3

CHAPTER 3

DESIGN OF DECANTER

3.4.1 Introduction

The decanter is used for the separation of two or more phases. It can be used for most
type of solid-liquid separation based on specific gravity or density. The decanter operates
mainly by centrifugal force and sedimentation. The decanter can separate even the finest
solid if its sedimentation rate in the carrier liquid is sufficiently high. The sedimentation rate
depends on the particle shape, particle size, viscosity on the latter and density. If the
density difference is high, then the specific gravity can provide sufficient driving force.

In basic terms, decanter centrifuge separates solids from one or two liquid phases
in one single continuous process. This is done using centrifugal forces that can be well
beyond 3000 times greater than gravity. When the mixture to such forces, the denser solid
particles are pressed outwards against the rotating bowl wall, while the less dense liquid
phase forms a concentric inner layer. The sediment formed by the solid particles is
continuously removed by the screw conveyor, which rotates at a different speed than the
bowl. The centrifugal force compact the solids and expels the surplus liquid. The dried
solids then discharge from the bowl. The clarified liquid phase or phases overflow the
dam plates situated at the opposite end of the bowl.
 4

3.4.2 Chemical Design Of Decanter

3.4.2.1 Properties of the Mixture

Physical properties play an important role in characterizing the operation of processes and
individual process unit. Thus before designing the decanter, all of the process variable
must be correctly measured. Liquid density are important in engineering design such as
sizing the decanter that contain the basic raw material and product for a plant. In addition,
density difference is main factor that makes on decanter separation work. Basically liquid
density is a temperature dependent. The modified form of the Rackett equation was
selected for correlation of saturated liquid density as function of temperature.

𝑇 𝑛
−(1− )
𝐷𝑒𝑛𝑠𝑖𝑡𝑦 = 𝐴𝐵 𝑇𝑐

Where
Density = Saturated liquid density, g/ml
A,B and n = Regression coefficient for chemical compound
T = Temperature, K
Tc = Critical temperature, K
Table 3.4.1: Modified form of the Rackett equation
Component A B n Tc Tmin Tmax Density@25 oC
Toluene 0.29999 0.27108 0.29889 591.79 178.18 591.79 865
Nitrogen 0.31205 0.28479 0.29250 126.10 63.15 126.10 -
Water 0.3471 0.274 0.28571 647.13 273.16 647.13 997
Paraxylene 0.27984 0.26003 0.271 616.26 286.41 616.26 858
Metaxylene 0.27866 0.25925 0.27243 617.05 225.30 617.05 861
Ortoxylene 0.2381 0.26083 0.27410 630.37 247.98 630.37 875
Source: Carl L Yaw
Decanter inlet
Component [email protected] K
Toluene 851.5
Nitrogen -
Water 992.6
Paraxylene 851
Metaxylene 848.9
Ortoxylene 725.5
 5

Density (𝛒) of a mixture of n liquids it can be simply calculated by average the pure-
component densities, weighting each on by the mass fraction of the component:

𝝆 = ∑(𝒙𝒊 . 𝝆𝒊 )
𝒊=𝟏
Therefore:
𝒌𝒈
𝝆𝒎(𝑳) = 𝟖𝟒𝟗. 𝟖
𝒎𝟑
𝒌𝒈
𝝆𝒎(𝑯) = 𝟗𝟗𝟐. 𝟔 𝟑
𝒎

Sample of calculation at S20 (Light phase):

Component Toluene Methanol Nitrogen H2O PX OX MX
Mass fraction 0.7380 0.0 0.0021 0.0008 0.2589 0.0001 0.0002
S20

Stream 20 (light phase)

𝒏

𝝆 = ∑(𝒙𝒊 . 𝝆𝒊 )
𝒊=𝟏
𝝆𝒎(𝑳) = (𝟎. 𝟕𝟑𝟖×𝟖𝟓𝟏. 𝟓 + 𝟎. 𝟎𝟎𝟐𝟏×𝟎 + 𝟎. 𝟎𝟎𝟎𝟖×𝟗𝟗𝟐. 𝟔 + 𝟎. 𝟐𝟓𝟖𝟗×𝟖𝟓𝟏 + 𝟎. 𝟎𝟎𝟎𝟏×𝟖𝟒𝟖. 𝟗
+ 𝟎. 𝟎𝟎𝟎𝟐×𝟕𝟐𝟓. 𝟓)
= 849.8 kg/m3

3.4.2.2 Viscosity of liquid mixture

Viscosity values will be needed for any design calculations involving the transport of fluids
or heat. It is difficult to predict the viscosity of mixtures of liquids. The viscosity of the
mixture may be higher or lower than the pure components.

