28350: Design Project For The Production of Dimethyl Carbonate From Ethylene Oxide

28350: Design Project for the Production of
Dimethyl Carbonate from Ethylene Oxide

Technical University of Denmark
Dane Hurley (s141089), Spardha Jhamb (s141287)

and Pawel Sobczyk (s143069)
Supervisor:
Amata Anantpinijwatna
June 2, 2015
Abstract
In this design project, the objective was to utilise carbon dioxide emissions from another
process in order to produce a more chemically important product. A process plant was
designed using two step reactions by utilising CO2 to react with ethylene oxide to produce
ethylene carbonate as an intermediate followed by a subsequent reaction with methanol
to produce DMC as well as the by-product ethylene glycol and any unconverted ethylene
carbonate. A hierarchical decomposition approach was used for designing the process,
by decomposing the synthesis into 12 hierarchical tasks. According to the design, the
process will produce on a yearly basis around 15,000 metric tonnes of DMC (10% of the
Chinese market, excluding the largest producer).
The first reaction was performed at a pressure of 103 bar, 448 K with a quaternary
ammonium halide catalyst in order to achieve a conversion of 99%. Additionally the
second reaction was performed at 1 bar, 423 K and using a smectite catalyst containing
magnesium in order to achieve a conversion of 92%.
The separation processes in the system involved two distillation columns and one
membrane separation using a thermodynamic model of Non-Random Two Liquid Model
(NRTL). A complete separation between ethylene glycol and dimethyl carbonate, was
achieved using a distillation column with 13 trays. Further, a pressure dependant
azeotrope between dimethyl carbonate and methanol was separated using a distilla-
tion column at 20 bar (in optimisation this was reduced to 8 bar) with 30 trays. The
dimethyl product was produced at a purity of 99.93% which is suitable for the market.
A membrane separation was also implemented to separate the non-pressure dependant
azeotrope formed between ethylene glycol and ethylene carbonate based on polarity.
Pure ethylene carbonate could be produced for sale and 99% ethylene glycol could also
be sold.
An economic evaluation determined our process to be extremely affordable, profitable
and viable with a payback time of 3.09 years. The process will still be profitable even if
there was an increase of 10% in raw materials, utilities and equipment costs. However
the azeotropic distillation resulted in 50% of the equipment costs and almost 80% of the
utility costs.
Heat integration and process optimisation were then performed in attempt to reduce
energy consumption and unnecessary costing according to the economic evaluation of the
process. In order to meet the design targets these limitations needed to be eliminated.
After performing optimisation on the process, the pay back time was able to be decreased
by 66.47% to 1.04 years. A heat integration was also implemented and found to not be
beneficial as it only decreases manufacturing costs by a maximum of 0.4%.
Finally, a sustainability analysis were performed in order to determine other limita-
tions in the process design. The main limitation is the decision to not recycle carbon
dioxide back to the second reactor as it is received for free from another process. How-
ever in order to significantly reduce our sustainability it would be required to recycle
this carbon dioxide. This is not yet implemented in our model but should be included
in further research.
Contents
1 Introduction 1
2 Task 1: Product Information 2

2.1 Dimethyl Carbonate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2 DMC Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
3 Task 2: Economic Viability 4

3.1 DMC Synthesis Routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.2 Production of Dimethyl Carbonate from Ethylene Oxide . . . . . . . . . . 4
3.3 Chemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.4 Simple Process Flow Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.5 Initial Economic Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4 Task 3: Hierarchical Decomposition 9

4.1 Level 1: Decisions on Batch versus Continuous . . . . . . . . . . . . . . . 9
4.2 Level 2: Decisions on Input-Output Structure . . . . . . . . . . . . . . . . 9
4.3 Level 3: Decisions on Recycle Structure and Reactor . . . . . . . . . . . . 9
4.4 Level 4: Decisions on Separation System . . . . . . . . . . . . . . . . . . . 10
5 Task 4: Mass Balance 12

5.1 Simple Model and Degrees of Freedom Analysis . . . . . . . . . . . . . . . 12
5.2 Flowsheet Decomposition and Calculation Sequence . . . . . . . . . . . . 14
5.3 Mass Balance Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
6 Task 5: Identification of Stream Conditions 18

6.1 Stream Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
6.2 Flash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
7 Task 6: Energy Balance 20

7.1 Simple Model Energy Balance . . . . . . . . . . . . . . . . . . . . . . . . . 20
7.2 Pumps and Compressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
8 Task 7: Rigorous Simulation 24

8.1 Thermodynamic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
8.2 Distillation Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
8.3 Membrane Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
8.4 Rigorous Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
8.4.1 Rigorous Model Mass Balance . . . . . . . . . . . . . . . . . . . . . 32
8.4.2 Rigorous Model Energy Balance . . . . . . . . . . . . . . . . . . . 33
9 Task 8: Equipment Sizing and Costing 34
9.1 Sizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
9.1.1 Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
9.1.2 Heat Transfer Equipment . . . . . . . . . . . . . . . . . . . . . . . 34
9.1.3 Vessels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
9.1.4 Distillation Columns . . . . . . . . . . . . . . . . . . . . . . . . . . 35
9.1.5 Membrane Separation . . . . . . . . . . . . . . . . . . . . . . . . . 36
9.2 Bare Module Costing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
9.3 Utility Costing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
10 Task 9: Economical Analysis 41

10.1 Capital Investment Calculations . . . . . . . . . . . . . . . . . . . . . . . 41
10.2 Net Present Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
10.3 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
10.4 Economic Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
11 Task 10: Heat Integration 46

11.1 Composite Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
11.2 Heat Exchanger Network . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
11.3 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
12 Task 11: Sustainability and Environmental Impact 53
13 Task 12: Optimisation 56

13.1 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
13.2 Base Case Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
13.2.1 Purge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
13.2.2 Recycling Ethylene Carbonate . . . . . . . . . . . . . . . . . . . . 59
13.2.3 Distillation Columns . . . . . . . . . . . . . . . . . . . . . . . . . . 61
13.2.4 Optimising Combination . . . . . . . . . . . . . . . . . . . . . . . . 66
14 Conclusion 68
15 List of Symbols 69
16 References 71
1 Introduction
Carbon dioxide (CO2 ) emissions are a great concern for the environment resulting in
many new technologies and policies in order to mitigate these emissions. This ongoing
issue has gained significant attention in recent years and poses a challenge for the 21st
century. CO2 utilization can be achieved in three ways; conversion into a fuel, a feedstock
for producing important chemical compounds and lastly via non-conversion utilization
such as for enhanced oil recovery and as a solvent.
One particularly important chemical which CO2 can be converted into is dimethyl
carbonate (DMC). DMC is a good replacement for toxic methylating agents and phos-
gene. Additionally, it can also be used as an octane booster in fuel as it significantly
increases the octane number due to the high oxygen content (53.3 wt%) in comparison
to other alkylcarbonates.
The objectives of this process design is to utilise carbon dioxide from another process
in order to produce a more chemically important product. Therefore we will attempt
to create a realistic process design that produces DMC from the raw material of CO2 .
Additionally the conceptional design will be broken down into 12 tasks. The tasks are
ordered according to a specific sequence, which needs to be followed for efficient use of
time and resources.
1
2 Task 1: Product Information
2.1 Dimethyl Carbonate
Dimethyl carbonate (DMC) is considered an environmental benign reagent as it is
biodegradable, non-mutagenic, non-corrosive and non-toxic for human health and the
environment. It has gained significant attention in recent years due to these properties
and its versatility as a chemical solvent and reagent.
In particular, DMC can be utilised in organic synthesises as a non-toxic substitute
for phosgene, which is a poisonous, corrosive carbonylating agent in isocyanate and poly-
carbonate syntheses. Further, it can also be used to replace dimethyl sulfate and methyl
halides which are methylating agents. DMC also has potential use as a solvent such
as a non-aqueous electrolyte component in lithium rechargeable batteries, alternative to
acetate esters and ketones in paints and adhesives or a blowing agent in polyurethane
foam after the CFC ban. DMC also has the ability to be used as an octane booster
in fuel as it significantly increases the octane number due to the high oxygen content
present in DMC (53.3 wt%) compared to other alkylcarbonates.
Therefore DMC is an environmentally benign chemical and has a diverse range of
uses according to its properties such as a methylating agent, carbonylating agent, octane
enhance, organic solvent and a chemical intermediate for many products. A summary
of the physical properties is displayed in Table 1.
Table 1: Physical properties of DMC
Abbreviation DMC
Name Dimethyl Carbonate
Formula OC(OCH3 )2
MW g/mol 90.08
Sp. Gr - 1.07
MP ◦ C 4.00
BP ◦C 90.35
Tc ◦C 283.85
Pc atm 47.37
ρc mol/L 3.80
Flash Point ◦C 18.30
Miscibility - Insoluble in water. Miscible with alcohol, ether.
2.2 DMC Production

In 2003, the world DMC production was around 80 kilotonnes/year. The largest produc-
ers in Europe are EniChem (8 kt/yr) and SNPE (4 kt/yr) as well as UBE in Japan (5
kt/yr) making a total of around 17 kilotonnes/year of DMC for the free market (Aresta,
2003). Additionally, each company uses their own technology according to the alterna-
2
tive synthesis routes in Section 3.1. The majority of the DMC produced is used as an
oxygenated solvent, polycarbonate production and a fuel additive.
In 2009, China had a combined capacity of over 200 kilotonnes/year of DMC from
more than 20 producers. The largest contributor was Shandong Shida Shenghua Chem-
ical Co., Ltd which accounted for around 30% of the total market with a capacity of
around 55 kilotonnes/year (Yuxiao, 2009). Other companies from China include Dongy-
ing Hi-tech Spring Chemical Industry Co., Ltd. and Hebei Chaoyang Chemical Co.,
Ltd.
Therefore according to the Chinese market we will make 15 kilotonnes/year. We
decided on this amount as it is 10% of the remaining market when excluding the largest
contributor. Additionally the DMC should be sold at a purity of greater than 99.7 %.
3
3 Task 2: Economic Viability
3.1 DMC Synthesis Routes
The phosgenation of methanol was the most common method to produce dimethyl car-
bonate until the 1980s. However, phosgene is a toxic reagent and therefore the process
has been abandoned and other synthesis routes have been investigated and are depicted
in Figure 1.
Figure 1: Various routes for production of DMC (Monteiro et al., 2009)
The five alternatives displayed include methanol oxy-carbonylation in liquid or gaseous

phase, carbonylation of methylnitrite, direct synthesis from CO2 and transesterifica-
tion of urea or ethylenecarbonate. Currently the world production is based on the
Enichem one-step liquid phase methanol oxycarbonylation and the UBE two-step gas-
phase methylnitratecarbonylation. However the direct methanol oxycarbonylation in the
gas phase and the alkylenetransesterification process are both becoming more attractive
for future technologies.
Further, the utilisation of carbon dioxide to produce commercially viable products is
becoming more important to industries today. Carbon dioxide is continuously released
into the atmosphere resulting in greenhouse environmental damages due to stratospheric
ozone depletion and could be limited by converting carbon dioxide into environmentally
friendly chemicals. Therefore the production of dimethylcarbonate from raw materials
of ethylene oxide and carbon dioxide is an important process that has the potential to
utilise carbon dioxide emissions whilst producing an environmental benign chemical.
3.2 Production of Dimethyl Carbonate from Ethylene Oxide

The synthesis of DMC from raw materials of ethylene oxide and carbon dioxide is per-
formed in two sequential reactions.
4
Figure 2: First reaction to product ethylene carbonate, R=H (Yang et al., 2010)
Figure 3: Second reaction to produce dimethyl carbonate and ethylene glycol, R=H (Yang et al.,
2010)
According to Peppel, the first reaction can be performed using a quaternary am-
monium halide catalyst in order to achieve a conversion of 98% of ethylene oxide with
excess carbon dioxide. Additionally, Bhanage et al. states that the second reaction can
be performed with a smectite catalyst containing magnesium in order to achieve a con-
version of 92% of ethylene carbonate with excess methanol. Both of these reactions has
a relatively high conversion as well as 100% selectivity towards their desired products.
This is beneficial in order to avoid side reactions that would produce additional contam-
inants in our process. The conditions required for each reaction as well as a comparison
with various other alternatives is displayed in Tables 2 and 3.
3.3 Chemical Properties

The raw materials of carbon dioxide, ethylene oxide and methanol contain impurities
depending on the source of each chemical. The carbon dioxide is assumed to be from a
petrochemical company disposing of the gas and therefore most likely contains methane
and other common gas impurities. For simplicity the carbon dioxide consists of 1%
impurities mainly consisting of methane. Additionally the ethylene oxide would be
obtained from a process starting with ethylene and therefore impurities of ethylene would
be present in the raw material. Therefore, 5% ethylene is considered to be present in
the ethylene oxide raw material. Further, methanol can be obtained in extremely high
purity of around 99.99% so it is assumed the methanol contains no impurities or they
are negligible and do not take part in the reactions.
Therefore the physical properties for the chemicals involved within the reactions as
well as the impurities present in the raw materials are summarised in Table 4. The
current market price for each of the chemical products is also included.
5
Table 2: Conversion Data for Reaction 1
EO EC EC Pressure Temperature Ratio Time By-
Reaction 1 Conversion (%) Selectivity (%) Yield (%) Catalyst (bar) (◦ C) CO2 :EO (hr) product Source
1 98 100 98 Quaternary ammonium halide 103 (1500 psig) 150-175 16:15 52 min - (Peppel, 1958)
2 98.9 98.9 97.8 K2 CO3 -KI/ZnO 165 150 Excess CO2 4 - (Chang et al., 2004)
3 97.2 99.7 96.9 KI/ZnO 165 150 Excess CO2 2 - (Chang et al., 2004)
4 85.6 100 85.6 Immobilised cobalt complex 125 110 2.4:1 4 - (Lu et al., 2004)
5 92.9 91.7 85.1 Amberlite IRA-400 80 110 Excess CO2 15 - (Cao et al., 2006)
6 81.9 64 52.4 MgO 80 150 Excess CO2 15 Cyclic crown ethers (Bhanage et al., 2002)
Table 3: Conversion Data for Reaction 2

