AIChE Spring Conference 2015_Distillation Troubleshooting_Article

2015 AIChE Spring Meeting __________________________________________________________

Quick and Easy Troubleshooting of a Packed Tower:

Thermal Imaging as a Novel Method

J. H. Barnard

(Corresponding Author)

Sasol Ltd, Secunda Synfuels Operations

PdP Kruger Street, Secunda, 2302

[email protected]

E. P. Du Toit



Copyright © All Rights Reserved, Sasol South Africa (Pty) Ltd, April 2015

Prepared for Presentation at

American Institute of Chemical Engineers

2015 Spring Meeting

11th Global Congress on Process Safety

Austin, Texas

April 27-30, 2015

UNPUBLISHED

AIChE shall not be responsible for statements or opinions contained

in papers or printed in its publications


Quick and Easy Troubleshooting of a Packed Tower:

Thermal Imaging as a Novel Method

J. H. Barnard

(Corresponding Author)



[email protected]

E. P. Du Toit



Keywords: thermal imaging, packed distillation tower, maldistribution, troubleshooting

Abstract

This paper presents thermal imaging as a successful troubleshooting method for packed

distillation towers. A distillation tower showed poor performance in the top section, indicated by

an unstable top temperature. The methodology of investigation combined surveys drawn from

the equipment’s available instrumentation on temperature, pressure and flow, together with

taking thermal scans to identify possible root causes of the poor performance. The thermal scans

explicitly uncovered severe maldistribution throughout the tower, ranging from the top packed

bed and pump around system all the way down the bottom of the tower. The thermal scans

clearly show the vapour channeling up one side of the tower with the liquid channeling down the

other side; there were no indications of flooding.

A damaged bed limiter, pump around liquid distributor and random packing were identified as

the main possible root causes of the poor performance prior to entering the tower. The tower

internals were inspected and it was found that the random packing in the top bed was fouled,

conglomerated and severely crushed. A support grid was installed as a bed limiter and some of

the random packing conglomerates were stuck in the grid, contributing to the maldistribution.

The maldistribution was completely rectified with the installation of a new bed limiter together

with new random packing. This study shows that thermal imaging can effectively be used to

troubleshoot packed towers on an industrial scale.


1. Introduction

The effects of maldistribution in packed towers are well understood in industry. It is however

difficult to determine the extent of the maldistribution within a packed bed while the tower is still

in operation. Even though traditional troubleshooting methods like pressure surveys and

studying the temperature profiles can partly reveal where the problem originates, these methods

require a high level of experience and is often time consuming. The interpretation of gathered

data by various analysts will also differ, and any theory can only be confirmed by

decommissioning the tower, leading to downtime and production loss.

This paper presents an alternative method of troubleshooting industrial scale packed towers.

An acid gas scrubber, operating with an unstable top section temperature profile, was the test

subject of this novel method. Traditional troubleshooting methods were applied with no

conclusive results. The tower was decommissioned; the top section was inspected and no major

defects were found. After commissioning the top temperature remained unstable.

For the investigation presented in this paper, an infra-red (IR) camera was used for further

troubleshooting and the images clearly revealed fluid channeling in the packed section of the

tower. If this was known beforehand, it would have been clear that an inspection of the top tray

section would be unnecessary.

Thermal imaging offers the opportunity to visually inspect the inside of an uninsulated packed

tower while it is still in operation. This eliminates unnecessary downtime and production losses.

It also allows for detecting the origin of the problem and a limited level of skill is involved for

capturing images for troubleshooting and interpretation.

2. Process Description

The purpose of the acid gas scrubber is to remove all the water and carbon dioxide from

ammonia, yielding ammonia of a higher concentration. In the tower the ammonium carbonation

reaction is used to ensure that all the carbon dioxide is removed from the ammonia before

sending the ammonia for further purification. The ammonia would be in excess and this causes

the carbon dioxide to reacts till completion. The excess ammonia is removed at the top of the

tower.

