You are on page 1of 45

Black oil is characterized by low GOR< 500 scf/stb and API of 30 (Todd)

Globally, oil wells produce about 220 million BWPD (barrels of water per day)roughly three
barrels of water for every barrel of oil. In older fields, the water "cut," or ratio-of-water-to-oil,
can be 95% or higher. (Hallburton). From literature and published data the data for computations
were obtained for the design. See table below

In designing three-phase separators, it is a good idea to examine both vertical and horizontal
configurations for the specific case, and decide on the one that meets the physical requirements,
stream specifications, and economic attractiveness. (Economides, 2009).Also the available plot
space (foot print) may be a factor in the choice of separator (Wayne D. Monnery and William Y.
Svrcek).

Table 1 Field Data Used for Separator Design

PARAMETERS
Qg 1.8 MMscf/ Gas rate
d
Qo 6000 bbl/d Oil rate
Qw 14000 bbl/d Water rate
o
'o 30 API Oil density
o 0.876 Oil specific gravity
o 20 cP Oil viscosity
w 1 cP Water viscosity
w 1.073 Water specific gravity
g 0.75 Gas specific gravity
Z 0.94 Gas compressibility
P 218 Psia Operating pressure
o
T' 70 F Operating temperature
o
T 530 R Operating temperature
tw 20 minutes Water retention time
to 20 minutes Oil retention time
g 0.015 cP Gas viscosity
dm,w 500 Micron Smallest water droplet size to be removed
dm,o 100 Micron Smallest oil droplet size to be removed
3
g 0.886 lb/ft Gas operating density
3
54.66 lb/ft Oil operating density
o
2
66.95 lb/ft3 Water operating density
w
5

Table 2 Retention tim for Liquid - Liquid Separation (Plisga, 2005)

Typical Retention Time for Liquid Liquid Separation


Type of Separator Retention Time
Hydrocarbon/ water separators
Above 35 API hydrocarbons 3 to 5 min
Below 35 API hydrocarbons
100F and above 5 to 10 min
80F 10 to 20 min
60F 20 to 30 min
Ethylene glycol/hydrocarbon separators 20 to 60 min
(cold separators)
Amine/hydrocarbon separators 20 to 30 min
Coalescers, hydrocarbon/ water separators
100F and above 5 to 10 min
80F 10 to 20 min
60F 20 to 30 min
Caustic/propane 30 to 45 min
Caustic/Heavy gasoline 30 to 90 min

Design
This section involves the designing and Sizing of horizontal and vertical three phase Separators.
Three Design approaches that are acceptable in Industry are considered
Three Phase (Gas - Oil - Water) Horizontal /Vertical Separator: As per "Petroleum and
Gas Field Processing -Hussein K. Abdel-Aal, Mohamed Aggour, M. A. Fahim"
Three Phase (Gas Oil-Water) Horizontal/ Vertical Separator: As per API 12J
Three Phase (Gas Oil-Water) Horizontal/ Vertical separators as per Monnery and
Svrcek Separator Design

Three Phase (Gas - Oil - Water) Horizontal /Vertical Separator: As per "Petroleum and
Gas Field Processing -Hussein K. Abdel-Aal, Mohamed Aggour, M. A. Fahim"

This approach considers the gas capacity constraint and the liquid retention time constraint
results in developing two equations; each relates the vessel diameter to its length. Analysis of the
two equations determines the equation that governs the design and that should be used to
determine possible combinations of diameters and lengths. For three-phase horizontal separators,
consideration of the settling of water droplets in oil results in a third equation that determines the
maximum diameter of the separator. Therefore, in determining the vessels diameterlength
combinations, the diameters selected must be equal to or less than the determined maximum
diameter. (H. K. Abdel-Aal and Mohamed Aggour and M. A. Fahim , 2003)
Three Phase (Gas - Oil - Water) Horizontal Separator: As per "Petroleum and Gas Field
Processing -Hussein K. Abdel-Aal, Mohamed Aggour, M. A. Fahim"

Step 1
Determine Aw/A
Aw 0.5Q w t w 0.5 14000 20
= = =0.35
A Qo t o +Qw t w 6000 ( 20 ) +14000(20)

Figure 1 (Plisga, 2005)

From the above graph, we determine Ho/D


Ho/D = 0.12
Determine Oil pad thickness Homax , use dm = 500m
3 2
1.28 10 t o ( wo ) d m 1.28 103 ( 20 ) ( 1.0730.876 ) 5002
H omax= = =63.04
o 20
Determine maximum allowable vessel diameter Dmax
H omax 63.04
D max = = =525
Ho 0.12
D

Determine the drag coefficient, Cd


The drag coefficient is calculated using iterative methods
1. Assume a value for Cd (a value of 0.34 could be used as a first assumption).
o g d m 0.5
2. Calculate the velocity, u, from u=0.01186[ ] ft/s
g Cd
g dm C d
3. Calculate Re from = 0.0049 [ ]
g
24 3
C d= + + 0.34
4. Calculate Cd using and compare to the assumed value.

5. If no match is obtained, use the calculated value of Cd and repeat steps 24 until convergence
is obtained.
For this design assumed Cd = 0.34,

Table 3 Results for Iterative Values of Drag Coefficient (Cd)

U Re Cd %
Approximate
error a
0.3400 100
1.589 45.989 1.3040 73.9
0.8119 23.168 1.9991 34.8
0.656 18.986 2.2926 12.8
0.6123 17.723 2.4069 4.7
0.5976 17.296 2.4489 1.7
0.5924 17.146 2.4643 0.62
0.5906 17.094 2.4697 0.22

The table above shows results from the iteration with percentage approximate error
computations.
Hence we use Cd= 2.4697
Determine the gas capacity constraints, dm = 100 m
54.6620.886
0.886 2.4697
=34.85 inft
100()0.5

0.5
TZ Q g g Cd
DL=420
P( )(
d m ( o g )
=420
)530 0.94 1.8
218 ( )
Determine the combinations of D and L that satisfy the gas capacity constraint. Select diameters
smaller than the determined maximum diameter and determine the corresponding effective
length. The result is shown below.
Table 4 D and L Combinations that Satisfy Gas Constraint

D/in L/ft D/in L/ft D/in L/ft D/in L/ft


60 0.58 120 0.29 180 0.19 240 0.15
66 0.53 126 0.28 186 0.19 246 0.14
72 0.48 132 0.26 192 0.18 252 0.14
78 0.45 138 0.25 198 0.18 258 0.14
84 0.41 144 0.24 204 0.17 264 0.13
90 0.39 150 0.23 210 0.17 270 0.13
96 0.36 156 0.22 216 0.16 276 0.13
102 0.34 162 0.22 222 0.16 282 0.12
108 0.32 168 0.21 228 0.15 288 0.12
114 0.31 174 0.20 234 0.15 294 0.12

Investigation of the results however, shows that for any selected diameter, the effective length is
too small (not realistic). Therefore, the gas capacity does not govern the design but rather liquid
capacity governs the design.
Determine the liquid capacity (retention time) constraint
D 2 L=1.429 ( Q o t o +Q w t w ) =1.429 ( 6000 20+14000 20 ) =571600 2 ft

Determine the combinations of D and L that satisfy the liquid capacity constraint. Select
diameters smaller than the determined maximum diameter and determine the corresponding
effective length.

Table 5 D and L that Satisfy Liquid Capacity Constraint

D/in Le/ft Ls/ft slenderne D/in Le/ft Ls/ft slenderne


ss ratio ss ratio
66 131.2 174.9 31.8 186 16.52 22.03 1.4
2 6
72 110.26 147.0 24.5 192 15.51 20.67 1.3
2
78 93.95 125.2 19.3 198 14.58 19.44 1.2
7
84 81.01 108.0 15.4 204 13.74 18.31 1.1
1
90 70.57 94.09 12.5 210 12.96 17.28 1.0
96 62.02 82.70 10.3 216 12.25 16.34 0.9
102 54.94 73.25 8.6 222 11.60 15.46 0.8
108 49.01 65.34 7.3 228 11.00 14.66 0.8
114 43.98 58.64 6.2 234 10.44 13.92 0.7
120 39.69 52.93 5.3 240 9.92 13.23 0.7
126 36.00 48.01 4.6 246 9.45 12.59 0.6
132 32.81 43.74 4.0 252 9.00 12.00 0.6
138 30.01 40.02 3.5 258 8.59 11.45 0.5
144 27.57 36.75 3.1 264 8.20 10.94 0.5
150 25.40 33.87 2.7 270 7.84 10.45 0.5
156 23.49 31.32 2.4 276 7.50 10.00 0.4
162 21.78 29.04 2.2 282 7.19 9.58 0.4
168 20.25 27.00 1.9 288 6.89 9.19 0.4
174 18.88 25.17 1.7 294 6.61 8.82 0.4
180 17.64 23.52 1.6 300 6.35 8.47 0.3

The most common slenderness ratio (SR) for horizontal separators: 3 SR 5 . Also As per

GPSA, typical horizontal L/D ratios are normally in the range of: 2.5 SR 5 .

