Documente Academic
Documente Profesional
Documente Cultură
g
= gas density, lb
m
/ft
3
l
= liquid density, lb
m
/ft
3
Calculating the settling velocity of the liquid droplets based on a uniform droplet size enables separator designers
to estimate the length or height liquid droplets will travel inside the separator before falling to the liquid collection
section. The time required for the gas and liquid mixture to reach equilibrium inside the separator and ensure
effective gas liquid separation is defined as the residence time. For the mixture to reach equilibrium a certain
storage capacity is required; the residence time enables separator designers to accurately estimate the storage
capacity of the separator
Basic fluid properties (such as pressure, temperature, gas, oil and water flow rates) are also essential to
designing and selecting the appropriate separator for a hydrocarbon stream. The separator designer has to be
provided data such as the minimum, maximum, and operating pressures as well as the, minimum, maximum, and
operating temperatures. The densities and viscosities of the fluids present in the inlet stream are also required. It is
SPE 167578 3
impossible to size an oilfield separator of any configuration and type if any of the data mentioned above is not
provided.
Finally, separator design cant only be based upon basic equations and physical separation concepts. The
designers ability to the select the most suitable design that is cost effective for the given production data is based
on experience and sound engineering judgments. Separator design procedures will allow for various choices and
combinations of diameter and length. For every separator design, a combination of diameter and length will exist
that minimizes the cost of the vessel and enable fluid stability within the separator. The slenderness ratio is a value
used by mechanical engineers to determine the strength of a structural element (column, strut, cylinder etc.). It is
defined as the ratio between the height or length of the structural element to the width or radius of the element. The
American Petroleum Institute (API) and separator designers have published slenderness ratio values that help to
identify that combination of separator cost, stability and strength. Table 2 shows the slenderness ratio guidelines
employed by the two design procedures used in this study. See Table 2 in the appendix.
Outline of the Computer Program
The use of computers to ensure accurate separator design procedures by modeling the basic equations into a
simulator to ensure reliability, greater efficiency, and reuse of the design procedures is summarized in a series of
operations shown in Fig. 3 and described briefly below. Refer to Fig. 3 in the appendix.
Input
This step is broken down into two parts: production data and experience. Prior to the start of designing any
separator, the designer requires the following production data:
Operating pressure
Maximum and minimum design pressures
Operating temperature
Maximum and minimum fluid flow rates.
Liquid droplet size to be removed
Physical characteristics of the inlet stream fluid (such as density, viscosity, specific gravity, API gravity,
GLR, WOR etc)
Operation
Employing the basic equations modeled into the program, the computer program is able to initiate separator
design using the production data provided by the purchaser. The separator sizing (see Fig. 3) step in the
computer program outline involves the use of equations developed by Svrcek and Monnery and Arnold and
Stewart for sizing horizontal and vertical oilfield separators. Conventional separator design procedures also employ
the extensive use of tables and charts developed by designers based on experience and sound engineering
judgments. The tables and charts used by both design procedures are modeled into the computer programs. The
charts and tables modeled into the 4 computer programs are shown in Appendix A. The major equations used in
this step are given below.
Horizontal Separator Design
Svrcek and Monnery:
J = _j
4 (v
H
+ v
S
)
0.5 n (L )
[
1 3
12_ (2)
d = separator diameter, in
g = acceleration due to gravity, ft/s
2
L/D = slenderness ration
V
H
= holdup volume, ft
3
V
S
= surge volume, ft
3
I
SS
=
v
H
+ v
S
A
T
- A
g
-(A
HL
+ A
LL
)
(3)
A
g
= separator area occupied by gas, ft
2
A
HL
= interface area occupied by heavy liquid, ft
2
A
T
= separator cross-sectional area, ft
2
A
LL
= interface area occupied by light liquid, ft
2
L
SS
= seam to seam separator length
4 SPE 167578
Refer to Fig. 8 for horizontal separator Dimensions
Arnold and Stewart:
J
mux
=
(h
c
)
mcx
[
(4)
d
max
= maximum separator internal diameter, in
(h
o
)
max
= maximum oil pad thickness, in
= fractional height of liquid inside the separator
Refer to Fig. 9 to estimate .
