Distillation Column Case Study

!"#"

HYDRAULIC STUDY OF ULE FRACTIONATION PLANT DISTILLATION COLUMNS (GLP-2) UNDER OPERATIONAL OCC SCENARIO

With the commissioning of the Occidente Cryogenic Complex (OCC), the Ulé Fractionation Plant (GLP-2) will receive from the extraction train of this complex, NGL at different conditions of composition (Lean, Average and Rich NGL), which generates the necessity of defining the highest NGL feed flow rate that plant can handle. From this point of view, it is required a hydraulic capacity study of each column in order to avoid operational problems such as flooding or weeping, and thereby ensure optimal operation of fractionation columns, meet quality product requirements demanded by the client and ensure the operational availability of fractionation towers. Feed flowrates for each of the evaluated scenarios are 27.5 MBPD of Lean NGL, 42.8 MBPD of Average NGL and 34.55 MBPD of Rich NGL.

Autor: MSc. Roberto Paz Gerencia de Procesamiento de Gas Occidente

!

Hydraulic Study Of Ule Fractionation Plant Distillation Columns (Glp-2) Under Operational Occ Scenario

RESUMEN

Ulé fractionation plant is located on “Intercomunal” Avenue, “Sector La Vaca”, Simón Bolívar Municipality in Zulia State. Its function is to split the NGL coming from Tia Juana 2 and 3 plants in propane, butanes mixture (n-butane and ibutane) and natural gasoline (C5 +). With the commissioning of the Occidente Cryogenic Complex (OCC), the Ulé Fractionation Plant (GLP-2) will receive from the extraction train of this complex, NGL at different conditions of composition (Lean, Average and Rich NGL), which generates the necessity of defining the highest NGL feed flow rate that plant can handle. That is why, it is required a hydraulic capacity study of each column in order to avoid operational problems such as flooding or weeping, and thereby ensure optimal operation of fractionation columns, meet quality product requirements demanded by the client and ensure the operational availability of fractionation towers. To achieve this objective, there were applied several assessments or sensitivities at different feed rates to GLP-2 by using Aspen Hysys 2006 process simulator to find the maximum power of fractionation columns, under different scenarios of OCC feed. From sensitivities results, it was obtained that the maximum feed flow rates for each of the evaluated scenarios for GLP-2 towers (under different OCC feeding scenarios) are 27.5 MBPD of Lean NGL, 42.8 MBPD of Average NGL and 34.55 MBPD of Rich NGL, without occurring hydraulic problems in fractionating columns. INTRODUCCION

Ulé Fractionation Plant is comprised by plants GLP-1, 2 and 3. It is currently operating the GLP-2 plant with a capacity by design of 46.0 MBPD of NGL from Tia Juana 2 and 3 plants, producing: propane, butanes mixture and natural gasoline. At present, GLP-2 processes about 16.0 MBPD as average.

Gerencia de Procesamiento de Gas Occidente


$

GLP-2 train consists of a depropanizer column (D8-504 or T5-B), a debutanizer column (D8-506 or T6-B), propane and butane condensers (D6-501 and D6-502 respectively), two feed drums (D8-501 A / B), two reflux drums (D8-505 and D8507) and two reboilers (D2-504 and D2-506). The propane production of the ULE facility is intended to meet local demand using up to 85% for that purpose, moreover, 12% of the propane production is sent back to the Tia Juana II and III to be used in the Mechanical Refrigeration System and the remaining production is exported to international markets. Due to future changes in the feeding of natural gas liquids (NGL) coming from Occidente Cryogenic Complex (OCC) to LPG Ulé in terms of its composition and feed flow, it is required a hydraulic capacity study of each column in order to avoid operational problems such as flooding or weeping, and thereby ensure optimal operation of fractionation columns, meet quality product requirements demanded by the client and ensure the operational availability of fractionation towers, after OCC commissioning. PROBLEM STATEMENT

On April 29, 1958, Ulé Fractionation Plant started operations with Creole Petroleum Corporation Company. After the nationalization period in 1975, Ulé became as part of the assets of the Venezuelan oil industry and began its operations with the company LAGOVEN S.A. After the unification of the Venezuelan oil subsidiary in 1997, the Ulé Fractionation Plant is transferred to PDVSA Gas, as part of the Western Gas Processing Management. With the commissioning of Western Cryogenic Complex, the extraction process will have the capacity to process 950 MMSCFD of gas which will produce 58.88 MBPD of natural gas liquids (Average NGL). The OCC inlet gas to be processed (extraction process), comes from various production blocks or zones in the region, resulting in different compositions, which are mentioned hereunder: 

Lean Gas (2.2 GPM).