Suitable separators are those with diameters between 126 D 144 . Therefore, the
recommended separator size can be either 126 in. by 48 ft, or 132 in. by 44ft or 138 in. by 40 ft
or 144 in by 36 ft.
Normally, smaller diameter and longer separators are less expensive than the larger diameter and
shorter separators. The choice of the diameter of separator is based on cost and availability.
We would go in for 126 in by 48 ft. The selected separator gas capacity is calculated by
substituting the values of D and L and calculating the value of Qg in the gas capacity constraint
equation. For a 126-in. by 48-ft separator, the gas capacity,

DLP 126 48 218


Qg = 0.5
= 0.5
=15412.13 MMscf / day
g Cd 0.886 2.4697
[( )( )] 422 T Z [ ] 422 530 0.94
o g d m 54.6620.886 100

. This is much larger than the production rate of 1.8 MMSCFD.

Three Phase (Gas - Oil - Water) Vertical Separator: As per "Petroleum and Gas Field
Processing - Hussein K. Abdel-Aal, Mohamed Aggour, M. A. Fahim"
Gas capacity is use to determine minimum diameter of the vessel
Liquid retention time constraints is used to determine the height of the vessel
Step 1
The minimum diameter that satisfies the water droplets settling constraint is determined

6686 6000 20
( w o )d2m= 2
=16290.76 2
(1.0730.876)500
6686 Q o o
D2min =

Dmin =127.64
The minimum diameter that satisfies the gas capacity constraint is determine as
0.5 0.5
g C
2
D =5058 Qg
min
TZ
P ( )( d
o g d m ) =5058 ( 1.8 ) ( 530 0.94
218 )( 0.886
54.6620.886

2.4697
100 ) =419.70 2

D min =20.49
The larger of the two minimum diameters determined above is then considered as the minimum
allowable vessel diameter.

Table 6 Comparison between Minimum Diameters due to gas and Liquid Constraint

Parameter Dmin/in
Minimum diameter due 20.49
to gas capacity
constraint
Minimum diameter due 127.64
to retention time
constraint
The minimum diameter due to the water droplets settling constraint is the larger. Hence liquid
capacity governs the design
Dmin =127.64
For various values of diameter larger than the minimum allowable vessel diameter, combinations
of diameters and liquid heights are determined as
( H o + H w ) D2=8.576 ( Q o t o +Q w tw ) =8.576 ( 6000 20+14000 20 )=3430400 3
( H o + H w ) D2=3430400 3
We find combinations of liquid height and Diameter that satisfy the above liquid capacity. For
1
diameters greater than 36, Seam to Seam length is determine by Ls= 12 (H o + H w +40)

The table below summarizes the results

Table 7 Results for Combinations D, Ls, and Hw + Ho that satisfy Liquid Capacity Constraints

D/in Ho+Hw/ Ls/ft Slenderne D/in Ho+Hw/ Slenderne 12Ls/


in ss ratio in ss ratio d
12Ls/d Ls/ft
60 952.888 87.7 17.5 180 105.876 27.2 1.8
9 5
66 787.511 74.5 13.5 186 99.1559 27.1 1.7
5 7
72 661.728 64.5 10.7 192 93.0555 27.1 1.7
4 6
78 563.839 56.8 8.7 198 87.5012 27.1 1.6
6 8
84 486.167 50.8 7.3 204 82.4298 27.2 1.6
8 3
90 423.506 46.1 6.2 210 77.7868 27.3 1.6
2 5
96 372.222 42.4 5.3 216 73.5253 27.5 1.5
2 8
102 329.719 39.3 4.6 222 69.6047 27.6 1.5
3 4
108 294.101 36.8 4.1 228 65.9895 27.8 1.5
5 4
114 263.958 34.8 3.7 234 62.6488 28.1 1.4
1 4
120 238.222 33.2 3.3 240 59.5555 28.3 1.4
2 6
126 216.074 31.8 3.0 246 56.6858 28.6 1.4
6 4
132 196.877 30.7 2.8 252 54.0186 28.8 1.4
9 4
138 180.130 29.8 2.6 258 51.5353 29.1 1.4
2 6
144 165.432 29.1 2.4 264 49.2194 29.4 1.3
1 7
150 152.462 28.5 2.3 270 47.0562 29.8 1.3
2 4
156 140.959 28.1 2.2 276 45.0325 30.1 1.3
9 6
162 130.711 27.7 2.1 282 43.1366 30.4 1.3
8 6
168 121.542 27.5 2.0 288 41.3580 30.8 1.3
2
174 113.304 27.3 1.9 294 39.6871 31.1 1.3
3 7

Slenderness ratio (SR) for vertical separators: 1.5 SR 3 . Also as per GPSA, typical vertical

slenderness ratios are normally in the range: 2 SR 4 .


From the above results recommended separators are those with diameter between
126 D 228 . The final selection would, therefore, depend on cost and availability. A
126-in. by 32-ft separator is probably the best choice. (H. K. Abdel-Aal and Mohamed Aggour
and M. A. Fahim , 2003)

Three Phase (Gas Oil-Water) Horizontal/ Vertical Separator: As per API 12J

This approach involve the determination of gas capacityGas capacities of separators by modified
Stokes Law. When using Stokes Law. The capacity is based on the principle of the minimum
droplet size that will settle out of a moving gas stream at a given velocity. The maximum
allowable superficial velocity of the gas at operating conditions is determined.

Three Phase (Gas Oil-Water) Horizontal Separator: As per API 12J

This employs the use of Souder Browns Equation

Table 8 K factor values for the Souder Brown's Equation (Plisga, 2005)

Typical K Factor Values


Separator Type K factor /ft/s
Horizontal (w/vertical pad) 0.40 to 0.50
Vertical or Horizontal (w/horizontal 0.18 to 0.35
pad) 0.35
Atm. Pressure 0.33
@300 psig 0.30
@600 psig 0.27
@900 psig 0.21
@1500 psig
Spherical 0.20 to 0.35
Wet Steam 0.25
Most Vapors under Vacuum 0.20
Salt and Caustic evaporators 0.15

K= 0.34
Step 1
The maximum gas superficial velocity is determined by
o g 0.5 54.6620.886 0.5
V a=k [ ] =0.34 [ ] =2.649 ft /s
g 0.886

Molar mass of gas is M g= g M air =0.75 29=21.75lb/mol


Qg M g 1.8 106 21.75 lb
Mass flow rate of gas = = =1.194
379.4 86400 379.4 86400 s
1.194 3
Gas flow rate n ft3/s Qg = g = 0.886 =1.348 ft / s

Q g 1.348
Minimum gas flow area = =0.5089 ft 2
V a 2.649

Minimum separator diameter Dmin =


0.5089 144
0.7854
=9.659

We determine the settling volume V


2
0.7854 D NLL H
v=
144 5.615
1440V
Liquid capacity of a separator (w) is given by w= t , where t= retention time

1440 2 1
w= 0.7854 D NLL H
20 144 5.615
Using t = 20 min Normal liquid level (NLL)
Maximum liquid load expected = 6000 + 14000 +60(extra for contingency) = 20060 bbl/d
2 20060 56150
D NLL H= =286826.8398
3927
We use generate different combination of separator diameter, normal liquid level and height to
find a suitable separator.
The table below shows the different combinations of D and L using half full (NLL =0.5)

Table 9 D and L combinations for Half Full Criterion

D/in L/ft Slenderness D/in L/ft Slendernes Qg/ ft3/s


Ratio(SR) s Ratio(SR)
60 318.70 63.7 120 79.67 8.0
66 263.39 47.9 126 72.27 6.9
72 221.32 36.9 132 65.85 6.0
78 188.58 29.0 138 60.25 5.2
84 162.60 23.2 144 55.33 4.6 299.60
90 141.64 18.9 150 50.99 4.1 325.08
96 124.49 15.6 156 47.14 3.6 351.61
102 110.28 13.0 162 43.72 3.2 379.18
108 98.36 10.9 168 40.65 2.9
114 88.28 9.3 174 37.89 2.6

Slenderness ratio (SR) for Horizontal separators: 3 SR 5 .From the table above the

recommended separators are those with diameter: 144 D 162 . Recommended


separators must be able to handle the expected gas capacity
That is the expected gas capacity must be less than the gas capacity of the selected separator
Expected gas capacity = 1.348 ft3/s
2
D 2
Gas capacity Qg =V a 4 =2.649 0.7854 D

From the table it can be seen that all the recommended separators have their gas capacity to be
greater than the expected.

Using 1/4 full criterion NLL= 0.25, the results is show below
Table 10 D and L combinations for One Quarter Full Criterion

D/in L/ft Slendernes D/in L/ft Slendernes Qg/ft3/s


s Ratio(SR) s Ratio(SR)
60 318.70 63.7 120 79.67 8.0
66 263.39 47.9 126 72.27 6.9
72 221.32 36.9 132 65.85 6.0
78 188.58 29.0 138 60.25 5.2
84 162.60 23.2 144 55.33 4.6 299.5955
90 141.64 18.9 150 50.99 4.1 325.082
96 124.49 15.6 156 47.14 3.6 351.6087
102 110.28 13.0 162 43.72 3.2 379.1756
108 98.36 10.9 168 40.65 2.9
114 88.28 9.3 174 37.89 2.6

Based on same reason as above the recommended separators are those with diameters:
144 D 162 .