EC DMC EG DMC EG Pressure Temperature Ratio Time By-
Reaction 2 Conversion (%) Selectivity (%) Selectivity (%) Yield (%) Yield (%) Catalyst (bar) (◦ C) MeOH:EC (hr) product Source
1 92 100 100 92 92 S-Mg-1 - 150 15:1 4 - (Bhanage et al., 2003)
2 93.4 92 90.3 86 84.3 BMImBr-AS 0.2g 10.7 180 8:1 6 - (Kim et al., 2010)
3 88 97 97 85.4 85.4 Amberlyst A-21 - 120 8:1 14 - (Dhuri and Mahajani, 2006)
4 95 88.4 87.4 84 83 DABCO Mild conditions 70 15:1 6 HEMC 3% (Yang et al., 2010)
5 96 85 -78 82 75 Poly-4-vinyl pyridine - 140 8:1 4 - (Jagtap et al., 2008)
6 - - - 81.2 - None 90 250 8:1 8 - (Feng et al., 2005)
7 90 90 91.1 81 82 [C4 DABCO]OH Mild conditions 70 10:1 4 HEMC 5% (Yang et al., 2010)
6
8 80.4 97.6 99.3 78.5 79.8 EMImCl 20.7 (300 psig) 160 8:1 6 - (Ju et al., 2007)
9 82 92.7 92.7 76 76 BMImBr-AS 10.7 160 8:1 6 - (Kim et al., 2010)
10 76.3 99.3 99.9 75.8 76.2 EMImCl 14 140 8:1 6 - (Ju et al., 2007)
11 85 88.1 90.8 74.9 77.2 201x7 Resin - 90 8:1 6 C4H8O4 (Cao et al., 2006)
12 68.9 97.7 100 67.3 69.1 K2 CO3 -KI/ZnO 165 150 4:1 2 - (Chang et al., 2004)
13 66.1 100 102 66.1 67.6 MgO 0.5 g 80 150 8:1 4 - (Bhanage et al., 2002)
14 75.2 87.8 89.9 66 67.6 Smectite-Mg-Ni-Na-K - 150 8:1 4 - (Bhanage et al., 2002)
15 82.3 73 77.2 60.1 63.5 MgO - 150 8:1 4 - (Bhanage et al., 2002)
Table 4: Properties for all chemicals involved in the process
Symbol M E CO2 EO MeOH DMC EG EC

Carbon Ethylene Dimethyl Ethylene Ethylene
Species Methane Ethylene Methanol
Dioxide Oxide Carbonate Glycol Carbonate
Formula CH4 C2 H4 CO2 C 2 H4 O CH3 OH OC(OCH3 )2 C2 H6 O2 (CH2 O)2 CO
MW g/mol 16.04 28.05 44.01 44.05 32.04 90.08 62.07 88.06
Sp. Gr. - - - - 0.82 0.79 1.07 1.11 1.32
BP ◦C -161.49 -103.74 -78.48 10.45 64.7 90.25 197.3 238
MP ◦C -182 -169 -78 -111 -98 4 -13 38
Liquid kJ/mol - - - -95.7 -239 - -460 -682.8
Hf ◦
Vapour kJ/mol -74.87 52.4 -393.52 -52.64 -205 - -388 -503
Tc ◦C -82.55 9.35 31.05 195.85 239.85 283.85 446.85 533
Pc atm 45.5 49.94 72.83 71.06 79.94 47.37 84.88 66.81
Price USD/tonne - - 0 1617.6 502.85 2000 1200 1500
7
3.4 Simple Process Flow Sheet
A process design for the synthesis of DMC starting from EO and CO2 is described in
Kongpanna et al. (2014) and is displayed in Figure 4.
Figure 4: Process flow diagram for the production of dimethyl carbonate from ethylene oxide
(Kongpanna et al., 2014)
We will use this flowsheet as a basis and verify using a Hierarchical Decomposition
(Section 4) to form our own simple flowsheet.
3.5 Initial Economic Potential

The initial economic potential is calculated according to the production amount of 15
kilotonnes/year of DMC, which was decided in section 2.2. Assuming 100% conversion
for both reactions and the market price for all chemicals sold, the initial economical
potential can be calculated (Table 5). The carbon dioxide was assumed to be free whilst
the ethylene oxide would be able to be obtained as a byproduct from another process and
be a cheaper price than on the market. Therefore, the market price for ethylene oxide
was 1600 USD/tonne so a value of 1000 USD/tonne was assumed in the calculations.
The initial economic potential equation can be calculated as follows:
EP0 = Cost of products sold (DMC, EG) − cost of raw materials (CO2 , EO, MeOH)
EP0 = $42, 402, 824.67 − $12, 702, 132.15 = $29, 700, 692.52
Table 5: Initial Economic Potential Flowrates and Price Summary

Impurities Reactants Products
Carbon Ethylene Dimethyl Ethylene Ethylene
Methane Ethylene Dioxide Oxide Methanol Carbonate Glycol Carbonate
MW (g/mol) 16.04 28.05 44.01 44.05 32.04 90.08 62.07 88.06
Mass (tonnes/year) 0.27 245.87 7328.57 7335.75 10671.4 15000 10335.69 14664.32
Moles (kmol/year) 16.82 8764.34 166522.53 166522.53 333045.06 166522.53 166522.53 166522.53
Market Price (USD/tonne) - - $0 $1,000 $502.85 $2,000 $1,200 $1,500
Cost (USD/year) - - $0 $7,336,020.34 $5,366,111.80 $30,000,000 $12,402,824.67 -
8
4 Task 3: Hierarchical Decomposition
4.1 Level 1: Decisions on Batch versus Continuous
The production of dimethyl carbonate can either be a batch or continuous process. The
process will have a high production rate with a continuous flowing supply of carbon
dioxide throughout the year meaning the whole years supply will not be produced in a
few days. Additionally the product lifetime for DMC will be long, as the product will
remain a chemical commodity in the market for many years due to its environmentally
friendly properties. The cost of the manufacturing process will also determine the value
of the product so it is required that as a producer a cost-effective process is designed.
The process can also be considered to not be sensitive to upsets and variations, as the
conversion for each reaction will not be significantly affected with slight variations of
temperature and pressure. Therefore, the manufacturing of dimethyl carbonate starting
from carbon dioxide and ethylene oxide is considered as a continuous process.
4.2 Level 2: Decisions on Input-Output Structure

The raw materials contain slight amounts of impurities with ethylene oxide containing
5% ethylene whilst the carbon dioxide contains 1% methane. The amount of impurities
are relatively small and do not take part in either of the reactions so it is not required
to remove them before the reactors. Both impurities have boiling points lower than the
reactants and products (Table 4) and are inert so can be recycled with a purge unit in
the gas phase to ensure there is no build up. Additionally neither of the reactions that
take place has 100% conversion of the reactants so the remaining unreacted reactants
will also be recycled in the gas phase. The intermediate reactant of ethylene carbonate is
required in the second reaction and needs to be separated from the unreacted reactants
using a flash before the second reactor. Ethylene glycol is also produced as a valuable
by product, which needs to be separated using a distillation column before being sold.
Further, the DMC forms an azeotrope with methanol so a pressurised distillation is
required to separate the two products.
Figure 5: Reactor 1 Figure 6: Reactor 2
4.3 Level 3: Decisions on Recycle Structure and Reactor

The process involves two reactors, one for each reaction, using different pressures, tem-
peratures and catalysts. Neither of the reactions have 100% conversion so there will be
9
a total of three recycle streams for the unreacted reactants. Ethylene oxide and carbon
dioxide will be recycled in the gaseous phase and will require a compressor to return the
pressure to the same as the reactor. The recycle stream of methanol returns at a higher
pressure than the reactor, due to subsequent pressurised distillation (Section 4.4), so
a valve will be required rather than a compressor. Additionally the recycle stream of
ethylene carbonate remains at the same pressure as the reactor so a pump is not needed.
However, Souza et al. (2013) mentions that ethylene carbonate is a reasonably valuable
product and could actually be sold rather than recycled. We therefore have chosen not to
recycle ethylene carbonate but will reevaluate this decision when optimising the process.
Figure 7: Recycle Stream 1 Figure 8: Recycle Stream 2
Both reactors will be isothermal as the heat of reaction in both cases is not negli-
gible so they cannot be adiabatic. The first reaction is kinetically controlled whilst the
second reaction is in equilibrium and can be shifted towards the products by increasing
the amount of methanol present by using the recycled stream in order to increase the
conversion.
4.4 Level 4: Decisions on Separation System

The reactor stream from the first reactor exists in two phases and can be separated using
a flash, with the unreacted reactants in the gaseous phase and the ethylene carbonate
product in the liquid phase due to the large difference between the boiling points of more
than 200◦ C (Table 2). The gaseous phase is then purged and compressed before being
returned to the reactor whilst the liquid phase continues to a pump before entering the
second reactor. The reactor stream from the second reactor exits as liquid which can
then separate all the products using a series of distillation columns. The ethylene glycol
and unreacted ethylene carbonate are first separated from the methanol and DMC. The
ethylene glycol and ethylene carbonate can then also easily be separated using another
distillation column. However the methanol and DMC is an azeotrope that requires a
pressurised distillation column of 20 bar (Kongpanna et al., 2014) in order to successfully
separate the two components. The methanol is then purged to avoid the build up of
impurities before being returned to the second reactor.
10
Figure 9: Separation 1 Figure 10: Separation 2
The final flowsheet formed after performing the hierarchical decomposition is dis-
played in Figure 11.
Figure 11: Flowsheet Formed from Hierarchal Decomposition
11
5 Task 4: Mass Balance
5.1 Simple Model and Degrees of Freedom Analysis
The flowsheet for the process was re-drawn using only mixers, reactors, dividers (split-
ters) and splitters (stream calculators) in order to solve the simple mass balance. The
simplified flow sheet was then analysed in order to located specified variables and is
displayed in Figure 12. If all the variables marked in blue are known then all the other
variables in the flowsheet can be calculated. Therefore the feed stream compositions,
the conversion and stoichiometry for the two reactors, the compound recoveries for the
four splitters and the purge factor for the two dividers need to be specified.
Figure 12: Simple Flowsheet with Specifications
Units such as pumps and compressors have not been considered as they do not affect
the mass balance. Additionally, the temperatures and pressures for each stream are fixed
as they are not required. It is also assumed that all vapour and liquid streams have ideal
equilibrium relationships and are at saturated conditions unless stated otherwise.
The equations governed for each module of the simple model can be calculated ac-
cording to Equations 1, 2, 3 and 4. However, the process simulator PROII will utilise
these equations in order to solve the mass balance in a more rigorous way.
Mixer
N
X M
fi,N M +1 = fi,j where NM = Number of Mixed streams (1)
j=1
12
Reactor
X
fi,j = fi,j + γr,i ηr,k fk,j (2)
r
Divider and Splitter

fi,j+1 = ξi,U fi,j where j = 1, N S − 1 (3)
X
fi,N S = (1 − ξi,U )fi,j where NS = Number of Streams (4)
j
where i is the component (i.e. EG, E, EO, M, MeOH, CO2 , DMC or EC), j is the stream
number, r is the reaction and k is the limiting reagent.
The degrees of freedom for the system is calculated in Table 6 as 90. Therefore,
90 variables need to be specified in order to solve the system. The process variables
specified are summarised in Table 7 and includes a total of 66 equipment parameters
and 3 stream component flowrates (i.e. 24 process variables). Since the degrees of
freedom are satisfied by the same number of specified variables, the remaining unknown
process variables can be calculated.
Table 6: Degree of Freedom Analysis
Component Equation Number Value

1 Mixer NC 8
2 Reactor NC 8
3 Splitter 2*NC 16
4 Divider 2*NC 16
5 Mixer NC 8
6 Reactor NC 8
7 Splitter 2*NC 16
8 Splitter 2*NC 16
9 Splitter 2*NC 16
10 Divider 2*NC 16
Total: NE 16*NC 128
Component Variables Number Value

Flowrates f1 -f19 19*NC 152
Reactor parameters γr1 , ηr1,k , γr2 , ηr2,k 2*(NC+1) 18
Stream parameters ξS NC 32
Divider parameters ξD NC 16
Total: NV 23*NC+2 218
Degrees of Freedom: NV-NE 13*NC+2 90
Specified Variables f1 , f2 , f9 and γr , ηk , ξS , ξD 7*NC +2 90

Unknowns f3 -f8 and f10 -f19 16*NC 128
13
Table 7: Process Specifications
Specified Value
Component Type Specifications
(EG, E, EO, M, MeOH, CO2 , DMC, EC)
Reaction stoichiometric EO + CO2 −−→ EC
2 Reactor
coefficients (γr ), γr1 =[0, 0, -1, 0, 0, -1, 0, 1]; ηr1,3 = 0.98; k=3
conversion (ηr,k ), EC + 2 MeOH −−→ DMC + EG
6 Reactor
key component(k) γr2 =[1, 0, 0, 0, -2, 0, 1, 1]; ηr2,8 = 0.92; k=8
4 Divider Stream - divider ξD4 = 0.5
10 Divider split fraction (ξD ) ξD10 = 0.5
3 Splitter ξS3 (0.00, 1.00, 1.00, 1.00, 0.00, 1.00, 0.00, 0.00)
7 Splitter Component split fractions in ξS7 (0.00, 1.00, 1.00, 1.00, 1.00, 1.00, 1.00, 0.00)
8 Splitter overhead product (ξS ) ξS8 (1.00, 1.00, 1.00, 1.00, 0.99, 1.00, 0.01, 1.00)
9 Splitter ξS9 (0.995, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00, 0.00)
Feed Stream 1 Component flow rates (f1 kmol/hr) f1 (0.00, 1.22, 23.09, 0.00, 0.00, 0.00, 0.00, 0.00)
5.2 Flowsheet Decomposition and Calculation Sequence

The simple process contains 2 partitions and 2 recycle loops. Therefore 2 tear streams
are needed and can be identified by using a signal flow graph (Figure 13) with a table
of nodes and precursors (Table 8).
Figure 13: Flowsheet Decomposition Stream Labelling
14
Table 8: Nodes and Precursers for Flowsheet Decomposition
Nodes Precursors (1) Prescursers (2)