For organic liquid mixtures a modified form of Souders’ equation can be used; using a mol
fraction weighted average value for the viscosity constant for the mixture, and the average
molecular weight.
𝑥1 𝐼1 + 𝑥2 𝐼2
log(log 10 µ𝑚 ) = 𝜌𝑚 [ ] ×10−3 − 2.9
𝑥1 𝑀1 + 𝑥2 𝑀2
Where,
µ𝑚 = viscosity of mixture,
𝜌𝑚 = density of mixture,
𝑥1 , 𝑥2 = mol fraction of components,
𝑀1 , 𝑀2 = molecular masses of components.
𝐼1 , 𝐼2 = Souders’ index, estimated from the group contributions given in Figure
3.1
 6

Figure 3.4.1: Souders’ index estimated from the group contributions

Sample of calculation at S20 (Light phase)

Sample calculation toluene souders index, I

Contribution from table 3.1:

7 carbon atoms 7 x 50.2 = 351.4
8 hydrogen atoms 8 x 2.7 = 21.6
3 double bonds 3(-15.5) = -46.5
1 six-membered ring = -21.1
1 side group = -9.0
Total, I = 296.4
Molecular weight = 92
 7

Hence,
Component Souder’s Index, I Molecular weight
Toluene 296.4 92
Nitrogen 74.0 14
Water 35.1 18
Paraxylene 346 106.16
Ortoxylene 346 106.16
Metaxylene 342 106.16

𝑥1 𝐼1 + 𝑥2 𝐼2 + 𝑥3 𝐼3 + 𝑥4 𝐼4
log(log 10 µ𝑚(𝐿) ) = 𝜌𝑚(𝐿) [ ] ×10−3 − 2.9
𝑥1 𝑀1 + 𝑥2 𝑀2 + 𝑥3 𝑀3 + 𝑥4 𝑀4

0.738×296.4 + 0.021×74 + 0.0008×35.1 + 0.2589×346 + 0.0001×346 + 0.0002×342

= 849.8 [ ]
0.738×92 + 0.021×14 + 0.0008×18 + 0.2589×106.16 + 0.0001×106.16 + 0.0002×106.16
×10−3 − 2.9
log(log 10 µ𝑚(𝐿) ) = −0.1478
log 10 µ𝑚(𝐿) = 0.7116
𝑚𝑁𝑠
µ𝑚(𝐿) = 0.5147
𝑚2

Sample of calculation at S21 (Heavy phase)

Because of S21 only exist pure component (water), hence viscosity of pure component
can be get on dynamic viscosity of water at viscosity table of water
𝑚𝑁𝑠
µ𝑚(𝐻) = 0.6652
𝑚2

3.4.2.3 Volumetric Flow rate

To find volumetric flow rate
𝑚 1 ℎ𝑟
𝑄= ×
𝜌𝑚 3600 𝑠
Mass flow rate and volumetric flow rate:
Stream/Flow rate Mass Flowrate, m(kg/hr) Volumetric Flowrate, Q (m3/s)
S20 (Light phase) 37 400 0.0122
S21 (Heavy phase) 12 620 0.0035
 8

Volumetric flow rate at light phase:

𝑚𝐿 1 ℎ𝑟
𝑄𝐿 = ×
𝜌𝑚(𝐿) 3600 𝑠
37 400 𝑘𝑔/ℎ𝑟 1 ℎ𝑟
𝑄𝐿 = 3
×
849.8 𝑘𝑔/𝑚 3600 𝑠
𝑚3
𝑄𝐿 = 0.0122
𝑠

3.4.2.4 Checking for Dispersed Phase

Part of the drop size prediction requires knowing which phase is dispersed and which is
continuous. Selker and SIeicher provide a useful correlation to predict which is the
dispersed phase based on phase volume ratios and density and viscosity of each phase.

𝑄𝐿 𝜌𝐿 µ𝐻 0.3
𝑋= ( )
𝑄𝐻 𝜌𝐻 µ𝐿
where QL is the volume of the light phase and QH the volume of the heavy phase is
consistent units. The following guidelines are suggested:

Table 3.4.2: Results in checking dispersed phase

X Result
< 0.3 Light phase always dispersed
0.3 – 0.5 Light phase probably dispersed
0.5 – 2.0 Phase inversion possible; design for the worst case
2.0 – 3.3 Heavy phase probably dispersed
> 3.3 Heavy phase always dispersed
(Source:

Therefore,
0.3
0.0122 𝑚3 /𝑠 849.8 𝑘𝑔/𝑚3 ∗ 0.1317 𝑚𝑁𝑠/𝑚2
𝑋= ( )
0.0035 𝑚3 /𝑠 1014.5 𝑘𝑔/𝑚3 ∗ 0.5147 𝑚𝑁𝑠/𝑚2
𝑋 = 2.12
Based on table of result, the heavy phase probably dispersed and light phase become a
continuous phase for this decanter design.
 9