%



Average Gas (2.6 GPM).



Rich Gas (3.26 GPM).

From these compositions, it will be obtained different qualities of the liquids extracted in the CCO that are classified as Lean NGL, Average NGL and Rich NGL, depending on the composition of the gas fed. The liquids extracted from gas will be distributed on the normal OCC operating scenario (two trains for extraction process and one train for fractionation process) to the different fractionation trains of the West division as follows: 

35.0 MBPD for the new OCC fractionation train.



18.0 MBPD for Ulé Fractionation Plant.



5.88 MBPD for Bajo Grande fractionation Plant and additionally 3.42 MBPD of stabilized condensates.

Since the NGL composition that will be fed to the Ulé Fractionation Plant with the OCC commissioning varies regarding the design composition of GLP-2 Plant, and to define the distribution scenario of the NGL production from the OCC extraction train in case of maintenance in the fractionation train of this complex, and that production would be distributed to the Ulé and Bajo Grande fractionation plants, it is necessary to evaluate the hydraulic behaviour of each fractionation tower internals in order to identify the maximum volumes that can be processed, which will allow to avoid operational issues such as flooding or weeping, guarantying the continuous and optimum plant operation, meet quality product requirements and ensure the operational availability of fractionation train, after OCC commissioning.



&

OBJETIVES MAIN OBJECTIVE

To evaluate the process hydraulic capacity of Ué Fractionation Plant under the scenario of future feeding from Occidente Cryogenic Complex, in order to meet the quality requirements of fractionated products and ensure operational continuity of fractionation towers with the OCC commissioning.

SPECIFIC OBJECTIVES •

To evaluate the hydraulic capacity of depropanizer and debutanizer towers (D8-504 and D8-506) of GLP-2 under OCC feed scenario.

•

To find the maximum flow feed rate to GLP-2 fractionation towers under different OCC feeding scenarios.

BASES AND ASSUMPTIONS

The bases and assumptions used during the study development are: •

To obtain the stream process physical - chemical parameters were used process simulator Aspen Hysys 2006.

•

The hydraulic evaluation of each GLP-2 fractionation towers was performed using Aspen Hysys 2006 simulation software, through the “Tray sizing” tool.

•

The considered NGL feed flows (minimum and maximum) for GLP-2 fractionation towers hydraulic evaluation were ranged from 10.0 to 42.0 MBPD.

•

The different NGL compositions (Lean, Average and Rich) from OCC Extraction Plant used for the sensitivities or hydraulic evaluations are as follow. (Source: Doc. 8922Y-000-PP-202 Rev. 0, Bases de Diseño Bloque I y II FEED 98% (OCC Project):



'

Lean NGL

Average NGL

Rich NGL

(% Molar)

(% Molar)

(% Molar)

Ethane

0.53

0.58

0.59

Propane

52.97

58.37

58.63

i-Butane

10.16

9.00

9.77

n-Butane

17.36

16.18

17.97

i-Pentane

5.69

5.40

5.18

n-Pentane

5.49

5.47

5.06

n-Hexane

4.49

3.29

2.17

n-Heptane

2.70

1.23

0.47

n-Octane

0.62

0.28

0.06

n-Nonane

0.0

0.16

0.04

n-Decane

0.0

0.03

-0.06

n-Undecane

0.0

0.0

0.06

Components

Table 1: NGL Molar Percentage from OCC

•

According to SIALAB reports, quality specifications that fractionated products such as propane and butane mixture must comply, are:

Propane Product (Depropanizer Column Top) Component

% Molar

Methane Ethane

Max. 0.5 Max. 5.1

Propane

Min. 90.0

i-Butane

Max. 1.82

n-Butane

Max. 0.28

Table 2: Maximum and minimum quality values for Depropanizer Column Top



(

Butane Mixture (Debutanizer Column Top) Component

% Molar

Propane

Máx. 2.5 Máx. 1.4 Máx. 0.35

i-Pentane n-Pentane

Table 3: Maximum and minimum quality values for Debutanizer Column Top

•

The natural gasoline must comply with a maximum Reid Vapor Pressure (RVP) of 15.0 psia.