3 7 Cut
4 3
5 4
7 5
8 4
10 8, 18 3, 10 Cut
11 10
12 11
15 11
17 15
18 17
It is evident that streams 3 and 10 need to be tear streams as they are both self-loops.
Therefore, one possible calculation sequence could be 2.Reactor, 3.Splitter, 4.Divider,
1.Mixer, 6.Reaction, 7.Splitter, 8.Splitter, 10.Divider, 5.Mixer, 9.Splitter. This can be
clarified by forming an incidence matrix consisting of of both partitions as displayed
in Table 9. An upper diagonal triangle is formed with the unknowns from the two
tear streams whilst everything else in the lower diagonal triangle is known or can be
calculated. Furthermore, the order of the modules matches the proposed calculation
sequence as desired.
Table 9: Incidence Matrix
Component Type f1 f2 f9 γr ηr,k ξS ξD f4 f5 f8 f6 f7 f3 f11 f15 f12 f17 f16 f18 f19 f10 f13 f14
2 Reactor x x (x) *
x * (x)
3 Splitter x * (x)
x * (x)
4 Divider x * (x)
1 Mixer x x * (x)
6 Reactor x x (x) *
x * (x)
7 Splitter x * (x)
x * (x)
8 Splitter x * (x)
x * (x)
10 Divider x * (x)
5 Mixer x * (x)
x * (x)
9 Splitter x * (x)
15
5.3 Mass Balance Summary
The simple process flowsheet was implemented in PROII (Figure 14) with the desired
calculation sequence. The process simulator was then able to perform a mass balance
with the results obtained in Table 10. On checking the conservation of mass the following
results were obtained:
Mass Balance = Mass Out (S6, S13, S14, S16, S19) − Mass In (S1, S2, S9)
= (771.61 + 1285.51 + 165.86 + 1961.00 + 4731.97)
− (1040.63 + 1723.45 + 6151.88)
= 8915.97061 − 8915.96999
= −0.00061 kg/hr
Therefore there is a loss of 0.00061 kg/hr which has a relative error of 6.89 × 10−6 %.
This is extremely small so the mass balance is considered satisfied. Additionally we get
the desired amount of DMC product as well as the desired purity for each sale product.
16
17
Figure 14: Mass Balance Flowsheet from PROII
Table 10: Mass Balance Values from PROII
Stream Name S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19
Molar Rate kmol/hr 24.06 39.41 81.69 59.06 36.44 18.22 18.22 22.63 191.99 362.11 341.29 22.63 20.71 1.91 318.67 23.69 294.98 147.49 147.49
Mass Rate kg/hr 1040.63 1723.45 3535.73 3535.72 1543.28 771.64 771.64 1992.44 6151.88 12876.3 12876.3 1451.36 1285.51 165.86 11424.94 1961.00 9463.94 4731.97 4731.97
Total Weight Comp. Rates kg/hr
Ethylene Glycol 0 0 0 0 0 0 0 0 0 0 1291.97 1291.97 1285.51 6.46 0 0 0 0 0
Ethylene 33.75 0 67.49 67.49 67.49 33.75 33.75 0 0 0 0 0 0 0 0 0 0 0 0
Ethylene Oxide 1006.88 0 1017.06 20.34 20.34 10.17 10.17 0 0 0 0 0 0 0 0 0 0 0 0
Methane 0 6.32 12.65 12.65 12.65 6.32 6.32 0 0 0 0 0 0 0 0 0 0 0 0
Methanol 0 0 0 0 0 0 0 0 6151.88 10874.43 9540.5 0 0 0 9540.5 95.41 9445.1 4722.55 4722.55
Carbon Dioxide 0 1717.13 2438.53 1442.8 1442.8 721.4 721.4 0 0 0 0 0 0 0 0 0 0 0 0
Dimethyl Carbonate 0 0 0 0 0 0 0 0 0 9.42 1884.43 0 0 0 1884.43 1865.59 18.84 9.42 9.42
Ethylene Carbonate 0 0 0 1992.44 0 0 0 1992.44 0 1992.44 159.4 159.4 0 159.4 0 0 0 0 0
6 Task 5: Identification of Stream Conditions
6.1 Stream Conditions
The mass balance flowsheet (Figure 14) can now be analysed and temperature and
pressure can be included as specified variables. The optimum temperatures and pressures
for each reactor were outlined in Section 3.2. However the temperatures and pressures
for the remaining streams and utilities are yet to be specified.
For the flash splitter, the temperature and pressure is chosen according to the desired
recoveries in the top and bottom products. In comparison the temperature and pressure
for each of the distillation column splitters can be calculated using PROII according to
either the bubble point or dew point of the top and bottom streams. If the stream was
desired to be liquid then the temperature and pressure would be for the bubble point
whilst if the stream was desired to be vapour it would be at the dew point. However
since the products from splitters 8 (SC3) and 9 (SC4) are being sold or being recycled the
dew point was desired for all streams as it is easier to store and transport the products
as a liquid.
A summary of all temperatures and pressures for each stream are indicated in Ta-
ble 11.
Table 11: Temperatures and pressures for all streams
Temperature Pressure
Stream (K) (bar) Phase Comments
1 298.15 23.10 Liquid At room temperature and bubble pressure of EO
2 298.15 68.02 Liquid At room temperature and bubble pressure of CO2
3 318.01 23.10 Mixed At Reactor Temperature and pressure
4 448.00 103.42 Vapor At Reactor Temperature and pressure
5 533.00 103.42 Vapor At flash drum temperature and pressure
8 533.00 103.42 Liquid At flash drum temperature and pressure
9 298.15 1.01 Liquid CH3 OH at room temperature and atm pressure
10 328.98 1.01 Liquid At Reactor Temperature and pressure
11 423.00 1.01 Vapor At Reactor Temperature and pressure
12 472.41 1.01 Liquid At bubble point temperature of EG and atm pressure
13 331.59 1.01 Liquid At bubble point temperature of DMC and atm pressure
14 470.13 1.01 Liquid At room temperature and atm pressure
16 315.13 1.01 Liquid At bubble point temperature of MeOH and atm pressure
18
6.2 Flash
The temperature for the flash was determined at the same pressure as reactor 1 of 10342
kPa. Various temperatures were tested using a reference flash as in Figure 15 in order
to determine the composition in the bottom stream of ethylene carbonate leading to the
next reactor.
Figure 15: Flash
Ideally only EC is required, however this is not possible using a flash, so a temperature
needs to be chosen in order to balance the cost of using a high temperature and the purity
of ethylene carbonate in the bottom product stream. A plot of the temperature of the
flash in comparison to the purity of ethylene carbonate is displayed in Figure 16.
Figure 16: Flash Temperature against Purity of Ethylene Carbonate in Bottom Product
A temperature of 578 K was chosen as increasing the temperature much higher

doesn’t result in any significant increase in purity.
19
7 Task 6: Energy Balance
7.1 Simple Model Energy Balance
The equations governed for the energy balance for each module of the simple model
can be calculated according to Equations 5, 6, and 7. The energy for each stream and
the heat duty for each unit can then additionally be calculated. However, the process
simulator PROII will utilise these equations in order to solve the energy balance in a
more rigorous way.
Mixer
N
X M
FN M +1 ∆hN M +1 + Qm = Fj ∆hj where NM = Number of Mixed streams (5)
j=1
Reactor
Fj+1 ∆hj+1 + Qr + Fj ∆Hr = Fj ∆hj (6)
Divider and Splitter

Fj+1 ∆hj+1 + Qs + Fj+2 ∆Hj+2 = Fj ∆hj (7)
A summary of the heat duties for each utility according to the simple flowsheet used
for the energy balance in Figure 17 is summarised in Table 12. Additionally a summary
of the enthalpy, temperature and pressure calculated using PROII for each stream is
indicated in Table 13.
Table 12: Heat Duty for Each Utility
Utility Name Duty (M*kJ/hr)

1. Mixer M1 0
2. Reactor R1 -1.447
3. Splitter F1 0
4. Divider SP1 0
5. Mixer M2 0
6. Rector R2 13.302
7. Splitter SC1 -12.926
8. Splitter SC2 0.003
9. Splitter SC3 -0.328
10. Splitter SP2 0
Total Duty -1.396
20
21
Figure 17: Simple Energy Balance Flowsheet on PROII
Table 13: Energy Balance Values using PROII
Stream Name S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19
Phase Liquid Liquid Mixed Mixed Liquid Vapor Vapor Vapor Liquid Mixed Vapor Liquid Liquid Liquid Liquid Liquid Liquid Liquid Liquid
Temperature K 298.15 298.15 318.01 448. 533. 533. 533. 533. 298.15 328.96 423. 472.41 331.58 470.12 510.86 315.13 362.96 315.13 315.13
Pressure kPa 2310. 6802.15 2310. 10342.14 10342.14 10342.14 10342.14 10342.14 101.33 101.33 101.33 101.33 101.33 101.33 101.33 101.33 101.33 101.33 101.33
Mole Rate kmol/hr 24.09 39.33 81.82 59.2 22.39 36.81 18.41 18.41 188.33 359.59 339.05 22.33 316.72 20.44 1.89 297.73 18.99 148.86 148.86
Mass Rate kg/hr 1042.02 1719.81 3554.81 3554.8 1968.84 1585.96 792.98 792.98 6034.65 12863.84 12863.84 1432.18 11431.66 1268.52 163.66 9720.7 1710.96 4860.35 4860.35
Enthalpy M*kJ/hr 0.05 0.14 0.62 1.11 0.85 0.85 0.43 0.43 0.37 1.73 15.32 0.75 1.64 0.69 0.07 1.02 0.29 0.51 0.51
An energy balance can now be performed according to the energy of the streams
entering and exiting the system as well as the the heat duty of the utilities.
Energy Balance = Energy Out (S7, S14, S15, S17, S18) − Energy In (S1, S2, S9) + Total Duty
= (0.426 + 0.691 + 0.065 + 0.291 + 0.509) − (0.052 + 0.142 + 0.426) + −1.396
= 1.98199 − 0.56529 − 1.39615
= 0.02055 M*kJ/hr
This value is positive indicating that energy needs to be put into the system. Additionally
this is the minimum amount of energy that the process will require.
7.2 Pumps and Compressors

According to the desired temperatures and pressures of each stream; heat exchangers,
pumps and compressors can now be implemented into the model.
Five heat exchangers are required in order to bring the feed streams (S1, S2, S9),
the bottom product from the flash (S5) and the recycled methanol (S19) to the reaction
temperatures for reactor 1 (160 ◦ C) and reactor 2 (150 ◦ C). Additionally 2 pumps were
installed to raise the pressure of the feed streams to the reactor pressure for reactor 1
(103.42 bar) whilst another pump was needed to increase the pressure of the stream
entering the pressurised azeotropic distillation (S13) to 20 bar. Two valves were also
required to reduce the pressure of the ethylene carbonate and recycled methanol to the
pressure of the second reactor (1 bar).
The product streams of DMC, EC and EG (S20, S21, S26) each require a heat
exchanger as well as one valve for DMC to ensure each stream exits the process close to
room temperature and pressure in order to be sold and stored.
In order to recycle the top product from the flash for the first reaction, a compressor
would be required for the CO2 vapour to ensure it was at supercritical conditions. This
would be highly uneconomical as it would require a significant amount of energy and is
unnecessary as we are receiving CO2 as a free raw material. However it is still justified
that CO2 is being utilised and is being converted to a useful product despite not recycling
any unused raw material. Additionally there is only 1.02% of EO in the recycle stream
which is not considered vital to recycle as we are receiving it at a cheaper price as a raw
material.
A total of 8 heat exchangers, 3 pumps and 3 valves are required for this process. An
updated flowsheet is displayed in Figure 18.
The inlet and outlet temperatures and duty for the heat exchangers is summarised
in Table 14 as well as the inlet and outlet pressures and duty for the pumps and valves
is summarised in Table 15.
22
Figure 18: Pumps and Compressors Flowsheet
Table 14: Heat Exchangers Temperatures and Duties
Heat Inlet Outlet Duty

Exchangers Temperature (K) Temperature (K) (kW)
E1 206.01 448 139.92
E2 201.89 448 342.19
E3 510.44 423 112.08
E4 298.15 423 555.5
E5 470.57 303.15 165.58
E6 494.25 303.15 13.89
E7 363.37 303.15 159.67
E8 333.04 423 3.56
Table 15: Pumps and Valves Pressures and Duties
Pumps and Inlet Outlet Duty

Valves Pressure (bar) Pressure (bar) (kW)
P1 2 125 7.39
P2 2 125 13.45
P3 1.01 20 2.39
V1 103.42 1.01 0
V2 20 1.01 0
V3 20 1.01 0
23
8 Task 7: Rigorous Simulation
8.1 Thermodynamic Model
Before implementing the rigorous model into our flowsheet, the thermodynamic model
for the system needs to be chosen. A flowsheet is displayed in Figure 19 that can be
followed in order to determine the thermodynamic model.
Figure 19: Choice of Thermodynamic Model
Since our system is non-ideal, the properties and components are known, and the
system is at a pressure above 12 bar due to the pressurised distillation column we need
to use the Non-Random Two Liquid model (NRTL) for our system.
8.2 Distillation Columns

A shortcut model was used to determine the number of trays required at infinite reflux
to accomplish a desired separation using the Fenske method. This method assumes
24
that the relative volatilities for the components are constant throughout the column
or a column section. This is a reasonable assumption for hydrocarbon systems which
are modelled by NRTL thermodynamics. Thus, for such systems useful information
about the component splits may be obtained. For two product columns, the shortcut
method can compute minimum reflux at infinite trays using the Underwood method.
The Fenske and Underwood results are interpolated with the Gilliland relationship to
produce a table of equilibrium trays versus reflux ratio for a practical operating range.
This feature helps determine the number of theoretical trays for the desired separation
and estimates of the product molar flow-rates for simulation of a rigorous distillation
column model.
The number of trays and feed tray from the shortcut model could then be imple-
mented in a reference distillation column as in Figure 20.
Figure 20: Short Cut Distillation
The first distillation column is separating methanol and DMC from EC and EG.
Therefore the heavy key is EC and the light key is DMC according to the boiling points.
A Txy diagram of the relationship between these two components is displayed in Fig-
ure 21 at 1 atm. It is evident that there is a clear distinction between the vapour and
liquid phases so we can achieve complete separation of the two components.
25
Figure 21: VLE Diagram for EG/DMC at 1 atm
In comparison, a separation is additionally required for the DMC and MeOH mixture.
A Txy diagram for these two components at 1 bar is displayed in Figure 22. It is evident
that an azeotrope is formed at 85 mol% MeOH and therefore this mixture can not
be separated to a higher purity at these conditions. However a Txy diagram of the
two components at 20 bar is displayed in Figure 23 and reveals that the azeotrope is
broken at this pressure indicating the azeotrope is pressure dependant. Therefore a
pressurised distillation column at 20 bar will be implemented in order to separate the
two components with one exit stream pure DMC and the other having a high purity of
MeOH.
26
Figure 22: VLE Diagram for DMC/MeOH Azeotrope at 1 atm Figure 23: VLE Diagram for DMC/MeOH Azeotrope at 20 atm
27
Table 16: Shortcut Distillation Column Specifications and Results
Input Output
Shortcut Light Heavy Specifications Flow Rate of Number of Position of Reflux Condensor Reboiler Total
Distillation Key Key ξ(LK, HK) Feed (kmol/hr) Trays Feed Tray Ratio Duty (M*kJ/hr) Duty (M*kJ/hr) Duty (M*kJ/hr)
1 DMC EG ξT 1 (1, 0) 52.8559 13 4 0.652 -1.8202 0.423 -1.3972
2 MeOH DMC ξT 2 (1, 0.05) 30.8276 30 3 1.971 -17.3542 17.9869 0.6327
Table 17: Rigorous Distillation Column Specifications and Results
Input Ouput
Rigorous Number of Position of Specifications Reflux Condensor Reboiler Total
Distillation Trays Feed Tray (LK, HK) Ratio Duty (M*kJ/hr) Duty (M*kJ/hr) Duty (M*kJ/hr)
1 13 4 (1, 0) 0.759 -1.9503 0.524 -1.4263
2 30 3 (1, 0.05) 20.073 -17.9765 18.6098 0.6333
8.3 Membrane Separation