Two sections of packing are used to enlarge the contact area for the reaction to take place. The

reaction is an exothermic reaction where excess heat must be removed and this is done by middle

and top pump around systems. Figure 1 shows a simplified process flow diagram of the acid gas

scrubber and supporting equipment.


DistilateReflux

Top Pump Around Pump

Top Pump Around Cooler

Middle Pump Around Pump

Middle Pump Around Cooler

Bottoms

Feed

Bottom Pump Around Pump

Bottoms Pump

Bottom Pump Around Cooler

Acid Gas Scrubber

Pac

kin

g Se

ctio

n 1

Packin

g Sectio

n 2

Trap Out Tray 2

Trap Out Tray 1

Trap Out Tray 3

CWS

CWR

CWS

CWR

CWS

CWR

Insp

ecte

d

Sectio

n

Side Draw

Figure 1: Simplified process flow diagram of the acid gas scrubber tower.

The overheads product from a preceding tower enters the tower below the first tray. Steam in the

feed to the tower needs to be condensed which is done by a bottom pump around system. The

bottoms pump around and cooler configuration is used for temperature control of the bottom

section.

The middle and top pump around systems are utilized to address the temperature dependency of

the formation reaction for ammonium carbonate; temperature is controlled to favor the formation

reaction. A stream from the Trap Out Tray 2 in the middle section is pumped through the middle

pump around cooler to attain the required temperature after which it is then reintroduced above

Packing Section 1.


The vapour leaving the middle pump around passes through a chimney tray into the upper

middle section of the tower. The top pump around system also aids in the formation of

ammonium carbonate. A stream from Trap Out Tray 3 is drawn off and pumped through the top

pump around cooler and reintroduced above the second section of random packing in the upper

middle section of the tower. A stream with a high purity of ammonia is used as a reflux and is

introduced right at the top of the tower. This stream also controls the temperature at the top of

the tower and washes down any the ammonium carbonate present. The ammonium carbonate is

removed from the tower at Trap Out Tray 1 by the side draw.

3. Problem Definition

The factory consists of two phases that operate in parallel. One of the acid gas scrubber towers

was experiencing an unstable top temperature. The top temperature of this tower was swinging

10 °C higher and 10 °C lower than the set point and this posed a risk contamination of the

overheads product to the downstream units. The top temperature is usually very stable and any

upset in this temperature can cause the overhead stream to be contaminated with the unwanted

acid gases. The tower was decommissioned, and no major defects were found.

4. Conventional Troubleshooting

The following steps were taken to try and alleviate the problem:

• It was found that the temperature transmitter used for the control of the top temperature

was operated outside its specified range. The range was corrected but this did not lead to

any improvement in the temperature control;

• The top temperature control loop was investigated. The control loop was found to be

functioning as intended. It had the same tuning parameters as for the acid gas scrubber

towers on the other phases;

• The temperature at the bottom of the top section was determined to be much higher than

on the other phases, indicating that the problem might actually be in the middle section of

the tower;

• The top reflux, fed to the top of the tower which plays an integral part in controlling the

top temperature, was investigated and it was found to be different when compared to the

other phases. This could also be detected by visual inspection as this was the only reflux

line that indicated ice formation on the outside of the line. The tower operation was

adjusted to allow for similar compositions of the reflux when compared to the other

phases. This resolved the ice formation on the line and did influence the temperature,

however the temperature at the bottom of the top section remained higher than normal;

• The performance of the tower was investigated, all pump around temperatures and tower

pressures were within normal operating limits; and

• Finally thermal scans of the tower were taken to determine if the packing inside the

middle section of the tower was still intact.


5. Thermal Imaging Results and Discussion

The following thermal scans were captured of two acid gas scrubber towers. Pictures of the

unstable tower (Tower 1) were compared to a tower showing stable operation (Tower 2).

Figure 2 shows the IR scans of the top section of the acid gas scrubber (Western side of the

tower), while Figure 3 is a simplified general arrangement drawing of the top section of the acid

gas scrubber.

Figure 2: Thermal scans taken of the top section of the acid gas scrubber towers. Tower 1 (on the Left)

is the tower showing the deviations from normal operations. Tower 2 (on the right) shows stable

operation.