3.4.2.5 Determining the Droplet Settling Velocity

Drop settling velocity is estimated from Stokes' Law using Newton's basic drag equation:
𝑔𝑑𝑑2 (𝜌𝑑 − 𝜌𝑐 )
𝑈𝑑 =
18 µ𝑐
where dd =drop diameter, m
g = the acceleration of gravity, 9.81 m/ss
ρd = density of the dispersed phase, kg/m3
ρc = density of the continuous phase, kg/m3
µc = viscosity of the continuous phase, Ns/m2.
Ud = settling (terminal) velocity of the dispersed phase droplets with diameter d,
m/s

The assumptions most often violated are:

1. The continuous phase is a quiescent fluid.

2. The droplet is a hard (rigid) sphere with no deformation or internal circulation.
3. The droplet moves in laminar flow
4. The droplet is large enough to ignore Brownian motion.
5. The droplet movement is not hindered by other droplets or by walls (vessel
surfaces).

Hence:
ρm(H) = ρd = 1014.5 kg/m3 ρm(L) = ρc = 849.8 kg/m3
µm(H) = µd = 0.1317 mNs/m2 µm(L) = µc = 0.5147 mNs/m2
QH = Qd = 0.0035 m3/s QL = Qc = 0.0122 m3/s

According to Walas, 150 µm is a common drop diameter for the design of decanters.
Taking dd = 150 µm
𝑔𝑑𝑑2 (𝜌𝑑 − 𝜌𝑐 )
𝑈𝑑 =
18 µ𝑐
𝑘𝑔 𝑘𝑔
𝑚2 (1014.5 3 − 849.8 3 )
𝑚 𝑚
= 9.81 ×(150×10−6 𝑚)2 ×
𝑠 −3 𝑁𝑠
18 ×0.0122 × 10
𝑚2
= 0.1655 ×10−2 𝑚/𝑠
The decanter vessel is sized on the basis that the velocity of the continuous phase
Must be less than settling velocity of the droplets of the dispersed phase. Sinnot 2005.
The positive sign means that the light heavy key move downward instead of upward.
 10

3.4.2.6 Determining the Decanter diameter

𝜋𝐷𝐿2
𝐴𝐿 =
8
𝑄𝐿
𝑉𝐿 =
𝐴𝐿
By substitute above equation
8 𝑄𝐿
𝑉𝐿 =
𝜋𝐷𝐿2

𝐷𝐿
𝜋
𝑅ℎ(𝐿) = 4
2+𝜋
4 𝑅ℎ(𝐿) 𝜌𝐿 𝑄𝐿
𝑅𝑒𝐿 =
µ𝐿
By substitute above equation
8𝜌𝐿 𝑉𝐿
𝑅𝑒𝐿 =
(2 + 𝜋)µ𝐿 𝐷𝐿

Similarly, for the heavy phase will form these equation

8 𝑄𝐻
𝑉𝐻 =
𝜋𝐷𝐻2
8𝜌𝐻 𝑄𝐻
𝑅𝑒𝐻 =
(2 + 𝜋)µ𝐻 𝐷𝐻

𝑅𝑒𝐿 ≤ 10 000
𝑅𝑒𝐻 ≤ 10 000

Thus,
8𝜌𝑚(𝐿) 𝑄𝐿
𝑅𝑒𝐿 =
(2 + 𝜋)µ𝐿 𝐷𝐿
𝑘𝑔 𝑚3
8×849.8 ×0.0122
10 000 = 𝑚3 𝑠
𝑘𝑔
(2 + 𝜋)×0.5147×10−3 𝑚. 𝑠 ×𝐷𝐿

𝐷𝐿 = 3.1341 𝑚

8𝜌𝐻 𝑄𝐻
𝑅𝑒𝐻 =
(2 + 𝜋)µ𝐻 𝐷𝐻
 11

𝑘𝑔 𝑚3
8×1014.5 3 ×0.0035 𝑠
10 000 = 𝑚
𝑘𝑔
(2 + 𝜋)×0.1317×10−3 𝑚. 𝑠 ×𝐷𝐻

𝐷𝐻 = 4.1949 𝑚 (13.7628 ft)

Therefore, the decanter diameter is 13.7628 ft (4.1949 m), which is rounded off
to 14.0 ft (4.2672 m).

3.4.2.7 Residence time

Decanters are essentially tanks which give sufficient residence time for the droplets of the
dispersed phase to rise to the interface between 2 to 5 min. A value of 10 per cent of the
decanter height is usually taken for design purpose.