Gas capacity:
I
SS
= I
c]]
+
d
12
(5)
L
SS
= seam to seam separator length
L
eff
= separator effective length, ft
Liquid capacity:
I
SS
=
4
3
I
c]]
(6)
Vertical Separator Design
Svrcek and Monnery:
J = __
4
q
n
g
_
1 2
12_ (7)
Q
g
= gas flow rate, ft
3
/min, MMscfd
v
g
= gas velocity, ft/s
E
1
= E
H
+ E
L
+E
R
+ E
BIN
+ E
A
+ E
.. (8)
H
T
= total vessel height, ft
H
H
= liquid hold up height
H
L
= height difference between the outlet of the light liquid and the heavy liquid interface, ft
H
R
= height difference between the outlet of the light liquid and the baffle, ft
H
BIN
= height difference between the baffle and inlet nozzle, ft
H
A
= height difference between baffle and liquid level, ft
H
D
= disengagement height, ft
(Refer to Fig. 7 in appendix to see vertical separator dimensions).
Arnold and Stewart:
Separator diameter based on liquid droplet size:
J
2
= Su4u [
z1
g
p
j[
p
l
- p
g
p
l
d
p
C
d
[ (10)
d
p
= droplet diameter, micron
p = pressure, psia
T = temperature,
o
R,
o
F
z = gas compressibility
Separator diameter based on water droplet size:
SPE 167578 5
J
2
= 669u _
c
c
(Su)d
p
] (11)
d
p
= droplet diameter, micron
Q
o
= oil flow rate, BPD
SG = difference in specific gravities of the water and oil
o
= viscosity of the oil, cp
Separator diameter based on oil droplet size:
J
2
= 669u _
w
w
(Su)d
p
] (12)
d
p
= droplet diameter, micron
Q
w
= water flow rate, BPD
SG = difference in specific gravities of the water and oil
w
= viscosity of the water, cp
For diameters 36 in
I
SS
=
h
c
+ h
w
+ 76
12
(13)
h
o
= height of oil pad, in
h
w
= height of water outlet to interface
For diameters > 36 in
I
SS
=
h
c
+ h
w
+ 40+76
12
(14)
The preliminary results obtained are void of the designers past experience on separator design. However
these results provide a baseline design that serve as a basis for the final design dimensions.
Output
Refining the generated preliminary designs using experience and sound engineering judgments results in a
second set of generated designs from which the designer could select the most suitable choice. API standard
dimensions for conventional oilfield separators are reported as follows: Separator diameter is given in multiples of
12 inches and the separator height is in multiples of 5 feet. The final design dimension selected should follow API
guidelines and satisfy all constraints of stability, strength and cost.
Case Study 1
This case study is adapted from one of the reviewed literature (Arnold and Stewart, 2008). It is used to validate the
results obtained in the literature and also to compare the results obtained from using two design procedures. Table
3 shows the input production data and Tables 4 and 5 displays the generated results. Tables 3 through 5 are
shown in the appendix.
The results displayed in Tables 4 and 5 using the Arnold and Stewart Design procedures are the same as the
results shown in the literature and this validates the computer program. The results using both design procedures
are similar and a final dimension can be selected that will satisfy both procedures and all other constraints. From
the results displayed in Table 4 and 5, most suitable separator dimensions are as follows:
Final Vertical Separator Dimensions:
Arnold and Stewart: 84-in (OD) X 20-ft (S/S) Length; SR = 2.9
Svrcek and Monnery: 84-in (OD) X 20-ft (S/S) Length; SR = 2.9
Final Horizontal Separator Dimensions:
Arnold and Stewart: 84-in (OD) X 25-ft (S/S) Length; SR = 3.6
Svrcek and Monnery: 90-in (OD) X 20-ft (S/S) Length; SR = 2.7
The final results are presented in API format and satisfy the constraints of cost, stability and strength. The
new dimensions ensure that the separator is properly configured for the appropriate installation of separator
internals while enabling dimensions to be reported in API format.