The mixture of EG and EC also needs to be separated in order for each component
to be sold. A VLE diagram of the mixture at 1 atm is displayed in Figure 24 and
indicates an azeotrope forms between these two components. However the azeotrope is
not pressure sensitive and hence some other method other than pressure distillation is
needed to separate the two components at high purity.
Figure 24: VLE Diagram for EG/EC at 1 atm
There are several possibilities for separating this mixture which could be either ex-
tractive distillation, reactive distillation or membrane separation. Extractive distillation
is an environmentally unfriendly process due to the use of toxic solvents as well as the
separation would require two or more columns. Use of a hydrolysis reactor could be a
possibility but the equilibrium reaction favours the formation of EC more than EG and
also generates CO2 . Hence, it is affective and environmentally friendly to make use of a
membrane. Additionally, there is a small volumetric flow rate of 0.101 m3 /hr meaning
it will not be that expensive as the area of the membrane will not be too large.
The US Patent 6,620,958 B2 has registered a method of separating mixtures of or-
ganic liquids specifically based on their polarity by making use of a low polarity or
non-polar polymeric membrane. It has been discovered that a liquid having reduced
polarity relative to a mixture of liquids having varying polarities can be selectively sep-
28
arated from the mixture using such a membrane. The patent says that the process is
suitable for removal of dialkyl carbonates from alcohol/water mixtures or diol/water
mixtures, the removal of alkyl pyridines from pyridine and the removal of esters from
reaction mixtures containing acids and alcohols as starting materials. The material used
for the polymeric membrane can be synthetic, preferably a polyisoprene membrane or
of naturally occurring latex which comprises of cis-1,4-polyisoprene and makes up 94%
of its weight.
Experimental results for the case of separation of DMC from MeOH showed that
DMC diffused faster than MeOH and after a period of 20 hours the MeOH weight
percentage on the same side of the membrane was 93.7%.
As the patent claims that the same procedure can be employed to separate an alkylene
glycol (ethylene glycol) from an organic oxygenate, example ester (ethylene carbonate), it
can be assumed that similar results would be obtained although currently, experimental
data is not available.
8.4 Rigorous Model

After implementing the flash, distillation columns and membrane separation as well as
pumps and compressors, the stream summary is displayed in Tables 18 & 19. The
rigorous model is displayed in Figure 25. Additionally the heat duties for each rigorous
component are displayed in Table 20.
29
Table 18: Stream Summary S1-S16
Stream Name S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16
Phase Liquid Liquid Liquid Liquid Mixed Vapor Mixed Mixed Vapor Liquid Mixed Mixed Liquid Vapor Mixed Mixed
Temperature K 192.47 190.8 206. 201.89 448. 448. 411.48 448. 533. 533. 510.44 423. 298.15 423. 411.64 423.
Pressure kPa 200. 200. 12500. 12500. 12500. 12500. 12500. 12500. 10342.14 10342.14 101.33 101.33 101.33 101.33 101.33 101.33
Enthalpy M*kJ/hr -0.16 -0.39 -0.14 -0.35 0.37 0.89 1.25 1.08 0.83 0.84 0.84 0.44 0.09 1.9 2.56 2.45
Molar Rate kmol/hr 24.3 55.97 24.3 55.97 24.3 55.97 80.27 57.65 35.53 22.11 22.11 22.11 45.64 45.64 73.04 52.75
Mass Rate kg/hr 1051.15 2447.52 1051.15 2447.52 1051.15 2447.52 3498.67 3498.66 1554.22 1944.44 1944.44 1944.44 1462.42 1462.42 3585.85 3585.85
Methane 0 8.9791 0 8.9791 0 8.9791 8.9791 8.9791 8.9775 0.0016 0.0016 0.0016 0 0 0.0033 0.0033
Ethylene 34.0881 0 34.0881 0 34.0881 0 34.0881 34.0881 34.0877 0.0004 0.0004 0.0004 0 0 0.0008 0.0008
Carbon Dioxide 0 2438.539 0 2438.539 0 2438.539 2438.539 1442.8015 1442.7699 0.0316 0.0316 0.0316 0 0 0.0632 0.0632
Ethylene Oxide 1017.0597 0 1017.0597 0 1017.0597 0 1017.0597 20.3412 17.5839 2.7573 2.7573 2.7573 0 0 5.5147 5.5147
Methanol 0 0 0 0 0 0 0 0 0 0 0 0 1462.4173 1462.4173 1624.9081 324.9816
Dimethyl Carbonate 0 0 0 0 0 0 0 0 0 0 0 0 0 0 13.7041 1840.9194
Ethylene Glycol 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0. 1259.0351
Ethylene Carbonate 0 0 0 0 0 0 0 1992.4523 50.7992 1941.6532 1941.6532 1941.6532 0 0 1941.6532 155.3323
Table 19: Stream Summary S17-S31

Stream Name S17 S18 S19 S20 S21 S22 S23 S24 S25 S26 S27 S28 S29 S30 S31
Phase Liquid Liquid Liquid Liquid Liquid Liquid Liquid Liquid Mixed Liquid Liquid Liquid Liquid Mixed Vapor
30
Temperature K 470.72 511.8 470.57 303.15 303.15 340.76 342.71 495.46 363.4 303.15 438.35 438.35 438.35 335.41 423.
Pressure kPa 101.33 101.33 101.33 101.33 101.33 101.33 2000. 2000. 101.33 101.33 2000. 2000. 2000. 101.33 101.33
Enthalpy M*kJ/hr 0.74 0.05 0.69 0.01 0.09 0.29 0.3 0.76 0.76 0.1 0.17 0.09 0.09 0.09 0.22
Molar Rate kmol/hr 22.03 1.56 20.47 1.56 20.47 30.73 30.73 20.15 20.15 20.15 10.57 5.29 5.29 5.29 5.29
Mass Rate kg/hr 1413.11 137.29 1275.82 137.29 1275.82 2172.74 2172.74 1814.77 1814.77 1814.77 357.97 178.99 178.99 178.99 178.99
Methane 0 0 0 0 0 0.0033 0.0033 0 0 0 0.0033 0.0016 0.0016 0.0016 0.0016
Ethylene 0. 0. 0. 0. 0. 0.0008 0.0008 0 0 0 0.0008 0.0004 0.0004 0.0004 0.0004
Carbon Dioxide 0. 0. 0. 0. 0. 0.0632 0.0632 0 0 0 0.0632 0.0316 0.0316 0.0316 0.0316
Ethylene Oxide 0. 0. 0. 0. 0. 5.5147 5.5147 0. 0. 0. 5.5147 2.7573 2.7573 2.7573 2.7573
Methanol 0. 0. 0. 0. 0. 324.9816 324.9816 0. 0. 0. 324.9816 162.4908 162.4908 162.4908 162.4908
Dimethyl Carbonate 0. 0. 0. 0. 0. 1840.9194 1840.9194 1813.5109 1813.5109 1813.5109 27.4085 13.7043 13.7043 13.7043 13.7043
Ethylene Glycol 1257.776 0 1257.776 0 1257.776 1.259 1.259 1.259 1.259 1.259 0. 0. 0. 0. 0.
Ethylene Carbonate 155.3318 137.2888 18.0429 137.2888 18.0429 0.0005 0.0005 0.0005 0.0005 0.0005 0. 0. 0. 0. 0.
31
Figure 25: Rigorous Model Flowsheet

Table 20: Heat duties for rigorous components
Unit Duty (M*kJ/hr)

P1 0.027
P2 0.048
P3 0.009
E1 0.504
E2 1.232
E3 0.403
E4 1.814
E5 0.046
E6 0.603
E7 0.661
E8 0.136
M1 0
M2 0
F1 0.585
R1 -2.1033
R2 -0.3914
T1 -1.426
T2 0.633
SP1 0
V1 0
V2 0
V3 0
SC1 0.011
Total 2.791
8.4.1 Rigorous Model Mass Balance

On checking the conservation of mass the following results were obtained:
Mass Balance = Mass Out (S9, S20, S21, S26, S28) − Mass In (S1, S2, S13)
= (1554.218 + 137.29 + 1275.82 + 1814.77 + 178.986)
− (1051.148 + 2447.518 + 1462.417)
= 4961.08327 − 4961.08244
= −0.0008317 kg/hr
Therefore there is a loss of 0.0008317 kg/hr with a relative error of 1.676 × 10−5 %
which is extremely small so the mass balance is still considered satisfied. However there
has been an 35.4% increase in loss from the simple mass balance.
32
8.4.2 Rigorous Model Energy Balance
On checking the energy balance the following results were obtained:
Energy Balance = Energy Out (S9, S20, S21, S26, S28) − Energy In (S1, S2, S13) + Total Duty
= (0.8289 + 0.0073 + 0.0897 + 0.103 + 0.0856)
− (−0.163 + −0.3939 + 0.09) + 2.791
= 1.1146 − −0.4669 + 2.7907
= 4.3721 M*kJ/hr
This is still a positive value indicating energy needs to be added to the system and is
larger than the simple model energy balance as expected. It is approximately more than
200 times larger revealing a significant difference once the rigorous model with pumps
and compressors is implemented.
33
9 Task 8: Equipment Sizing and Costing
9.1 Sizing
The sizing of all the physical equipment was performed in terms of height, cross-sectional
area, volume and capacity after consolidating the values of temperatures, pressures and
flow-rates of all the streams and the duties of all the equipment. All sizing calculations
and assumptions were performed according to Biegler et al. (1997).
9.1.1 Pumps
A pump efficiency of 0.5 and a motor efficiency of 0.9 was chosen according to Biegler
et al.. The capacity of the pump in GPM was determined from the volumetric flow-rate
of the incoming stream, which was multiplied by the pump outlet pressure to calculate
the C/H factor.
Table 21: Sizing Pumps as Processing Equipment
Equipment Work Capacity Pressure C/H Factor

Pumps Type (kW) (GPM) (PSI) (GPM*PSI)
P1 Centrifugal 3.70 4.76 1812.97 8.64E+03
P2 Centrifugal 6.73 13.28 1812.97 2.41E+04
P3 Centrifugal 1.20 9.40 290.08 2.73E+03
9.1.2 Heat Transfer Equipment

The heat transfer coefficient multiplied by area (UA) values were obtained from PRO II
while an overall heat transfer coefficient (U) was estimated from Biegler et al.. A value
of 0.43 kW/m2 K was used for the heat exchangers using steam as the utility and a value
of 0.57 kW/m2 K was used for the heat exchangers using cooling water as the utility. The
U values were chosen according to the assumption that the system is operating using
”Fuel Oil No.2 and Steam” and ”Organic solvents and water” for heating and cooling
respectively. All the heat-exchangers were assumed to be of shell and tube except for
the reboilers which are kettle. Hence, the area of the heat exchanger could then be
calculated.
9.1.3 Vessels
The flash and the two CSTR reactors were sized as vessels. It is assumed the flash has
a holdup time of 5 minutes whilst the residence time for each reactor could be used to
calculate the space velocity. The space velocity was then multiplied by the volumetric
flow rate of the incoming stream obtained to achieve the volume of the vessel. The
optimum aspect ratio for length:diameter is assumed to be 4:1 according to if the bottom
and top caps are four times as expensive as the sides. If the diameter is greater than 1.2
m the equipment will be sized as a horizontal vessel.
34
Table 22: Sizing Heat Exchangers as Processing Equipment
Heat Equipment Duty U UA A

Exchangers Type Type (kW) (kW/m2 K) (kW/K) (m2 )
E1 Heating Shell & Tube 140.00 0.43 2.23 5.24
E3 Cooling Shell & Tube 111.94 0.57 0.70 1.24
CT1 Condenser Shell & Tube 554.44 0.57 12.96 22.83
RT1 Reboiler Kettle 149.44 0.43 2.82 6.64
CT2 Condenser Shell & Tube 5560.83 0.57 43.54 76.73
RT2 Reboiler Kettle 5273.61 0.43 177.89 417.99
Table 23: Sizing Reactors and Flash as Pressure Vessels
Equipment Space Volumetric Flow V D L

Reactors Type Velocity (hr−1 ) Rate (m3 /hr) (m3 ) (m) (m)
R1 Vertical 1.15 4.32 3.75 1.06 4.24
R2 Horizontal 0.25 3.50 13.98 1.64 6.58
F1 Vertical 12 3.44 0.29 0.45 1.80
9.1.4 Distillation Columns

A typical value of tray spacing of 24 inches (0.6096 m) was chosen and implemented in
PROII for the distillation columns. The length of the distillation column was calculated
according to Equation 8 where the additional 6 m accounts for the extra feed space,
disengagement space (top and bottom) and skirt height.
Distillation column height (m) = (N − 1) × Tray spacing + 6 (8)
The diameters of the trays were obtained from PRO II according to the largest diameter
tray inside the column. The maximum height allowed is 60 m otherwise the column
needs to be split into multiple columns.
Table 24: Sizing Distillation Columns as Pressure Vessels
Equipment Tray D L
Distillation Type Trays Spacing (m) (m) (m)
T1 Vertical 13 0.6096 0.6588 13.3152
T2 Horizontal 30 0.6096 1.7392 23.6784
35
9.1.5 Membrane Separation
A typical flux rate of 75 gallons/ft2 day (0.127 m3 /m2 h) was chosen for the membrane
separation (AMTA, 2015). The area of the membrane can then be calculated according
to the flow-rate of the incoming stream obtained from PRO II divided by the flux-rate
as displayed in Equation 9. The cost of a typical membrane separation is assumed to
be 300 AC/m2 and 300 A C/m2 to install it. Since the flow rate of the incoming stream is
extremely small the membrane separation is not too expensive. The results are displayed
in Table 25.
A = Q/J (9)
Table 25: Membrane Sizing and Costing
Flux Volumetric Flow Area Uninstalled Total

Membrane (m3 /m2 h) Rate (m3 /hr) (m2 ) Cost Installation BMC
SC1 0.127 0.101 0.793 $255.75 $322.43 $578.18
9.2 Bare Module Costing

For preliminary design equations, the equipment bare module costs increase nonlinearly
with equipment size or capacity. Therefore the relationship between cost and size can
be described according to the correlations for processing equipment and pressure vessels
used by Guthrie as displayed in equations 10 and 11.
BC = C0 (S/S0 )α for processing equipment (10)

α β
BC = C0 (L/L0 ) (D/D0 ) for pressure vessels (11)
The correlations for each variable as well as pressure correction factors (MPF) and
module factors (MF) are calculated according to Biegler et al.. Hence the bare module
cost can be modified by the following factors:
Uninstalled Cost = (BC)(MPF) (12)

Total Installed Cost = BC(MPF+MF-1) (13)
Installation = BC(MF-1) (14)
Updated Bare Module Cost = UF(BC)(MPF+MF-1) (15)
The calculation procedure for each equipment type is displayed in Tables 26, 27, 28,
29 and 25. All costs are given in United States Dollar (USD).
36
Table 26: Costing Pumps as Processing Equipment
P Uninstalled Total Updated