Using a criterion of 1/3 full NLL=1/3, the results is shown below

Table 11 D and L combinations for One Third Full Criterion

D/in L/ft Slenderness D/in L/ft Slenderness Qg/ft3/s


Ratio(SR) Ratio(SR)
60 239.02 47.8 120 59.76 6.0
66 197.54 35.9 126 54.20 5.2
72 165.99 27.7 132 49.38 4.5 251.74
78 141.43 21.8 138 45.18 3.9 275.15
84 121.95 17.4 144 41.50 3.5 299.60
90 106.23 14.2 150 38.24 3.1 325.08
96 93.37 11.7 156 35.36 2.7
102 82.71 9.7 162 32.79 2.4
108 73.77 8.2 168 30.49 2.2
114 66.21 7.0 174 28.42 2.0

Based on same reason as above the recommended separators are those with diameters:
132 D 150

Using a criterion of 60% full, NLL = 6/10

Table 12 D and L combinations for 60% Full Criterion


D/in L/ft Slenderness Qg/ft3/s D/in L/ft Slenderness
Ratio(SR) Ratio(SR)
60 132.79 26.6 126 30.11 2.9
66 109.74 20.0 132 27.44 2.5
72 92.22 15.4 138 25.10 2.2
78 78.57 12.1 144 23.05 1.9
84 67.75 9.7 150 21.25 1.7
90 59.02 7.9 156 19.64 1.5
96 51.87 6.5 162 18.22 1.3
102 45.95 5.4 168 16.94 1.2
108 40.98 4.6 168.52 174 15.79 1.1
114 36.78 3.9 187.77
120 33.20 3.3 208.05

Based on same reason as above the recommended separators are those with diameters:
108 D 120

Using a criterion of 70% full, NLL = 7/10

Table 13 D and L combinations for 70% Full Criterion

D/in L/ft Slendernes Qg/ft3/s D/in L/ft Slendernes


s Ratio(SR) s Ratio(SR)
60.00 113.82 22.8 120.00 28.46 2.8
66.00 94.07 17.1 126.00 25.81 2.5
72.00 79.04 13.2 132.00 23.52 2.1
78.00 67.35 10.4 138.00 21.52 1.9
84.00 58.07 8.3 144.00 19.76 1.6
90.00 50.59 6.7 150.00 18.21 1.5
96.00 44.46 5.6 156.00 16.84 1.3
102.00 39.38 4.6 150.32 162.00 15.61 1.2
108.00 35.13 3.9 168.52 168.00 14.52 1.0
114.00 31.53 3.3 187.77 174.00 13.53 0.9

Based on same reason as above the recommended separators are those with diameters:
102 D 114

Using a criterion of 80% full, NLL = 8/10


Table 14 D and L combinations for 80% Full Criterion

D/in L/ft Slenderness Qg/ft3/s D/in L/ft Slenderness


Ratio(SR) Ratio(SR)
60 99.59 19.9 120 24.90 2.5
66 82.31 15.0 126 22.58 2.2
72 69.16 11.5 132 20.58 1.9
78 58.93 9.1 138 18.83 1.6
84 50.81 7.3 144 17.29 1.4
90 44.26 5.9 150 15.93 1.3
96 38.90 4.9 133.15 156 14.73 1.1
102 34.46 4.1 150.32 162 13.66 1.0
108 30.74 3.4 168.52 168 12.70 0.9
114 27.59 2.9 174 11.84 0.8

Based on same reason as above the recommended separators are those with diameters:
96 D 108

Three Phase (Gas Oil-Water) Vertical Separator: As per API 12J

This employs the use of Souder Browns Equation

K= 0.34 refer to table 8.


Step 1
The maximum gas superficial velocity is determined by
o g 0.5 54.6620.886 0.5
V a=k [ ] =0.34 [ ] =2.649 ft /s
g 0.886

Molar mass of gas is M g= g M air =0.75 29=21.75lb/mol


Qg M g 6
1.8 10 21.75 lb
Mass flow rate of gas = 379.4 86400 = 379.4 86400 =1.194 s

1.194 3
Qg = = =1.348 ft / s
Gas flow rate n ft3/s g 0.886
Q g 1.348
Minimum gas flow area = =0.5089 ft 2
V a 2.649

Minimum separator diameter D min =


0.5089 144
0.7854
=9.659

We determine the settling volume V


2
0.7854 D NLL H
v=
144 5.615
1440V
Liquid capacity of a separator (w) is given by w= t , where t= retention time

1440 2 1
w= D NLL H
20 144 5.615
Using t = 20 min Normal liquid level (NLL)
Maximum liquid load expected = 6000 + 14000 +600(extra for contingency) = 20600 bbl/d
20600 56150
D 2 NLL H= =294545.001
3927
We generate different combination of separator diameter, normal liquid level and height to find a
suitable separator. The table below shows the different combinations of D and L using half full
(NLL =0.5)

Table 15 D and L combinations for Half Full Criterion

D/in H/ft SR D/in H/ft SR Qg


60 163.63778 32.7275557 30.933417 2.6898623 275.149
138 5 9 4
66 135.23783 24.588697 28.409336 2.3674447 299.595
144 5 1 5
72 113.63735 18.9395577 26.182044 2.0945635
150 5 6 325.082
78 96.827088 14.896475 24.206771 1.8620593 351.608
156 9 8 7
84 83.488662 11.9269518 22.446883 1.6627320 379.175
162 2 9 6
90 72.727901 9.69705353 20.872165 1.4908689 407.782
168 6 7 8
96 63.921007 7.9901259 19.457524 1.3418982
174 2 2
102 56.622069 6.66141984 18.181975 1.2121316
180 4 9
108 50.505487 5.61172079 17.027864 1.0985719
186 6 1
114 45.329024 4.77147626 15.980251 0.9987657
192 8 4
120 40.909445 4.09094446 15.026425 0.9106924
198 9 8
126 37.106072 3.53391164 14.155517 0.8326774
204 2 8
132 33.809458 3.07358712 0.7633249
210 13.358186 1
12.626371
216 8 0.7014651
11.953088 0.6461128
222 3 8
11.332256 0.5964345
228 1 3
10.758565
234 3 0.5517213

1.5 SR 3 .From the table above the


Slenderness ratio (SR) for vertical separators:
recommended separators are those with diameter: 138 D 168 . Recommended
separators must be able to handle the expected gas capacity
That is the expected gas capacity must be less than the gas capacity of the selected separator
Expected gas capacity = 1.348 ft3/s
2
D
Gas capacity Q g =V a =2.649 0.7854 D2
4
From the table it can be seen that all the recommended separators have their gas capacity to be
greater than the expected.

Using full criterion NLL= 0.25, the results is show below

Table 16 D and L combinations for One Quarter Full Criterion

D/in H/ft SR/ D/in H/ft SR Qg


60 327.27555 65.4551 150 52.3640 4.189127
7 1 9
66 270.47566 49.1773 156 48.4135 3.724119
7 9 4
72 227.27469 37.8791 162 44.8937 3.325464
2 2 7
78 193.65417 29.7929 168 41.7443 2.981738 407.782
6 5 3 8
84 166.97732 23.8539 174 38.9150 2.683796 437.430
5 5 3
90 145.45580 19.3941 180 36.3639 2.424263 468.118
3 1 5
96 127.84201 15.9802 186 34.0557 2.197144 499.846
4 5 3
102 113.24413 13.3228 192 31.9605 1.997531 532.614
7 4 3
108 101.01097 11.2234 198 30.0528 1.821385 566.422
4 4 5 8
114 90.658048 9.54295 204 28.3110 1.665355 601.271
9 3 3 6
120 81.818889 8.18188 210 26.7163 1.52665 637.160
2 9 7 7
126 74.212144 7.06782 216 25.2527 1.40293
4 3 4
132 67.618916 6.14717 222 23.9061 1.292226
7 4 8
138 61.866834 5.37972 228 22.6645 1.192869
9 5 1
144 56.818673 4.73488 234 21.5171 1.103443
9 3
Based on same reason as above the recommended separators are those with diameters:
168 D 210 .

Using a criterion of 1/3 full NLL=1/3, the results is shown below

Table 17 D and L combinations for One Third Full Criterion

D/in H/ft SR D/in H/ft SR Qg/ft3/s


60 327.27555 65.4551 150 52.3640 4.18912
7 1 9 7
66 270.47566 49.1773 156 48.4135 3.72411
7 9 4 9
72 227.27469 37.8791 162 44.8937 3.32546
2 2 7 4
78 193.65417 29.7929 168 41.7443 2.98173 407.782
6 5 3 8 8
84 166.97732 23.8539 174 38.9150 2.68379 437.430
5 5 6 3
90 145.45580 19.3941 180 36.3639 2.42426 468.118
3 1 5 3
96 127.84201 15.9802 186 34.0557 2.19714 499.846
4 5 3 4
102 113.24413 13.3228 192 31.9605 1.99753 532.614
7 4 1 3
108 101.01097 11.2234 198 30.0528 1.82138 566.422
4 4 5 5 8
114 90.658048 9.54295 204 28.3110 1.66535 601.271
9 3 3 5 6
120 81.818889 8.18188 210 26.7163 1.52665 637.160
2 9 7 7
126 74.2121447.06782 216 25.2527 1.40293
4 3 4
132 67.6189166.14717 222 23.9061 1.29222
7 4 8 6
138 61.8668345.37972 228 22.6645 1.19286
9 5 1 9
144 56.8186734.73488 234 21.5171 1.10344
9 3 3
Based on same reason as above the recommended separators are those with diameters:
168 D 210 .

