordor intelligence. https​://www.mordo​rinte​ Upper Saddle River
llige​nce.com/indus​try-repor ​ts/globa​l-malei​c-anhyd​r ide-marke​ Tvedt TJ, Holloway RT (1997) Consensus on operating practices for
t-indus​try. Accessed 11 Mar 2019 the control of feedwater and boilerwater chemistry in modern
Musa OM (2016) Handbook of maleic anhydride based materials: syn- industrial boilers (ASME I00367). NACE International. (dúvida
thesis, properties and applications. Springer, Switzerland se tem cidade)
Pereira CP, Prata DM, Santos LS, Monteiro LPC (2018) Development UNCTAD – United Nations Conference on Trade and Development
of eco-efficiency comparison index through eco-indicators for (2004) A manual for the preparers and users of eco-efficiency
industrial applications. Braz J Chem Eng 35:69–89 indicators. United Nations, New York
Peters MS, Timmerhaus KD, West RE (2003) Plant design and eco- Westerink EJ, Westerterp KR (1988) Safe design of cooled tubular
nomics for chemical engineers, 5th edn. The McGraw-Hill Com- reactors for exothermic multiple reactions: multiple-reaction net-
panies Inc., New York works. Chem Eng Sci 43:1051–1069
PwC – PricewaterhouseCoopers (2018) Brazil: corporate—deductions.
PwC. http://taxsu​mmari​es.pwc.com/ID/Brazi​l-Corpo​rate-Deduc​ Publisher’s Note Springer Nature remains neutral with regard to
tions​. Accessed 08 May 2018 jurisdictional claims in published maps and institutional affiliations.
Rase HF (2010) Handbook of commercial catalysts: heterogeneous
catalysts. CRC Press LLC, Florida
Seider WD, Seader JD, Lewin DR, Widagdo S (2008) Product and
process design principles, 3rd edn. Wiley, New Jersey

Affiliations

Patrick V. Mangili1 · Pedro G. Junqueira1 · Lizandro S. Santos1 · Diego M. Prata1

* Patrick V. Mangili Diego M. Prata


patrickmangili@gmail.com pratadiego@gmail.com
Pedro G. Junqueira 1
Department of Chemical and Petroleum Engineering,
pedrojunqueira@id.uff.br
Universidade Federal Fluminense, Niterói, RJ 24210‑240,
Lizandro S. Santos Brazil
lizandrossantos@gmail.com

13

You might also like