•

The technical specifications of the internal fractionation GLP-2 columns were taken from the manual “CREOLE PETROLEUM CORPORATION LPG

EXPANSION

PROJECT

ULE

FRACTIONATION

PLANT

PROCESS DESIGN SPECIFICATIONS” . •

The behaviour of a distillation column is efficient in a range of 40 to 90% of flooding, according to Process Design PDVSA Manual , PDVSA MDP04-CF-14 (Tray efficiency).

•

The downcomer inlet velocity should be limited to a maximum of 0.15 m/s (0.5 ft/s). For foaming systems, lower inlet velocities should be used (in the order of 0.06 m/s (0.2 ft/s)). The maximum outlet velocity for sloped or stepped downcomers should be twice the inlet velocity or 0.6 ft/s (0.18 m/s), whichever is less, according to the PDVSA Process Design Manual, PDVSA MDP-04-CF-12 (Valve Tray Type).

•

The final dry tray pressure drop will generally fall in the range of 1 to 4 in. (25 to 100 mm) of hot liquid, according to PDVSA Design Manual, PDVSA MDP-04-CF-12 (Valve Tray Type).

•

The following tables show some hydraulic parameters for fractionation towers, to take into account according to PDVSA Design Manual , PDVSA MDP-04-CF-12 (Valve Tray Type). Gerencia de Procesamiento de Gas Occidente


)

•

In the next chart it can see the recommended residence time in the downcomers, depending on the hydrocarbon molecular weight. (Applied Process Design Ludwing).

for

Chemical

and

Petrochemical

Plants,


E.


*

METHODOLOGY

This research was done according with the following steps:

+,-.,

+ / 01 2 - , / 3 4

+ /012-,/34 3 7 8 8 3

5-2/6-,/34

3 9 : . - , / 3 4 - 2 8 3 4 6 /, / 3 4 +

- NGL FEED - PROCESS DATA - HYSYS SIMULATOR

- PRESSURE DROP - FLOODING FACTOR % - DOWNCOMER VELOCITY - RESIDENCE TIME

2 3 = : . / 4 ; 4; 2 723= 7::6 .-,:

NO

8345:.;:<

- OPERATIONAL PARAMETER ANALYSIS - SYSTEM MODIFICATION PROPOSAL

NO

: 46

YES

: 46

YES

- QUALITY PLAN 83092>< 0-?/010 7 : : 6 723= .-,:

RESULTS AND ANALYSIS •

Feed Flow of 42.0 MBPD to GLP-2 (Lean, Average and Rich NGL).

Regarding to the GLP-2 fractionation columns hydraulic evaluation at a feed rate of 42.0 MBPD under the NGL schemes from OCC, it can be observed the following: 

Dry Tray Pressure Drop.

Figure 1 shows the dry tray pressure drop for the depropanizer column (D8504), for the scenario that processes 42.0 MBPD of NGL, under various schemes of composition from OCC (Lean, Average and Rich NGL). Gerencia de Procesamiento de Gas Occidente

#"


1,8500 Lean NGL Aver age NGL

1,8000

Rich NGL

) n i ( p 1,7500 o r D e r 1,7000 u s s e r P 1,6500 d i u q i L 1,6000 y r D

1,5500 1,5000 0

5

10

15

20

25

30

35

40

Tray Number

Figure 1: Pressure Drop in Tray Section of D8-504

In Figure 1, the pressure drop in the internal section of the depropanizer tower (D8-504) is within the PDVSA established range in its standard MDP-04-CF-12 (valve trays type), from 1.0 to 4.0 inches of hot liquid. The range of pressure drops varies from 1.5 to 1.83 inches. For the debutanizer tower D8-506, Figure 2 shows the variation of dry tray pressure drop for 42.0 MBPD of Lean, Average and Rich NGL. 7,0000 Lean NGL

6,0000

Average NGL Rich NGL

5,0000

) n i ( p 4,0000 o r D e r u 3,0000 s s e r P d i 2,0000 u q i L y r 1,0000 D

0,0000 0

5

10

15

20

25

30

Tray Number

Figure 2: Pressure Drop in Tray Section of D8-506



##

For the debutanizer tower (D8-506), the pressure drop along the internal section, Figure 2 shows that only the scenario with Average NGL feed is within the limits described in the PDVSA standard, these values are in a range of 1.3 to 3.0 inches. While the other two feeding schemes (Lean and Rich NGL) exceed the values of pressure drop in trays sections from the top and bottom in the column, in the case of Lean NGL for example, only the first two internal (trays) are among the values described by the standard (3.71 - 3.94 inches of liquid) and the middle part of trays below the feed tray (tray 19). The top column internal values range from 4.0 to 6.0 inches of liquid. The Rich NGL ranges between 2.5 and 4.5 inches of liquid, so it neither complies with the PDVSA standard MDP-04-CF-12 (valve tray type). 