𝐻𝐷 = 0.1 𝐷
= 0.1 (4.2672𝑚)
= 0.4267𝑚

𝐻𝐷
𝑅𝑒𝑠𝑖𝑑𝑒𝑛𝑐𝑒 𝑡𝑖𝑚𝑒 =
𝑈𝐷
0.4267 𝑚
= 𝑚
0.1655 ×10−2
𝑠
= 258 𝑠𝑒𝑐𝑜𝑛𝑑𝑠 = 4.3 𝑚𝑖𝑛𝑠

3.4.2.8 Length of Decanter

Next step is to calculate the length of the decanter

8 𝑄𝐻
𝑉𝐻 =
𝜋𝐷2
𝑚3
8 ×0.0035 𝑠
=
𝜋 ×(4.2672 𝑚)2
𝑚
= 0.4895 ×10−3
𝑠

The settling length

𝑉𝐻 𝐷
𝐿𝑠 =
2 𝑈𝑑
 12

𝑚
0.4895 ×10−3 𝑠 ×4.2672 𝑚
= 𝑚
2 × 0.1655 ×10−2 𝑠

= 0.6311 m

The interfacial area required for coalescence

2 𝑄𝐻 𝑡𝑅
𝐴𝑖 =
𝐻𝐷
𝑚3
2×0.035 𝑠 ×258 𝑠
=
0.4267 𝑚
= 42.3248 𝑚2

Dispersion length is
𝐴𝑖
𝐿𝐷 =
𝐷
42.3248 𝑚2
=
4.2672 𝑚
= 10.581 m

Thus, the total decanter length is

𝐿 = 𝐿𝑠 + 𝐿𝐷
= 0.6311 𝑚 + 10.581 𝑚
= 11.2123 𝑚

Checking length to diameter ratio

𝐿 11.2123 𝑚
= = 2.628
𝐷 4.2672 𝑚

Gravity decanters normally are specified with a length-to-diameter ratio must greater than
2 to maximize the phase boundary between the two settled layers (Perry R. H., 2008). In
this case, a length-to-diameter ratio of is 2.63 which is satisfied this condition.
 13

3.4.2.9 Decanter Geometry

Therefore,
1⁄
𝐼 = 2[(2.1336𝑚)2 − (11.2123𝑚)2 ] 2

= 22.01 𝑚

𝐴𝐼 = 22.01 ×11.2123 𝑚
= 246.7827 𝑚2

3.4.2.10 Check Turbulence Level

In a design horizontal decanter, the continuous phase is flowing perpendicularly to the
settling of the droplets. This movement creates turbulence which interferes with the settling
process according to Rousseau, R. W. (1987). If the cross flow were fully laminar, there
would be no problem, but this is usually impractical. The degree of turbulence is best
expressed by the Reynolds number NRE where:
𝜐𝐷𝐻 𝜌𝑐
𝑁𝑅𝐸 =
µ𝑐

υ is the continuous-phase crossflow velocity

 14

𝑄𝑐
𝜐=
𝐴𝑐

The following guidelines summarize an experience to hold successful decanter design

Table 3.4.3: Effects of Reynolds Number

Reynolds Number Reynolds Number
<5000 Little probem
5000 – 20 000 Some hindrance
20 000 – 50 000 Major problem may exist
>50 000 Expect poor separation
(Source: Rousseau, R. W. 1987).
𝑚3
0.0122 𝑠
𝜐=
𝜋(0.07412𝑚)2
4
𝑚
= 2.8275
𝑠

𝜐𝐷ℎ 𝜌𝑐
𝑁𝑅𝐸 =
µ𝑐
𝑚 𝑘𝑔
2.8275 ×4.1949 𝑚×849.8 3
𝑠 𝑚
=
𝑁𝑠
0.5147 2
𝑚
= 19 583 (some hindrance)

3.4.2.11 Inlet Diameter

Stainless steel is used for the inlet and outlet of decanter. Optimum diameter for this
process is given by the following equation (Sinnott, R. and Towler, G., 2010).

𝑑𝑜𝑝𝑡𝑖𝑚𝑢𝑚 = 260𝐺 0.53 𝜌−0.37

𝑘𝑔 1ℎ𝑟
𝐹𝑙𝑜𝑤𝑟𝑎𝑡𝑒, 𝐺 = (125 700 ) ( ) = 34.9167 𝑘𝑔/𝑠
ℎ𝑟 3600𝑠

𝑑𝑜𝑝𝑡𝑖𝑚𝑢𝑚 = 260(34.9167)0.53 (6.683)−0.37 = 846.33 𝑚𝑚

3.4.2.12 Outlet Diameter

For Light Phase outlet diameter,
𝑘𝑔 1 ℎ𝑟
𝐹𝑙𝑜𝑤𝑟𝑎𝑡𝑒, 𝐺 = (37 400 )( ) = 10.3889 𝑘𝑔/𝑠
ℎ𝑟 3600 𝑠