Case Study 2
This case study is also adapted from one of the reviewed literature (Arnold and Stewart, 2008). This case study
is used to further validate the computer program and also to compare the results obtained from using the two
design procedures. Table 6 shows the production input data and Tables 7 and 8 displays the generated results.
Tables 6 through 8 are displayed in the appendix.
Final Vertical Separator Dimensions:
Arnold and Stewart: 72-in (OD) X 20-ft (S/S) Length; SR = 3.3
Svrcek and Monnery: 72-in (OD) X 20-ft (S/S) Length; SR = 3.3
Final Horizontal Separator Dimensions:
Arnold and Stewart: 84-in (OD) X 25-ft (S/S) Length; SR = 3.6
Svrcek and Monnery: 84-in (OD) X 25-ft (S/S) Length; SR = 2.7
The final results are presented in API format and satisfy the constraints of cost, stability and strength. It can be
observed that the results shown in Tables 7 and 8 are similar and further validate the computer program. The
results also show a similar trend with the results obtained using case study one. An increase in diameter
corresponds to a decrease in separator diameter height or length and this is due to the direction of the travelling
particle. If the given design data for a vertical separator is accurate, the results always show a decrease in
separator height as the separator diameter increases. If the separator diameter the liquid particles (i.e. liquid, oil
and water) have to travel is sufficiently large, the height required for particle settling would reduce. This trend is
illustrated in Figs. 4 and 5.
The final separator dimensions shown above are as a result of the design procedures modeled into a simulator
and designer experience. Estimating separator size for a given production data using only the basic equations or
design experience could lead to a bottleneck of the entire crude oil processing facility. Hence it is very important to
apply both in separator design.
Conclusions
1. A user-friendly computer program was developed to facilitate the design of three-phase oilfield vertical and
horizontal separators using two design procedures (Arnold and Stewart, 2008 and Svrcek and Monnery, 1994).
Automating the design process enable separator designers save time and ensures the reliability and reuse of
the design procedure.
2. The design procedure proposed by Arnold and Stewart is more suitable for the design of separators. This is
because it involves the use of less design variables and constants, and does not rely heavily on the use of
charts and graphs to estimate vessel dimensions. Less guesswork is involved, computation time is reduced,
and the calculation procedures are easy to understand and less prone to human errors.
3. Horizontal separators provide greater fluid stability for the same production input data. The results show that
the predicted horizontal separator dimensions will allow for higher separation efficiency than vertical separators
for the same production input data.
Recommendations
1. The liquid particle sizes to be removed from the inlet stream have to be specified. During the course of this
study, it was observed that failure to specify the particle size to be removed led to separator dimensions that
were not feasible. The results shown in this study are based on the following liquid particle sizes: liquids = 100
microns; oil = 200 microns; and water = 500 microns.
2. Due to uncertainty within the reservoir the production input data cant remain the same forever and changes
would always occur with the original design data. It is suggested that the developed programs should be used
with a reservoir simulator program to ensure separator dimensions that will be equipped to handle such
changes in the production data.
References
Spec. 12J, AP1, API Specification for Oil and Gas Separators, seventh edition, 1989. Washington, DC: API
Arnold, K. and Stewart, M. 2008. Surface Production Operations Vol. 1, third edition: Gulf Publishing Company.
Arnold, K. and Stewart, M. 1989. Surface Production Operations Vol. 1: Design of Oil-Handling Systems and
Facilities, first edition: Gulf Publishing Company.
Arnold, K. and Stewart, M. 1986. Surface Production Operations Volume 1: Gulf Publishing Company.
Arnold, K. and Stewart, M. 1984. Designing Oil and Gas Production Systems. World Oil Journal, 199 (7) 87-98.
Halvorson, M. 2010: Visual Basic 2010 Step by Step: Microsoft Press.
Bradley H. B. 1987. Petroleum Engineering Handbook: Oil and Gas Separators, Chapter 12 (Smith, H. V).