Flooding Factor.

Figures 3 and 4 show the flooding factors of GLP-2 fractionation columns for the different scenarios evaluated in this study.

70,00

65,00

60,00 % r o t c 55,00 a F g n i d 50,00 o o l F

Lean NGL

45,00


40,00

35,00 0

5

10

15

20

25

30

35

40

Tray Number

Figure 3: Flooding Factor for Column D8-504



#!

120,00 110,00 100,00 % r o t c a F g n i d o o l F

90,00 80,00 70,00 60,00

Lean NGL

50,00

Average NGL 40,00

Rich NGL

30,00 0

5

10

15

20

25

30

Tray Numbe r

Figure 4: Flooding Factor for Column D8-506

On Figure 3 it is shown that the flooding factor of the depropanizer column for the evaluated scenarios has an efficient behavior, according to the standards PDVSA MDP-04-CF-14 (Tray Efficiency) and PDVSA MDP-04-CF-12 (Valve

Tray), which states that a flooding factor range between 40% and 90% is an efficient behavior for a column. The reported values are between 43% and 63% for Rich, Lean and Average NGL. Regarding the debutanizer columns, Figure 4 shows that only the Average NGL scenario complies with the standards having a range of 53% to 73% of flooding, the rest of the feed scenarios show a flooding factor that exceeds the standard all along column internals. For the Rich NGL scenario the section of the column below the feed tray has flooding factor ranging from 80% to 74%, whereas the section above ranges from 90% to 82%. For the Lean NGL case, the whole column is flooded 95% to 113%. 

Downcomer velocity.

Figures 5 and 6 show the downcomer velocities for the GLP-2 fractionation columns for the different composition scenarios evaluated in this study.


#$

Hydraulic Study Of Ule Fractionation Plant Distillation Columns (Glp-2) Under Operational Occ Scenario 0,8000 Lean NGL Average NGL

0,7000

Rich NGL

) s / t f ( 0,6000 y t i c o l e V 0,5000 r e m o c n 0,4000 w o D

0,3000

0,2000 0

5

10

15

20

25

30

35

40

Tray Number

Figure 5: Downcomer Velocities for Column D8-504 0,6500 0,6000 0,5500

) s / t f ( 0,5000 y t i c o 0,4500 l e V r e 0,4000 m o c n 0,3500 w o D

Lean NGL

0,3000

Aver age NGL Rich NGL

0,2500 0,2000 0

5

10

15

20

25

30

Tray Numbe r

Figure 6: Downcomer Velocities for Column D8-506

In figure 5, it can be seen that downcomer velocity ranges from 0.21 to 0.67 ft/s which indicates that velocities in the bottom section would be exceeded in the bottom section of the depropanizer (D8-504) for the three feed scenarios of the CCO. According to the standard PDVSA MDP-04-CF-12 (Valve Trays) the downcomer entrance velocity should be limited to 0.15 m/s (0.5 ft/s) maximum. The top section of the column complies with the standard (0.41 ft/s).



#%

In the debutanizer column (D8-506), downcomer velocites, Figure 6, displays velocities ranging from 0.63 to 0.25 ft/s. These high velocities are found in the bottom of the column, being the case of the Lean NGL scenario for which they were found. According to the results, downcomer velocities only comply for the Rich and Average scenarios. 

Residence Time.