Trays

Distillate Outlet

Top Pump Around ReturnLiquid

Distributor

Man Hole

Vent

Reflux Distributor

Figure 3: A simplified internals schematic drawing of the top section of the acid gas scrubber tower.


From Figure 2 it can be seen that there is a different temperature distribution in Tower 1 when

compared to Tower 2. The average operating temperature of the section just above the liquid

distributor is 45 °C (the temperature seen in Tower 2), Tower 1 shows a maximum temperature

of 52.5 °C. Towards the top of the tower the same temperature differences can be seen with

Tower 1 operating 6.9 °C higher than Tower 2. The bottom right side of Tower 1 in Figure 2

also shows a cold spot which is the first sign of liquid/vapour maldistribution.

Figure 4 below shows the IR scans of the top pump around return section of the acid gas

scrubber (Western side of the tower), while Figure 5 is a simplified general arrangement drawing

of the middle pump around return section of the acid gas scrubber.

Figure 4: Thermal scans taken of the top pump around return section of the acid gas scrubber towers.

Tower 1 (on the Left) is the tower showing the deviations from normal operations. Tower 2 (on the

Right) shows stable operation.



Distributor

Packing Section 2

Man Hole

Man Hole

Figure 5: A simplified internals schematic drawing of the top pump around return section of the acid gas

scrubber tower.

From Figure 4 it can be seen that there is a large temperature gradient in the cross-section view

of Tower 1 with a maximum cross-sectional temperature difference of 27 °C. When this is

compared to the same section of Tower 2 the maximum cross-sectional temperature is only

2.1 °C indicating a more uniform radial temperature distribution. Visual inspection of the

thermal scans clearly indicates that the liquid and vapour is unevenly spread through the packing

in Tower 1.

The position of the cold spot when compared to the general arrangement drawing of this section

(Figure 5) shows that the liquid/vapour maldistribution originates from the top pump around

return line. The preliminary conclusion was made that this was due to the liquid distributor

being damaged or fouled, causing the liquid/vapour maldistribution starting from this section or

channeling of liquid and vapour through the random packing section. Packed towers are very

sensitive to distribution and maldistribution is detrimental to packing efficiency (Kister; 1990:

35).

In order to investigate the temperature effect the maldistribution has on the entire tower, thermal

scans were taken from a distance away, allowing for visual inspection of both the top section of

the tower as well as the entire middle pump around. Figure 6 shows the IR scans of the top

section as well as the entire middle pump around section of the acid gas scrubber (South-Eastern

side of the tower). Figure 7 is a simplified general arrangement drawing of the top and middle

pump around section of the acid gas scrubbers.


Figure 6: Thermal scans taken of the top as well as the top pump around section of the acid gas scrubber

towers. Tower 1 (on the Left) is the tower showing the deviations from normal operations. Tower 2 (on

the Right) shows stable operation.

From Figure 6 it can be seen that there is a definite maldistribution when comparing Tower 1

with Tower 2 can be seen. The South-Eastern side of Tower 1 is much colder than the same side

of Tower 2. This maldistribution propagates through the entire middle pump around of Tower 1.

Tower 2 also shows a much better temperature distribution through-out tower.

This indicates that there is a maldistribution of the liquid and vapour throughout the entire top

packing (Packing Section 2) of the tower – the cause of the temperature instabilities seen in the

top section.


Trays

Distillate Outlet


Distributor

Man Hole

Vent

Reflux Distributor

Packing Section 2

Packing Section 1

Man Hole

Chimney Stand Pipe

Top Pump Around Outlet

Trap Out Tray 3

Figure 7: A simplified internals schematic drawing of the top and top pump around section of the acid

gas scrubber tower.

It was decided to take more scans from different angles around the towers to try and match what

was seen on Figure 6. Figure 8 and Figure 9 were taken from different angles and the

maldistribution can be seen on all of these scans.


Figure 8: Thermal scans taken of the top as well as the middle section of the acid gas scrubber towers.

