Department of Chemical and Biomolecular Engineering

NATIONAL UNIVERSITY OF SINGAPORE

CN4123R: FINAL YEAR DESIGN PROJECT

AY2019/2020 Semester 2
SECTION REPORT
SECTION 2 – MF SEPARATION

Name: Rachel Chow Pei Ting

Date: March 2020

1
EXECUTIVE SUMMARY

The aim of this project is to design a plant based in Singapore to produce Formic Acid (FA)
through the hydrolysis of Methyl Formate (MF). MF is produced on-site, via the
carbonylation of the starting materials, methanol (MeOH) and Carbon Monoxide (CO). The
focus of this report would be on the separation of MF. This section receives its feed from the
liquid outlet of the upstream carbonylation reactor, and this feed contains mainly CO, MF,
MeOH and water. The MF separation process is simulated on ASPEN HYSYS.

The fluid package chosen for this simulation is HYSYS UNIQUAC (liquid) – PR (vapour),
as it has the lowest root mean square error when each fluid package model was compared
with the experimental data.

The operating parameters in the distillation column were optimized with the objective of
minimizing the Total Annualized Cost (TAC), while taking into account the actual tray
efficiency of 53.8%. The TAC was calculated using the Aspen Process Economic Analyzer
(APEA) tool in HYSYS. The optimized simulation was found to have a TAC of $10,229,999
per year.
The finalized distillation column has 45 stages, with the feed stage at stage 26. The column
has a diameter of 2.78m and a height of 36.5m. The column internals were designed with
justifications, following the guidelines available in literature.

Using values extracted from HYSYS, and together with manual calculations, hydraulic
checks were conducted to ensure that every tray is operating satisfactorily. In addition, a
continuous evaporative crystallizer was proposed to recover the sodium methoxide catalyst,
which will be dissolved in MeOH.
The final design of this section is presented in the Process Flow Diagram (PFD) found in
Appendix A. The auxiliary units in this section were also designed and the equipment
specification sheet can be found in Appendix B.

2
Contents
EXECUTIVE SUMMARY .................................................................................................................. 2
1. INTRODUCTION ............................................................................................................................ 5
1.1 Problem statement ....................................................................................................................... 5
1.2 Purpose of MF separation ........................................................................................................... 5
1.3 Background ................................................................................................................................. 5
2. DESIGN BASIS ............................................................................................................................... 5
2.1 Section 2’s scope......................................................................................................................... 5
2.2 Separation method ...................................................................................................................... 6
2.3 Selection of Fluid Package .......................................................................................................... 6
3. COLUMN DESIGN ......................................................................................................................... 8
3.1 Type of column – Tray vs Packed ............................................................................................... 8
3.2 Type of trays ............................................................................................................................... 8
3.3 Optimization of distillation column ............................................................................................ 8
3.3.2 Optimization Methodology .................................................................................................. 9
3.3.3 Condenser pressure .............................................................................................................. 9
3.3.4 Number of stages with optimal feed stage ............................................................................ 9
3.3.5 Feed temperature ................................................................................................................ 10
3.4 Results of preliminary optimization .......................................................................................... 10
3.5 Accounting for tray efficiency & determining actual number of stages .................................... 11
3.6 Final optimization results .......................................................................................................... 11
4. DISTILLATION COLUMN SPECIFICATIONS ........................................................................... 11
4.1 Column diameter ....................................................................................................................... 11
4.2 Column internals design ............................................................................................................ 12
5. COLUMN HYDRAULICS............................................................................................................. 13
5.1 Weep point ................................................................................................................................ 13
5.2 Entrainment ............................................................................................................................... 14
5.3 Tray pressure drop (∆P) ............................................................................................................ 15
5.4 Downcomer liquid backup ........................................................................................................ 16
5.5 Downcomer residence time ....................................................................................................... 16
5.6 Final Hydraulic Plot .................................................................................................................. 17
5.7 Turndown ratio ......................................................................................................................... 18
6. DISTILLATION COLUMN DESIGN ........................................................................................... 18
7. DESIGN OF AUXILIARY UNITS ................................................................................................ 19
7.1 Reboiler .................................................................................................................................... 19
7.1.1 Reboiler type ...................................................................................................................... 19

3
 7.1.2 Reboiler design .................................................................................................................. 19
7.2 Condenser ................................................................................................................................. 20
7.2.1 Condenser type................................................................................................................... 20
7.2.2 Condenser design ............................................................................................................... 20
7.3 Reflux drum design ................................................................................................................... 20
7.4 Reflux pump design .................................................................................................................. 20
7.5 Cooler (E-200) design ............................................................................................................... 21
7.6 Flash separator (V-200) design ................................................................................................. 21
7.7 Feed pump (P-200) design ........................................................................................................ 22
7.8 Compressor (K-200) design ...................................................................................................... 22
7.9 Compressor (K-201) design ...................................................................................................... 22
8. SEPARATING SODIUM METHOXIDE CATALYST FROM MeOH ......................................... 23
8.1 Sodium Methoxide .................................................................................................................... 23
8.2 Proposed solution to separate sodium methoxide catalyst from MeOH: Crystallization ........... 23
9. LIMITATIONS, ASSUMPTIONS AND OTHER POSSIBLE IMPLEMENTATIONS ................ 24
9.1 Optimization ............................................................................................................................. 24
9.2 Possible contamination of MeOH ............................................................................................. 24
9.3 Crystallization of sodium methoxide from MeOH .................................................................... 24
10. REFERENCES ............................................................................................................................. 25
Appendix A: PFD and Stream table .................................................................................................... 26
Appendix B: Equipment specification sheet ....................................................................................... 28

4
1. INTRODUCTION
1.1 Problem statement
The aim of this project is to design a plant to produce 70 ktonnes of Formic Acid (FA) per
annum with an operating time of 8000 hours per annum and a product purity of 95.0 wt%
This sectional report (Section 2) aims to design a separation column to obtain MF with a
purity of 96 wt%. MF acts as an intermediate in the formation of FA, for the further
downstream hydrolysis reaction.
1.2 Purpose of MF separation
This section receives its feed from the liquid outlet of the upstream carbonylation reactor, and
this feed mainly contains CO, MF, MeOH and water. It is essential that the MF is separated
out before entering the hydrolysis reactor as the second reaction (hydrolysis) is reversible
(Rangaiah et al., 2019). If the MF is not separated and the products from carbonylation
reactor directly enters the hydrolysis reactor, the high concentration of MeOH will cause the
hydrolysis reaction to be pushed backwards. Furthermore, without the separation of MF, the
high amount of MeOH will likely require a larger reactor for the hydrolysis reaction.
1.3 Background
Methyl formate (MF), also known as methyl methanoate, is a formate ester and has a
molecular formula of HCOOCH3. It appears as a clear, colourless and flammable liquid with
a boiling point of 31.5℃. (ChEBI, 2017). Industrially, MF is produced by the base-catalyzed
carbonylation reaction of MeOH and CO and it is used as a refrigerant, fumigant, insecticide,
blowing agent for foams as well as in the manufacture of other formic acid derivatives (ACS,
2014).

2. DESIGN BASIS
2.1 Section 2’s scope
After a reduction in pressure, the outlet liquid stream of the carbonylation reactor will be the
start of Section 2’s scope. The components in this stream includes: CO, MF, MeOH (with
sodium methoxide dissolved), H2O and a small amount of FA and N2.

The amount of N2 in this stream is very little, which is probably due to the fact that N2 is an
inert gas, which is unlikely to have affinity with any of the compounds present. The amount
of CO, however, is noticed to be higher as it is likely to have solubility in MeOH or MF. The
small amount of FA present is from a recycling stream that is coming from Section 5 and
enters the carbonylation reactor, which subsequently enters Section 2.