Richardson, Texas, SPE
SPE 167578 7
Souders, M. J. and Brown, G.G. 1934.Design of Fractionating Columns: Entrainment and Capacity. Industrial and
Engineering Chemistry Journal, 26 (1): 98 100.
Svrcek, W.Y. and Monnery, W.D. 1993. Design Two-phase separator within the Right Limits, Chemical
Engineering Progress, 89 (10): 53-60.
Svrcek, W.Y. and Monnery, W.D. 1994. Successfully Specify Three-phase Separators .Chemical Engineering
Progress 90 (9): 29-40.
Walkenbach, J. 2010: Excel 2010 Power Programming with VBA: Wiley Publishing Inc.
Nomenclature
A
g
= separator area occupied by gas, ft
2
A
HL
= Interface area occupied by heavy liquid, ft
2
A
LL
= Interface area occupied by light liquid, ft
2
A
T
= separator cross-sectional area, ft
2
C
d
= drag coefficient
d = separator diameter, in
d
max
= maximum separator internal diameter, in
d
p
= droplet diameter, micron
F
B
= buoyancy force, N, lb
f
F
D
= drag force, N, lb
f
F
G
= gravity force, N, lb
f
F
i
= sum of forces acting on the particle, N, lb
f
H
A
= height difference between baffle and liquid level, inches, ft
H
BIN
= height difference between the baffle and inlet nozzle, inches, ft
H
D
= disengagement height, ft
H
H
= liquid hold up height
H
L
= height difference between the outlet of the light liquid and the heavy liquid interface, ft, inches
H
R
= height difference between the outlet of the light liquid and the baffle, ft, inches
H
T
= total vessel height, ft
h
o
= height of oil pad, in
h
w
= height of water outlet to interface
(h
o
)
max
= maximum oil pad thickness, in
g = acceleration due to gravity, ft/s
2
L
SS
= seam to seam separator length
L
eff
= separator effective length, ft
L/D = slenderness ration
p = pressure, psia
Q
g
= gas flow rate, ft
3
/min, MMscfd
Q
o
= oil flow rate, BPD
Q
w
= water flow rate, BPD
S
g
= gas specific gravity
SG
o
= oil specific gravity
SG
w
= water specific gravity
T = temperature,
o
R,
o
F
(t
r
)
o
= oil retention time, min
(t
r
)
w
= water retention time, min
V
H
= holdup volume, ft
3
V
S
= surge volume, ft
3
v
g
= gas velocity, ft/s
v
t
= settling velocity, ft/s
z = gas compressibility
Greek Symbols
SG = difference in specific gravities of the water and oil
g
= gas density, lb
m
/ft
3
l
= liquid density, lb
m
/ft
3
g
= gas viscosity, cp
HL
= viscosity of the heavy liquid, cp.
8 SPE 167578
LL
= viscosity of the light liquid, cp.
o
= viscosity of the oil, cp
w
= water viscosity, cp
Appendix A
Table 1: Comparisons between the three types of separators (Petroleum Engineering Handbook: Oil and Gas Separators, Chapter 12)
Consideration Horizontal Vertical Spherical
Efficiency of separation 1 2 3
Stabilization of separated fluids 1 2 3
SPE 167578 9
Adaptability to varying conditions 1 2 3
Flexibility of operation 2 1 3
Capacity (same diameter) 1 2 3
Cost per unit capacity 1 2 3
Ability to handle foreign materials 3 1 2
Adaptability to portable use 1 3 2
Ability to handle foamy oil 1 2 3
Ease of installation 2 3 1
Ease of inspection and maintenance 1 3 2
*Ratings (1) most favorable (2) intermediate (3) least favorable