Figures 7 and 8 show the residence time of the tray sections of the depropanizer column (D8-504), in the case where 42.0 MBPD of NGL are processed for the different compositions scenarios from the CCO (Lean, Average and Rich NGL) and for the debutanizer column D8-506. 4,0000 3,5000 Lean NGL

) s ( e 3,0000 m i T e 2,5000 c n e d i s 2,0000 e R


1,5000 1,0000 0

5

10

15

20

25

30

35

40

Tray Number

Figure 7: Residence Time on Tray for Section for Column D8-504 2,9000 2,7000 Lean NGL

) s ( e 2,5000 m i T e 2,3000 c n e d i s 2,1000 e R


1,9000 1,7000 0

5

10

15

20

25

30

Tray Number

Figure 7: Residence Time on Tray for Section for Column D8-504 Gerencia de Procesamiento de Gas Occidente

#&


Assuming that hydrocarbon fractionating in the GLP-2 columns has a medium molecular weight at the bottom and light at the top, in figure 7 can be seen that the residence time on the internals of the depropanizer column is in the range of 1.13-1.26 seconds in the bottom section, this lowering of the residence time is due to multiple Flow Paths (4). Regarding the top section it can be mentioned that the residence time is 3.2 – 3.6 seconds. According to recommendations from Applied Process Design for Chemical and Petrochemical Plants (E. Ludwig), the residence time in the downcomer of a column should be approximately 4 seconds for medium molecular weight hydrocarbons and 3 seconds for light molecular weight hydrocarbons. In Figure 9, the debutanizer column has a residence time oscillating between top and bottom from 1.9 to 2.8 seconds. •

Maximum Allowable Feed Flow Rate to the GLP-2 columns under different CCO Feed Scenarios. 

Dry Liquid Pressure Drop.

In Figures 9 and 10 show the dry liquid pressure drop in the tray section for the depropanizer (D8-504) and debutanizer (D8-506) columns; for the case where NGL feeds are processed depending of the maximum allowed capacity of the distillation columns, considering the different composition schemes for the CCO (Lean, Average and Rich NGL). Hydraulic evaluation to find the maximum feed rate to the fractionation columns of GLP-2 under the different CCO Feed scenarios without hydraulic problems occurring in the fractionation columns, it is found that for each evaluated scenario are 27.5 MBPD for the Lean NGL Case, 42.8 MBPD for the Average NGL Case and 34.55 MBPD for the Rich NGL Case.


#'


1,8500 Lean NGL 27,5 MBPD

1,8000

Average NGL 42,8 MBPD Rich NGL 34,55 MBPD

n 1,7500 i ( p o r 1,7000 D e r u 1,6500 s s e r P 1,6000 d i u q i L 1,5500 y r D 1,5000

1,4500 1,4000 0

5

10

15

20

25

30

35

40

Tray Number

Figure 9: Pressure Drop on Tray Section D8-504

On Figure 9 can be seen the pressure drop on the internals of the depropanizer column (D8-504) for different flow rates and feed compositions, each of the scenarios is under the ranges established on the standard PDVSA MDP-04-CF12

(Valve Trays). The highest range of pressure drop was found for the

Average NGL scenario since is the one with the highest feed rate. In this case, the pressure drop is 1.68 to 1.8 in.

3,0000

Lean NGL 27,5 MBPD ) 2,5000 n i ( p o r D e r u 2,0000 s s e r P d i u q 1,5000 i L y r D

Average NGL 42,8 MBPD Rich NGL 34,55

1,0000 0

5

10

15

20

25

30

Tray Number

Figure 10: Pressure Drop on Tray Section D8-506 Gerencia de Procesamiento de Gas Occidente

#(


On Figure 10 are shown the debutanizer (D8-506) pressure drop for each of the flow rate and composition scenarios, the highest pressure drops on the internals of the debutanizer column were found for the Rich NGL case since it has the greatest amount heavy ends on the feed stream. The range of the pressure drops found was 1.79 to 3.02 in 

Flooding Factor.

Figures 11 and 12 show the flooding factors of the GLP-2 columns for the different feed scenarios evaluated in the study. Lean NGL 27,5 MBPD 65,00

Average NGL 42,8 MBP D Rich NGL 34,55 MBPD

60,00 55,00 r o t c 50,00 a F g 45,00 n i d o o 40,00 l F

35,00 30,00 25,00 0

5

10

15

20

25

30

35

40

Tray Number

Figure 11: Flooding Factor of Column D8-506

80,00

Lean NGL 27,5 MB PD Average NGL 42,8 MBPD

75,00

Rich NGL 34,55 MBPD

) % ( 70,00 r o t c a F 65,00 g n i d o o 60,00 l F

55,00

50,00 0

5

10

15

20

25

30

Tray Number

Figure 12: Flooding Factor of Column D8-506 Gerencia de Procesamiento de Gas Occidente