The boiling points of the main components are shown in Table 1 below.
Table 1: Boiling points of the main components in Section 2’s feed stream

Component Boiling point K values Relative volatility

CO -192℃ 2181.7
MF 32℃ Light key 4.1535 𝐾𝑀𝐹
≈7
MeOH 63℃ Heavy key 0.59608 𝐾𝑚𝑒𝑡ℎ𝑎𝑛𝑜𝑙
H2O 100℃ 0.21154

5
The outlet liquid stream from the carbonylation reactor will pass through a flash separator to
purge out some of the CO and N2. This vapour stream containing CO and N2 will eventually
be recycled back into the carbonylation reactor.
2.2 Separation method
To separate MF, distillation is the selected method of separation as the components have
different boiling points. Distillation is by far the most economical separation method and it is
widely used in many industries, especially in large-scale applications.
The liquid output stream of the flash separator will be the inlet feed to the distillation column.
A distillation column with a partial condenser will be used due to the presence of the non-
condensable CO gas.
The design of this MF separation distillation column will be the focus of this report.

2.3 Selection of Fluid Package

It is essential to choose a suitable fluid package for the separation process as it will impact
the accuracy of the HYSYS simulation obtained.
In the distillation column, the key components present are CO, MF, MeOH and water. The
light key is MF and the heavy key is MeOH. This is determined based on their respective
boiling points. Since the separation in a distillation process is heavily dependent on the light
and heavy key, the VLE experimental data for the MF – MeOH binary system was obtained
and evaluated with the fluid packages. The VLE experimental data was obtained using
ASPEN PLUS and plotted using Excel, as shown in Figure 1 below. The VLE plot expresses
the bubble-point and the dew-point of the equilibrium binary mixture at constant pressure.

MF-MeOH VLE data

340
335
330
Temp. (K)

325
320
315
310
305
300
0 0.2 0.4 0.6 0.8 1
x methyl formate
bubble dew

Figure 1: MF-MeOH VLE data

Note that according to Figure 1 above, it can be observed that the system exhibits some form
of azeotropic behaviour (i.e. same compositions in vapour and liquid phase) as the mole
fraction of MF reaches 1. Ideally, a very high purity of MF (~99.9 wt%) is desired at the
distillate of the MF separation column which will be favourable for the downstream
hydrolysis reaction. However, the presence of this azeotrope limits the MF purity that can be
6
obtained in a single distillation column. Hence, a MF purity of 96 wt% is targeted instead
when designing the MF separation column.

All of the components are polar and soluble in one another. The total pressure of the system
is not expected to exceed 10 bars. Based on the procedure to select fluid package suggested
by Carlson (1996), the suitable shortlisted activity coefficient fluid packages for this system
are NRTL and UNIQUAC.
The root mean square error (RMSE) between different fluid packages and the experimental
data is calculated as shown in Table 2 below.
Table 2: RMSE of different fluid packages

HYSYS fluid package RMSE ComThermo fluid package RMSE

NRTL-Ideal 0.619 PR-NRTL 0.887
UNIQUAC-Ideal 0.600 PR-UNIQUAC 0.772
UNIQUAC-PR 0.454 RK-UNIQUAC 0.772
UNIQUAC-RK 0.506
UNIQUAC-SRK 0.458

UNIQUAC-PR is the chosen fluid package used in this section as it has the lowest RMSE,
hence the most desirable as it exhibits behaviour which is closer to the experimental data.
Figure 2 below shows the VLE data of UNIQUAC-PR in comparison with experimental VLE
data.

TXY plot
340 y = -0.5773x6 + 167.11x5 - 311.59x4 + 190.1x3 - 59.09x2 - 18.338x +
335 337.55
330 R² = 0.9999
Temp. (K)

325
320
315
310
305
y = 316.24x6 - 1137.6x5 + 1687.5x4 - 1350x3 + 641.07x2 - 189.82x + 337.53
300 R² = 1
0 0.2 0.4 0.6 0.8 1
x methyl formate
bubble (liquid) - Expt dew (gas) - Expt
bubble (liquid) - fluid package dew (gas) - fluid package
Poly. (bubble (liquid) - fluid package) Poly. (dew (gas) - fluid package)

Figure 2: VLE data of UNIQUAC-PR fluid package with experimental VLE data obtained

From Figure 2 above, it can be seen that the VLE data of UNIQUAC-PR fluid package fits
very well into the MF-MeOH experimental data. Furthermore, using PR (Peng Robinson)
EOS to fit with the activity coefficient fluid package so as to account for the vapour phase is
suitable as PR is the most enhanced model in Aspen HYSYS and it has the largest
applicability range in terms of temperature and pressure (Aspen Technology, 2006).

7
3. COLUMN DESIGN
3.1 Type of column – Tray vs Packed
There are two main types of distillation columns – Tray and Packed. Table 3 below lists out
briefly the differences between the two types of columns.
Table 3: Conditions favouring Tray and Packed columns (Premkumar, 2020)

Conditions favouring Tray columns: Conditions favouring Packed columns:

- Allow wider range of liquid flow rates - More suitable for low liquid hold-up
- Suitable for large diameters (>0.6m) - Lower pressure drop (important for
- More predictable performance vacuum distillation)
- Dirty service (tray columns easier to clean) - Suitable for smaller diameters (<0.6m)
- Presence of thermal/mechanical stress due to - Corrosive services (wider range of
large temperature changes which may lead materials to choose for the packings such
to cracked packings as plastic, ceramic etc)

As the feed stream to the MF Separation distillation column has a relatively high liquid flow
rate and it is operating at atmospheric pressure without the need to minimize the pressure
drop between the top and bottom of the column, a tray distillation column is used.
The Material of Construction (MOC) used for the column is Stainless Steel 304, which is
more corrosion-resistant. Carbon steel is more likely to corrode and cause MeOH
contamination than stainless steel, especially when MeOH absorbs moisture easily. Though
stainless steel incurs a higher capital cost than carbon steel, it has the advantage of a lower
life cycle maintenance cost and reduced likelihood of MeOH contamination (Methanol
Institute, n.d.).
3.2 Type of trays
The 4 common types of trays are: Sieve tray, bubble cap tray, dual flow tray and valve tray.
Sieve tray will be used in this case as it is the cheapest and simplest to fabricate with low
pressure drop and comparable performance with the other types of trays.
3.3 Optimization of distillation column
3.3.1 Objective of optimization
The optimization objective adopted in my case is to minimize the Total Annualized Cost
(TAC). This is because cost is an extremely important and is a decisive factor in practical
industries. Furthermore, calculating the TAC is convenient with the help of HYSYS tools.
The TAC can be calculated as follows (Towler, 2020):
𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝑐𝑜𝑠𝑡
𝑇𝐴𝐶 = + 𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝑐𝑜𝑠𝑡
𝑃𝑎𝑦𝑏𝑎𝑐𝑘 𝑝𝑒𝑟𝑖𝑜𝑑

A payback period of 3 years will be used and operating cost is the utility cost. Operating
hours per year will be 8000. The capital cost is taken from Aspen Process Economic
Analyzer (APEA) in HYSYS, whereas the operating cost for each utility is calculated as
follows:

8
 $ ℎ𝑟
𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝑐𝑜𝑠𝑡 ($/𝑦𝑒𝑎𝑟) = 𝐶𝑜𝑠𝑡 𝑓𝑎𝑐𝑡𝑜𝑟( ) × 𝑝𝑜𝑤𝑒𝑟(𝑘𝑊) × 8000( )
𝑘𝑊 × ℎ𝑟 𝑦𝑒𝑎𝑟
Note that the TAC calculated for optimization here is not final and will only be used for
optimization purpose to study the trend of TAC with respect to the variables to be optimized.
The specifications of the distillation column in the base case (in the interim report) is used as
a starting point for optimization.

3.3.2 Optimization Methodology

3.3.2.1 Variables to optimize
After conducting sensitivity analysis to determine the most sensitive variable (i.e. a certain %
change of the variable causes the greatest % change in TAC), the following cost-based
variables will be optimized in this order and reiterated to obtain the most optimized values:

1. Condenser pressure
2. Number of stages together with optimal feed stage and reflux ratio
3. Feed temperature
3.3.2.2 Separation targets
The MF purity at the distillate is targeted at 96 wt%, as further increase in purity will lead to
problems in converging the column, which is likely attributed due to the azeotrope exhibited
by the MF and MeOH mixture, as explained in Section 2.3. The MF recovery is targeted at
90%.