Pumps Material Fm (kPa) F0 MPF S C0 S0 α MF UF
Cost BC
Installation Installed Cost BMC
P1 Stainless 1.93 1812.97 8.21 15.84 8.64E+03 650 2000 0.36 3.38 5.53 $1,100.60 $17,430.05 $2,619.43 $20,049.48 $110,956.46
P2 Stainless 1.93 1812.97 8.21 15.84 2.41E+04 1500 20000 0.64 3.38 5.53 $1,689.36 $26,754.08 $4,020.67 $30,774.75 $170,311.48
P3 Stainless 1.93 290.08 1.50 2.90 2.73E+03 650 2000 0.36 3.38 5.53 $726.73 $2,103.89 $1,729.62 $3,833.51 $21,215.16
Table 27: Costing Heat Exchangers as Processing Equipment
Heat P S Uninstalled Total Updated

Exchangers Type Fd (PSIG) Fp Fm MPF (ft2 ) C0 S0 α MF UF BC Cost Installation Installed cost BMC
E1 U tube 0.85 1812.97 0.55 2.50 3.50 56.43 300 5.5 0.024 1.83 5.53 $317.24 $1,110.34 $263.31 $1,373.65 $7,601.96
E2 U tube 0.85 1812.97 0.55 3.10 4.34 136.32 5000 400 0.65 3.29 5.53 $2,483.71 $10,779.32 $5,687.70 $16,467.02 $91,130.64
E3 U tube 0.85 14.70 0 2.50 2.13 13.34 300 5.5 0.024 1.83 5.53 $306.44 $651.20 $254.35 $905.54 $5,011.40
E4 U tube 0.85 14.70 0 3.10 2.64 166.95 5000 400 0.65 3.29 5.53 $2,833.45 $7,466.13 $6,488.59 $13,954.73 $77,227.28
E5 U tube 0.85 14.70 0 2.50 2.13 5.16 300 5.5 0.024 1.83 5.53 $299.54 $636.52 $248.62 $885.14 $4,898.49
E6 U tube 0.85 14.70 0 2.50 2.13 80.33 300 5.5 0.024 1.83 5.53 $319.94 $679.87 $265.55 $945.43 $5,232.11
37
E7 U tube 0.85 14.70 0 3.10 2.64 187.13 5000 400 0.65 3.29 5.53 $3,051.56 $8,040.86 $6,988.07 $15,028.93 $83,172.04
E8 U tube 0.85 14.70 0 2.50 2.13 15.07 300 5.5 0.024 1.83 5.53 $307.35 $653.11 $255.10 $908.21 $5,026.17
CT1 U tube 0.85 14.70 0 3.10 2.64 245.74 5000 400 0.65 3.29 5.53 $3,642.86 $9,598.95 $8,342.16 $17,941.10 $99,288.41
RT1 Kettle Reboiler 1.35 14.70 0 2.50 3.38 71.42 300 5.5 0.024 1.83 5.53 $319.04 $1,076.76 $264.80 $1,341.56 $7,424.38
CT2 U tube 0.85 290.08 0.1 3.26 3.10 825.89 5000 400 0.65 3.29 5.53 $8,009.95 $24,806.80 $18,342.78 $43,149.58 $238,795.41
RT2 Kettle Reboiler 1.35 290.08 0.1 3.74 5.42 4499.25 5000 400 0.65 3.29 5.53 $24,108.60 $130,689.07 $55,208.70 $185,897.77 $1,028,782.53
Table 28: Costing Reactors and Flash as Pressure Vessel
P Uninstalled Total Updated

Reactors Equipment Fm (PSIG) Fp MPF C0 L0 D0 α β MF UF BC Cost Installation Installed Cost BMC
R1 Vertical 3.67 1812.97 4.84 17.78 1000 4 3 0.81 1.05 4.23 5.53 $3,207.58 $57,026.85 $10,360.47 $67,387.32 $372,930.23
R2 Horizontal 3.67 14.70 1.02 3.74 690 4 3 0.78 0.98 3.18 5.53 $4,568.50 $17,071.96 $9,959.33 $27,031.28 $149,594.65
F1 Vertical 3.67 1500.00 3.71 13.63 1000 4 3 0.81 1.05 4.23 5.53 $651.77 $8,882.95 $2,105.22 $10,988.17 $60,809.96
Table 29: Costing Distillation Columns as Pressure Vessel
Tray Tray Uninstalled Total Updated

Distillation Material Fm Spacing (inch) Fs Type Ft MPF C0 L0 D0 α β MF UF BC Cost Installation Installed Cost BMC
T1 Stainless Steel 1.7 24 1 Sieve 0 2.7 180 10 2 0.97 1.45 1 5.53 $841.92 $2,273.19 $0.00 $2,273.19 $12,580.15
T2 Stainless Steel 1.7 24 1 Sieve 0 2.7 180 10 2 0.97 1.45 1 5.53 $6,013.00 $16,235.11 $0.00 $16,235.11 $89,847.20
38
Figure 26: Cost for each Equipment Type Figure 27: Equipment Cost for each Unit
9.3 Utility Costing
The utility costs for our process includes electricity, steam and cooling water according
to the Chinese market prices (Daily and Wong). The required equipment that involves
these additional costs are the pumps, flash, heat exchangers, condensers and reboilers.
The utility cost for each unit is summarised in Tables 30, 31 and 32.
Table 30: Utility Cost for Electricity
Work Electricity Utility

Unit (kW) Cost ($/kWh) Cost ($/hr)
P1 7.39 $0.09 $0.67
P2 13.45 $0.09 $1.22
P3 2.4 $0.09 $0.22
F1 162.55 $0.09 $14.79
Total ($/hr) $16.91
Table 31: Utility Cost for Cooling Water
Cooling CW Utility
Unit Water (kg/hr) Cost ($/m3 ) Cost ($/hr)
E3 19274.65 $0.03 $0.56
E5 2197.22 $0.03 $0.06
E6 28808.83 $0.03 $0.84
E7 31395.46 $0.03 $0.92
CT1 95364.74 $0.03 $2.79
CT2 956302.37 $0.03 $28.00
Total ($/hr) $33.18
Table 32: Utility Cost for Steam
Steam Steam Utility

Unit (kmol/hr) Cost ($/mol) Cost ($/hr)
E1 13.88 $0.00036 $5.00
E2 33.94 $0.00036 $12.23
E4 49.97 $0.00036 $18.00
E8 3.82 $0.00036 $1.38
RT1 17.51 $0.00036 $6.31
RT2 617.61 $0.00036 $222.53
Total ($/hr) $265.45
A comparison of the different utility costs is displayed in Figure 28. It is evident

that the majority of utility costs come from steam which is around 84% followed by
cooling water making up 10.52%. If we further analyse the utility costs according to
each unit as displayed in Figure 29, we can see that 70.52% of the utility costs is from
the second distillation reboiler followed by 8.87% from the second distillation condenser.
39
Therefore according to these results it is evident that in order to reduce the utility costs,
the second distillation needs to be optimised, in particular the reboiler.
Figure 28: Comparison of Cost for each Utility Figure 29: Utility Cost for each Unit
40
10 Task 9: Economical Analysis
10.1 Capital Investment Calculations
The total bare module cost for the process can now be calculated by summing the bare
module cost of each component from section 9.3. Additionally the fixed capital invest-
ment (FCI) and total capital investment (CI) can be calculated according to equations 16
and 17 respectively.
FCI = Manufacturing Capital + Nonmanufacturing Capital

= Total BMC + Contingency + Nonmanufacturing Capital (16)
CI = FCI + Working Capital (17)
where the contingency is 25% of the total BMC and the non manufacturing capital is
40% of the total BMC according to Biegler et al.. The manufacturing capital includes
the total bare module cost of the equipment as well as the contingency whilst the non-
manufacturing capital includes buildings, service and land. The results are displayed in
Table 33.
Table 33: Summary of Costs
Cost Value ($)

Total Bare Module Cost $2,642,414.30
Manufacturing Capital 25% BMC $660,603.57
Nonmanufacturing Capital 40% BMC $1,056,965.72
Fixed Capital Investment (FCI) $4,359,983.59
Working Capital 1 Month MC $2,969,658.41
Total Capital Investment (CI) $7,329,642.00
The working capital represents the funds required to operate the plant if there are
delays in payment and maintenance of inventories including meeting the initial payroll
and expenses. The working capital is calculated according to one month of the total
manufacturing costs.
The total manufacturing costs include all expenses that are made over the lifetime
of the plant for one year. This involves the daily operation of the plant as well as
indirect expenses such as taxes, insurance and depreciation. A summary of the expenses
included in the total manufacturing cost and how they were calculated are provided in
Table 34. The revenue, profit and net profit are additionally calculated. The SARE
expenses represents the Sales, Administration, Research & Engineering expenses, which
is typically 10% of sales.
A pie chart displaying the distribution of the manufacturing costs is displayed in
Figure 30. It is evident that the most expense is towards the catalysts (50.06%), raw
material (40.11%) and utilities (7.08%).
Therefore in order to decrease manufacturing costs significantly, the catalysts are the
main contributing factor, in particular the second reaction catalyst (47.42%). Either a
41
Table 34: Manufacturing Costs, Revenue and Profit Calculations
Manufacturing Costs Item Amount Price $/yr %

EO 1051.15 kg/hr $1.00 $/kg $8,409,183.20 23.60%
1. Raw Material MeOH 1462.42 kg/hr $0.50 $/kg $5,883,012.31 16.51%
Catalyst 1 52.56 kg/hr $7 $/kg $939,989.01 2.64%
2. Other Material Catalyst 2 2204.85 kg/hr $3 $/kg $16,900,175.25 47.42%
3. Maintenance 5% of FCI $217,999.18 0.61%
4. Labor 5 $32,640 $/yr $163,200.00 0.46%
5. Managers 1 $38,400 $/yr $38,400.00 0.11%
6. Lab analysis 1 $36,480 $/yr $36,480.00 0.10%
7. Insurance 2% of FCI $87,199.67 0.24%
8. Supplies 2% of FCI $87,199.67 0.24%
9. Depreciation 8% of FCI $348,798.69 0.98%
Steam 736723 mol/hr $0.00036 $/mol $2,123,563.38 5.96%
Electricity 185.79 kW $0.09 $/kWh $135,255.12 0.38%
10. Utilities
Cooling Water 1133.34 m3 /hr $0.03 $/m3 $265,445.42 0.74%
Total Manufacturing Cost (MC) $35,635,900.91
1. Sale of Product DMC 1805.63 kg/hr $2.00 $/kg $28,890,153.60 67.52%

EG 1275.82 kg/hr $1.20 $/kg $12,247,862.40 28.63%
2. Sale of By-Product EC 137.29 kg/hr $1.50 $/kg $1,647,465.60 3.85%
Revenue (R) $42,785,481.60
Profit (P) $7,149,580.69

SARE Expenses 10% of R $4,278,548.16
Net Proft (NP) $2,871,032.53
Figure 30: Manufacturing Cost Analysis
42
cheaper catalyst, less amount or more repetitions would be required in order to reduce
the cost. However this could sacrifice the conversion of the reaction and vary costs for
subsequent equipment. Further process analysis would be required to determine whether
this would be beneficial towards the process design and will be analysed in optimisation
of the process.
Further, the price of the raw materials depends on the current market, whilst the
amount of raw materials depends on the amount of DMC we want to produce. Therefore
this could be improved by finding cheaper suppliers of our raw materials as well as
ensuring that our reactors have the best optimal conditions.
In order to decrease the utilities cost a heat integration could be performed between
heat exchanger units. Additionally as mentioned earlier, the majority of these costs are
from the azeotropic distillation column which could undergo optimisation to reduce any
unnecessary expenses.
10.2 Net Present Value

The net present value (NPV) provides the present value of all payments according to a
given rate of return and project lifetime. It can be calculated according to Equation 18.
NPV = −CI + NP(1 − (1 + i)−n )/i (18)

where i = rate of return (%) and n = project lifetime (years)
To calculate the profitability of the process we can analysis the net present value
(NPV) with a given rate of return and project lifetime, the rate of return (ROR) for
NPV=0 and a given project lifetime, and the payback time (PBT) required for NPV=0
and a given rate of return. The results are displayed in Table 35.
Table 35: Net Present Value (NPV) Calculations
Specify Specify Calculate

Method Item Value Item Value Item Value
Net Present Value (NPV) i 10.0% n (years) 10 NPV $10,311,610.07
Rate of Return (ROR) n (years) 10 NPV $0.00 i 37.6%
Pay Back Time (PBT) i 10.0% NPV $0.00 n (years) 3.09
From the results it is evident that after 10 years the net present value would reach
$10,311,610.07 and we would break even after 3.09 years. Additionally we would break
even after 10 years, even if the rate of return was 37.6%, which would never be the case.
Additionally the change in the net present value over the project lifetime can be
represented as a cumulate cash flow diagram as displayed in Figure 31. It is also evident
in the cumulative cash flow figure that we will break even at 3.09 years according to the
intersection of the curve with the projected lifetime axis.
43
Figure 31: Cumulative Cash Flow
10.3 Sensitivity Analysis

A sensitivity analysis for the change in net present value, payback time and rate of return
after 10 years with a percentage change in the price of the raw material, the product, the
cost of utilities and the cost of equipment ranging between -10% and 10% was carried
out as displayed in Figure 32. It is seen that the NPV is most sensitive to the sales price
of the product (DMC) where an increase in 10% of the price would increase the NPV
by more than 150%. On the other hand, it is seen that the NPV is least sensitive to
the cost of the equipment, where a decrease in 10% of the equipment cost results in an
increase in NPV by only 8.7%.
Additionally the only way this process would not be profitable after 10 years would
be if the price of DMC decreased by more than 6% which is unlikely considering the
continuous increase in utilisation of this product due to its properties and versatility.
44
Table 36: Sensitivity Analysis
Variable % Change NPV ROR PBT

-10.00% $16,134,181.44 51.67% 2.22
-5.00% $13,222,895.75 44.65% 2.59
EO 5.00% $7,400,324.38 30.31% 3.84
10.00% $4,489,038.70 22.78% 5.06
-10.00% $13,975,491.57 46.47% 2.48
-5.00% $12,143,550.82 42.03% 2.75
MeOH 5.00% $8,479,669.32 33.02% 3.53
10.00% $6,647,728.57 28.40% 4.10
-10.00% -$5,664,963.81 -14.93% -
-5.00% $2,323,323.13 16.96% 6.59
DMC 5.00% $18,299,897.01 56.25% 2.03
10.00% $26,288,183.94 74.36% 1.51
-10.00% $11,883,696.51 41.40% 2.80
-5.00% $11,097,653.29 39.48% 2.94
Utilities 5.00% $9,525,566.85 35.62% 3.26
10.00% $8,739,523.63 33.67% 3.46
-10.00% $11,209,218.67 41.43% 2.79
-5.00% $10,760,414.37 39.44% 2.94
Equipment 5.00% $9,862,805.77 35.77% 3.25
10.00% $9,414,001.47 34.07% 3.42
Figure 32: Sensitivity Analysis
45
10.4 Economic Evaluation
According to the results of the preceding sections, it is evident that our process is
extremely affordable, profitable and viable. A payback time can be achieved after 3.09
years and will be profitable even if there is a increase of 10% in raw materials, utilities
and equipment costs.
However, this process could undergo optimisation in order to further improve the
equipment costs, utility costs, manufacturing costs and consequently the profitability
of this process. The most significant areas that should be considered in optimisation
include the azeotropic distillation, heat integration and the second reactor catalyst.
The azeotropic distillation condenser and reboiler results in almost 50% of the equip-
ment costs and almost 80% of the utility costs. Therefore by optimising this distillation
column the costs in these areas could be significantly reduced. However since it is an
azeotropic distillation it should be noted that it will require more equipment and utility
costs than the other distillation column. Further, heat integration could be performed
over the heat exchangers to remove any other unnecessary utility costs. Finally the
second reactor catalyst, due to its large amount and expense accounts for almost 50% of
the manufacturing costs. Therefore an alternative catalyst with a less amount, expense
and repetitions could be investigated.
11 Task 10: Heat Integration