Using a criterion of 60% full, NLL = 6/10

Table 18 D and L combinations for 60% Full Criterion

D/in H/ft SR Qg/ft3/s D/in H/ft SR Qg/ft3/s


60 136.364 27.2729 150 21.8183 1.74547 325.082
8 6 7
66 112.6982 20.4905 156 20.1723 1.55171 351.6087
8 1 6
72 94.6977 15.7829 162 18.7057 1.38561
9 6 4
78 80.6892 12.4137 168 17.3934 1.24239
4 3 7 1
84 69.5738 9.93912 174 16.2146 1.118249
9 6
90 60.6065 8.08087 180 15.1516 1.01011
8 8 5
96 53.2675 6.65843 186 14.1898 0.91547
1 8 9 7
102 47.1850 5.551183 192 13.3168 0.83230
6 8 5
108 42.0879 4.67643 198 12.5220 0.75891
1 4 2
114 37.7741 3.97623 204 11.79626 0.69389
9 8
120 34.0912 3.40912 210 11.13182 0.63610
4
126 30.9217 2.94492 229.377 216 10.5219 0.58455
3 6 8 8 4
132 28.1745 2.56132 251.743 222 9.96090 0.53842
5 3 5 7 7
138 25.7778 2.24155 275.149 228 9.44354 0.49702
5 2 4 7 9
144 23.6744 1.97287 299.595 234 8.96547 0.45976
5 1 5 1 8
Based on same reason as above, the recommended separators are those with diameters:
126 D 156 .

Using a criterion of 70% full, NLL = 7/10

Table 19 D and L combinations for 70% Full Criterion

D/in H/ft SR Qg/ft3/s D/in H/ft SR Qg/ft3/s


60 116.884 23.3768 150 18.7014 1.49611 325.08
1 3 6 7 2
66 96.5984 17.5633 156 17.2905 1.33004
5 5 5 2
72 81.1695 13.5282 162 16.0334 1.18766
3 6 9 6
78 69.1622 10.6403 168 14.9086 1.06490
1 4 9 6
84 59.6347 8.51925 174 13.8982 0.95849
6 1 3 9
90 51.9485 6.92646 180 12.9871 0.86580
7 3 8
96 45.6578 5.70723 186 12.1627 0.78469
6 3 6 4
102 40.4443 4.75815 192 11.4144 0.71340
3 7 7 4
108 36.0753 4.00837 198 10.7331 0.65049
5 2 6 5
114 32.3778 3.40819 204 10.1110 0.59477
7 7 8
120 29.2210 2.92210 208.052 210 9.54156 0.54523
3 3 5 1 2
126 26.5043 2.52422 229.377 216 9.01883 0.50104
4 3 8 7 6
132 24.1496 2.19541 251.743 222 8.53792 0.46150
1 9 5 9
138 22.0953 1.92133 275.149 228 8.09446 0.42602
4 9 5
144 20.2923 1.69103 299.595 234 7.68469 0.39408
8 2 5 7
Based on same reason as above the recommended separators are those with diameters:
120 D 150 .

Using a criterion of 80% full, NLL = 8/10

Table 20 D and L combinations for 80% Full Criterion

D/in H/ft SR Qg/ft3/s D/in H/ft SR


60 102.273 20.4547 150 16.3637 1.30910
6 2 8 2
66 84.5236 15.3679 156 15.1292 1.16378
5 4 3 7
72 71.0233 11.8372 162 14.0293 1.03920
4 2 8
78 60.5169 9.31029 168 13.0451 0.93179
3 7 3
84 52.1804 7.45434 174 12.1609 0.83868
1 5 5 6
90 45.4549 6.06065 180 11.3637 0.75758
4 8 3 2
96 39.9506 4.99382 186 10.6424 0.68660
3 9 2 7
102 35.3887 4.16338 192 9.98765 0.62422
9 7 7 9
108 31.5659 3.50732 198 9.39151 0.56918
3 5 6 3
114 28.3306 2.98217 187.767 204 8.84719 0.52042
4 3 3 8 3
120 25.5684 2.55684 208.052 210 8.34886 0.47707
5 6 8
126 23.1913 2.20869 229.377 216 7.89148 0.43841
5 8 2 6
132 21.1309 1.92099 251.743 222 7.47068 0.40382
1 2 5 1
138 19.3333 1.68116 275.149 228 7.08266 0.37277
9 4 4 2
144 17.7558 1.47965 299.595 234 6.72410 0.34482
4 3 5 3 6
Based on same reason as above the recommended separators are those with diameters:
114 D 144 . ( American Petroleum Institute , 1989)

Monnery and Svrcek Separator Design

Monnery and Svrcek (1994) suggested that for horizontal separators, different devices can be
used to control the interface level such as a boot, a weir, or the combination of a bucket and weir.
A boot is used when the volume of heavy liquid is <1520 wt%, while a weir is used when the
volume is much greater. The bucket and weir type design is used when the interface level control
may be difficult, such as heavy oil, or when large amounts of emulsions or paraffins are present
(Arnold and Stewart, 1986).

This section looks at the design of three phase separators using procedures introduced by
Monnery and Svrcek (1994).

Vertical separators as per Monnery and Svrcek Separator Design 1994


Figure 2 Vertical Three phase Separator Monnery and Svrcek 1994 (Economides, 2009)

Table 21 Symbols Used in this section (Economides, 2009)

Symbol Nomenclature
A Vertical vessel cross-sectional area, ft2
AD Downcomer cross-sectional area, ft2
AL Equals A - AD
D Vessel diameter, ft or in
H Height, ft
HA Liquid Level above Baffle, in or ft
HBN Liquid height from above to feed nozzle, ft
HD Disengagement Height, ft
HH Holdup Height, ft
HL Height from Liquid interface to light nozzle, ft
HR Height from liquid nozzle to baffle, ft
HS Surge Height, ft
HT Total Vertical Separator Height, ft
WD Downcomer chord width, in

Table 22 Ks Values For some Sytems (Economides, 2009)

KS Values for Some Systems (Monnery and Svrcek 1994)


Light Phase Heavy Phase Minimum Droplet KS
Diameter, m
Hydrocarbons
SG at 60F < 0.85 Water or Caustic 127 0.333
SG at 60F < 0.85 Water or caustic 89 0.163
Water Furfural 89 0.163
Methylethyl Water 89 0.163
ketone
sec-Butyl alcohol Water 89 0.163
Methyl isobutyl Water 89 0.163
ketone
Nonyl alcohol Water 89 0.163

K= 0.34 refer to table 8 LL = Light Liquid (Oil) HL = Heavy Liquid V= Vapour (Gas)
Step 1
Determine terminal velocity

V T =k O

g
g
=0.34

55.6620.886
0.886
=2.649
ft
s
For conservative design
ft
V V =0.75V T =0.7 2.649=1.987
s
Determine the mass flowrate of the fluids (W)

1.8 10 6 ft 3 0.886 lb/ft


Gas mass flowrate W V = 24 h
=66450 lb/h

bbl 622.4 lb
6000 5.615
Oil mass flowrate W = d ft lb
=76730
24 hours h
bbl lb
14000 62.4 5.615
Water mass flowrate W HL= d ft
=204390 lb/h
24
Step 2
Determine the density of the liquid mixture of oil and water at the separator operating conditions
by
W W HL 76730 204390
L= O + w =54.662 + 66.955 =63.600lb/ft 3
W +W HL W +W HL 76730+204390 76730+204390
Step 3
The vapour volumetric flowrate is determined by
WV 66450
Qg = = =20.833 ft 3 /s
3600 g 3600 0.886
Step 4
Calculate the vessel internal diameter, Di
4 Q g 0.5 4 20.833 0.5
D i=( ) =( ) =13.349 ft
VV 1.987
With mist eliminator, add 3 to 6in to Di to accommodate a support ring. For this design 5in
(0.416 ft) was added.
D=13.349+0.416=13.765 ft

But we round it up to the next half foot increment D=14 ft=168


However if the design was without a mist extractor no need for an addition of an allowance

Step 5
The settling velocities are determined, from table Ks = 0.163
66.95554.662
V HL =k s HL =0.163 =4.02 min
0.5
HL 66.95554.662
V =k s =0.163 =2.004 min
HL 1

Step 6
Calculate light/heavy liquid volumetric flow rate
W 76730
Q = = =23.395 ft 3 /min
60 60 66.955
W HL 204390
QHL = = =50.877 ft 3 /min
60 HL 60 66.955

Step 7
Calculate the settling times (ts)
HL and HH are taken to be 1ft (minimum values)
12 H L 12 1
t s , HL= = =2.985 min
V HL 4.02
12 H H 12 1
s ,= = =5.988 min
V 2.004
t
Step 8
Calculate the baffle plate Area, AL
The vertical vessel cross-sectional Area is determined as follows
2
D2 (14)
A= = =153.938 ft 2
4 4
Calculate the downcomer cross-Sectional area, AD
Q + QHL
A D=7.48 60
G
G = Baffle liquid load(gallons per hour per square feet gph/ft2 )and High liquid level above the interface
HL+HR
Minimum value of HR =1 ft = 9 in, HL = 1 ft =12 in
Hence HL+HR = 12 +12 =24 in
Using the above values G is obtained from the graph below

Figure 3Obtain G from the Downcomer Allowable Flow (Monnery and Svrcek, 1994) (Economides, 2009)

From graph G = 9800 gph/ft2


23.395+50.877
A D=7.48 60 =3.401 ft 2 take WD = 4 in, and find x
9800
WD 4
x= = =0.024
D 168
Determine y from
a+cx + e x2 + g x 3+ i x 4
y=
1+bx+ d x 2 + f x 3 +h x 4
a= 4.755930103
b= 3.924091
c= 0.174875
d= 6.358805
e= 5.668973
f= 4.018448
g= 4.916411
h= 1.801705
i= 0.145348.