3.3.3 Condenser pressure

Condenser pressure is the most sensitive variable, hence it will be the first variable to be
optimized. Furthermore, condenser pressure is considered as the most important control
variable in a distillation column (Kister, 1992), which can affect the temperatures of
vaporization, volatilities and almost all the processes that occur in the column.
Reboiler pressure can be calculated once the condenser pressure is fixed, using an estimation
of 0.1 psi (0.689 kPa) per tray (Seider et al., 2010) and 10 kPa of pressure drop is assumed
across the condenser and reboiler each.
𝑃𝑟𝑒𝑏𝑜𝑖𝑙𝑒𝑟 = 𝑃𝑐𝑜𝑛𝑑𝑒𝑛𝑠𝑒𝑟 [𝑘𝑃𝑎] + 𝑁𝑜. 𝑜𝑓 𝑡𝑟𝑎𝑦𝑠 × ∆𝑃𝑝𝑒𝑟 𝑡𝑟𝑎𝑦 + ∆𝑃𝑐𝑜𝑛𝑑𝑒𝑛𝑠𝑒𝑟 + ∆𝑃𝑟𝑒𝑏𝑜𝑖𝑙𝑒𝑟
𝑃𝑟𝑒𝑏𝑜𝑖𝑙𝑒𝑟 [𝑘𝑃𝑎] = 𝑃𝑐𝑜𝑛𝑑𝑒𝑛𝑠𝑒𝑟 [𝑘𝑃𝑎] + 𝑁𝑜. 𝑜𝑓 𝑡𝑟𝑎𝑦𝑠 × 0.689 + 10 + 10

According to Seider et al. (2010), the condenser pressure should be set at 20 to 30 psia (138
to 207 kPa) to avoid vacuum operation.
3.3.4 Number of stages with optimal feed stage
The next variable to be optimized is the number of stages with the optimal feed stage. The
increase in number of trays will help in the separation process and produces a lower reboiler
duty. However, an increase in number of trays causes a higher pressure drop across the
column due to the pressure drop associated in each tray, leading to a higher pressure and
temperature in the reboiler. For every number of trays, the feed stage will be optimized to
give the lowest reflux ratio, which will minimize the cost.

9
3.3.5 Feed temperature
The next variable to be optimized is the feed temperature. The feed temperature is important
in minimizing cost of heating/cooling. It affects the amount of flashing of the feed when the
feed enters the distillation column. For example, when the feed is a cold liquid, more reboiler
duty will be needed to heat it up in the column.

It was realized that although preheating the feed will result in lower reboiler duty, additional
capital and utility cost is required to preheat the feed. Both the preheater and the reboiler uses
low pressure steam as the heating medium. The preheater is eventually removed as it would
result in an overall lower TAC. The feed will enter the column as a subcooled liquid.
3.4 Results of preliminary optimization

TAC vs Condenser pressure TAC vs Number of stages

13000000 10650000
12500000 10600000

TAC ($/year)
TAC ($/year)

10550000
12000000
10500000
11500000
10450000
11000000 10400000
10500000 10350000
10000000 10300000
138 140 145 150 151 152 154 155 160 180 200 14 20 22 23 24 25 30
Condenser Pressure (kPa) Number of stages
Figure 3: Graph of TAC vs Condenser pressure Figure 4: Graph of TAC vs Number of stages
Optimal Condenser pressure: 152 kPa Optimal no. of stages: 22, feed stage: 16

TAC vs Feed temperature

11200000
11000000
TAC ($/year)

10800000
10600000
10400000
10200000
10000000
9800000
20.05 30 40 50 52 54 64 67
Feed temperature (℃)
Figure 5: Graph of TAC vs Feed temperature
Optimal feed temperature: 20.05℃

Table 4 below shows a summary of the results obtained after the first round of optimization.
Here, the tray efficiency is not yet accounted for.
Table 4: Summary of optimized variables after first round of iteration, as compared to the base case variables

Variable Base case values Optimized values

(after first iteration)
Condenser pressure 138 kPa 152 kPa
Number of stages & feed stage 14 stages 22 stages
feed stage = 12 feed stage = 16

10
 Reflux ratio 3.9 1.1
Feed temperature 67℃ 20℃
TAC ($/year) 13,054,252 10,283,782

3.5 Accounting for tray efficiency & determining actual number of stages
HYSYS simulation assumes that the distillation column is operating at 100% efficiency,
which is not true in reality. The overall column efficiency can be calculated from O’Connell
correlation (Towler, 2020):
𝑇𝑟𝑎𝑦 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑐𝑦 (𝐸𝑜 ) = 51 − 32.5 lg(𝜇𝑎 𝛼𝑎 )

𝜇𝑎 : Molar average liquid viscosity (mNs/m2)

𝛼𝑎 : Average relative volatility of the light key

𝐾𝑀𝐹 𝑡𝑜𝑝 𝐾𝑀𝐹 𝑏𝑜𝑡𝑡𝑜𝑚

𝛼𝑀𝐹−𝑚𝑒𝑡ℎ𝑎𝑛𝑜𝑙 = √𝛼𝑡𝑜𝑝 × 𝛼𝑏𝑜𝑡𝑡𝑜𝑚 = √ ×
𝐾𝑚𝑒𝑡ℎ𝑎𝑛𝑜𝑙 𝑡𝑜𝑝 𝐾𝑚𝑒𝑡ℎ𝑎𝑛𝑜𝑙 𝑏𝑜𝑡𝑡𝑜𝑚

1.019 5.529
=√ × = 2.8711
0.8044 0.8497

𝜇𝑙𝑖𝑞𝑢𝑖𝑑 = √𝜇𝑡𝑜𝑝 × 𝜇𝑏𝑜𝑡𝑡𝑜𝑚 = √0.2789 × 0.2938 = 0.2863

∴ 𝐸𝑜 = 51 − 32.5 lg(𝜇𝑙𝑖𝑞𝑢𝑖𝑑 𝛼𝑀𝐹−𝑚𝑒𝑡ℎ𝑎𝑛𝑜𝑙 ) = 51 − 32.5 lg(0.2863 × 2.8711) = 53.77%

Considering a 10% overdesign allowance,

110
𝐴𝑐𝑡𝑢𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑠𝑡𝑎𝑔𝑒𝑠 = × 22 = 𝟒𝟓
53.77
3.6 Final optimization results
Table 5 below summarizes the final optimized variables after the tray efficiency were
inputted into HYSYS and accounted for. Several rounds of iterations and adjustments were
conducted with the help of HYSYS Optimizer tool. The TAC calculated here is not final.
Table 5: Summary of final optimization results

Variable Final optimized values

(after tray efficiency accounted for)
Condenser pressure 152 kPa
Number of stages & feed stage 45 stages, feed stage = 26
Reflux ratio 1.2
Feed temperature 20℃
TAC ($/year) 10,229,999

4. DISTILLATION COLUMN SPECIFICATIONS

4.1 Column diameter
The diameter of the distillation column is primarily dependent on the vapour flow rate. It is
an essential parameter to ensure that the column will not flood. For instance, when the vapour

11
flow rate in the column is high, designing a relatively small column diameter will cause
excessive entrainment to occur, leading to poor separation.

It can be obtained from HYSYS. The column diameter is increased until the operating point
falls below the maximum entrainment. A column diameter of 2.783 m is used to converge the
HYSYS internals with no errors. Table 6 below describes how the column diameter can be
obtained.
Table 6: Formula to calculate column diameter (Seider et al., 2010)