11.1 Composite Curves
The goal of performing a heat integration and developing a heat exchanger network for
energy recovery is to minimise the annualised cost of the equipment plus the annual cost
of utilities. We decided to perform a heat integration over the entire system to determine
the maximum amount of utility costs that could be saved, despite it not being physically
possible to implement. The procedure for performing the heat integration is provided in
Biegler et al. (1997).
A summary of the heat exchangers indicating the hot or cold stream label as well as
the inlet and outlet temperatures and flowrate multiplied by heat capacity is provided
in Table 37.
Table 37: Hot and Cold Streams for Heat Integration
Heat Inlet Outlet Heat FCp

Exchangers Stream Temperature (K) Temperature (K) Out (kW) (kW/◦ C)
E1 C1 206.01 448 -139.92 0.58
E2 C2 201.89 448 -342.19 1.39
E3 H1 510.44 423 112.08 1.28
E4 C3 298.15 423 -555.5 4.45
E5 H2 470.57 303.15 165.58 0.99
E6 H3 494.25 303.15 13.89 0.07
E7 H4 363.37 303.15 159.67 2.65
E8 C4 333.04 423 -3.56 0.04
46
The problem can be partitioned into temperature intervals with a fixed minimum
temperature difference of 10◦ C. This is the minimum temperature driving force that
we are willing to have in any of the heat exchangers that will be present in the final
network. Vertical lines in Table 38 represent the hot and cold stream present in each
temperature interval range. The amount of heat available is tabulated according to the
hot streams and the amount of heat required is tabulated according to the cold streams
in each interval. Cascaded heat columns can then be formed for the hot streams and
cold streams which are then plotted according to Figure 33. It is important to note that
the heat flows start at zero on the right and increase as they move to the left.
The figure conveys a plot of temperature against heat flow and could represent the
temperature profiles within a heat exchanger. However we specified that the hot stream
must be at least 10◦ C hotter than the cold stream in any of our heat exchangers. There-
fore the hot curve needs to be shifted to the left to avoid the intersection of the two
curves and satisfy this condition.
47
Table 38: Heat Integration for ∆Tmin = 10◦ C
Composite Hot Streams Streams in Interval Temperatures Streams in Interval Composite Cold Streams Grand Composite
Avail Heat Casc’d Heat H1 H2 H3 H4 Hot Cold C1 C2 C3 C4 Req’d Heat Casc’d Heat Net Heat Casc’d Heat Adj Casc’d
0.00 — 510.44 (500.439) 0 589.94
20.75 20.75
20.75 — 494.25 (484.252) 20.75 610.69
32.08 32.08
52.83 — 470.57 (460.566) 52.83 642.77
29.45 29.45
82.28 (458) 448 — — 0.00 82.28 672.22
58.59 49.21 9.37
140.87 (433) 423 — — 49.21 91.66 681.60
23.44 64.57 -41.14
164.31 — 423 (413) 113.79 50.52 640.46
63.31 385.07 -321.76
227.62 — 363.37 (353.368) 498.86 -271.24 318.70
75.50 131.30 -55.80
48
303.12 (343.035) 333.04 — 630.16 -327.04 262.90

129.54 223.89 -94.35
432.66 (308.15) 298.15 — 854.05 -421.39 168.55
18.57 9.84 8.72
451.22 — — — 303.15 (293.15) 863.89 -412.67 177.27
0.00 171.55 -171.55
451.22 (216.005) 206.01 — 1035.44 -584.22 5.72
0.00 5.72 -5.72
451.22 (211.888) 201.89 — 1041.17 -589.94 0.00
Figure 33: Cascaded Composite Hot and Cold Heat Flows
A grand composite curve can be produced for the problem by creating the last three
columns of Table 38 where the cascaded heat is the amount of heat the problem has
available from the hot streams over that required by the cold streams as we move from
the higher temperatures to the cold ones. The negative numbers indicate that not
enough heat is produced by the hot stream to satisfy the need of the cold streams above
this entry. Therefore the most negative number is the amount of heat that must be
supplied by the hot utilities. This amount can be added to the hot streams to create
an adjusted hot stream and consequently an adjusted grand composite cascade. Now
no entry in the cascaded column is negative. The adjusted hot stream composite curve
is indicated in Figure 33 and reveals that the driving force condition of 10◦ C is now
satisfied. Additionally the grand composite curve is displayed in Figure 34.
Figure 34: The Grand Composite Curve
49
This curve conveys whether each interval is producing or consuming heat. It is
evident there is no pinch point as the zero entry in the adjusted cascade heat is for the
last temperature interval as indicated on the curve. Therefore only heat is added to the
system and no heat needs to be removed as all the heat intervals are above this point.
11.2 Heat Exchanger Network

A heat exchanger network can now be developed that will use precisely the minimum
utilities computed for it and nowhere in this network will any temperature driving force
be less than 10◦ C. This can be produced using a seven step procedure:
1. A minimum temperature driving force, ∆Tmin , is selected to be 10◦ C, as mentioned

earlier. This value was chosen in order to ensure there is a balance between heat
exchanger areas and the minimum utilities required.
2. The minimum utility use is computed according to the minimum driving force.
Since only heat needs to be added to the system and no heat needs to be removed,
589.94 kW is required as heat input from hot utilities.
3. The grand composite curve can then be analysed to determined the utilities to use
and their amounts. Since our problem only involves cooling water or steam, we
will use steam for heating the utilities.
4. There is no pinch point in our problem which is supported by step 2 as only heating
is required. Therefore the problem cannot be partitioned into a hot part and a
cold part.
5. The number of exchanges for each partition can then be estimated as N − 1, where
N is the number of streams in that part of the problem. Since there is only one
partition with 8 streams, 7 exchanges in total will be required.
6. A network can then be invented using all insights available. The heat exchanger
network created is displayed in Figure 35.
In order to drop the temperatures of H2, H3 and H4 to 303.15 K, only C1 and
C2 are cold enough. The other cold streams are either too hot or will violate the
minimum driving force requirement. However, the duty and FCp value of these two
cold streams is not large enough to reduce the entire temperature interval without
violating the minimum driving force. Therefore the hot streams first need to be
cooled down by alternative cold streams. Since C4 only requires a small amount
of energy it will not be useful in cooling H2, H3 or H4 and instead is heated by
H1.
C3 is the stream requiring the largest duty and has the largest FCp so it can be
used to initially cool down H3 and H4. Firstly, H4 is cooled down by 90 kW
to 329.4 K and H2 by 114 kW to 355.3 K, to satisfy the minimum driving force
condition of 10 K. Next we cool down H1 to the required temperature of 423 K
50
using 108.52 kW from C3. C3 therefore needs an additional 242.98 kW to reach
the desired temperature of 423 K by using steam.
After initial cooling, it is now feasible to cool down H2 and H4 to 303.15 K using
C2. When the exchange is performed C2 will reach a temperature of 289.08 K.
Therefore 220.94 kW is required by steam for C2 to reach 448 K. Finally, C1 is
used to cool down H3, and an additional 126.03 kW needs to be provided by steam
to heat the stream to 448 K.
Upon checking the number of heat exchanges, it matches the initial estimation
of 7 with an additional 3 heat exchanges from hot utilities. The amount of heat
required by the steam reaches a total of 589.94 kW as expected.
7. Remove heat cycles if possible. There are no loops in the system and the energy
used by the stream matches the calculations, indicating this is one possible heat
exchange network.
11.3 Evaluation
Upon performing a heat integration over the entire process it is evident that only heating
is required for 589.94 kW, using steam, which equates to $21.07 per hour. This is reduced
from $39 per hour which is a 45.98% decrease. However, in comparison to the entire
utility cost it is only a reduction of 5.68%. Therefore despite this heat integration being
physically impossible for our process as the heat exchangers are too far away from each
other, it is additionally only reducing the manufacturing costs by 0.4%. The cost of
implementing the heat network would most likely be more expensive than the savings.
However, a physically possible heat exchange network could be implemented between
the two cooling and one heating heat exchangers for the streams mixing before the second
reactor. This was additionally investigated and found to reduce the cost of the three
heat exchangers by 24.96%. But once again it only reduced the utility costs by 5.06%
and consequently the manufacturing costs by 0.36%.
Overall, neither a heat integration over the entire process or over a physically pos-
sible region would significantly impact on the costs of the process and are not worth
implementing.
51
Figure 35: Heat Exchanger Network
52
12 Task 11: Sustainability and Environmental Impact
Earlier, only the economical (capital and operating) aspects were considered while de-
signing the process and whilst environmental issues have not been accounted for. In order
to describe and quantify the potential environmental impact (PEI) through a chemical
process, the Waste Reduction Algorithm (WAR) has been developed. This has been
used to conduct the sustainability analysis of the project. The fundamental equation of
WAR algorithm is given by:
0 = Iin − Iout + Igen (19)
where, Iin denotes the PEI entering the system, Iout the PEI leaving the system and Igen ,
the PEI generated within the system. The impact can be calculated using the following
equation:
X
I= Mj Ψij (20)
j
where Mj is the mass flow rate of chemical j and Ψij is the chemical and category
specific impact (PEI/mass). The categories over which the PEI are measured have
been considered over eight environmental concerns namely human toxicity (ingestion
and inhalation) (HTPI and HTPE), aquatic toxicity (ATP), terrestrial toxicity (TTP),
global warming (GWP), ozone depletion (ODP), acid rain formation (AP) and smog
formation (PCOP).
The environmental impact was compared for both the base case design and two
alternative cases:
1. The purge ratio after the second reaction is decreased from a value of 0.5 to 0.025
2. 90% of the unreacted carbon-dioxide released from the flash is recycled (Figure ??)
Figure 36: Carbon Dioxide Recycle Loop
The results for the effluent streams and feed streams and displayed in Table 39.
Additionally the impact generated and hence the PEI per unit mass of the product
(DMC) for each case is displayed in Table 40.
53
Table 39: Specifications of the Feed and Effluent Streams (Non-Product) in Base Case and Alternative Designs
Effluent
Feed Base Case Alternative Case 1 Alternative Case 2
Stream S1 S2 S13 S9 S28 S9 S28 S40 S28
Temperature (K) 192.47 190.8 298.15 533 438.35 533 425.35 533 383.72
Pressure (bar) 2 2 1.01 103.42 20 103.42 20 103.42 8
Mass Flow Rate (kg/hr) 1051.15 2447.52 1462.42 1554.22 188.12 1554.22 12.05 190.59 11.59
Mass Fractions
Methane 0 0.0037 0 0.0058 0 0.0058 0.0001 0.022 0.0005
Ethylene 0.0324 0 0 0.0219 0 0.0219 0 0.1758 0.0002
Carbon Dioxide 0 0.9963 0 0.9283 0.0002 0.9283 0.0026 0.7563 0.0018
Ethylene Oxide 0.9676 0 0 0.0113 0.0147 0.0113 0.2283 0.0095 0.1816
Methanol 0 0 1 0 0.8638 0 0.6742 0 0.7154
Dimethyl Carbonate 0 0 0 0 0.1214 0 0.0948 0 0.1006
Ethylene Glycol 0 0 0 0 0 0 0 0 0
Ethylene Carbonate 0 0 0 0.0327 0 0.0327 0 0.0364 0
Table 40: PEI for the Base Case and Alternative Designs Calculated Using WAR Algorithm
54
Base Case Alternative Case 1 Alternative Case 2 Combination of 1 & 2

PEI Impact PEI/kg Impact PEI/kg Impact PEI/kg Impact PEI/kg
Categories Generated of product Generated of product % Change Generated of product % Change Generated of product % Change
HTPI -4963.56 -2.737 -4975.22 -2.7573 0.23% -5055.2 -2.8016 1.85% -5055.32 -2.7876 1.85%
HTPE -328.73 -0.1813 -328.92 -0.1823 0.06% -334.34 -0.1853 1.71% -334.34 -0.1844 1.71%
ATP -47.85 -0.0264 -48.06 -0.0266 0.44% -97.17 -0.0539 103.07% -97.3 -0.0537 103.34%
TTP -4963.56 -2.737 -4975.22 -2.7573 0.23% -5055.2 -2.8016 1.85% -5055.32 -2.7876 1.85%
GWP -0.32 -0.0002 -0.32 -0.0002 0.00% -0.75 -0.0004 134.38% -0.75 -0.0004 134.38%
ODP 0 0 0 0 0.00% 0 0 0.00% 0 0 0.00%
PCOP -2684.56 -1.4803 -2768.2 -1.5342 3.12% -2795.29 -1.5492 4.12% -2794.53 -1.541 4.10%
AP 0 0 0 0 0.00% 0 0 0.00% 0 0 0.00%
For the base-case design it was seen that the impact generated is negative for all
the eight categories which means that the process design is moving in the direction of
achieving sustainability. The ozone depletion potential and acidification potential are
both zero for the process. The human toxicity potential which contributes majority to
the PEI for the base case is significantly reduced from the feed. The main concern of
mitigating carbon-dioxide is indicated by the global warming potential (GWP), which
is also reduced in comparison to the feed.
Hence, comparing the the base-case with the 1st alternative case of decreasing the
purge ratio, it is seen that only the photochemical oxidation potential is decreased by
3.11%. However, in the comparison of the base-case with the 2nd alternative case it is
seen that the there is a significant drop in the Global Warming Potential (by 134.37%)
and the Aquatic Toxicity Potential (103.34%). This was due to the 90% decrease in
carbon-dioxide released into the atmosphere. However, taking into consideration the
economical aspects, the recycle of carbon-dioxide would require an extra compressor
and a heat-exchanger to bring it from the flash conditions to the reactor conditions.
Typically, 61 kWh of electricity is required to compress 1 tonne of CO2 . Therefore, an
extra operating cost for the compression of 1.3 tonne of CO2 (amount recycled per hour)
is 7.13 USD/hr.
Considering that the plant generates electricity by burning coal, the amount of CO2
that is being let out into the atmosphere in the base case design was 1.4 tonnes/hour.
Considering a typical value of 1.0959 kg of CO2 produced per kWh of power consumed
(EIA, 2015), 0.2 tonnes/hour of CO2 is produced. Hence, if the CO2 is not recycled the
process would be generating more CO2 than utilising.
Therefore in order to increase the sustainability of our process it would be wise to
recycle the carbon dioxide to reduce the Global Warming Potential.
55
13 Task 12: Optimisation
13.1 Problem Formulation
A process optimisation problem can be formulated for our process design. A general
mixed integer nonliner program can be formulated in the form:
minf (x, y)
x,y
s.t. h(x, y) = 0
g(x, y) ≤ 0
x ∈ Rn , y ∈ {0, 1}t
where f (x, y) is the objective function, h(x, y, ) = 0 are the equations that describe the
performance of the system and the inequality constraints g(x, y) ≤ 0 can define process
specifications or constraints.
We will attempt to create an optimisation problem according to the distillation
columns. Since the reboilers account for a major proportion of the operating costs due to
the use of steam, the objective of the problem will be to minimise the reboiler heat duty.
A distillation column labelled with variables is displayed in Figure 37 with N number
of trays. It is evident there is a countercurrent flow between the liquid and vapour that
provides the enrichment of the light key components in the top product and the heavy
key in the bottom. It is also provided that there is a condensor to condense the vapour
and a reboiler to boil up the liquid. For example, in the second distillation column we
Figure 37: A Distillation Column Labelled with Variables
want to recover most of the DMC in the bottom product and all of the methanol in the
top product. The feed of chemical components to the distillation column are as follows:
ς = {Methane, Ethylene, CO2 , EO, MeOH, DMC, EG, EC} (21)
56
We assume a total condenser and a partial reboiler, and that the liquid and vapour
phases are in equilibrium. Therefore we can construct a tray by tray distillation column
model using the mass-equilibrium-summation-heat equations as follows.
Total Mass Balances
B + V0 − L1 = 0 (22)
Li + Vi − Li+1 − Vi−1 = 0, i ∈ [1, N ], i ∈
/ δ, (23)
Li + Vi − Li+1 − Vi−1 − F = 0, i ∈ [1, N ], i ∈
/ δ, (24)
LN +1 + D − VN = 0 (25)
Component Mass Balances
Bx0,j + V0 y0,j − L1 x1,j = 0, j ∈ ς (26)