𝑑𝑜𝑝𝑡𝑖𝑚𝑢𝑚 = 260(10.3889)0.53 (849.8)−0.37 = 74.12 𝑚𝑚

 15

For Heavy Phase outlet diameter,

𝑘𝑔 1 ℎ𝑟
𝐹𝑙𝑜𝑤𝑟𝑎𝑡𝑒, 𝐺 = (12 620 )( ) = 3.5056 𝑘𝑔/𝑠
ℎ𝑟 3600 𝑠

𝑑𝑜𝑝𝑡𝑖𝑚𝑢𝑚 = 260(3.5056)0.53 (992.6)−0.37 = 39.35 𝑚𝑚

3.4.2.13 Piping arrangement

The values of Z1, Z2, and Z3 are determined where the height of every piping are arranged
based on Figure 3.4.2 and figure 3.4.3. The height of pipe arrangement can be done by
making pressure balance. Assumption is done by neglecting the friction loss in the pipes.
The correlation in piping arrangement as follows:

Take the length as height of decanter for horizontal decanter design. Take the position of
the interface as half-way up the vessel and the light liquid off-take as at 90 per cent of
the vessel height, then

Figure 3.4.2: Vertical decanter

Figure 3.4.3: Horizontal decanter

 16

(𝑧1− 𝑧3 )𝜌1 𝑔 + 𝑧3 𝜌2 𝑔 = 𝑧2 𝜌2 𝑔

Hence,
(𝑧1 − 𝑧3 )𝜌1
𝑧2 = + 𝑧3
𝜌2
𝑧1 = 0.9 ×4.2672𝑚 = 3.8405𝑚
𝑧3 = 0.5 ×4.2672𝑚 = 2.1336𝑚
(3.8405 − 2.1336)𝑚 𝑘𝑔
𝑧2 = ×849.8 3 + 2.1336 𝑚 = 3.5949 𝑚
𝑘𝑔 𝑚
992.6 3
𝑚

3.4.3 Summary of Decanter Chemical Design

Droplet settling velocity, uD 0.1655 x10-2 m/s
Decanter diameter, D 4.2672 m
Decanter length, L 11.2123 m
Area of interface (Assuming flat interface), Ai 42.3248 m2
Height of the dispersion band, HD 0.4267m
Residence time, t 4.3 mins
Inlet diameter 846.33 mm
Light phase outlet diameter 74.12 mm
Heavy phase outlet diameter 39.35 mm
 17

3.4.4 Mechanical Design

The mechanical design of chemical plant is a particular interest to chemical engineer. The
purpose of mechanical design is to determine the dimensions and construction of the
equipment, operating pressure and temperature, material construction and equipment
dimensions.

The requirements needed in designing a decanter are:

1. Design pressure
2. Design temperature
3. Material selection
4. Corrosion Allowance
5. Design stress
6. Cylindrical section
7. Domed head
8. Dead weight of vessel.
9. Vessel support.
10. Flange design

3.4.5 General Design Considerations for Pressure Vessels

3.4.5.1 Design Pressure

A vessel must be designed to resist the maximum pressure that can withstand the process
involved. Decanter is a vessel under internal pressure, so that the design pressure is taken
at the relief valve is set.

Top stream pressure = 333.5 kPa

Bottom stream pressure = 333.5 kPa

Taking operating pressure at 333.5 kPa = 3.335 x 105 N/m2

Normally, 5-10% above the normal working pressure usually been considered in order to
avoid spurious operation during minor process upsets. So, by taking 10% above the
operating pressure, therefore, the design pressure for this decanter, Pi :

𝑁
𝑃𝑖 = 3.335 × 105 ×1.1
𝑚2
𝑁 𝑁
= 3.669 × 105 2 = 0.3967
𝑚 𝑚𝑚2
 18

3.4.5.2 Design Temperature

In designing the decanter, material temperature plays an important role and influenced the
maximum allowable stress. The strength of metals decreases with the increasing of
temperature. Decanter operating temperature:

Top stream temperature = 38.35 oC

Bottom stream temperature = 38.35 oC

The design temperature at which the design stress is calculated should be taken
as the maximum working temperature of the material with suitable allowance for any
hesitation involved in predicting the vessel wall temperature. By taking 10% above the
operating temperature, therefore, design temperature is:

𝐷𝑒𝑠𝑖𝑔𝑛 𝑡𝑒𝑚𝑝𝑒𝑟𝑎𝑡𝑢𝑟𝑒, 𝑇𝑖 = 38.35 °𝐶 ×1.1

= 42.19 °𝐶

3.4.5.3 Material Selection

Material of construction is important in designing the decanter. Many factors have been
considered in selecting the material of decanter construction. The material chosen should
be economically fulfilling both the process and mechanical requirements. Since the
decanter involve with non-corrosive chemical such as xylene isomers, toluene, nitrogen
gas and water, therefore the ordinary material can be used for purpose financial saving

In this design, carbon-manganese steel is applied to reduce the risk. Carbon-manganese

steel has excellent ductility, permitting, many cold forming operation. The carbon-
manganese steel can be readily cast, wrought and machined and the problem occurs in
welding can be overcome using correct welding material and procedure. One of the
advantage the price is quite low for carbon-manganese steel compare to the stainless
steel. Carbon-manganese steel also have tensile strength value close to the stainless steel
make stronger material suitable for construction decanter.
 19

3.4.5.4 Design Stress (Nominal Design Strength)

Each material of construction such as stainless steel or carbon steel has their own
maximum allowable stress. The allowable stress means the maximum stress that the
material could withstand without failure. For design purposes it is necessary to decide a
value for the maximum allowable stress that can be accepted in the material construction.