Parameters Description
𝐹𝐿𝐺 : Flow ratio parameter 𝐿 𝜌𝐺
𝐿: Liquid mass flow rate 𝐹𝐿𝐺 = √
𝐺 𝜌𝐿
𝐺: Vapour mass flow rate
𝐶𝑠𝑏 : Flooding parameter Can be taken from Figure 19.4 in Seider et al. (2010), after knowing
the values of 𝐹𝐿𝐺 and tray spacing
𝐹𝑆𝑇 : Surface tension factor 𝜎 0.20
𝐹𝑆𝑇 = ( )
𝜎: Surface tension (dyne/cm) 20
𝐹𝐹 : Foaming factor 𝐹𝐹 = 1, for non-foaming system
𝐹𝐻𝐴 : Hole factor area For sieve trays,
𝐴ℎ 𝑇𝑜𝑡𝑎𝑙 ℎ𝑜𝑙𝑒 𝑎𝑟𝑒𝑎 𝑜𝑛 𝑡𝑟𝑎𝑦 𝐴 𝐴ℎ
: 𝐹𝐻𝐴 = 5 ( ℎ) + 0.5 for 0.06 ≤ ≤ 0.10
𝐴𝑎 𝐴𝑐𝑡𝑖𝑣𝑒 𝑎𝑟𝑒𝑎 𝑜𝑓 𝑡𝑟𝑎𝑦 𝐴𝑎 𝐴𝑎
𝐴ℎ
𝐹𝐻𝐴 = 1 for ≥ 0.10
𝐴𝑎
C: Empirical capacity parameter 𝐶 = 𝐶𝑠𝑏 𝐹𝑆𝑇 𝐹𝐹 𝐹𝐻𝐴
𝑈𝑓 : Flooding vapour velocity 𝜌𝐿 − 𝜌𝑉
𝑈𝑓 = 𝐶 √
𝜌𝑉
𝐴𝑑
:
𝐷𝑜𝑤𝑛𝑐𝑜𝑚𝑒𝑟 𝑎𝑟𝑒𝑎 0.1 𝐹𝐿𝐺 ≤ 1.0
𝐴𝑇 𝐶𝑟𝑜𝑠𝑠 𝑠𝑒𝑐𝑡𝑖𝑜𝑛𝑎𝑙 𝑎𝑟𝑒𝑎 𝐴𝑑 (𝐹 − 0.1)
= {0.1 + 𝐿𝐺 0.1 ≤ 𝐹𝐿𝐺 ≤ 1.0 }
𝐴𝑇 9
0.2 𝐹𝐿𝐺 ≥ 1.0
𝐷𝑇 : Tower inside diameter (m) 4𝐺
𝐷𝑇 =
𝑓: % flooding at which column 𝐴
(𝑓𝑈𝑓 )𝜋(1 − 𝑑 )𝜌𝐺
𝐴𝑇
operates (typically 0.75 to 0.85)

4.2 Column internals design

Table 7 below describes the column internals design.
Table 7: Column internals design

Column internals Description

Tray spacing Tray spacing are usually available in 18, 24 and 36 inches.
Trays 1 to 25  24 inches.
Trays 26 to 45  36 inches. (Tray spacing is increased from the HYSYS default of 24
inches to solve the problem of entrainment)
Number of tray The number of tray passes is determined by the liquid flow rate and the column
passes diameter. Single pass will be used for this column.
Tray material Stainless steel 304
Tray thickness Gauge 10 (3.4 mm), as set by HYSYS. The typical tray thickness used is 3 mm for
stainless steel (Towler, 2020).
Type of tray Sieve trays. It is the cheapest with comparable performance and lowest pressure drop.
Hole diameter For non-fouling system, 5 mm is recommended (Towler, 2020).

12
 Weir height The weir height determines the volume of liquid on a tray and is essential in
determining column efficiency. As weir height increases, the column efficiency
increases but pressure drop per tray also increases.
For columns operating above atmospheric pressure, the weir height will usually be
between 40 and 90 mm. (Towler, 2020). The default weir height used in HYSYS is
50.8 mm (Stages 1 to 25) and 76.2 mm (Stages 26 to 45).
Downcomer In Stages 1 to 25, the default downcomer clearance in HYSYS is 38.1mm.
clearance It is acceptable as according to KLM (2011), the recommended downcomer clearance
is 1.5 in (38.1mm) and a value less than 1 inch is not recommended.
In Stages 26 to 45, the liquid load is higher and the default downcomer clearance in
HYSYS is increased to 63.5mm to minimize head loss.
Weir length Towler (2020) recommends a side weir length of 77% of column diameter, which is
equivalent to 2.14 m.
Downcomer width In my simulation. the default downcomer width in HYSYS is around 502 mm.
Hole area to Active Stages 1 to 25: 0.1
area Stages 26 to 45: 0.075
Recommended value of hole area to active area is 0.5 – 0.15 (Separation Technology,
2012).

5. COLUMN HYDRAULICS
In this section, the weep point, entrainment, tray pressure drop, downcomer liquid backup
and downcomer residence time will be discussed. The purpose of this section is to ensure that
the column is operating satisfactorily and every tray is in an acceptable mass transfer regime.
5.1 Weep point
Weeping occurs when the vapour flow in the column is insufficient to maintain a liquid level
on the tray (Towler, 2020). Weep point refers to the lower limit when liquid leakage through
the tray holes become excessive. Excessive weeping can lead to dumping, whereby the liquid
on all trays will crash to the base of the column (Costello, 2019). At the weep point, vapour
velocity is at its minimum. The weep points for each stage as well as the actual vapour
velocity in the column are calculated, as described in Table 8 below.
Table 8: Description of weep point calculations (Towler, 2020)

Parameters Description
2
ℎ𝑜𝑤 : height of liquid crest over weir (m) 𝐿𝑤 3
𝑙𝑤 : weir length (m) ℎ𝑜𝑤 = 750 [ ]
𝜌𝑙 𝑙𝑤
𝐿𝑤 : Liquid mass flow rate (kg/s)
𝜌𝑙 : Density of liquid (kg/m3)
Constant 𝐾2 , which can be found after calculating 𝐾2 can be found from Figure 11.37 in (Towler, 2020),
the value of ℎ𝑤 + ℎ𝑜𝑤 after the value of ℎ𝑤 + ℎ𝑜𝑤 is calculated.
𝑢ℎ : Minimum design vapour velocity (m/s) [𝐾2 − 0.90(25.4 − 𝑑ℎ )]
𝑢ℎ =
𝑑ℎ : Hole diameter (mm) (𝜌𝑣 )0.5
When 𝑢𝑎𝑐𝑡𝑢𝑎𝑙 > 𝑢ℎ for all stages, no weeping occurs.

The results obtained are shown in Figure 6 below.

13
 Vapour velocity vs Stages
22
20
Vapour velocity (m/s) 18
16
14
12
10
8
6
4
1 5 9 13 17 21 25 29 33 37 41 45
Stages
Minimum design vapour velocity (weeping velocity)
Actual vapour velocity in column

Figure 6: Comparison of the minimum design vapour velocity required and the actual vapour velocity in column

Since the weeping velocity is lower than the minimum design velocity for all stages, no
weeping occurs in the column under normal operating conditions.

5.2 Entrainment
Entrainment occurs when the vapour flow rate is too high, causing the liquid to be carried up
in the by the vapour flow, which would reduce the actual operating efficiency. The tray
spacing for stages 26 to 45 was increased to 36 inches from the original 24 inches to solve the
problem of entrainment. The calculation steps to obtain the fractional entrainment (𝜑 ) are
shown in Table 9 below.
Table 9: Description of entrainment calculations (Towler, 2020)

Parameters Description
𝐹𝐿𝑉 : Liquid-vapour flow factor 𝐿𝑤 𝜌𝑉
𝐿𝑤 : Liquid mass flow rate (kg/s) 𝐹𝐿𝑉 = √
𝑉𝑤 𝜌𝐿
𝑉𝑤 : Vapour mass flow rate (kg/s)
𝐾1 correction K1 can be obtained from Figure 11.34 in Towler (2020)
𝑢𝑓 : Flooding vapour velocity (m/s) 𝜌𝐿 − 𝜌𝑉
𝑢𝑓 = 𝐾1 √
𝜌𝑉
% flooding 𝑢𝑛 (𝑎𝑐𝑡𝑢𝑎𝑙 𝑣𝑒𝑙𝑜𝑐𝑖𝑡𝑦 𝑏𝑎𝑠𝑒𝑑 𝑜𝑛 𝑛𝑒𝑡 𝑎𝑟𝑒𝑎)
% flooding = 𝑢𝑓

𝜑: fractional entrainment (kg/kg gross After % flooding and 𝐹𝐿𝑉 are known, 𝜑 can be obtained from
liquid flow) Figure 11.36 in Towler (2020).

A summary of the results obtained is shown in Table 10 below.

Table 10: Summary of fractional entrainment 𝜑 obtained at Stages 1, 26 and 45.