Solving resulted in y=0.002419


AD 2
But y= A and hence A D= Ay=153.938 0.002419=0.372 ft
2 2
The larger of the two AD is always used , 3.401 ft >0.372 ft
A L= A A D=153.9383.401=150.537 ft 2

Step 9
Calculate the residence time tr
H A 1 150.537
r ,= L L = =6.435 min
Q 23.395
t
H L A 1 153.938
t r , HL= = =3.025 min
Q HL 50.877
s , ,the liquid separation is controlling the design .The diameter neeed
r ,<t s , HL t r , HL <t be increasedthen
If t
The procedure must be repeated from step 7

Table 23 Comparison of Settling Times and Residence Times for Light and Heavy Component

Parameter ts/min tr/min

HL 2.993 3.025

LL 5.988 6.35

From the above table tr > ts.Hence we proceed


Step 10
Calculate height of liquid above outlet (Hold up height) tH =25 min
Q t H 23.395 25
H R= = =3.885 ft
AL 150.537
Calculate the surge height, ts = 5 min
( Q +Q ) (23.395+50.877)
H S =t S HL =5 =2.412 ft
A 153.938

Step 11
Calculate volumetric flowrate of mixture( gas, oil and water)
Q +Q 23.395+50.877
Qm=Q g + HL =20.833+ =22.071 ft 3 /s
60 60
The mixture density is also determined
23.395+50.877
= =0.056
74.272+20.833 60
m= L + V ( 1 )=63.600 ( 0.056 ) +0.886 ( 10.056 )=4.398
Step 12
The minimum value of HA is assumed to be 6 in (0.5 ft)
HBN = 0.5dN + greater of (2 ft or Hs + 0.5 ft). Here dN is the inlet or outlet vapor/liquid nozzle
diameter in ft and can be calculated by
1 0.5
4 Qm
dN (60 ) (
m 2 =
4 22.071
60 )
4.398 =0.991 ft=11.892 .

dN is taken to be 2 ft dN = 2 ft
HBN = 0.5dN + greater of (2 ft or Hs + 0.5 ft)
H BN =0.5 d N + greater of ( H s+ 0.5 )=0.5 (2 )+1 greater than ( 2.412+ 0.5 )=4.412 ft=4.5 ft

HD = 0.5 Dor minimum of 3 ft + 0.5dN (without mist eliminator), or 2 ft + 0.5dN


(with mist eliminator).
H D =2+0.5 d N =2+ 0.5 ( 2 ) =3 ft minimum value

H D =0.5 D=0.5 14=7 ft


Calculate the vessel total height
H T =H H + H L + H R + H A + H BN + H D +1.5 ( for mist eliminator )=1+1+4 +0.5+ 4.5+7+1.5=19.5 ft
The last step is to make sure that the ratio of HT/D is within a reasonable range (1.56.0).
HT/D = 19/14 = 1.4 < 1.5. HR is increase to 6ft and HD = 9 ft
H 23.5
H T =1+ 1+ 6+0.5+ 4.5+9+1.5=23.5 fthence T = =1.7
D 14
'' '
Hence the recommended separator is 14ft by 23.5 ft (168 23 , 6 ' ')
Horizontal separators as per Monnery and Svrcek Separator Design

Figure 4 Three-phase horizontal separator with a weir (Monnery and Svrcek, 1994) (Economides, 2009)

Table 24 Symbols used for Horizontal Separator Design (Economides, 2009)

Symbol Nomenclature
Dv Vessel diameter, ft or in
dN Inlet or outlet nozzle diameter, ft or in
H Height, ft
HD Disengagement height, ft
HH Holdup height, ft
HLIN HLL to inlet nozzle centerline height, ft
HLLL Low liquid Level
HLL High Liquid Level
HS Surge height, ft
HT Total vertical height, ft
NLL Normal liquid level

Table 25 Low Liquid Level (LLL) for Different Vessel Diameters and Operating Conditions (Economides, 2009)

Vessel Vertical LLL, in Horizontal LLL,


Diameter,ft <300 psia >300 psia in

4 25 6 9
6 25 6 10
8 15 6 11
10 6 6 12
12 6 6 13
16 6 6 15

Table 26 Liquid Holdup and Surge Times (Wayne D. Monnery and William Y. Svrcek)

Liquid Holdup and Surge Times


Service Holdup time, Surge time, min
min (NLL-HLL)
(NLL-LLL)
A. Unit drum 10 5
B. Separators
1. Feed to column 5 3
2. Feed to other drum or tankage
a. With pump or through exchanger 5 2
b. Without pump 2 1
3. Feed to fired heater 10 3
C. Reflux or product accumulator
1. Reflux only 3 2
2. Reflux and Product (Based on reflux 3min 3+ 2+
plus appropriate holdup time of overhead
product (as per B 1-3) )
D. Column bottoms
1. Feed to another column 5 2
2. Feed to other drum or tankage
a. With pump through exchanger 5 2
b. Without pump 2 1
3. Feed to fired reboiler(Based on reboiler 5-8 2-4
vapor expressed as liquid 3min plus
appropriate holdup time for the bottom
product(as per D 1,2))
E. Compressor suction/interstage scrubber
3 min between HLL(High liquid alarm) and high level shutdown
10 min from bottom tangent line to high liquid alarm
F. Fuel gas knock-out drum
20 ft slug in the incoming fuel gas line between NLL and high level shutdown
G. Flare Knock-out drum
20 to 30 min to HLL
Multiply by the following factors (optional)
Personnel Factor Instrumentation Factor
Experienced 1.0 Well instrumented 1.0

Trained 1.2 Standard Instrumented 1.2

Inexperienced 1.5 Poorly Instrumented 1.5

K= 0.34 refer to table 8 LL = Light Liquid (Oil) HL = Heavy Liquid V= Vapour (Gas)


V T =k O
g
g
=0.34

55.6620.886
0.886
=2.649
ft
s
For conservative design
ft
V V =0.75V T =0.7 2.649=1.987
s

We determine the mass flowrate of the fluids (W)


6 3
1.8 10 ft 0.886 lb/ft
Gas mass flowrate W V = 24 h
=66450 lb/h

bbl 622.4 lb
6000 5.615
Oil mass flowrate W = d ft lb
=76730
24 hours h
bbl lb
14000 62.4 5.615
Water mass flowrate W HL= d ft
=204390 lb/h
24

We determine the density of the liquid mixture of oil and water at the separator operating
conditions by
W W HL 76730 204390
L= O + w =54.662 + 66.955 =63.600lb/ft 3
W +W HL W +W HL 76730+204390 76730+204390
The vapour volumetric flowrate is determined by
WV 66450
Qg = = =20.833 ft 3 /s
3600 g 3600 0.886

Hold up time = 5 mins and surge time = 3 min


The hold up and surge volumes are calculated as,
V H =t H Q =60 5 23.395=7018.5 ft 3

V S =t S Q =60 3 23.395=4211.1 ft 3
Obtain L/D from table below
L/D Ratio Guidelines (Monnery and Svrcek, 1994)

Table 27 L/D ratios for different Separator Operating Conditions (Economides, 2009)

Vessel operating L/D


pressure, psig
0 < p < 250 1.5 3
250 < p < 500 3.0 4.0
p > 500 4.0 6.0

L/D = 3
The diameter of the vessel is determined
S
V H +V
1
3
16 (7018.5+4211.1) 13
16( 0.6 ( L ) ) =( ) =31.67 ft
D 0.6 3

D=
The total crossectional area is determined
2
D 2 ( 31.67 )
AT = = =787.75 ft
4 4
The vapour space height HV is set to the larger of 0.2D or 2ft (1 ft if there is no mist eliminator).
H V =0.2 ( 31.67 )=6.334 ft
H V 6.334
x= = =0.2
D 31.67
Determine y
a+cx + e x2 + g x 3+ i x 4
y=
1+bx+ d x 2 + f x 3 +h x 4
a= 4.755930103
b= 3.924091
c= 0.174875
d= 6.358805
e= 5.668973
f= 4.018448
g= 4.916411
h= 1.801705
i= 0.145348.