#)


Figure 11 shows the Flooding Factors for the evaluated scenarios for the depropanizer column. The highest flooding factor is found for the Average NGL composition in a range of 57 to 62%, in this case the behavior is considered efficient, according to the standards PDVSA MDP-04-CF-14 (Tray Efficiency) and PDVSA MDP-04-CF-12 (Valve Trays), which indicates that a range form 40 to 90% of flooding factors is considered efficient for a column. While the Lean NGL feed case shows an inefficient behavior with a range from 35.7 to 39.4% in the bottom sections and part of the top sections of the column, due to the low contents of heavy ends in the column feed. Although other authors recommend an efficiency range for a fractionation column between 10 to 90% (A Working Guide to Process Equipment, N. Lieberman ), the authors of this work consider that low efficiency during operation of a fractionation column could lead to weeping and flooding. On Figure 12, it can be seen that the scenario that presents the highest flooding factor is the Rich NGL case, and thus having the highest efficiency value, which were found in the range of 65.9 to 76.2%. 

Downcomer Velocity.

Figures 13 and 14 show the estimated downcomer velocities for the GLP2 fractionation columns. On Figure 13, it can be seen that downcomer velocity for the different evaluated cases are in a range from 0.65 to 0.24 ft/s for a Average NGL feed, which indicates that velocity values in the bottom section of the depropanizer column (D8-504), would exceed the recommended values of the standard PDVSA MDP-04-CF-12 (Valve Trays) which states a maximum of 0.5 ft/s.

The

velocities in the top sections were found to be according to the standard.


#*


0,8000 Lean NGL 27,5 MBPD Average NGL 42,8 MBPD

0,7000

Rich NGL 34,55 MBPD

) s / t 0,6000 f ( y t i c o 0,5000 l e V r e m0,4000 o c n w0,3000 o D

0,2000 0,1000 0

5

10

15

20

25

30

35

40

Tray Number

Figure 13: Downcomer Velocity Column D8-504

0,4300 0,4100 0,3900 ) s / 0,3700 t f ( y t i 0,3500 c o l e V 0,3300 r e m0,3100 o c n w 0,2900 o D

Lean NGL 27,5 MBPD Av erage NGL 42,8 MBPD

0,2700

Rich NGL 34,55 MBPD

0,2500 0,2300 0

5

10

15

20

25

30

Tray Number

Figure 14: Downcomer Velocity Column D8-506

Regarding downcomer velocities on the debutanizer column (D8-506), Figure 14 shows that the highest velocities were reached under the Rich OCC feed scenario ranging from 0.40 to 0.23 ft/s. These high downcomer velocities could lead to liquid carryover and tray flooding.


!"



Hydraulic Study Of Ule Fractionation Plant Distillation Columns (Glp-2) Under Operational Occ Scenario Residence Time.

Figures 15 and 16 show the residence time for each tray on the evaluated columns. 5,0000 4,5000 ) 4,0000 s ( e 3,5000 m i T e 3,0000 c n e d i 2,5000 s e R 2,0000

Lean NGL 27,5 MBPD Average NGL 42,8 MBPD Rich NGL 34,55 MBPD

1,5000 1,0000 0

5

10

15

20

25

30

35

40

Tray Number

Figure 15: Residence Time on Trays Section Column D8-504 2,9000 2,8000 2,7000 ) s 2,6000 ( e m2,5000 i T e 2,4000 c n e 2,3000 d i s e 2,2000 R

Lean NGL 27,5 MBPD Average NGL 42,8 MBPD Rich NGL 34,55 MBPD

2,1000 2,0000 1,9000 0

5

10

15

20

25

30

Tray Number

Figure 16: Residence Time on Trays Section Column D8-506

On Figure 15 can be seen that the residence time on the internals of the depropanizer column for the different evaluated scenarios. The residence time ranges from 1.4 to 1.7 seconds in the bottom section, while in the top section it ranges from 4.55 to 4.3 seconds. From the three evaluated scenarios, the highest residence time values were found for Lean Feed case. On Figure 16, it Gerencia de Procesamiento de Gas Occidente

!#


can be seen that the debutanizer column has residence time oscillating from bottom to top section 1.9 to 2.8 seconds for the Average NGL Feed case. •

Top Product Quality of the Depropanizer Column (D8-504).