Stage 1 26 45
Fractional entrainment 𝝋 0.0077 0.0037 0.0041

14
As the fractional entrainment in each stage is less than 0.1, the entrainment effects on column
efficiency is small.

5.3 Tray pressure drop (∆P)

HYSYS assumes the same ∆P for each tray, which gives rise to a linear ∆P profile across the
column. Initially, the ∆P used for each tray is 0.1 psi (0.689 kPa). However, the actual ∆P
may not be the same for every tray. Hence, the ∆P per tray calculated by HYSYS internals is
extracted and inputted into the column parameters. Figure 7 below shows the ∆P per tray
extracted from HYSYS, as compared to the original ∆P of 0.1 psi per tray.

Graph of pressure drop per tray vs stage

1.8
Pressure drop per tray (kPa)

1.6
1.4
1.2
1
0.8
0.6
0.4
1 5 9 13 17 21 25 29 33 37 41 45
Stage
Initial pressure drop per tray
Actual pressure drop per tray calculated by HYSYS

Figure 7: Comparison of the initial ∆P per tray of 0.1 psi (0.689 kPa) and the values of ∆P per tray calculated
by HYSYS

After inputting the actual ∆P per tray, the new pressure for the reboiler is calculated to be
217.5 kPa, which is higher as compared to the original reboiler pressure (203 kPa). Hence, a
higher reboiler duty will be required. Table 11 below describes how the ∆P per tray can be
calculated.
Table 11: Calculations for ∆P per tray (Towler, 2020)

Parameters Description
𝐴ℎ 2
: Fraction of hole area over perforated area 𝐴ℎ 𝑑ℎ
𝐴𝑝 = 0.9 [ ]
𝐴𝑝 𝑙𝑝
𝑑ℎ : Hole diameter
𝑙𝑝 : Hole pitch (distance between hole
centres)
𝐶𝑜 : Orifice coefficient Value of 𝐶𝑜 can be taken from Figure 11.49 in Towler (2020),
𝑝𝑙𝑎𝑡𝑒 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝐴
after the values of and 𝐴ℎ × 100 are known
ℎ𝑜𝑙𝑒 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟 𝑝

ℎ𝑟 : Residual head 12.5 × 103

ℎ𝑟 =
𝜌𝐿
ℎ𝑑 : Dry tray pressure drop (mm liquid) 𝑢ℎ 2 𝜌𝑉
ℎ𝑑 = 51 [ ]
𝑢ℎ : Min. vapour design velocity (m/s) 𝐶𝑜 𝜌𝐿
calculated earlier in Section 5.1.

15
 ℎ𝑡 : Total tray pressure drop (mm liquid) ℎ𝑡 = ℎ𝑑 + (ℎ𝑤 + ℎ𝑜𝑤 ) + ℎ𝑟
∆𝑃𝑡 : Total tray pressure drop (Pa) ∆𝑃𝑡 = 9.81 × 10−3 ℎ𝑡 𝜌𝐿

5.4 Downcomer liquid backup

In order to avoid flooding in the column, it is important to ensure that the downcomer area
and tray spacing is appropriate such that the level of liquid in the downcomer is below the top
of the outlet weir on the tray above (Towler, 2020). The downcomer backup (ℎ𝑏 ) is extracted
from HYSYS. Figure 8 below shows the downcomer backup and the value of 0.5(𝑙𝑡 + ℎ𝑤 ).

Graph of Height vs Stage

0.55
0.5
0.45
0.4
Height (m)

0.35
0.3
0.25
0.2
0.15
0.1
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45
Stage
0.5(lt+hw) Downcomer backup (extracted from HYSYS)

Figure 8: Downcomer backup in each stage and the value of 0.5(𝑙𝑡 + ℎ𝑤 ) calculated

Since ℎ𝑏 < 0.5(𝑙𝑡 + ℎ𝑤 ) for all stages, the tray spacing used in the column is acceptable.
Table 12 below describes the calculation steps taken by HYSYS to obtain the downcomer
backup for each stage.
Table 12: Calculations for downcomer backup (Towler, 2020)

Parameters Description
ℎ𝑎𝑝 : Height of bottom edge of apron above the tray (m) ℎ𝑎𝑝 = ℎ𝑤 − (5 𝑡𝑜 10 𝑚𝑚)
2
𝐴𝑎𝑝 : Clearance area under downcomer (m ) 𝐴𝑎𝑝 = ℎ𝑎𝑝 𝑙𝑤
𝑙𝑤 : Weir length (m)
ℎ𝑑𝑐 : Downcomer head loss 𝐿𝑤𝑑
ℎ𝑑𝑐 = 166 [ ]
𝐿𝑤𝑑 : Liquid flow rate in downcomer (kg/s) 𝜌𝐿 𝐴𝑚
𝐴𝑚 : Downcomer area (𝐴𝑑 ) or clearance area under downcomer (𝐴𝑎𝑝 ),
whichever is smaller.
ℎ𝑏 : Downcomer backup (mm) ℎ𝑏 = (ℎ𝑤 + ℎ𝑜𝑤 ) + ℎ𝑡 + ℎ𝑑𝑐
If ℎ𝑏 < 0.5(𝑙𝑡 + ℎ𝑤 ) for all stages, the tray spacing is suitable to avoid flooding. (𝑙𝑡 : tray spacing)

5.5 Downcomer residence time

In order to prevent liquid from being carried under the downcomer, a residence time of at
least 3 seconds is recommended in the downcomer for the entrained vapour to disengage

16
from the liquid stream (Towler, 2020). The downcomer residence time for each stage is
extracted from HYSYS and shown in Figure 9 below.

Graph of downcomer residence time vs stage

11
Downcomer residence time (s)

10
9
8
7
6
5
4
3
2
1
1 5 9 13 17 21 25 29 33 37 41 45
Stage
Minimum recommended downcomer residence time
Downcomer residence time in column (extracted from HYSYS)

Figure 9: Comparison of downcomer residence time and the minimum recommended downcomer residence time

Figure 9 above illustrates that the downcomer residence time for each stage is at least 3
seconds. Therefore, the downcomer residence time in this column is sufficient for adequate
vapour disengagement.
The downcomer residence time can be calculated, as shown in Table 13 below.
Table 13: Calculation of downcomer residence time (Towler, 2020)

Parameter Description
𝑡𝑟 : Downcomer residence time (s) 𝐴𝑑 ℎ𝑏 𝜌𝐿
𝑡𝑟 =
𝐿𝑤𝑑
5.6 Final Hydraulic Plot

Figure x: Hydraulic Plot

Figure 10: Hydraulic Plot

The final hydraulic plot obtained from HYSYS is shown in Figure 10 above. For all stages,
the operating point is within the allowable area and there are no problems such as weeping or

17
flooding occurring in the distillation column. The hydraulic checks in Section 5.1 to 5.5 also
proves that the column is operating satisfactorily.

5.7 Turndown ratio

The turndown ratio is defined as the maximum possible vapour load (below flooding limit) to
the minimum possible vapour load (above weeping limit) the column can handle before
undesirable effects such as flooding or weeping occurs. The normal operating throughput is a
safe margin away from the relevant flooding and weeping limit. A sufficient turndown ratio
is desired to ensure flexibility in operation as the column will be able to handle a wide range
of vapour flow rates (KLM, 2011).
Table 14: Minimum and maximum possible feed flow rates

Minimum feed flow rate (kg/h) 8.216 × 104 Turndown % = ~39%

Current feed flow rate (kg/h) 1.355 × 105
Maximum feed flow rate (kg/h) 1.489 × 105 Turnup % = ~10%

Table 14 above shows the minimum and maximum feed flow rates to the column. From
Table 14, the turndown ratio is calculated to be 1.81, which is deemed to be satisfactory.
When the feed flow rate is increased to beyond 11% of the original feed, flooding occurs in
the column as the column diameter is not sized big enough to handle the increase in vapour
flow rate. When the feed flow rate is decreased to beyond 39%, weeping occurs due to the
decrease in vapour flow available in the column.

One way to improve the turndown performance is to change the type of tray, such as to
bubble cap tray which will have a higher cost. Another way is to change the reflux ratio.
When flooding occurs due to an increase in flow rate, the reflux ratio can be reduced to
reduce the amount of liquid in column. Similarly, when weeping occurs due to a decrease in
flow rate, the reflux ratio can be increased to increase the liquid flow rate in the column and
the amount of vapour produced by the reboiler.