Solving gives y = 0.139392

Determine crossectional area of vapour phase


2
A V = y A T =0.139392787.75=109.81 ft
Calculate the low liquid level in the light liquid compartment by
H LLL ( ) =0.5 D ( ft ) +7=0.5 (31.67 )+ 7=22.835 23
Round HLLL up to the nearest inch.
H LLL 23
Determine x= D = 12 31.67 =0.06052
2 3 4
a+cx + e x + g x + i x
Determine y y=
1+bx+ d x 2 + f x 3 +h x 4

Solving results in y = 0.020986


2
Determine A LLL= y A T =0.020986 787.75=16.532 ft

Calculate the weir height Hw


H W =DH V =31.676.334=25.336 ft

Calculate minimum length of liht liquid compartment to accommodate the light liquid hold up
/surge
V H +V S 7018.5+4211.1
L2 = = =16.98 ft
A T A V A LLL 787.75109.8116.532
Set interface at height Hw/2
H 25.336
H HL =H = W = =12.668 ft
2 2
Calculate crossectional area of heavy liquid by
H HL 12.668
x= = =0.4 and solving for y = 0.370818
D 31.67
2
A HL= y AT =0.370818 787.75=292.112ft
Crossectional area of light liquid
A =A T AV A HL=787.75109.81292.112=385.828 ft 2
Step 10
Calculate the rising velocity of oil droplets from water (VLH) and that of water droplet from oil
(VHL)
66.95554.662
V HL =k s HL =0.163 =4.02 min
0.5
HL 66.95554.662
V LH =k s =0.163 =2.004 min
HL 1
Calculate the settling times

12 H 12 12.668
t s , HL= = =37.81 min
V HL 4.02
12 H HL 1212.668
s ,= = =75.86 min
V 2.004
t
Determine L1
t S , Q t S , HL Q
HL
,
A HL A

L1=MAX
t S , Q 75.86 50.877 t Q 37.81 23.395
HL
= =13.21 ft > S , HL = =2.29 ft
A HL 292.112 A 385.828

Hence L1 = 13.21 ft = 13.5 ft

Total length L = L1 + L2 = 13.5 + 16.98 = 30.48 = 30.5 ft


L/D = 30.5/31.67 = 0.963 < 1.5
So we decrease D in steps of 6 in to obtain a good L/D ratio
L 30.5
1.5 = =2.1 6.0
D 174
12
Hence recommended separator is 17430.5 ft (174' ' 30 ' , 6 ' ') (Economides, 2009) (Plisga,
2005) (Saeid Mokhatab, 2006)
Operational problems

Paraffin

Coalescing plates in the liquid section and mesh pad mist extractors in the gas section are
particularly prone to clogging by accumulations of paraffin waxes. Manways, hand-holes, and
nozzles should be provided to allow steam, solvent, or other types of cleaning of the separator
internals. Also, the bulk temperature of the liquid should always be kept above the cloud point of
the crude oil to prevent paraffin wax formation in the separators (Saeid Mokhatab, 2006).

Foam

Foaming crudes offer a special problem in sizing separators. Foam is a mixture of gas dispersed
in a liquid and having a density less than the liquid but greater than the gas. Greater interface
area and longer retention time are needed to remove the gas from the liquid. Horizontal
separators normally give the largest interface area. Retention times of as high as 15 minutes may
be necessary. However, a retention time of 2 to 5 minutes is sufficient in most cases for the
separators to handle foaming crudes. ( American Petroleum Institute , 1989). Foam depressants
will often be effective in increasing the capacity of a given separator. Foam can be reduced by
(1) using a defoaming pack, (2) using defoaming chemicals, and (3) utilizing heat to break it
down. (Saeid Mokhatab, 2006)

Sand

Sand can be very troublesome in separators by causing cutout of valve trim, plugging of
separator internals, and accumulation in the bottom of the separator, thus leading to level control
problems. Traditionally, sand has only been removed once it has collected in the main production
separators. However, removal of sand upstream of these separators reduces sand problems to a
minimum, giving substantial operational benefits. To meet these needs, the Mozley Wellspin
desander has been developed to remove sand effectively in simple, compact systems based on
solid/liquid hydrocyclones, which remove the sand before it enters the separator (NATCO,
2002b). It should be noted that sand problems may be solved by using a filter or desanding
cyclone before the separator; however, filters will quickly block in sandy service and are not
often used.

Sloshing

More and more separators are installed on floaters and other installations subject to motion, e.g.
Tension Leg Platforms. The performance of these separators can be adversely affected by the
motion imposed by waves and wind. The accompanying sloshing will compromise liquid
handling capacity, separation efficiency and the functioning of level instrumentation. A common
way to mitigate the negative impact of sloshing is to install perforated baffles to limit the liquid
motion and damp the waves. However, the layout of these baffles and the selection of the net free
area is critical for their success. With decreasing Net Free Area (NFA) the damping will become
more effective, but in the case of oil/water mixtures the increasing fluid velocity through the
holes may lead to redispersion. (Shell, 2007)

Liquid Carryover and Gas Blowby

Liquid carryover occurs when free liquid escapes with the gas phase and can indicate high liquid
level, damage to vessel internals, foam, improper design, plugged liquid outlets, or a flow rate
that exceeds the design rate of the vessel. Gas blowby occurs when free gas escapes with the
liquid phase and can be an indication of low liquid level, vortexing, or level control failure.

Emulsions

Emulsions can be particularly troublesome in the operation of three-phase separators. Over a


period of time an accumulation of emulsified materials and/or other impurities usually will form
at the interface of the water and oil phases. In addition to adverse effects on the liquid level
control, this accumulation will also decrease the effective oil or water retention time in the
separator, with a resultant decrease in wateroil separation efficiency. The addition of chemicals
and/or heat often minimizes this difficulty. Frequently, it is possible to appreciably lower the
settling time necessary for oilwater separation by either the application of heat in the liquid
section of the separator (heat can be added through recycling) or the addition of demulsifying
chemicals.

Scales

Since water injection is being done, there is a likelihood of scale formation. If injected fluids are
not compatible with formation water, scales could results. In this case the field is characterize by
high water cut with high concentration of Barite ion.
Mixing of two incompatible waters usually due to water flooding processes, most commonly
formation water rich in cations such as barium, calcium and/or strontium, mixing with sulfate
rich seawater, goes to the precipitation of sulfate scale which are deposited in the wellbore,
production facilities(Separators) and near wellbore. This mechanism is referred to as
incompatible mixing. Below is the chemical equation for the process. (Yassin, Study of Scale
Formation in Oil Reservoir During Water injection-A review, 2007)
Ba2+ (aq) (Sr2+ or Ca2+) + SO42-(aq) BaSO4(S) (SrSO4 or CaSO4)
Hence there is a need to make provisions for scale control and scale removal

Bader gave a detailed possible locations of scale deposits throughout the flow path of water. The
Figure below depicts the possible locations.
Figure 5 Possible Locations for Scale Deposition (J. Moghadasi, 2007)

Case 1: At the surface water injection facility where incompatible sources of water are
mixed prior to injection.
Case 2: In injection wells where the injected water starts to mix with the reservoir
formation water.
Case 3: Downhole in the reservoir where the injected water displaces reservoir formation
water.
Case 4: Downhole in the reservoir where the mixed injected water and formation water
are about to reach the range of producing wells.
Case 5: Downhole in the reservoir where the mixed (injected and formation) waters are
within the range of producing wells.
Case 6: At the connection of a branched zone where each branch produces different
water.
Case 7: At the manifold of a producing zone where water is produced from different
blocks within the same producing zone.
Case 8: At topside facility where produced fluids are mixed from different production
zones to separate oil and gas from produced waters, or in pipelines that transport
produced fluids to on-shore processing facilities.
Case 9: At disposal wells where produced water is injected for final disposal (Bader,
2006) (J. Moghadasi, 2007)
Problems caused by scale

One major problem caused by scales is corrosion to both downhole and surface equipment. One
important condition for corrosion is that materials surface must be dampened by an electrolyte,
which is a water solution that can conduct an electric current. Naturally occurring reservoir water
contains dissolve salt which makes it a good electrolyte. Some salts do dissolve with a decrease
in temperature and pressure rather than the general principle of precipitations with increasing
temperature and pressure hence creating a good electrolyte. Other factors contributing to
corrosion are pH, salt concentration, and oxygen concentration, along with the velocity of the
flowing medium. Also as a result of turbulence flow of fluid during production, scale deposits or
fragments that are carried in the flowing medium tend to scratch and erode the surfaces of
tubulars and equipment, causing a reduction in the corrosion resistance ability of materials
(Catherine Houska,CSI, TMR Consulting) (Mona El-Said, 2008).

Scales also cause equipment damage including coating and damaging downhole completions
equipment, surface valves and causing restriction to flow in flow conduits by reducing the
internal radius of tubulars due to the growth or the increase of scale thickness. Once scales are
formed, if not treated will continue to grow thicker. Calcium carbonate scale in production tubing
can grow and obstruct over 40% of the flowing area of the tubular whiles restricting access to
lower sections for workovers.