Design stress, f = 180 N/mm2

Tensile strength =460 N/mm2

3.4.6 The Design of Thin-Walled Vessels under Internal Pressure

3.4.6.1 Cylindrical Shell

The minimum wall thickness is required to highlight in order to ensure the vessel to
withstand the weight and loads. The vessel should not be less than value from the table
below:

Vessel diameter (m) Minimum thickness (mm)

1 5
1-2 7
2-2.5 9
2.5-3.0 10
3.0-3.5 12
Source: Sinnot, R. K. (2003)

Inner Diameter, Di = 4.2672 m = 4267 mm

Taking welded joint factor, J = 1.0 implies that the joint is equally as strong as the virgin
plate.

𝑃𝑖 𝐷𝑖
Minimum thickness, 𝑒=
2𝐽𝑓 − 𝑃𝑖

𝑁
0.3967 2 × 4267 mm
= 𝑚𝑚
𝑁 𝑁
(2 ×1 × 180 2 ) − 0.3967
𝑚𝑚 𝑚𝑚2

= 4.7071 mm
 20

Figure 3.4.4: Extrapolation vessel diameter with minimum thickness

By doing extrapolation figure 3.2 vessel diameter with minimum thickness, the lowest
thickness of the cylindrical shell will be 14.84mm when the vessel diameter is 4.267 m.
Rounding up to the nearest decimal point will be 15 mm

3.4.6.2 Heads and Closures

It is necessary to choose the suitable closure since decanter is a closed cylindrical vessel
with a domed head. In order to choose hemispherical, ellipsoidal or torispherical head, it
must be depends on design pressure.

Hemispherical head Ellipsoidal head Torispherical head

 Strongest shape (able to  Most economical closure  Normally used at

resist about twice the at pressure above 15 bar. operating pressure up to
pressure of a torispherical 15 bar.
head of the same
thickness.)  Can be used at high
pressure.
 Cost > torispherical head
 If operating pressure is
above 10 bar, cost should
be compared with an
equivalent ellipsoidal
head.

(Source:Coulson & Richardson’s Chemical Engineering Volume 6)

 21

Based on the table above, ellipsoidal head is chosen since it does not require high
pressure to operate and its reasonable price rather than hemispherical head.

Standard ellipsoidal head

Steps:
1. Ratio major: minor axes = 2: 1
2. Minimum thickness is determined, e required for ellipsoidal
𝑃𝑖 𝐷𝑖
𝑒=
2𝐽𝑓 − 0.2𝑃𝑖
𝑁
0.39672 × 4267 mm
= 𝑚𝑚
𝑁 𝑁
(2 ×1 × 180 ) − (0.2×0.3967 )
𝑚𝑚2 𝑚𝑚2
= 4.703 𝑚𝑚

Thus, thickness of the ellipsoidal heads is taken as 12mm same as thickness of the wall
for safety purpose.

3.4.6.3 Dynamic Wind Pressure

For cylindrical column, semi-empirical formula can be applied to estimate the wind
pressure.

𝑃𝑤 = 0.05𝑈𝑤2

Where,

Pw = wind pressure, N/m2

UW = wind velocity, km/h

Therefore;

𝑃𝑤 = 0.05(160)2

𝑁
= 1280
𝑚2

Loading per unit length of column, Fw

𝐹𝑤 = 𝑃𝑤 𝐷𝑒𝑓𝑓

Where
 22

Deff = effective column diameter

𝐷𝑒𝑓𝑓 = 𝑃𝑤 𝐷𝑒𝑓𝑓

= 𝐷𝑖 + 2(𝑡𝑠ℎ𝑒𝑙𝑙 )

= 4.267 + 2(15 ×10−3 ) = 4.297 𝑚

Thus,
𝑁
𝐹𝑤 = 1280 ×4.297 𝑚
𝑚2

𝑁
= 5500.2
𝑚

3.4.6.4 Manhole

This manhole is used for cleaning and inspection purposes. In order to ease to inspect and
cleaning purpose, manhole is designed with standard size of manhole which is from 300
mm to 500 mm. For this decanter, diameter of 400 mm is applied in order to ease for the
maintenance.