6. DISTILLATION COLUMN DESIGN

Table 15 below shows the specifications for the distillation column.
Table 15: Specifications for distillation column

Specifications Description
Column diameter 2.783 m
MOC Stainless Steel 304 will be used, due to reduced likelihood of corrosion by MeOH
and a lower life cycle maintenance. (Methanol Institute, n.d.).
Column height Height of column is dependent on number of trays/stages. An additional allowance
of 1.2m is added to the top for vapour disengagement and 1.8m is added at the
bottom for liquid to collect at bottom (KLM, 2011).

𝐶𝑜𝑙𝑢𝑚𝑛 ℎ𝑒𝑖𝑔ℎ𝑡 = 𝑁𝑜. 𝑜𝑓 𝑠𝑡𝑎𝑔𝑒𝑠 × 𝑡𝑟𝑎𝑦 𝑠𝑝𝑎𝑐𝑖𝑛𝑔 + 1.2 + 1.8

= 25 × 0.6096 + 20 × 0.9144 + 1.2 + 1.8
= 36.528 𝑚
A column height of less than 53m is preferred due to wind load and foundation
considerations (KLM, 2011). Hence, a column height of 36.528m is acceptable.

18
 Length to Diameter 𝐿 36.528
𝑟𝑎𝑡𝑖𝑜 = = 13.1
ratio (L/D ratio) of 𝐷 2.783
column The length to diameter ratio of column should not be more than 30 and preferably
less than 20 (KLM, 2011). Hence, an L/D ratio of 13.1 is acceptable.
Manhole diameter Manhole diameter used is typically 0.6 m (Towler, 2020). No additional height is
added for manhole spacing, since the minimum tray spacing in the column is
0.6096m, which is sufficient for easy access to maintenance.

7. DESIGN OF AUXILIARY UNITS

7.1 Reboiler
7.1.1 Reboiler type
There are 3 main types of reboilers: kettle reboiler, horizontal thermosiphon reboiler and
vertical thermosiphon reboiler. Table 16 below briefly compares these 3 types of reboiler.
Table 16: Comparison between kettle and thermosiphon reboilers (Towler, 2020)

Type of Advantages Disadvantages

reboiler
Kettle - Can be used for wide range of operating - Relatively expensive as a mechanical
reboiler pressure and for vacuum operations pump is needed.
Horizontal - Less prone to fouling than kettle reboiler - Not suitable for high viscosity fluids or
thermosiphon - Lower operating costs because no pump high vacuum operations. Normally not
required. Flow is driven by density difference. used for pressures below 30 kPa.
- Easier to maintain than vertical thermosiphon
Vertical - Less prone to fouling than kettle reboiler. - Not suitable for high viscosity fluids or
thermosiphon - Lower operating costs because no pump high vacuum operations. Normally not
required. Flow is driven by density difference. used for pressures below 30 kPa.
- Additional height is required in order to
mount the reboiler, hence increasing cost
for the additional support structure.

The reboiler chosen is the horizontal thermosiphon reboiler because it is the cheapest,
without the need to raise the column so as to mount the reboiler.
7.1.2 Reboiler design
Low Pressure (LP) steam is used as the heating medium for the reboiler. The heating medium
(LP steam) is in the tube and the process fluid is in the shell. The calculations for the reboiler
design is shown in Table 17 below.
Table 17: Reboiler design

Parameters Calculations
Power (Q) 𝑄 = 1.647 × 107 𝑊
Log mean temperature difference (∆𝑇𝑙𝑚 ) (∆𝑇1 ) − (∆𝑇2 )
∆𝑇𝑙𝑚 = = 66.22℃
∆𝑇
ln (∆𝑇1 )
∆𝑇1: Difference between tube inlet and shell outlet. 2

∆𝑇2 : Difference between tube outlet and shell inlet.

Overall heat transfer coefficient (U) A value of U=1000 W/m2 ℃ is used. (Towler, 2020)
Area of equipment (A) 𝑄
𝐴 = 𝑈×∆𝑇 = 249 m2
𝑙𝑚

19
7.2 Condenser
7.2.1 Condenser type
Shell and tube exchangers is the most common type used in industries. It provides a large
surface area and is generally less expensive. The shell and tube (fixed tubes) will be used for
the condenser.
7.2.2 Condenser design
Chilled water is used as the cooling medium for the condenser. Though the temperature of the
distillate is at 40℃, cooling water is not used as the temperature of the fluid to be cooled by
the condenser is not high enough for the minimum temperature approach to be met. The
cooling medium (chilled water) is in the tube and the process fluid is in the shell. The
calculations for the condenser design is shown in Table 18 below.
Table 18: Condenser design

Parameters Calculations
Power (Q) 𝑄 = 9.228 × 106 𝑊
Log mean temperature difference (∆𝑇𝑙𝑚 ) (∆𝑇1 ) − (∆𝑇2 )
∆𝑇𝑙𝑚 = = 21.51℃
∆𝑇
ln (∆𝑇1 )
∆𝑇1: Difference between tube inlet and shell outlet. 2

∆𝑇2 : Difference between tube outlet and shell inlet.

Overall heat transfer coefficient (U) A value of U=700 W/m2℃ is used. (Towler, 2020)
Area of equipment (A) 𝑄
𝐴 = 𝑈×∆𝑇 = 613 m2
𝑙𝑚

7.3 Reflux drum design

The reflux drum should be sized to give sufficient hold-up time (around 5 to 15 min) for good
control in column operation (Towler, 2020). In this case, the reflux drum will be sized to hold
liquid up to 5 minutes at 50% liquid level. The reflux drum is placed around 10m above the
ground. Table 19 below shows the calculations and formulas for reflux drum design.
Table 19: Reflux drum design

Parameters Calculations
Volumetric flow rate Q (m3/s) 0.011 m3/s
Hold-up time 𝑡ℎ𝑜𝑙𝑑−𝑢𝑝 (s) 5 min = 300 seconds
Volume of reflux drum 𝑉𝑑𝑟𝑢𝑚 (m3) 𝑉𝑑𝑟𝑢𝑚 = 2 × 𝑄 × 𝑡ℎ𝑜𝑙𝑑−𝑢𝑝 = 6.60 𝑚3
Reflux drum dimensions Using a length to diameter ratio of 3:1,
Length = 4.23 m
Diameter = 1.41 m

7.4 Reflux pump design

The reflux pump has to provide sufficient head in order to pump the liquid back into the
distillation column. Table 20 below shows the calculations and formulas for reflux pump
design. The reflux pump is placed at ground level.

20
 Table 20: Reflux pump design

Parameters Calculations
Head ∆𝑃
+ 𝑧 = 26 𝑚
𝜌𝑔
Efficiency 𝜖 Assume 75% efficiency
Pump input power 𝜌𝑄𝑔𝐻
𝑃= = 3455 𝑊
𝜖
Type of pump Centrifugal.
According to Figure 5.10 in Towler (2020), based on the head required and flow rate, a
single-stage centrifugal pump is suitable.
Centrifugal pumps are more compact as compared to positive displacement pumps, and
benefit from having a lower maintenance requirements and costs (Castle pumps, n.d.)

7.5 Cooler (E-200) design

A cooler (E-200) is placed before the flash separator, as after a series of trial and error, it is
found that reducing the temperature of feed into the flash can reduce the amount of MF lost
through the vapour output at the top of the flash, which is desirable. Chilled water is used as
the cooling medium. The cooling medium (chilled water) is in the tube and the process fluid
is in the shell. The calculations for the cooler design is shown in Table 21 below.
Table 21: Cooler (E-200) design

Parameters Calculations
Power (Q) 𝑄 = 7.481 × 106 𝑊
Log mean temperature difference (∆𝑇𝑙𝑚 ) (∆𝑇1 ) − (∆𝑇2 )
∆𝑇𝑙𝑚 = = 17.52℃
∆𝑇
ln (∆𝑇1 )
∆𝑇1: Difference between tube inlet and shell outlet. 2

∆𝑇2 : Difference between tube outlet and shell inlet.

Overall heat transfer coefficient (U) A value of U=750 W/m2℃ is used. (Towler, 2020)
Area of equipment (A) 𝑄
𝐴= = 569 m2
𝑈×∆𝑇𝑙𝑚

7.6 Flash separator (V-200) design

A flash separator was added to remove most of the N2 and CO before the distillation column
as it was found that adding it could prevent the condenser temperature from becoming too
low, hence avoiding the use of refrigerant which is relatively costly.