On table 4 are presented the product qualities of the column D8-504 (depropanizer) of GLP-2, for each of the cases studied (Lean, Average and Rich NGL): ANALYZED STREAMS Component Ethane (% Molar) Propane (%Molar) i-Butane (%Molar) n-Butane (%Molar)

Quality Plan

Lean NGL

Average NGL

Rich NGL

Máx. 5.1

0.98

0.77

0.98

Min. 90

97.46

97.64

97.44

Máx. 1.82

1.43

1.43

1.43

Máx. 0.28

0.13

0.15

0.15

Table 4: D8-504 Propane Product Molar Porcentages

•

Topa and Bottom Product Quality of the Debutanizer Column (D8-506). On

tables 5 and 6 are shown the qualities of the t op and bottom products of the column D8-506 (debutanizer) of GLP-2, it can also be seen that the Reid Vapor Pressure (RVP) of the non-stabilized of the Gasoline Stream. ANALYZED STREAMS Component Propane (% Molar) i-Butane (%Molar) n-Butane (%Molar) i-Pentane (% Molar) n-Pentane (%Molar)

Quality Plan

Lean NGL

Average NGL

Rich NGL

Máx. 2.5

0.34

0.33

0.29

Min. 10

34.34

33.12

32.61

Min. 30

63.28

65.41

65.42

Máx. 1.4

1.9

0.99

1.53

Máx. 0.35

0.15

0.15

0.15

Table 5: D8-506 Butane Mixture Molar Percentages Gerencia de Procesamiento de Gas Occidente

!!


ANALYZED STREAMS Property RVP (psia)

Quality Plan

Lean NGL

Average NGL

Rich NGL

Máx. 15.0

11.72

13.32

14.8

Table 6: D8-506 Non-stabilized Gasoline Reid Vapor Pressure

When comparing the maximum and minimum values of the molar percentages of the contaminant components of the top of the depropanizer (ethane, i-butane and n-butane) and the debutanizer (propane, i-pentane and n-pentane) columns of GLP-2 (D8-504 and D8-506) to the values found on the different CCO feed simulations, the Propane Product Stream (top of D8-504) is within plant design specifications regarding methane content (0.0% molar Lean, Average, and Rich NGL), ethane (0.98 – 0.77 – 0.98% molar Lean, Average and Rich NGL), ibutane (1.43% molar for the three scenarios) and n-butane (0.13 – 0.15 – 0.15% molar Lean, Average and Rich NGL). Regarding top product quality for the debutanizer column (D8-506), the butane mixture also complies with plant design specifications, propane (0.34 – 0.33 – 0.29% molar Lean, Average and Rich NGL), n-pentane (0.15% for the three scenarios). It must be noted that i-pentane content in the butane mixture has a molar concentration of 1.9 – 0.99 and 1.53%, that complies with quality specifications of the Lean NGL GLP-2 feed scenario.


!$


CONCLUSIONS •

Top and bottom product quality of the depropanizer (D8-504) and debutanizer (D8-506) columns of GLP-2 comply with SIALAB specifications for the evaluated NGL compositions.

•

For the 42.0 MBPD NGL form the CCO scenario (Lean, Average and Rich), the depropanizer column (D8-504) complies with recommendations of the standard PDVSA MDP-04-CF-12 (Valve Trays) , regarding tray pressure drop of 1.0 to 4.0 in of liquid.

•

Accordingly, for the feed scenario of 42.0 MBPD of OCC NGL, the debutanizer column (D8-506), the pressure drop along column internals are only standard allowable values for the Average NGL case, these values ranged from 1.3 to 3.0 in

•

It was observed that the flooding factors of the depropanizer column (D8504) for the evaluated scenarios display an efficient behavior according the standards PDVSA MDP-04-CF-14 (Tray Efficiency) and PDVSA MDP-04CF-12 (Valve Trays) , which indicates that a range of 40 to 90% of flooding

for a column is considered efficient column behavior (42.0 MBPD of NGL fractionation scenario). •

For the debutanizer column (D8-506), only the Average NGL scenario complies with the standards having a flooding factor ranging from 53 to 73% (42.0 MBPD of NGL fractionation scenario).