3.4.7 Design of Vessels Subject to Combined Loading

3.4.7.1 Dead Weight of Vessel

For a steel vessel, the approximate weight of a cylindrical vessel with domed ends, and
uniform wall thickness, can be estimated from the following equation given by Sinnot 1999:
For a steel vessel, total weight of shell, excluding internal fittings is given as:

𝑊𝑣 = 240𝐶𝑣 𝐷𝑚 (𝐻𝑣 + 0.8𝐷𝑚 )𝑡

Where,

Cv: 1.08 for vessels with only a few internal fittings

: 1.15 for distillation columns, or similar vessels, with several manways, and with
plate support rings, or equivalent fittings. (Sinnot 1999)

Hv = 11.2123 m

𝐷𝑚 = 𝐷𝑖 + 𝑡

= 4267 𝑚𝑚 + 15 𝑚𝑚 = 4282 𝑚𝑚 = 4.282 𝑚

 23

Therefore,

𝑊𝑣 = 240×1.08×4.282𝑚×[11.2123𝑚 + (0.8×4.282𝑚)]×15𝑚𝑚

= 243 700 𝑁 = 243.7 𝑘𝑁

Hence,

𝑊𝑇 = 𝑊𝑣 = 243.7 𝑘𝑁

3.4.8 VESSEL SUPPORT

The method used to support a vessel will depend on the size, shape and weight of the
vessel, design pressure and temperature and vessel location and arrangement. The
supports must be designed to carry the weight of the vessel and contents, and any
superimposed loads. (Sinnot, 1999)

3.4.8.1 Skirt support

Skirt support is chosen since the decanter has been designed as a vertical vessel.
Supports will support loads on the vessel wall, and the design need to be checked to
ensure that the resulting stress concentrations are below the maximum allowable design
stress.

𝜋 2
Approximate weight = 𝐷 𝐻 𝜌 𝑔
4 𝑖 𝑣 𝐿

𝜋 𝑘𝑔 𝑚
= ×(4.267𝑚)2 ×11.2123𝑚×849 3 ×9.81 2
4 𝑚 𝑠

= 1 335 385 𝑁 = 1335 𝑘𝑁

𝑊𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑣𝑒𝑠𝑠𝑒𝑙 = 243.7 𝑘𝑁

𝑇𝑜𝑡𝑎𝑙 𝑤𝑒𝑖𝑔ℎ𝑡 = 1335 𝑘𝑁 + 243.7 𝑘𝑁

= 1 578.7 𝑘𝑁

3.4.8.2 Bending moment at skirt base

(𝐻𝑣 + 𝐻𝑠𝑘𝑖𝑟𝑡 )2
𝑀𝑠 = 𝐹𝑤 [ ]
2
 24

𝑁 (11.2123 m + 1m)2
= 5500.2 [ ]
𝑚 2

= 410 151 𝑁𝑚 𝑜𝑟 410.2 𝑘𝑁𝑚

Bending stress in skirt

4𝑀𝑠
𝜎𝑏𝑠 =
𝜋(𝐷𝑠 + 𝑡𝑠 )𝑡𝑠 𝐷𝑠

Where;

Ms = maximum bending moment, evaluated at the base of the skirt.

Ds = inside diameter of the skirt at the base.

ts = skirt thickness

Therefore,

4 ×410 151 𝑁𝑚 ×103

𝜎𝑏𝑠 =
𝜋(4267𝑚𝑚 + 15 𝑚𝑚)(4267𝑚𝑚)(15𝑚𝑚)

𝑁
= 1.91
𝑚𝑚2

3.4.8.3 Dead weight stress in skirt

𝑊
𝜎𝑤𝑠 (𝑡𝑒𝑠𝑡) =
𝜋(𝐷𝑠 + 𝑡𝑠 )𝑡𝑠

Where,

W = total weight of the vessel and contents

𝑊
=
𝜋(𝐷𝑠 + 𝑡𝑠 )𝑡𝑠

1578.7 𝑘𝑁 ×103
=
𝜋(4267𝑚𝑚 + 15 𝑚𝑚)(4267𝑚𝑚)

𝑁
= 0.0275
𝑚𝑚2

𝑊𝑡𝑜𝑡𝑎𝑙 𝑣𝑒𝑠𝑠𝑒𝑙
𝜎𝑤𝑠 (𝑜𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔) =
𝜋(𝐷𝑠 + 𝑡𝑠 )𝑡𝑠
 25

243.7 𝑘𝑁 ×103
=
𝜋(4267𝑚𝑚 + 15 𝑚𝑚)(4267𝑚𝑚)

𝑁
= 4.2456×10−3
𝑚𝑚2

𝑁 𝑁 𝑁
𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝜎𝑠 (𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑣𝑒) = 1.91 2
+ 0.0275 2
= 1.9375
𝑚𝑚 𝑚𝑚 𝑚𝑚2