The flash separator will be sized to hold liquid up to 5 minutes at 50% liquid level. Table 22
below shows the calculations and formulas for flash separator design.
Table 22: Flash separator (V-200) design

Parameters Calculations
Volumetric flow rate Q (m3/s) 4.496 × 10−2 m3/s
Hold-up time 𝑡ℎ𝑜𝑙𝑑−𝑢𝑝 (s) 5 min = 300 seconds
Volume of flash separator 𝑉𝑓𝑙𝑎𝑠ℎ (m3) 𝑉𝑓𝑙𝑎𝑠ℎ = 2 × 𝑄 × 𝑡ℎ𝑜𝑙𝑑−𝑢𝑝 = 26.976 𝑚3
Reflux drum dimensions Using a length to diameter ratio of 4:1,
Length = 8.20 m
Diameter = 2.05 m

21
7.7 Feed pump (P-200) design
Table 23 below shows the calculations and formulas for the feed pump design.
Table 23: Feed pump (P-200) design

Parameters Calculations
Head ∆𝑃
11.35 m (This value is taken from HYSYS, which can be calculated by 𝜌𝑔)
Efficiency 𝜖 Assume 75% efficiency
Pump input power 𝜌𝑄𝑔𝐻
𝑃= = 5585 𝑊
𝜖
Type of pump Centrifugal pump, due to a relatively high flow rate (~162 m3/h), and also
based on Figure 5.10 in Towler (2020).

7.8 Compressor (K-200) design

Compressor K-200 is added into the recycling stream (containing mostly MF) coming from
Section 5 into Section 2. Table 24 below shows the calculations and formulas for the
compressor design.
Table 24: Compressor (K-200) design

Parameters Calculations
Efficiency 𝜖 Assume 75% polytropic efficiency
Compressor input 𝑃 = 6566𝑊
power P This power is extracted from HYSYS. It can be calculated as follows (Towler, 2020):
𝑛−1 𝑛−1
𝑛 𝑃 𝑛 𝑅𝑇 𝑛 𝑃 𝑛
Work done (J/kg) = 𝑃1 𝑉1 𝑛−1 [[𝑃2] − 1] = 𝑍 𝑀 1 𝑛−1 [[𝑃2] − 1]
1 𝑤 1

where: Z=compressibility factor, R=8.314 JK-1 mol-1, 𝑇1 =inlet temperature, 𝑇2 =outlet

temperature, 𝑃1 =inlet pressure, 𝑃2 =outlet pressure, n=depends on design and operation.
Power = Work done (J/kg) × Rate (kg/s)
Type of According to Figure 10.80 in Towler (2020), a reciprocating compressor will be used
compressor for a relatively low feed flow rate of 528.5 m3/h and a discharge pressure of 1.52 bar
Compression ratio 1.337

7.9 Compressor (K-201) design

Table 25 below shows the calculations and formulas for the compressor design.
Table 25: Compressor (K-201) design

Parameters Calculations
Efficiency 𝜖 Assume 75% polytropic efficiency
Compressor input power P 𝑃 = 18440 𝑊
Type of compressor According to Figure 10.80 in Towler (2020), a reciprocating compressor
will be used due to a feed flow rate of 1710 m3/h and a discharge pressure
of 1.52 bar
Compression ratio 1.266

22
8. SEPARATING SODIUM METHOXIDE CATALYST FROM MeOH
8.1 Sodium Methoxide
For simplicity purpose, the catalyst used for the carbonylation reaction (Sodium Methoxide,
CH3ONa) is not simulated in HYSYS. However, it exists in reality and is dissolved in MeOH
in the liquid outlet stream of the carbonylation reactor.

Over time, this catalyst may get deactivated or poisoned and needs to be replaced. It is thus
necessary to separate out this catalyst from MeOH. Also, sodium methoxide reacts violently
with water and poses as a fire hazard upon contact with moisture (ThermoFisher, 2018).
Since MeOH absorbs moisture very easily, the catalyst is likely to ignite if left dissolved in
the MeOH. The deactivated/poisoned catalyst should be stored in dry and closed container
and disposed to an approved waste disposal plant whereas catalyst which can be reused will
be recycled back into the carbonylation reactor.
8.2 Proposed solution to separate sodium methoxide catalyst from MeOH: Crystallization
The catalyst will be separated from MeOH via crystallization. During crystallization, the
solution can be cooled to supersaturation (Cooling crystallization) or evaporated to
supersaturation (Evaporative crystallization).
Cooling crystallization is often used when the solubility of the product in the solvent
decreases significantly with decreasing temperature (SoliQz, n.d), whereas evaporative
crystallization is often used when the solubility of product in solvent increases only slightly
with increasing temperature. In most cases, the application of cooling crystallization is
limited as the decrease in solubility with decreasing temperature is not sufficient to make the
method practical (Seider, et al.,2010). Hence, evaporative crystallizers are more common in
industries.
A continuous evaporative crystallization will be employed. The two common types of
continuous evaporative crystallizers are the draft tube baffle crystallizer and the forced
circulation crystallizer. Table 26 below very briefly states the advantages of these two
crystallizers.
Table 26: Advantages of the draft tube baffle and forced circulation crystallizers (Swenson Technology, n.d.)

Type of crystallizer Advantages

Draft Tube baffle Capable of producing consistent and large sized crystals. Requires minimal
space requirement, adaptable to most corrosion-resistant MOC.
Forced circulation Suitable for high rates of evaporation and solutions with relatively high
viscosity. It is most economically favourable per yield of product formed.

A forced circulation crystallizer will be used to separate sodium methoxide from MeOH,
since the feed rate is relatively high and is least expensive. The crystallizer will be placed
along the stream S-207-1, at the bottom outlet stream of the distillation column and before
recycling it back to Section 1.

23
9. LIMITATIONS, ASSUMPTIONS AND OTHER POSSIBLE IMPLEMENTATIONS
9.1 Optimization
During the optimization of the distillation column, it was noticed that the Economic Analyzer
tool in HYSYS does not always give consistent values. This may affect the optimization
results obtained and as such, it may be possible to further reduce the TAC value for this
simulation without compromising the separation performance.
The tray efficiency was not recalculated after each round of optimization as they are assumed
to be remain around the same, since the liquid viscosity and the relative volatility should not
change much.
Also, after sizing of the internals in the distillation column and after the actual pressure drop
per tray calculated by HYSYS internals was inputted into the column parameters, the optimal
point for each variable optimized earlier is likely to change.
9.2 Possible contamination of MeOH
MeOH absorbs moisture from air very easily, and this may be particularly harmful in coastal
areas whereby dissolved chloride salts can be carried by moist air and enters the distillation
column, which will affect the purity of products obtained. Also, the MeOH may also dry the
air present in the vapour space of the column, hence contaminating the MeOH itself
(Methanol Institute, n.d.). This is a significant issue as the feed stream to the MF separation
column contains around 73 wt% of MeOH, which is a relatively high amount. As such,
instead of using carbon steel, all equipment which is in contact with MeOH in this design
uses stainless steel 304.
In order to maintain high purity levels, inert gases such as N2 can be introduced into the flash
separator and column at regular intervals (also known as Nitrogen blanketing) to remove any
moisture present and eliminate water absorption.
9.3 Crystallization of sodium methoxide from MeOH
Information regarding the crystallization of sodium methoxide from MeOH is not widely
available in literature. Therefore, uncertainties exists and the performance of the
crystallization process cannot be effectively predicted.

Besides evaporative crystallization, other types of crystallization, such as anti-solvent

crystallization, can also be implemented by adding organic solvents which are insoluble in
sodium methoxide, such as n-hexane and dimethyl carbonate, into the solution.
(Kaewsubdejsiri et al., 2018).

24
10. REFERENCES
ACS (2014, January 27). Methyl formate. Retrieved from
https://www.acs.org/content/acs/en/molecule-of-the-week/archive/m/methyl-formate.html
Aspen Technology (2006, October 17). Aspen HYSYS Property Packages.