•

Downcomer velocities were found to range from 0.21 to 0.67 ft/s, this indicates that velocities on the bottom section of the depropanizer column (D8-504) for the three feed scenarios would exceed standards recommendations. According to standard PDVSA MDP-04-CF-12 (Valve Trays) the downcomer entrance velocity should be limited to maximum of

0.15 m/s (0.5 ft/s). The top section of the column was found to comply to standard (42.0 MBPD of NGL fractionation scenario). •

Downcomer velocities on the debutanizer columna (D8-506) were found to range from 0.63 to 0.25 ft/s. These high velocities were found on the Gerencia de Procesamiento de Gas Occidente

!%

Hydraulic Study Of Ule Fractionation Plant Distillation Columns (Glp-2) Under Operational Occ Scenario bottom section of the column for the Lean NGL case (42.0 MBPD of NGL fractionation scenario).

•

Residence time of the internals of the depropanizer column (D8-504) ranged from 1.13 to 1.26 seconds on the bottom sections, this residence time distribution is due to presence of multiple Flow Paths (4). The top section of the column has a residence time ranging from 3.2 to 3.6 seconds (42.0 MBPD of NGL fractionation scenario).

•

The debutanizer column (D8-506) has a residence time oscillation on the bottom and top sections from 1.9 to 2.8 seconds (42.0 MBPD of NGL fractionation scenario).

•

The maximum feed rate to the GLP-2 fractionation columns for the different CCO feed scenarios without having column hydraulic problems were found to be 27.5 MBPD for the Lean NGL case, 42.8 MBPD for the Average NGL case and 34.55 MBPD for the Rich NGL Case.

•

Pressure drop on the internals of the depropanizer column (D8-504) for the different feed rates and compositions evaluated comply with the standard PDVSA MDP-04-CF-12 (Valve Trays) . The highest pressure drop range

was found for the Average NGL feed scenario since it has the highest feed rate. The values found for the pressure drop ranged from 1.68 to 1.8 in. •

For the debutanizer column (D8-506), the pressure drop was evaluated for the different scenarios of flowrate and feed composition and the highest were found for the Rich NGL case since it has the greates amount of heavyends content on the feed. The found pressure drop ranged from 1.79 to 3.02 in.

•

The highest flooding factors of the depropanizer column (D8-504) ranged from 57 to 62% for the Average NGL case. For the Lean NGL case, the flooding factor as an inefficient behavior having a range from 35.7 to 39.4% all along the bottom section and part of the top section of the column.

•

For the debutanizer column, D8-506, the highest flooding factors were found for the Rich NGL case. This case was the one with highest efficiency on which the flooding factor were found to range from 65.9 to 76.2%. Gerencia de Procesamiento de Gas Occidente

!&

•

Hydraulic Study Of Ule Fractionation Plant Distillation Columns (Glp-2) Under Operational Occ Scenario Downcomer velocities for the different cases evaluated ranged from 0.65 to 0.24 ft/s for an Average NGL feed composition; this indicates that downcomer velocities on the bottom section of the depropanizer column (D8-504) would exceed recommended values.

The top section of the

column velocity values comply with the standards. •

For the debutanizer column (D8-506) the downcomer velocities were found to reach a maximum for the Rich NGL scenario, on which the velocities ranged from 0.40 to 0.23 ft/s.

•

Residence time of the internals of the depropanizer column (D8-504) ranged from 1.4 to 1.7 seconds for the bottom section. For the top section, this parameter was found to range from 4.55 to 4.3 seconds. From the three evaluated scenarios the highest reported values were found of the Lean NGL feed.

•

The debutanizer column (D8-506) has a residence time oscillating from top to bottom section 1.9 to 2.8 seconds for the Average LGN feed.


!'


REFERENCES •

CREOLE PETROLEUM CORPORATION LPG EXPANSION PROJECT ULE FRACTIONATION PLANT PROCESS DESIGN SPECIFICATIONS.

•

MANUAL DE DISEÑO DE PROCESOS DE PDVSA, PDVSA MDP-04-CF14 (EFICIENCIA DE PLATOS).

•

MANUAL DE DISEÑO DE PROCESOS DE PDVSA, PDVSA MDP-04-CF12 (PLATOS TIPO VÁLVULA).

•

APPLIED PROCESS DESIGN FOR CHEMICAL AND PETROCHEMICAL PLANTS, E. LUDWING.

•

A WORKING GUIDE TO PROCESS EQUIPMENT, N. LIEBERMAN.


Distillation Column Case Study

Recommend Documents