Formation damage around wellbores of both injection and producing wells is caused by scale
formation. Scales can develop around wellbore and cause a reduction in near wellbore porosity
and permeability and even block perforations (Mike Crabtree). Injection water at the wellhead is
usually of lower temperature as compared to reservoir temperature and when this water travels
along the injection pathway and well string the water cools the surrounding formations whiles
gaining heat (temperature) and pressure increment. If the water is saturated at surface conditions
with salts whose solubility decreases with increasing temperatures (e.g. anhydrite), scale may
form along the well-string. Also scale precipitation from the injection water may happen behind
the mixing zone resulting from pressure and temperature changes. This is particularly true of
waters containing salts whose solubility decreases with increasing temperature and decreasing
pressure (Yassin, Study of Scale Formation in Oil Reservoir During Water injection-A review,
2007).

Figure 6 Problems Caused by Scales


Scale Removal

Scale removal techniques must be quick, non-damaging to the separator and effective at
preventing re-precipitation. Best scale removal techniques depends on a knowledge of the type of
and quantity of scale and its physical composition and texture. A poor choice of scale removal
method can results in the promotion of rapid re-occurrence of scale or complex problems such as
formation of other precipitate as a result of the fluids used for treatment. In tubulars and surface
facilities, scale strength and texture play significant roles in the choice of removal technique.
Strengths and textures vary from delicate, brittle whiskers or crystals with high micro-porosity,
to rock-like, low-permeability, low-porosity layers. Scale purity affects its resistance to removal
methods. Scale may occur as single-mineral phases, but is more commonly a mixture of similar,
compatible compounds. For example, pure barium sulphate is hard and of low porosity and thus
not permitting penetration of treatment chemicals.. Mixtures of barium sulfate, often with
strontium sulfate, calcium sulfate or even calcium carbonate, will frequently yield to a variety of
removal methods, both chemical and mechanical. (Mike Crabtree, David Eslinger, Phil Fletcher
Matt Miller, Ashley Johnson, George King, 1999).

Chemical techniques

The first and often used approach to scale removal is chemical technique, especially when scale
is not easily accessible or exists where conventional mechanical removal methods are ineffective
or expensive to deploy.

Unlike salts of carbonate which are acid soluble, sulfate compounds of Barium, Strontium, or
Calcium are acid insoluble (example calcium Carbonate is soluble in Hydrochloric acid). A
major ionic component of seawater is sulfate ion (SO42-) which reacts with the following ions
(Ba2+, Sr2+ and Ca2+) which are naturally found in formation water depending on the fields
geological history. The above mention scales are sparingly soluble in water and as a result can
precipitate out and form deposits (Sulaiman, 2014).Therefore acid treatment cannot be use to
remedy barium scale deposit in separators.

Insoluble scale can be removed by the use of different scale dissolver chemicals. These are
chelate agents such as ethylene-diamine-tetra-acetic acid (EDTA) or diethylene-triamine-penta-
acetic acid (DTPA). A chelating agent is a complex molecule which breaks up the scale by
isolating and tying up the metallic ions in the scale. The application of these dissolvers is time
consuming and is dependent on high temperatures and circulation for optimum effect.
Calcium sulfates are soluble in many chelate dissolvers and is therefore the easiest sulfate scale
to handle. In contrast, barium sulfate is more difficult to handle, being very hard. (Tungesvik,
2013). Variations in the chemical structure of EDTA can also be used to remove non-carbonate
scale such as calcium sulphate and calcium/barium sulphate. Examples:
EDTA U105: used to dissolve CaCO3, iron carbonate and or iron oxide.
EDTA U104; used to dissolve wide variety of scale such as CaSO4 and even mixed scale
that is hard. (M.D. Garba, 2014)
Figure 7 EDTA compound structure (M.D. Garba, 2014)

Conventional mechanical method

This technique offers a wide range of tools, applicable depending on the areas of the scale
deposits, but also provides some limitation of ranges applicability. Therefore, selecting the good
method depends on the well position and the scale deposition. Mechanical methods are among
the successful methods in scale removal in tubular and surface facilities. Earlier, explosives
(Shock (e.g. String Shot) 1 to a maximum of 4 strands of 90 grain RDX) were used to rattle
pipe and break the brittle scale, but can easily cause damage to tubular, internal structures of
surface facilities. Very hard scale cannot be removed by explosive or penetrate by chemicals.
Thus, conventional is the next alternative option, this is most used by engineering tools and
machines. Conventional mechanical Method

Milling with carefully selected mills

Thick scales, especially those in tubulars, are often too strong for safe explosive removal and
have too little porosity for effective chemical treatments in a reasonable time frame. For these
deposits, removal usually requires techniques developed for drilling rock and milling steel.
Impact bits and milling technology have been developed to run on coiled tubing inside tubulars
using a variety of chipping bits and milling configurations. The downhole power source is
typically a hydraulic motor or a hammer-type impact tool. Motors are fluid-powered, stator and
rotor combinations that turn the bit. Because scale is rarely deposited evenly on the tubing wall,
milling power requirements vary enormously. When motors cannot supply the power needed for
the bit to cut the scale, the motor stalls and the milling process stops. As a result, scale-removal
rates vary with the type of scale and application, but generally range from about 5 to over 30
linear feet [1.5 to over 9 m] of scale removed from the tubular per hour of milling. The variation
in milling speed depends on the match between the type of deposit and the combination of motor
and mill.

Under-reaming under restrictions (under-reamers are slower than mills).


Impact tools like motors and mills usually need fullbore access and seldom clean scale
completely to the steel walls. For such partial access situations, under-reaming mills can increase
the effective diameter by moving the milling blades outward in response to pump pressure and
rate. Under-reaming mills are effective, but remove scale only at about half the rate of a typical
mill.

Jetting with Solid Beads (e.g. abrasives) damage potential is moderate to high.

Downhole fluid-jetting systems, such as Halliburtons Hydroblast and BJ-NOWSCOs RotoJet


system, have been available for many years to remove scales in production tubing and
perforations. Such tools use multiple jet orifices or an indexed jetting head to achieve full
wellbore coverage. These tools can be used with chemical washes to attack soluble deposits
wherever placement is critical to prevent bullheading reagent losses. Water jetting can be
effective on soft scale, such as halite, and debris or fill, but experience shows that it is less
effective on some forms of medium to hard scale such as calcite and barium sulfate
.
At surface pressure, water jetting removes scale by cavitation, whereby small bubbles form in the
fluid jet stream. These bubbles are created by the large pressure release as fluid passes through
the jet nozzle. The bubbles collapse on impact with scale, causing a forceful almost explosive
erosive effect.

Abrasive slurries; Adding a small concentration of solids, 1% to 5% by weight, to a water jet can
drastically improve its ability to cut through scale. Water jets using abrasive sand are widely used
in the construction and demolition industries for cutting reinforced concrete, and even in
demilitarization for cutting live ammunitions without generating heat or an ignition source.
This technique also shows superior cutting performance in calcium carbonate scale over water
jetting alone
.
Unfortunately, using abrasives such as sand can damage steel tubulars. When scale is completely
removed from tubing, the abrasive jet erodes the steel as efficiently as it does the scale. Should
the jetting tool stall, there is a significant risk of the abrasive jet perforating the steel tubing.
An abrasive jet that cuts scale without damaging tubing must exploit the difference in hardness
between wellbore scale and the underlying steel. One of the key differences between wellbore
scale and tubular steel is that while scale is brittle, steel is prone to ductile failure. A sharp sand
particle will erode the surface of ductile material by a cutting and plowing action. On the other
hand, a hard round particle will bounce off the surface, removing only a small volume of steel
and leaving an impact crater. Scale exhibits brittle failure, so the impact of a hard particle
fractures the scale and ultimately causes substrate disintegration. Scale breakdown is
independent of particle shape. Choosing round rather than sharp, angular particles promotes scale
erosion while reducing damage to steel tubulars. (M.D. Garba, 2014) (Mike Crabtree, David
Eslinger, Phil Fletcher Matt Miller, Ashley Johnson, George King, 1999) (King, 2009)
Removing hard scale

The Blaster Services system features three scale removal techniques that can be applied to a
wide range of scale problems.
The Scale Blasting technique combines the use of Sterling beads abrasive with new
jetting tool for hard-scale removals.
The Bridge Blasting technique uses a powered milling head and abrasive jetting, when
scale completely plugs the tubular.
The Jet Blasting technique uses the new jetting tool with nonabrasive fluids for soft scale
removal.
For hard scales like iron, strontium and barium sulfate, nonabrasive fluid jetting and chemical
treatments are inadequate. The controlled-erosive action of the Sterling Beads abrasive has been
successful in removing every type of scale in tubing, including the most difficult barium sulfate
scales, at rates up to 100 ft/hr [30 m/hr] or more.

The Scale Blasting technique is a particularly good option when the scale encountered in the well
is insoluble, unknown or of variable hardness. The system also provides a safe method to remove
scale from downhole completion equipment. Rate of penetration (ROP) is controlled using a drift
ring that ensures full tubing-diameter cleaning with minimum damage to the steel surface. (Mike
Crabtree, David Eslinger, Phil Fletcher Matt Miller, Ashley Johnson, George King, 1999)

Scale Inhibition

The cost of removing scale is very high and non-profitable and can be as high as 2.5 million
Dollars, while the cost of differed production is even higher. In most cases prevention of scale
formation is preferable through chemical inhibitions to maintain good production. This can range
from basic dilution method to more advanced methods of threshold inhibitions.