Li xi,j + Vi yi,j − Li+1 xi+1 , j − Vi−1,j yi−1,j = 0, j ∈ ς, i ∈ [1, N ], i ∈
/δ (27)
Li xi,j + Vi yi,j − Li+1 xi+1 , j − Vi−1,j yi−1,j − F xF,j = 0, j ∈ ς, i ∈
/δ (28)
(LN +1 + D)xN +1,J − VN yN,j = 0, j ∈ ς (29)
xN +1,j − yN,j = 0, j ∈ ς (30)
Enthalpy Balances
BHB + V0 HV,0 − L1 HL,1 − QR = 0 (31)

Li HL,i + Vi HV,i − Li+1 HL,i+1 − Vi−1 HV,i−1 = 0, i ∈ [1, N ], i ∈
/ δ, (32)
Li HL,i + Vi HV,i − Li+1 HL,i+1 − Vi−1 HV,i−1 − F HF = 0, i ∈
/ δ, (33)
VN HV,N − (LN +1 + D)HL,D − QC = 0 (34)
Summation, Enthalpy and Equilibrium Relations

m
X m
X
yi,j − xi,j = 0, i = 0, ..., N + 1 (35)
j=1 j=1
yi,j − Ki,j (Ti , P, xi )xi,j = 0, j ∈ ς, i = 0, ..., N + 1 (36)

HL,i = ϕL (xi , Ti ), HV,i = ϕV (yi , Ti ), i = 1, ..., N (37)
HB = ϕL (x0 , T0 ), HF = ϕV (xF , TF ), HN +1 = ϕL (xN +1 , TN +1 , i = 1, ..., N (38)
where all the symbols are given in the List of Symbols.

The column is operated at constant pressure of 8 bar with no pressure drop. Addi-
tionally we want only 5% of the mole fraction of MeOH to be DMC in the top leaving
57
the column. Therefore the problem can be given by:
minQR
s.t.(22) − (38)
xtop,HK ≤ 0.05xtop,LK
Li , Vi , Ti ≥ 0, i = 1, ..., N + 1,
D, QR , QC ≥ 0
yi,j , xi,j ∈ [0, 1], j ∈ ς, i = 1, ..., N + 1
This problem is extremely difficult to solve by hand and instead we will attempt to
be performed using the base case model and PROII.
13.2 Base Case Design

13.2.1 Purge
At the start of the process we decided to purge 50% of the recycle streams and reevaluate
this decision upon optimisation of the process. The only purge we have in the system
is for the recovery of methanol and dimethyl carbonate. In the purge stream we need
to minimise the loss of valuable reactant and product molecules. The smaller the purge
split, the more impurities of ethylene oxide, methane and ethylene would build up in
the system, but less dimethyl carbonate and methanol would be lost. However this will
come at a cost as the smaller the fraction we remove, the larger the flow of the recycle
stream consequently resulting in larger equipment sizes. Moreover, since we would be
recycling methanol it would reduce the amount of raw material required.
Since the recycle stream is almost 99% composed of valuable reactants or products
we will recycle 97.5% and remove 2.5% in order to avoid the build up of impurities. This
results in 3% of the stream entering the second reactor is ethylene oxide. Therefore, we
assume that a purge any higher than this value would cause the build up of too much
impurities in the reactor that could affect the catalyst and the reaction conditions. A
comparison of the component amounts, costs and net present value for the base case and
the purge optimisation can then analysed according to Tables 41 , 42 and 43 respectively.
Table 41: Comparison of Component Amounts for Purge Optimisation
Amount (kg/hr)
Component Base Case Optimised % Change
Reactant MeOH 1462.42 1308.05 -10.56%
Product DMC 1805.63 1827.33 1.20%
It is evident in Table 41 that by increasing the purge, the amount of methanol re-
quired would decrease by 10.56% whilst the amount of DMC produced increases by
1.20%. This results in a decrease in manufacturing costs (1.69%) and an increase in
revenue (0.81%) according to Table 42. Further since there is an increase in the recycled
58
Table 42: Comparison of Costs for Purge Optimisation
Purge
Cost Base Case Optimised % Change
Total Bare Module Cost (BMC) $2,642,414.30 $2,698,485.44 2.12%
Total Capital Investment (CI) $7,329,642.00 $7,371,848.63 0.58%
Total Manufacturing Cost (MC) $35,635,900.91 $35,032,171.87 -1.69%
Utility Cost $2,524,263.92 $2,525,791.68 0.06%
Revenue (R) $42,785,481.60 $43,132,652.80 0.81%
Net Profit (NP) $2,871,032.53 $3,787,215.65 31.91%
stream flow the equipment costs increase (2.12%) for the recycle loop equipment. Addi-
tionally since there is a larger amount of recycled stream to be cooled, the cooling water
increases whilst the required steam to heat the methanol stream decreases leading to
an increase (0.06%) in the utility costs. Overall there is a net profit increase of 31.91%
which is significantly larger.
The net present value, rate of return and pay back time are also compared in Table 43.
It is evident that the purge significantly increases the net present value after 10 years at
10% interest rate resulting in a 54.18% increase. Additionally the rate of return for the
net present value to be zero after 10 years has to be 50.51% which would never be the
case. Further, the pay back time reduces to 2.27 which is a decrease of 26.56%.
Table 43: Comparison of NPV (10 years and 10% interest) for Purge Optimisation
Purge
Method Base Case Optimised % Change
Present Value (NPV) $10,311,610.07 $15,898,952.10 54.18%
Rate of Return (ROR) 37.56% 50.51% 34.50%
Pay Back Time (PBT) 3.09 2.27 -26.56%
Therefore by increasing the purge fraction recycled to 97.5% we can significantly

optimise our process and reduce the pay back time to 2.27 years, which is a remarkable
difference and should be implemented in the process design.
13.2.2 Recycling Ethylene Carbonate

Our process produces pure ethylene carbonate, which is being sold as it is quite a valuable
product. However it could be recycled instead to produce additional dimethyl carbonate
as it is a more valuable product. This could influence equipment costs and utility costs,
in particular the second reactor as well as the amount of catalyst required. Since it
is pure ethylene carbonate no purge stream is necessary. A process diagram with the
recycle stream is displayed in Figure 38.
A comparison of the component amounts, costs and net present value for the base
case and the EC recycle optimisation can then analysed according to Tables 44 , 45
and 46 respectively.
59
Figure 38: Recycle Loop for Ethylene Carbonate
Table 44: Comparison of Component Amounts for EC Recycle Optimisation
Amount (kg/hr)
Reactant MeOH 1462.42 1573.69 7.61%
Other Material Catalyst 2 2204.85 2372.61 7.61%
DMC 1805.63 1943.02 7.61%
Products EG 1275.82 1372.89 7.61%
Table 45: Comparison of Costs for EC Recycle Optimisation
EC Recycle
Costs Base Case Optimised % Change
Total Bare Module Cost (BMC) $2,642,414.30 $2,661,897.09 0.74%
Total Capital Investment (CI) $7,329,642.00 $4,392,130.20 -40.08%
Total Manufacturing Cost (MC) $35,635,900.91 $37,393,155.39 4.93%
Utility Cost $2,524,263.92 $2,542,552.10 0.72%
Revenue (R) $42,785,481.60 $44,268,105.28 3.47%
Net Profit (NP) $2,871,032.53 $2,448,139.36 -14.73%
60
In Table 44 the amount of methanol, catalyst, DMC and ethylene glycol all increase
by 7.61% due to recycling all the ethylene carbonate. Additionally, because of the
increase in catalyst and the reactant, methanol, the manufacturing cost increases by
4.93% whilst the increase in DMC and ethylene glycol only increases the revenue by
3.47% indicated in Table 45. Due to an increase in flow entering the second partition
of the process, there is an increase in equipment cost (0.74%) and utility cost (0.72%).
This results in an overall net profit decrease of 14.73%.
If we further analyse the net present value adjustment in Table 46 it is evident that
recycling the ethylene carbonate is not beneficial to our process. The net present value
after 10 years would decrease by 26.93%, the rate of return would decrease by 19.34%
and the pay back time would increase to 3.84 years which is an increase of 24.46%.
Table 46: Comparison of NPV (10 years and 10% interest) for EC Recycle Optimisation
EC Recycle
Present Value (NPV) $10,311,610.07 $7,534,530.12 -26.93%
Rate of Return (ROR) 37.56% 30.29% -19.34%
Pay Back Time (PBT) 3.09 3.84 24.26%
Therefore, recycling the ethylene carbonate should overall result in beneficial results
for our process by increasing the amount of DMC produced. However since a large
amount of catalyst is required for the second reaction according to how much ethylene
carbonate reacts, the cost of the catalyst prevents this adjustment from optimising the
process. Overall the recycle stream can only be implemented if the second reaction is
optimised to find a cheaper catalyst or one that requires less amount.
13.2.3 Distillation Columns

According to the economical evaluation conclusion in Section 10.4 it is extremely vital
to optimise the distillation columns, in particularly the azeotropic distillation column,
which results in almost 50% of the equipment costs and almost 80% of the utility costs. In
order to achieve this the pressure required to break the azeotrope is first analysed before
attempting to reevaluate the number of stages and duties required for each column.
The pressure of 20 bar was chosen to break the azeotrope according to Kongpanna
et al. (2014). However a curve plotting the pressure of the stream entering the distillation
column against the composition of methanol in the azeotrope can be analysed in order
to determine the minimum pressure required to break the azeotrope. This is displayed
in Figure 39.
According to this figure the minimum pressure required to break the azeotrope is
only 7.5 bar. Therefore the pressure for the azeotrope distillation could instead be
any pressure above this value. We decided to choose a pressure of 8 bar to ensure the
azeotrope is broken in our process. After implementing the pressure change, the two
distillation columns could then be optimised to reduce the number of stages as well as
61
Figure 39: Pressure Dependency of the DMC/MeOH Azeotrope
the duties and then the optimum feed tray location could be determined using driving
force.
The optimum feed tray location can be determined according the the composition
of the more volatile component in the separation at the maximum driving force. Equa-
tion 39 indicates how this can be calculated.
Nf = N (1 − xLK ) (39)
where Nf is the feed tray location, N is the total number of trays and xLK is the mole
fraction of the light key where the driving force is a maximum.
The driving force curves for each distillation separation are displayed in Figure 40
and 41.
It is evident that the maximum driving force for the DMC/EG separation results
in a mole fraction of DMC of 6% and equates to an optimum feed tray location of 11.
Similarly it can be seen that the maximum driving force for the MeOH/DMC separation
occurs when the mole fraction of MeOH is 22% and equates to an optimum feed tray
location of 13.
The results for the both distillation columns, involving the optimisation of pressure,
feed tray location, number of trays and duties is displayed in Tables 47 and 48
62
Figure 40: Driving Force for DMC/EG Separation at 1 bar
Figure 41: Driving Force for MeOH/DMC Separation at 8 bar
63
Table 47: Optimised Distillation Column Results for DMC/EG
Input Ouput
Distillation Number of Position of Reflux Condensor Reboiler
Column 1 Trays Feed Tray Ratio Duty (M*kJ/hr) Duty (M*kJ/hr)
Base Case 13 4 0.759 -1.9503 0.524
Optimised 11 11 0.568 -1.7391 0.3067
% Change -15.38% - -25.23% -10.83% -41.47%
For the first distillation column (Table 47), we were able to reduce the number of
stages by two. Additionally the reflux ratio decreased (25.23%) and both the condenser
and reboiler decreased in duty.
Table 48: Optimised Distillation Column Results for MeOH/DMC
Input Ouput
Distillation Number of Position of Reflux Condensor Reboiler
Column 2 Trays Feed Tray Ratio Duty (M*kJ/hr) Duty (M*kJ/hr)
Base Case 30 3 20.073 -17.9765 18.6098
Optimised 16 13 8.086 -2.9327 3.3527
% Change -46.67% - -59.72% -83.69% -81.98%
For the second distillation column we were able to significantly reduce everything in
comparison to the base model. In particular the condenser and reboiler duties both sig-
nificantly decreased by more than 80%. This will greatly affect the utility and equipment
costs.
A comparison of the costs and net present value for the base case and the distillation
optimisation can then analysed according to Tables 49 and 50 respectively.
Table 49: Comparison of Costs for Distillation Columns Optimisation
Distillation
Total Bare Module Cost (BMC) $2,642,414.30 $1,566,502.73 -40.72%
Utility Cost $2,524,263.92 $907,642.88 -64.04%
Revenue (R) $42,785,481.60 $42,785,481.60 0.00%
Net Profit (NP) $2,871,032.53 $4,789,446.77 66.82%
It is evident that by optimising the distillation columns we can significantly decrease

our costs and improve our profit. Our equipment costs decrease by 40.72% as well
as our utility costs by 64.04% which is what we aimed to decrease by optimising the
columns. Further our manufacturing costs decrease due to the utility costs decreasing.
The revenue however does not change as we have not altered the amount of products we
are producing. Overall the net profit significantly increases by 66.82%.
A comparison of the effect of this optimisation on utility cost for each equipment unit
64
is evident in Figures 42 and 43. The reboiler and condenser for the second distillation
significantly reduce in cost and ultimately contribute a less percentage to the overall
cost.
Figure 42: Base Case Utility Cost for each Unit Figure 43: Optimised Utility Cost for each Unit
Further a comparison of the effect of this optimisation on equipment cost for each
unit is evident in Figures 44 and 45. Similarly, the reboiler and condenser for the second
distillation equipment costs significantly reduce and contribute a less percentage to the
overall cost.
65
Figure 44: Base Case Equipment Unit Cost Figure 45: Optimised Equipment Unit Cost
The net present value, rate of return and pay back time are also compared in Table ??.
It is evident that optimisation the distillation columns also significantly increases the
net present value after 10 years at 10% interest rate resulting in a 133.08% increase.
Additionally the rate of return for the net present value to be zero after 10 years has
to be 135.99% which would be impossible. Further, the pay back time reduces to 1.25
which is a decrease of 59.46%.
Table 50: Comparison of NPV (10 years and 10% interest) for Distillation Column Optimisation
Distillation
Pay Back Time (PBT) 3.09 1.25 -59.46%
Therefore by optimising the distillation columns we can significantly reduce equip-

ment costs and utility costs and is able to reduce the payback time to 1.25 years. This
optimisation should clearly be implemented in our process
13.2.4 Optimising Combination