Tower 1 (on the right) is the tower showing the deviations from normal operations. Tower 2 (on the left)

shows stable operation.

Figure 9: Thermal scans taken of the top as well as the middle pump around section of the acid gas

scrubber towers. Tower 1 (on the right) is the tower showing the deviations from normal operations.

Tower 2 (on the left) shows stable operation.

The vapour liquid maldistribution can be seen on both Figure 8 and Figure 9. Figure 8 shows the

hotter side of Tower 1 while Figure 9 shows the colder side of Tower 1. In both these pictures

Tower 2 shows an evenly distributed temperature through-out the entire tower.

From the thermal scans the three main deviations possibly contributing to the observed vapour

liquid maldistribution were identified as:

Uneven or damaged top pump around liquid distributor.

Blocked or damaged random packing.

Blocked or damaged bed limiter.


6. Internal Inspections and Findings

Tower 1 was opened during the routine statutory maintenance turnaround. The top pump around

liquid distributer was inspected and no deviations were found.

Figure 10 and Figure 11 below show findings from the inspection of the packing as well as the

hold down grid.

Figure 10: Damaged random packing removed from the packing sections.

Figure 11: Support grid installed as a bed limiter and damaged packing and fouling causing blockages in

the support grid.

Figure 10 shows the damaged and fouled random packing which was found in packing section 2.

The maintenance team struggled to remove the packing due to the packing being crushed and

sticking together. Fouled and damaged random packing leads to liquid and vapour

maldistribution through packing sections and can cause the severe temperature gradient observed

in the packing section. Liquid distribution plays a vital role in the effective operation of a

packed tower. Poor distribution reduces the effective wetted packing and promotes liquid

channeling (Xu, 2000). The random packing was replaced with new packing.


Figure 11 shows the random packing being stuck in the support grid, installed as a bed limiter,

contributing to the maldistribution seen on the IR images. Upon further investigation it was

found that support grid was installed when the original hold down grid was unavailable. A hold

down grid should not be used with plastic packing. Since the hold down grid rests directly on

packing it can lead to the plastic packing being compressed (Kister; 1990: 223). The support

grid was removed and replaced with a new bed limiter seen in Figure 12.

Figure 12: New bed limiter before it was installed.

7. Conclusion From the case study presented it is clear that thermal imaging proves to be a viable option in

packed tower troubleshooting. The benefit of being able to visually inspect the flow pattern

inside a packed tower while it is in operation cannot be overstated. Regular visual inspections of

this kind even allows for identifying problems even before it has any major effect on production.

These thermal scans can be used to identify problems with equipment before they are

decommissioned for statutory maintenance. Thus proving to be a significant aid in planning

ahead and procurement of spares to ensure that downtime is minimized.

Furthermore, troubleshooting packed towers in the traditional way, using pressure and

temperature surveys, are time consuming and more regular than not requires a great deal of

expertise. However, thermal imaging offers the opportunity to perform a temperature-based

visual inspection on the flow patterns inside the tower. Very little expertise is required for

capturing the images and defects can easily be identified by anyone that regularly inspects the

thermal scans for a specific tower. The thermal scans can also aid in improving the

understanding of flow patterns within a tower.

One major drawback for the use of thermal imaging is that the tower’s lagging/insulation would

need to be removed before the thermal image can be taken. However, depending on the type of

lagging/insulation and the size of the column, the costs of removal and replacement of

lagging/insulation to understand what the current operation in the tower is can be insignificant

when compared to product loss or increased utility usage.


8. Acknowledgements

We gratefully acknowledge the support and input of our supervisor Carla Cronje, without which

this study could not have been completed. We thank Jakes van der Walt for his help and reliance

on us regarding the use and availability of the infrared camera. Furthermore we would also like

to thank the various anonymous reviewers for their feedback.

9. References

[1] Kister, HZ (1990) Distillation Operation, McGraw-Hill, New York

[2] Xu, SX (2000) “Quantitatively Measure and Assess Maldistribution in Industrial

Packed Towers”

AIChE Spring Conference 2015_Distillation Troubleshooting_Article

Documents