Packed Bed Performance Analytics
Packed Bed Performance Analytics
Packed Bed Performance Analytics
R
efinery and chemical plant fraction. If the measured liq-
operations depend heav- uid retention density is divided
ily on distillation and sep- by the process liquid density at
aration towers. Tower gamma bed conditions (the liquid den-
scanning is well established in A B C sity at the actual operating tem-
the process industries as a qual- perature and pressure), liquid
itative tool to help troubleshoot Figure 1 Typical gamma scanline hold-up or liquid volume frac-
towers. Advances in data anal- orientations: A for diameters 1.5 m tion can be established. A com-
ysis have led to a quantitative and larger; B for diameters 0.5-1.5 parison of the liquid hold-up
approach in expressing gamma m; C for diameters less than 0.5 m fraction to the packing operat-
scan data in numerical terms ing capacity curves provides an
easily understood by process and operations objective appraisal of current operating capacity.
engineers. It is always easier to understand and discuss
For packed towers, a grid scan of three or four technical issues when quantitative information
equidistant scans crossing through the beds of can be used to compare operational parameters
packing would typically be used to investigate the with engineering design. This advanced analysis
quality of liquid distribution. The conventional provides a new method for extracting quantita-
approach to ‘analysing’ a gamma scan has been tive information from gamma scan data to diag-
to visualise how well the scan data from the indi- nose and characterise the operation of distillation
vidual scans matched each other or how well they and separation towers. It is our goal that the use
‘overlaid’ with each other. This is a totally subjec- of the advanced analytics presented improves
tive analysis lacking consistency, open to varying the value of gamma scan data and facilitates
interpretation and does not translate well from improvements in the operation of mass transfer
tower to tower. Therefore, the resulting conclu- equipment.
sions from this approach can be very ambiguous Figure 1 shows the most typical orientations
regarding the magnitude of any detected liquid for conducting a gamma scan on a packed tower.
maldistribution. These scans are conventionally called grid scans.
An alternative analytical approach, termed Some of the process information that can be
PackView, has been developed whereby a relative obtained for a gamma scan of a packed tower
density scale is calculated from data that the grid includes:
scan provides. The density scale begins at the den- • Condition of packed beds – elevation of pack-
sity of the dry or non-operating packing. The den- ing, depth of packing elements, uniformity of
sity scale displays the calculated density of liquid packing elements, and so on
retained in the bed of packing based on the scan • Collector and distributor liquid levels:
data results. Damaged? Overflowing?
Another calculation by which to put the liquid • Base or bottoms liquid level
distribution into perspective and to get a measure • Flooding present?
on the useful capacity of the packing is to calcu- • Foaming present?
late the liquid hold-up fraction or liquid volume • Excess liquid retained in packing?
this quantitative analysis does not give any insight extent or severity of any liquid maldistribution
into the severity or quantity of liquid maldistribu- in terms of liquid density.
tion. Therefore, the resulting conclusions from this
purely qualitative approach can be very ambigu- PackView Example 1
ous regarding the presence and magnitude of any Figure 3a shows the gamma scan results from
liquid maldistribution. a small-diameter packed tower (only two scan-
lines performed due to the small diameter).
Advanced analytics: PackView The tower diagram on the right side of Figure 3a
An advanced analytical analysis for gamma shows where the bed of packing was supposed
scan data from packed columns was developed to be as per the reference tower drawing. There
to consistently analyse gamma scan data and was a reduction in radiation counts from the
reach a conclusive result. Densities are calcu- clear vapour region at the expected elevation for
lated based on tower dimensions, scan path the top of the bed. After a short distance into the
length (variable χ) and gamma scan response bed the radiation counts decreased. So the ques-
through the packing. The results are used to tion was, what was the operating condition of the
superimpose a density scale onto the scan data. bed? Were the lower counts (higher density) liq-
The baseline of the density scale is the dry bulk uid hold-up (flooding) at the bottom of the bed?
or non-operating packing density. Calculated Or was the radiation count at the bottom of the
densities greater than the dry packing density bed ‘normal’ and something had happened to the
represent liquid and/or solids retained in the packing in the top of the bed? A visual or qualita-
bed of packing. Figures 3b, 4b and 5b show tive evaluation of the radiation counts could not
examples of this liquid retention density scale. answer these questions with confidence.
As with the qualitative gamma scan analy- Figure 3b shows the gamma scan results from
sis, if the multiple scanlines have matching liq- Figure 3a with the liquid retention scale added.
uid retention densities then the implication Note that the density of the dry packing is 160
is the liquid distribution is good. However, if kg/m3 and the top of the bed has a density essen-
there is a difference between the scan- tially equal to this. However, the tower was
lines, the retention density gives a numer- operating and there was liquid travelling down
ical comparison from which to gauge the through the packing. Where is the additional
density from the retained liquid? The overall ing while the bottom section was grid packing. The
density at the top of the packing should be the scan results showed the bottom section to be very
combination of retained liquid and packing den- dense, as the radiation intensity was less than five
sity. Since the scan results show the overall den- counts, essentially background (no radiation pass-
sity to be nearly equal to the dry packing, either ing through the tower). Based on this result, the
there was no liquid (obviously not the case) or diagnosis was that the grid section was coked and/
packing was missing. This was a bed of random or flooding with liquid, but plant management was
packing. The PackView analysis provided con- not convinced. Grid packing is dense so a ques-
clusive evidence that portions of the packing tion was asked concerning whether the radiation
were missing. The high density in the bottom of source was too small. In other words, some radia-
the bed was likely crushed packing retaining an tion would pass through the packing from using a
excess of liquid. Eventually, entry into the tower larger radiation source and then perhaps the tower
confirmed these results. would not appear to be flooding.
Figure 4b shows the gamma scan results from
PackView Example 2 Figure 4a with the liquid retention scale added.
Figure 4a shows the gamma scan results from a The dry density of the grid packing was 255 kg/
crude vacuum tower (the fourth scanline was not m3. The density of the process material inside
performed due to limited access). One of the big- the grid packing, based on the scan results, was
gest problems with crude vacuum tower operation calculated to be approximately 300 kg/m3. Grid
is managing the coking (fouling) of the wash bed. packing is very open packing and the typical liq-
In this example, plant staff suspected the wash uid rate on a vacuum wash bed is very low. A
bed had coked up so the gamma scan was car- typical liquid retention density for grid packing
ried out to assess the situation. As shown in the in this service is 80-100 kg/m3. The density cal-
tower diagram on the right side of Figure 4a, this culated from the scan data was far above this, so
wash bed consisted of two different types of pack- the grid was coked and/or flooding. Therefore,
ing. The top section was typical structured pack- the scan radiation source used on this tower
Figure 6 (A) Initial gamma scan results showing what appears to be liquid maldistribution through non-uniformity of
the scanlines through the bed (B) Gamma scan results, enhanced with liquid retention scale, showing large density
differences among the scanlines. One side of the tower (blue scanline) shows almost dry packing as the liquid
retention density is almost the same as the dry packing density