The alterations in purge and distillation columns both significantly optimise our process
design. Therefore a combination of the two optimisations can be implemented to yield
the following results in Tables 51, 52 and 53.
It is evident that only the purge optimisation contributes to increasing the amount
of product produced and decreasing the amount of methanol required. However both
contribute to a decrease in equipment cost, manufacturing cost, utility cost and net
66
Table 51: Comparison of Component Amounts for Combined Optimisation
Amount (kg/hr)
Reactant MeOH 1462.42 1308.43 -10.56%
Product DMC 1805.63 1827.18 1.20%
profit. An overall net profit is more than double the amount of the base case when both
optimisations are combined.
Table 52: Comparison of Costs for Combined Optimisation
Combined
Total Bare Module Cost (BMC) $2,642,414.30 $1,606,897.19 -39.19%
Utility Cost $2,524,263.92 $870,763.55 -65.50%
Revenue (R) $42,785,481.60 $43,137,008.40 0.82%
Net Profitt (NP) $2,871,032.53 $5,750,811.98 100.30%
Further, if we analyse the net preset value after 10 years we can see that is has almost
tripled whilst the pay back time has reduced by 66.47% to 1.04 years.
Table 53: Comparison of NPV (10 years and 10% interest) for Combined Optimisation
Combined
Pay Back Time (PBT) 3.09 1.04 -66.47%
Therefore by combining the purge and distillation optimisation we can significantly

improve our process design reducing the payback time to 1.04 years which is extremely
good. Additionally we were able to reduce all unnecessary costs whilst still producing
the required amount of DMC for the market
67
14 Conclusion
Overall, we were able to produce an extremely profitable and feasible process design after
following the 12 step conceptual design procedure. We were able to to achieve a pay-
back time of 1.09 years after performing optimisation with a net profit of $4,789,446.77.
Additionally all products produced were at the required purity for sale. Some additional
values are provided in Table 54.
Table 54: Conclusion Values
Cost of Product/kg of Product $2.00 $/kg

kg of Raw Material (EO)/kg of Product 0.58
kg of Raw Material (MeOH)/kg of Product 0.81
Amount of Steam/kg of Product 0.13 mol/kg
Amount of Cooling Water/kg of product 0.17 m3 kg
Amount of Waste/kg of product 0.87
CO2 Emission/kg of product 0.80
Consumption of CO2 /kg of product 1.35
68
15 List of Symbols
BC Bare Cost
BMC Total Bare Module Cost
C0 Basic Cost
Ci Cold Stream i
CI Total Capital Investment
CO2 Carbon-dioxide
Cp Heat capacity
CTi Condenser of distillation coulmn i
D Diameter
D/B Distillate/Bottoms Flow Rate
DMC Dimethyl Carbonate
E Ethylene
EC Ethylene Carbonate
EG Ethylene Glycol
Ei Heat-Exchanger i
EO Ethylene Oxide
EP0 Initial Economic Potential
F Feed flow rate
FCI Fixed Capital Investment
fi Vector of flow-rates for stream i
Fi Flash i
FNM Flow-rate of stream formed by mixing N streams
GPM Gallons per minute
HF Feed enthalpy
Hi Hot Stream i
HLi /HVi i enthalpy of liquid/vapor leaving tray i
Igen PEI generated in the system
Iin PEI into the system
Iout PEI out of the system
Ki,j Nonlinear vapor/liquid equilibrium constant
L Length
Li /Vi Flow rate of liquid/vapor leaving tray i
lk /hk ∈ ς Light/heavy key components that determine the separation.
M Methane
MC Total Manufacturing Cost
MeOH Methanol
MF Module Factor
Mi Mass flow-rate of stream i
MPF Material and Pressure Factor
NE Number of equations
Nf Feed Tray
69
NP Net Profit
NPV Net Present Value
NV Number of variables
P Pressure in the column
PBT Pay-back time
Pi Pump i
Qr Heat
QR /QC Heat load on reboiler/condenser
R Revenue
Ri Reactor
ROR Rate of return
RTi Reboiler of Distillation Column i
S ∈ [0,1] Set of feed tray locations in column, numbered from the bottom
S0 Basic Capacity
SCi Splitter as a Stream Calculator
SPi Divider
Ti Temperature of tray i
UF Updating factor
V Volume
Vi Valve i
xF Feed composition
xi,j Mole fraction j in liquid leaving tray i
xLK Liquid composition of light key
yi,j Mole fraction j in vapor leaving tray i
α Exponent less than 1 in the power law expression to estimate the cost
γr,i Stoichiometric coefficient for component i
∆ hj Enthalpy of stream
∆ Hr Enthalpy of reaction
ηr,k Conversion of the reaction with respect to the key component k
D,i Recovery of component i from the divider
s,i Recovery of component i from the splitter
φV /φL Nonlinear vapor/liquid enthalpy function
Ψij Category specific impact for stream j in chemical category i
70
16 References
AMTA. Membrane filtration (mf/uf). http://www.amtaorg.com/wp-content/
uploads/2_membranefiltration.pdf, 2015.
Aresta, M. Carbon Dioxide Recovery and Utilization. Springer, 2003. ISBN

9781402014093.
Bhanage, B. M., Fujita, S.-i., He, Y., Ikushima, Y., Shirai, M., Torii, K., and Arai, M.
Concurrent synthesis of dimethyl carbonate and ethylene glycol via transesterification
of ethylene carbonate and methanol using smectite catalysts containing mg and/or ni.
Catalysis letters, 83(3-4):137–141, 2002.
Bhanage, B. M., Fujita, S.-i., Ikushima, Y., Torii, K., and Arai, M. Synthesis of dimethyl
carbonate and glycols from carbon dioxide, epoxides and methanol using heteroge-
neous mg containing smectite catalysts: effect of reaction variables on activity and
selectivity performance. Green Chemistry, 5(1):71–75, 2003.
Biegler, L., Grossmann, I., and Westerberg, A. Systematic methods of chemical process
design. Prentice-Hall international series in the physical and chemical engineering
sciences. Prentice Hall PTR, 1997. ISBN 9780134924229.
Cao, M., Meng, Y., and Lu, Y. Synthesis of dimethyl carbonate (dmc) from co2, ethylene
oxide and methanol using heterogeneous anion exchange resins as catalysts. Reaction
Kinetics and Catalysis Letters, 88(2):251–259, 2006.
Chang, Y., Jiang, T., Han, B., Liu, Z., Wu, W., Gao, L., Li, J., Gao, H., Zhao, G., and
Huang, J. One-pot synthesis of dimethyl carbonate and glycols from supercritical co
2, ethylene oxide or propylene oxide, and methanol. Applied Catalysis A: General, 263
(2):179–186, 2004.
Daily, S. Tiered power bill debated. http://english.sz.gov.cn/ln/201205/

t20120517_1914423.htm, May 2012.
Dhuri, S. M. and Mahajani, V. V. Studies in transesterification of ethylene carbonate

to dimethyl carbonate over amberlyst a-21 catalyst. Journal of Chemical Technology
and Biotechnology, 81(1):62–69, 2006.
EIA. How much carbon dioxide is produced per kilowatthour when generating electricity
with fossil fuels? http://www.eia.gov/tools/faqs/faq.cfm?id=74&t=11, March
2015.
Feng, X.-J., Li, X.-G., He, R., and Zhou, H. A new process for synthesis of dimethyl
carbonate from ethylene carbonate and methanol without any catalyst under super-
critical conditions. Chinese Chemical Letters, 16(6):767–770, 2005.
71
Jagtap, S. R., Bhor, M. D., and Bhanage, B. M. Synthesis of dimethyl carbonate via
transesterification of ethylene carbonate with methanol using poly-4-vinyl pyridine as
a novel base catalyst. Catalysis Communications, 9(9):1928–1931, 2008.
Ju, H.-Y., Manju, M. D., Park, D.-W., Choe, Y., and Park, S.-W. Performance of ionic
liquid as catalysts in the synthesis of dimethyl carbonate from ethylene carbonate and
methanol. Reaction Kinetics and Catalysis Letters, 90(1):3–9, 2007.
Kim, D.-W., Kim, C.-W., Koh, J.-C., and Park, D.-W. Synthesis of dimethyl carbonate
from ethylene carbonate and methanol using immobilized ionic liquid on amorphous
silica. Journal of Industrial and Engineering Chemistry, 16(3):474–478, 2010.
Kongpanna, P., Pavarajarn, V., Gani, R., and Assabumrungrat, S. Techno-economic

evaluation of different co 2-based processes for dimethyl carbonate production. Chem-
ical Engineering Research and Design, 93:496–510, 2014.
Lu, X.-B., Xiu, J.-H., He, R., Jin, K., Luo, L.-M., and Feng, X.-J. Chemical fixation of
co 2 to ethylene carbonate under supercritical conditions: continuous and selective.
Applied Catalysis A: General, 275(1):73–78, 2004.
Monteiro, J. G. M.-S., Araújo, O. d. Q. F., and de Medeiros, J. L. Sustainability metrics

for eco-technologies assessment, part i: preliminary screening. Clean Technologies and
Environmental Policy, 11(2):209–214, 2009.
Peppel, W. Preparation and properties of the alkylene carbonates. Industrial & Engi-
neering Chemistry, 50(5):767–770, 1958.
Souza, L. F., Ferreira, P. R., de Medeiros, J. L., Alves, R. M., and Araújo, O. Q.
Production of dmc from co2 via indirect route: Technical–economical–environmental
assessment and analysis. ACS Sustainable Chemistry & Engineering, 2(1):62–69, 2013.
Wong, F. China steam coal prices up 1st time in 11 months, stoke

recovery hopes. http://www.reuters.com/article/2013/10/17/
china-coal-prices-idUSL3N0I71MO20131017, October 2013.
Yang, Z.-Z., He, L.-N., Dou, X.-Y., and Chanfreau, S. Dimethyl carbonate synthe-
sis catalyzed by dabco-derived basic ionic liquids via transesterification of ethylene
carbonate with methanol. Tetrahedron Letters, 51(21):2931–2934, 2010.
Yuxiao, S. Dmc market potential needs to be developed.(organics). China Chemical

Reporter, 2009.
72

28350: Design Project For The Production of Dimethyl Carbonate From Ethylene Oxide

Uploaded by

Copyright:

Available Formats

28350: Design Project For The Production of Dimethyl Carbonate From Ethylene Oxide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

28350: Design Project For The Production of Dimethyl Carbonate From Ethylene Oxide

Uploaded by

Copyright:

Available Formats

28350: Design Project for the Production of

Dimethyl Carbonate from Ethylene Oxide

Dane Hurley (s141089), Spardha Jhamb (s141287)

2 Task 1: Product Information 2

3 Task 2: Economic Viability 4

4 Task 3: Hierarchical Decomposition 9

5 Task 4: Mass Balance 12

6 Task 5: Identification of Stream Conditions 18

7 Task 6: Energy Balance 20

8 Task 7: Rigorous Simulation 24

10 Task 9: Economical Analysis 41

11 Task 10: Heat Integration 46

12 Task 11: Sustainability and Environmental Impact 53

13 Task 12: Optimisation 56

Table 1: Physical properties of DMC

2.2 DMC Production

Figure 1: Various routes for production of DMC (Monteiro et al., 2009)

The five alternatives displayed include methanol oxy-carbonylation in liquid or gaseous

3.2 Production of Dimethyl Carbonate from Ethylene Oxide

3.3 Chemical Properties

Table 3: Conversion Data for Reaction 2

Symbol M E CO2 EO MeOH DMC EG EC

3.5 Initial Economic Potential

Table 5: Initial Economic Potential Flowrates and Price Summary

4.2 Level 2: Decisions on Input-Output Structure

Figure 5: Reactor 1 Figure 6: Reactor 2

4.3 Level 3: Decisions on Recycle Structure and Reactor

Figure 7: Recycle Stream 1 Figure 8: Recycle Stream 2

4.4 Level 4: Decisions on Separation System

Figure 11: Flowsheet Formed from Hierarchal Decomposition

Figure 12: Simple Flowsheet with Specifications

Divider and Splitter

Table 6: Degree of Freedom Analysis

Component Equation Number Value

Component Variables Number Value

Degrees of Freedom: NV-NE 13*NC+2 90

Specified Variables f1 , f2 , f9 and γr , ηk , ξS , ξD 7*NC +2 90

5.2 Flowsheet Decomposition and Calculation Sequence

Figure 13: Flowsheet Decomposition Stream Labelling

Nodes Precursors (1) Prescursers (2)

Table 9: Incidence Matrix

Figure 14: Mass Balance Flowsheet from PROII

Table 10: Mass Balance Values from PROII

Figure 15: Flash

A temperature of 578 K was chosen as increasing the temperature much higher

Divider and Splitter

Utility Name Duty (M*kJ/hr)

Figure 17: Simple Energy Balance Flowsheet on PROII

Table 13: Energy Balance Values using PROII

7.2 Pumps and Compressors

Table 14: Heat Exchangers Temperatures and Duties

Heat Inlet Outlet Duty

Table 15: Pumps and Valves Pressures and Duties

Pumps and Inlet Outlet Duty

Figure 19: Choice of Thermodynamic Model

8.2 Distillation Columns

Figure 20: Short Cut Distillation

Table 16: Shortcut Distillation Column Specifications and Results

8.3 Membrane Separation