𝑁 𝑁 𝑁
𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝜎𝑠 (𝑡𝑒𝑛𝑠𝑖𝑙𝑒) = 1.91 2
− 4.2456×10−3 2
= 1.9058
𝑚𝑚 𝑚𝑚 𝑚𝑚2

3.4.8.4 Base ring and anchor bolts

Approximate pitch circle diameter, Db

𝐷𝑏 = 𝐷𝑖 + 2(𝑡𝑑 )

= 4267 𝑚𝑚 + 2(5 𝑚𝑚)

= 4277 𝑚𝑚

Circumference of bolt circle = 4277 π

Number of bolt required, at minimum recommended bolt spacing

4277 𝜋
= = 22.4
600

3.4.8.5 Selection of Gasket

Gaskets are used to make a leak-tight joint between two surfaces. It is impractical to
machine flanges to the degree of surface finish that would be required to make a
satisfactory seal under pressure without a gasket. Gaskets are made from ‘semi-plastic’
materials; which will deform and flow under load to fill the surface irregularities between
the flanges faces, yet retain sufficient elasticity to take up the changes in the flange
alignment that occur under load. (Sinnot, 1999)

A vegetable fibre or synthetic rubber gasket are recommended for this vessel since they
can be used at temperatures of
 26

3.4.8.6 Flange Face

The raised face, narrow faced which is probably the most commonly used types of flange
are used for all the flanges.

3.4.8.7 Flange Design

Standard flanges are available in a range of types, sizes and materials; and are used
extensively for pipes, nozzles and other attachments to pressure vessels. The proportions
of standard flanges are set out in the various codes and standards. (Sinnot, 1999)

3.4.9 SUMMARY OF MECHANICAL DESIGN

Orientation Vertical

Material of construction Carbon-manganese steel

Operating pressure 333.5 kPa

Operating temperature 38.35 oC

Thickness of decanter 15 mm

Type of head Ellipsoidal

Thickness of head 15 mm

Total dead weight of decanter 243.7 kN

Type of support Skirt support

Inlet diameter 846.33 mm

Light phase outlet diameter 74.12 mm

Heavy phase outlet diameter 39.35 mm

 27

REFERENCES

1. Hibbeler, R. C. (2008). Mechanics of Material, 7th Edition. Singapore: Prentice Hall.

2. Hooper, W. B. (1979). Decantation in Handbook of Separation Process for Chemical

Engineers, 3rd edition, Schweitzer: McGraw-Hill.

3. Lakshmanan, N et al, (2009). Basic Wind Speed map of India with Long-term Hourly
Wind Data. Structural Engineering Research Centre, 911-922.

4. Perry, R. (2008). Perry's Chemical engineers' handbook (6th ed.). New York: McGraw-
Hill

5. Sinnot, R. K. (2003). Coulson and Richardson Chemical Engineering Volume

6:Chemical Engineering Design, 3rd Edition. Butterworth- Heinemann Series in
Chemical Engineering.

6. Smith, R. (1995). Chemical Process Design. McGraw-Hill International Editions.

7. Turton, R., Bailie, R. C., Whiting, W. B., Shaeiwitz, J. A., Bhattacharyya, D. (2012).
Analysis, Synthesis and Design of Chemical Processes. Fourth edition. Pearson.

8. W.Y Svrcek, W. M. (2004). Design Two phase Separator within the right limits.
Fluids/Solids Handling.

9. Sinnott, R. K., & Coulson, J. M. (2005). Coulson & Richardson's chemical engineering.
Oxford: Butterworth-Heinemann.

10. Rousseau, R. W. (1987). Handbook of separation process technology. New York:

J. Wiley.
 28

APPENDIX

Sample calculation nitrogen souders index, I

Contribution from table 3.1:

2 nitrogen atoms 2 x 37.0 = 74.0
Total, I = 74.0
Molecular weight = 14

Sample calculation toluene souders index, I

Contribution from table 3.1:

1 oxygen atoms 1 x 29.7 = 29.7
2 hydrogen atoms 2 x 2.7 = 5.4
Total, I = 35.1
Molecular weight = 18

Sample calculation paraxylene and ortoxylene souders index, I

Contribution from table 3.1:

8 carbon atoms 8 x 50.2 = 401.6
8 hydrogen atoms 10 x 2.7 = 27
3 double bonds 3(-15.5) = -46.5
1 six-membered ring = -21.1
2 side group 2 x -9.0 = -18.0
Para/orto position =3
Total, I = 346.0
Molecular weight = 106.16

Sample calculation paraxylene and ortoxylene souders index, I

 29

Contribution from table 3.1:

8 carbon atoms 8 x 50.2 = 401.6
8 hydrogen atoms 10 x 2.7 = 27
3 double bonds 3(-15.5) = -46.5
1 six-membered ring = -21.1
2 side group 2 x -9.0 = -18.0
Meta position = -1
Total, I = 342.0
Molecular weight = 106.16