Castle pumps (n.d.), Positive Displacement vs Centrifugal Pumps. Retrieved from

https://www.castlepumps.com/info-hub/positive-displacement-vs-centrifugal-pumps/

Carlson, E. C. (1996, October). Don't Gamble With Physical Properties For Simulations.
ChEBI (2017, March 8). CHEBI:77699 - methyl formate. Retrieved from
https://www.ebi.ac.uk/chebi/searchId.do?chebiId=CHEBI:77699

Chua, W.X., da Cunha, S., Rangaiah, G.P., Hidajat, K., Design and Optimization of Kemira-
Leonard Process for Formic Acid Production. Chem. Eng. Sci. 2019, 100021.

Costello (2019). Factors affecting distillation column operation. Retrieved from

https://www.rccostello.com/distil/distilop.htm
Kaewsubdejsiri, B., Mueansichai, T., Appamana, W. & Assabumrungrat, S. (2018), Solubility of
Sodium Methoxide in Pure and Mixed Solvents. The 8th International TIChE Conference.
Kister, H. Z. (1992). Distillation design. New York: McGraw-Hill.

KLM Technology Group (2011). Distillation Column Selection and Sizing. Retrieved from
http://kolmetz.com/pdf/EDG/ENGINEERING%20DESIGN%20GUIDELINES%20-%20di
stillation%20column%20-%20Rev%2004%20web.pdf
Methanol Institute (n.d.). Atmospheric Above Ground Tank Storage of MeOH. Retrieved from
https://www.MeOH.org/

Premkumar Er. (2020). Physical Separation Processes. Petroleum Refining. Chapter 3 of 8.

Separation Technology (2012, May 7). Distillation Column Tray Selection & Sizing. Retrieved
from http://seperationtechnology.com/distillation-column-tray-selection-1/
Seider, W. D., Seader, J. D., Lewin, D. R., & Widagdo, S. (2010). Product and process design
principles: synthesis, analysis and design. Chichester: John Wiley.
Sinnott, R., & Towler, G. (2020). Chemical engineering design. Kidlington, Oxford, United
Kingdom: Butterworth-Heinemann, an imprint of Elsevier.

SoliQz (n.d.). Melt crystallization & Hydraulic wash column. Retrieved from
https://www.soliqz.com/crystallization/types-of-crystallization/

Swenson Technology (n.d.). Crystallizers. Retrieved from

http://www.swensontechnology.com/crystallizers/
Thermo Fisher Scientific (2018, January 19). Safety Data Sheet. Retrieved from
https://www.fishersci.com/store/msds?partNumber=AC168600250&productDescription=S
ODIUM+METHOXIDE+30WT%25+25ML&vendorId=VN00032119&countryCode=US
&language=en

25
Appendix A: PFD and Stream table

Purge

S-212

To Section 1
S-211

S-210

From Section 5
S-500 S-501 S-208
K-200
CHWR

T-200C
S-204 S-209

K-201
CHWS
CHWR
T-200
From Section 1 E-200 V-200 Reflux Pump
S-201 S-202
To Section 3
CHWS S-206

S-203
Utilities T-200
P-200
CHWS - Chilled Water Supply
S-205
CHWR - Chilled Water Return
LPS - Low Pressure Steam Supply
LPSC - Low Pressure Steam Condensate

Legend

Centrifugal Pump Distillation

Column
LPS
Shell-and-Tube Heat
Exchanger Crystallizer
Utility
T-200R
Reflux Drum Manhole S-207

Horizontal Thermosyphon Reciprocating LPSC

Reboiler Compressor
To Section 1
S-207-1

26
Physical properties:

S-201 S-202 S-203 S-204 S-205 S-206 S-207 S-208 S-209 S-210 S-211 S-212 S-500 S-501
Vapour Fraction 0.102 0.0215 0 1 0 0 0 1 1 1 1 1 1 1
Temperature [℃] 58.05 20 20 20 20.05 40 84.18 40 41.30 47.04 47.04 47.04 46.37 62.01
Pressure [kPa] 155 120 120 120 213 152 218 152 152 152 152 152 114 152
Molar Flow [kgmole/h] 3935 3935 3850 84.7 3850 524 3318 8.67 84.7 117 44.9 71.7 23.2 23.2
Mass Flow [kg/h] 138367 138367 135497 2870 135496 30375 104641 479 2870 4666 1797 2870 1317 1317

Compositions (mole fraction):

S-201 S-202 S-203 S-204 S-205 S-206 S-207 S-208 S-209 S-210 S-211 S-212 S-500 S-501
CO 0.0149 0.0149 0.000236 0.682 0.000236 5.28E-05 4.67E-55 0.102 0.682 0.504 0.504 0.504 0.00688 0.00688
N2 0.00111 0.00111 3.56E-07 0.0517 3.56E-07 7.23E-08 1.05E-86 0.000154 0.0517 0.0376 0.0376 0.0376 0 0
MF 0.141 0.141 0.140 0.173 0.140 0.928 0.0140 0.846 0.173 0.367 0.367 0.367 0.894 0.894
MeOH 0.784 0.784 0.799 0.0919 0.799 0.0724 0.916 0.0526 0.0919 0.0859 0.0859 0.0859 0.0764 0.0764
H2O 0.0566 0.0566 0.0578 0.00171 0.0578 1.75E-09 0.067054 1.16E-10 0.00171 0.00544 0.00544 0.00544 0.0211 0.0211
FA 0.00248 0.00248 0.00253 3.07E-05 0.00253 3.11E-12 0.00294 1.46E-13 3.07E-05 0.000246 0.000246 0.000246 0.00113 0.00113

27
Appendix B: Equipment specification sheet
T-200 Distillation Column T-200 Distillation Column Internals
Section 1: Stages 1 to 25; Section 2: Stages 26 to 45
Column Name MF Separation Tray type Sieve
Column type Tray Number of passes 1
Column Diameter 2.78 m Tray spacing Section 1: 0.6096 m
Section 2: 0.9144 m
Column Height 36.5 m Tray thickness 3.404 mm
Number of stages 45 Weir height Section 1: 50.8 mm
Section 2: 76.2 mm
Feed stage 26 Side weir length 2.14 m
Operating pressure 152 kPa to 217.5 kPa Downcomer clearance Section 1: 38.1 mm
Section 2: 63.5 mm
Operating temperature 40℃ to 84.2℃ Hole diameter 5 mm
Manhole diameter 0.6 m Number of holes Section 1: 23348
Section 2: 17511
Number of manholes 3 Hole area to active area Section 1: 0.1
Section 2: 0.0750
Stages with manhole 1, 26, 45 Active area 4.584 m2
Material of Construction SS 304 Net area 5.332 m2
Material of Construction SS 304

Vessels
Identification Type Material Volume Length Diameter Orientation
V-200 Flash separator SS 304 26.98 m3 8.20 m 2.05 m Vertical
V-201 Reflux drum CS 6.60 m3 4.23 m 1.41 m Horizontal

Heat Exchangers
Identification E-200 Column reboiler Column condenser
Type Fixed tube Horizontal thermosyphon Fixed tube
Material Shell: SS 304 Tube: CS Shell: SS 304 Tube: CS Shell: CS Tube: CS
Heat transfer area 569 m2 249 m2 613 m2
Tube fluid Chilled water LP steam Chilled water
Shell fluid Process stream S-201 Bottoms, from T-200 Top stream, from T-200
Power 7481 kW 1.647 × 104 kW 9228 kW
Inlet temperature Tube: 10℃ Shell: 58.05℃ Tube: 148℃ Shell: 79.2 ℃ Tube: 10℃ Shell: 44.8℃
Outlet temperature Tube: 30℃ Shell: 20℃ Tube: 147.9℃ Shell: 84.2℃ Tube: 30℃ Shell: 40℃

Pumps
Identification Type Material Power Head Suction Discharge Efficiency
Pressure Pressure
P-200 Feed pump Centrifugal SS 304 5585W 11.35m 120 kPa 213.2 kPa 75%
Reflux pump Centrifugal CS 3455W 26m 152 kPa 162 kPa 75%

Compressors
Identification Type Material Power Capacity Suction Discharge Polytropic
Pressure Pressure Efficiency
K-200 Reciprocating CS 6566W 528.5 m3/h 113.7 kPa 152 kPa 75%
K-201 Reciprocating CS 18440W 1710 m3/h 120 kPa 152 kPa 75%

28