Threshold inhibitors

Threshold inhibitors reacts chemically with the scale crystal nucleation side thereby reducing the
crystal growth. Threshold inhibitors are very effective, inhibiting scale mineral formation at
concentrations of 100 time less than balance stoichiometric ratio which significantly reduces cost
of treatment. Most of these compounds are of phosphate components. Examples:
Inorganic polyphosphate compounds
Organic Phosphate esters
Organic Phosphonates
Organic Aminophosphatees
Organic Polymers
These compound minimize scale deposition. They do so by combination of crystal dispersion
and stabilization of scale. (M.D. Garba, 2014) (Mike Crabtree, David Eslinger, Phil Fletcher
Matt Miller, Ashley Johnson, George King, 1999)
For Effective Scale Removal in Separators and Surface Facilities, all internals shall be removable
and installable through the manways and handways provided without welding.
Separator Internal Design

As per API standard and NORSOK standard, the following guidelines are followed when
installing and making choices for separator internal designs

All internals shall be removable and installable through the manways without welding.
Increased internal pressure drop caused by fouling shall be considered for all relevant
scenarios, including, but not necessarily limited to drain pipes (from gas outlet
arrangement).
On floating installations motion/wave-dampening internals shall be installed.
The separator shall operate such that undesired channelling,non-plug flow or short circuit
flows do not occur in both liquid and gas phase. For each project, a Computational Fluid
Dynamics (CFD) analysis for the entire separator shall be carried out to visualize
undesired flow patterns.

Inlet arrangement

The inlet arrangement shall have a documented design, and have a documented high
efficiency towards prevention of foam and emulsion and minimize generation of small
droplets.
The inlet arrangement shall be able to handle slug flow, if expected.
A flow straightening device should be located downstream, the inlet device shall cover
the full sectional area. Care should be taken to avoid plugging due to fouling (scale,
solids and asphaltenes).

Gas phase internals

Gas outlet nozzles shall be provided with demister arrangements. Pressure tappings
should be installed upstream and downstream gas outlet device, for continuous
measurement of differential pressure to detect clogging.
The liquid collected by the gas outlet device are collected and drained by a drain pipe to
the bottom of the separator. The pipe must be submerged below the Level Alarm Low
Low (trip level) (LALL). Sufficient drainage head must be assured, so that liquid carry-
over through the drain pipe does not occur under any circumstances. The drainage is
normally internal, into the vessel bottom, but should be routed externally in case of
insufficient drainage height. The total differential pressure over the demisting section,
measured in liquid height, shall not be more than 50% of the available drainage height
related to LAHH.

Liquid phase internals

Internals in the liquid phase should be minimised due to potential clogging.


Internals for improved efficiency of liquid / liquid separation shall be applied based on
knowledge about emulsifying and separability characteristics, and a thorough
understanding of fouling tendencies (through all relevant mechanisms such as scaling,
sand accumulation and deposits of heavy hydrocarbons).
Liquid outlet nozzles shall be provided with vortex breakers.
The bottom of the separator shall be possible to inspect with as little dismantling of
internals as possible for access.

Solids removal

All separators shall, as a minimum, have nozzles for sand removal installed.
When solid production is likely the following requirements apply:
All separators shall have sand pans.
Internal jet water headers shall have nozzles for high volume / low driving pressure with
a fan spray pattern which is overlapping between each nozzle
The headers shall be spaced sufficiently for efficient sand removal by unidirectional jets
towards the sand pan.
Sand removal shall be based upon sand fluidisation rather than sand displacement.
Consideration shall be given towards efficient sand removal along the entire length of the
separator.
Nozzles to be protected against damage or misalignment by human activity during
normal vessel inspection/maintenance.

Nozzles

Manways shall be minimum 24 and installed to prevent trapped volumes when the separator is
drained.
Instrument nozzles shall be located outside the sand accumulation area.
Gas nozzles to flare shall be located such that liquid carry-over is avoided.
Separator inlet and outlet process nozzles (gas, oil, water) shall be sized a minimum of one
standard dimension larger than the connected pipe work. (NORSOK Standard, 2001)

Relief Devices

All separators, regardless of size or pressure, shall be provided with pressure protective devices
and set in accordance with ASME Code requirements. Multiple pressure relieving devices such
as a pressure relief valve in conjunction with a rupture disk may be used to provide the necessary
relieving capacity. The relief valve is normally set at the MAWP. The rupture disk is normally
selected to relieve above the set pressure of the relief valve. The pressure relief devices need not
be provided by the separator manufacturer, but over-pressure protection shall be provided prior
to placing the separator in service. The purchaser should determine who has the responsibility to
furnish relief devices

Mist Extraction Section


The mist extractor of the coalescing section can be one of several designs (a series of vanes,
woven wire mesh pad or a centrifugal device). The mist extractor removes from the gas stream
the small droplets (normally down to 10 micron diameter) of liquid before the gas leaves the
vessel. Liquid carryover is normally less than 0.1 gallon per MMSCF

Corrosion Guidelines

The following guidelines are recommended for determining corrosion considerations for an
applicable vessel.

Well streams that contain water as a liquid and any or all of the following gases are considered to
be corrosive and are due consideration under these specifications:
Oxygen (O2)
Carbon dioxide (CO2)
Hydrogen sulfide (H2S).

The following guidelines are not mandatory but may be used to judge the extent of the corrosive
environment, with respect to carbon steels.

Oxygen:
less than 0.005 ppm in natural brine - non-corrosive
from 0.005 ppm to 0.025 ppm requires - consideration
Greater than 0.025 ppm in natural brine - corrosive.
Carbon dioxide:
less than 600 ppm in natural brine - non-corrosive
from 600 ppm to 1200 ppm requires consideration
Greater than 1200 ppm in natural brine - corrosive.

Hydrogen sulfide.
No lower limit of hydrogen sulfide has been identified as being non-corrosive. With
hydrogen sulfide presence, the environment should be considered corrosive.
NACE MR 0175 (latest edition) should be used for all cases of hydrogen sulfide content
for judgment of the possibility of sulfide stress cracking (SSC) and is extracted as
follows: Systems operating below 65 psia total pressure or below 0.05 psi H2S partial
pressure are outside the scope of this standard.
Should alloy steel or stainless steel be used, other forms of corrosion should be
considered such as, but not limited to, chloride stress cracking.
Some of the other factors that influence corrosion in a given vessel include: temperature,
pressure, fluid velocities, metal stress and heat treatment, vessel surface condition, and
time

Corrosive Environment Practices


If the environment is judged as being subject to Sulphide Stress Cracking (SSC) from the criteria
of NACE MR 0175 as stated above, then all provisions of this NACE standard as apply to the
vessel materials and construction shall be followed.
If the environment is judged as corrosive from any of the other criteria stated above, the intent of
this specification will be met provided any one or combination of the following practices are
used.
An allowance for corrosion to the vessel parts may be made according to the ASME Code.

Good Practices Regarding Corrosion Allowances.

Either sacrificial or impressed current anodes may be used, providing that the area of the
corrosion attack can physically be protected by use of these anodes.
Corrosion effects may be controlled with holiday-free internal coatings on all exposed metal
surfaces. NACE RP 0181 and NACE RP 0178 present guidelines and procedures for coating
vessels such as oil and gas separators.
Corrosion effects may be disregarded provided they can be shown to a negligible or entirely
absent on a historical basis. However, the system should be monitored periodically for possible
new corrosion.
Corrosion effects may be reasonably controlled with chemical inhibitor treatments.
Post weld heat treatment is recommended for carbon steel vessels for use in acid gas (containing
hydrogen sulfide and/or carbon dioxide) service. Post weld heat treatment may be required by
ASME Code regardless of corrosion considerations. (American Petroleum Institute, 2009)
( American Petroleum Institute , 1989)

In summary for reasons stated above the following Internal devices are recommended for
installation in the chosen separators.
Diverter plate
Half pipe
Inlet Vane distributor
Inlet Cyclone
Slotted tee distributor
Tangential inlet with annular ring
Deflector baffle
All the above separator design makes provisions and allowances for installation of the above
devices. The dimensions in terms of thickness is specified in the detailed design above. Monnery
and Svrcek 1994 makes room for installation of the above internal device.

Water Handling
Water production presents serious operating, economic, and environmental problems. Production
of water with the crude oil or natural gas reduces the productivity of the well due to the increased
pressure losses throughout the production system. Production of water containing high
concentrations of ions (barium) also results in serious corrosion problems, which add to the cost
of the operation. In most situations, the produced water has no value and should be disposed of.
In other situations, the produced water may be used for water flooding or reservoir pressure
maintenance. The produced water, collected from the separation, emulsion treatment, and
desalting systems, contains hydrocarbon concentrations that are too high for environmentally
safe disposal and hence needed to be treated which brings added operational cost of both
treatment and disposal. The produced water can be treated to meet environmental specifications
before disposal. Filters, precipitators, skim tanks and vessels, plate coalescers, serpentine pipe
packs and other technologies including chemical additions are used for treating produced water
to specification before disposal. (H. K. Abdel-Aal and Mohamed Aggour and M. A. Fahim ,
2003)

You might also like