Figure 1: Three-phase horizontal separator
Table 2: Slenderness ratio guidelines
Svrcek and Monnery Arnold and Stewart
Pressure Slenderness ratio
Horizontal Slenderness ratio Vertical Slenderness ratio
0 < p 250 1.5 -3.0
250 < p 500 3.0 - 4.0
3.0 - 5.0 1.5 - 3.0 500 < p 4.0 - 6.0
10 SPE 167578
Figure 2: Forces acting on a moving liquid droplet in a gas phase
F
D
F
B
F
G
F
i
F
B
= buoyanc
F
D
= drag forc
F
G
= gravitati
SPE 167578 11
Figure 3: Computer Program Outline
Table 3: Production Input Data
Table 4: Vertical Separator Dimensions
Table 5: Horizontal Separator Dimensions
Oil flow rate, Qo 5000 BPD
Water flow rate, Qw 3000 BPD
Gas flow rate, Qg 5 MMscfd
Operating pressure, p 100 psia
Operating Temperature, T 90
o
F
API 30
o
API
Water specific gravity, (SG)W 1.07
Gas specific gravity, (SG)g 0.6
Oil specific gravity, (SG)o 0.876
Water retention time (tr)w 10 mins
Oil retention time (tr)o 10 mins
Oil viscosity, o 10 cp
Water viscosity, w 1 cp
Gas density, g 0.3 lbm/ft
3
Water density, w 62.11 lbm/ft
3
Oil density, o 54.7 lbm/ft
3
Drag Co-efficient, Cd 2.01
Surge time 5 mins
Arnold and Stewart Svrcek and Monnery
d (inches)
LSS
(feet) SR d (inches) LSS (feet) SR
84 18.2 2.6 60 18.5 3.7
90 17.7 2.4 66 16.7 3
96 17.4 2.2 72 15.5 2.6
102 17.2 2 78 14.6 2.2
108 17.1 1.9 84 14 2
Arnold and Stewart Svrcek and Monnery
d (inches) LSS (feet) SR d (inches) LSS (feet) SR
84 21.5 3.1 90 16.5 2.13
96 16.4 2.1 96 15.4 1.93
108 13 1.4 102 12.8 1.51
12 SPE 167578
Table 6: Input production data
Table 7: Vertical Separator Dimensions
Table 8: Horizontal Separator Dimensions
Arnold and Stewart Svrcek and Monnery
d (inches) LSS (feet) SR d (inches) LSS (feet) SR
60 36.2 7.3 78 25 3.8
72 25.2 2.6 84 25 3.6
84 18.5 3.1
Figure 4: Vertical Separator Dimensions Figure 5: Horizontal Separator Dimensions
Oil flow rate, Qo 3900 BPD
Water flow rate, Qw 3000 BPD
Gas flow rate, Qg 1.6 MMscfd
Operating pressure, p 455 psia
Operating Temperature, T 90oF
API 30o API
Water specific gravity, (SG)w 1.07
Gas specific gravity, (SG)g 0.6
Oil specific gravity, (SG)o 0.876
Water retention time (tr)w 10 mins
Oil retention time (tr)o 10 mins
Oil viscosity, o 10 cp
Water viscosity, w 1 cp
Gas density, g 1.43 lbm/ft3
Water density, w 62.11 lbm/ft3
Oil density, o 54.41 lbm/ft3
Drag Co-efficient, Cd 1.93
Surge time 10 mins
Arnold and Stewart Svrcek and Monnery
d (inches)
LSS
(feet) SR d (inches) LSS (feet) SR
72 18.6 3.1 60 18.7 3.1
78 17.7 2.7 66 17.5 2.6
84 17.1 2.4 72 16.6 2.2
90 16.7 2.2 78 16.0 2.0
96 16.5 2.1 84 15.6 1.8
SPE 167578 13
Figure 6: Down-comer allowable flow chart (Svrcek and Monnery 1994)
Figure 8: Horizontal separator (Monnery and Svrcek 1994)
Figure 7: Horizontal separator (Monnery and Svrcek 1994)
Table 9: Empirical Constants for Cylindrical height and area conversions (Svrcek and Monnery, 1993)
WD/D A/AT
a = -4.7593 X 10
-5
b = 3.924091 c =0.174875
d = -6.358805 e = 5.6668975 f= 4.018448
g = -4.916411 h = -1.801705 i = 0.145348
14 SPE 167578
Figure 7: Coefficient for a cylinder half filled with liquid (Arnold and Stewart, 2008)