Hydrocarbon Processing - April 2022 - GP-20

PROCESS OPTIMIZATION
Due to a very different temperature
in its connecting piping, the measured
pressure drop by a PDT in an ASU
cannot be used to judge packing
hydraulics permanence directly.
FIG. 1. Pressure drop of the top bed against F-factor.
Predicted dynamic pressure drop, kPa
Predicted gas static head in HP, kPa
Total pressure drop, kPa
Measured pressure drop by PDT, kPa
TABLE 1. Pressure drop for the HP columns at the same load
Unit 1
1.111
2.852
3.963
2.96
Unit 2
1.111
2.852
3.963
3.16
able to assume that the gas density and the subsequent static
pressure drop do not vary with the F-factor, as shown in FIG. 1.
Using two pressure transmitters, the pressure drop in a column
can be measured by subtracting the top pressure from the
bottom one. The advantage of this arrangement is its simplicity,
as the two transmitters (top and bottom) do not need to
be connected by a vertical line. However, a disadvantage can
be inaccuracy. In this instance, the typical error of 0.1% on the
pressure measurement for a column operating at 550 kPa is
expected to be +/- 0.55 kPa for both the top and the bottom
measurements. Therefore, the pressure drop difference will
have an error of +/- 1.1 kPa. A measurement of 3.0 kPa (as
shown in TABLE 1) with an error of +/- 1.1 kPa is useless. This
method is inadequate, and not recommended in this case.4
differential pressure method is preferred.
A proprietary transmitter modelc
FIG. 2. PDT taps and piping connected to the HP column in the
cold box (other equipment is omitted for clarity). Note: the dotted
lines indicate the contour of the cold box.
dicted by the hydraulic rating tool for both units. It can also
be seen that the gas static head contributes to > 70% of the total
pressure drop in the column, as the gas density is between
22.7 kg/m3
and 22.4 kg/m3
in the 14-m tall columns.
The HP columns investigated consist of two beds, and FIG. 1
illustrates three different types of pressure drop for the top bed
against F-factor, which can be calculated according to Eq. 2:
F-factor = (V/A) × √ρg
where,
V stands for volumetric flowrate of gas (m3
A for the tower cross-sectional area (m2
ρg for gas density (kg/m3
).
The change in gas density is < 1% when the operating pressure
of the HP column varies between 545 kPa and 551 kPa
(the result of change in throughput). Therefore, it is reason20
MARCH/APRIL 2022 | GasProcessingNews.com
/sec),
) and
(2)
was used in the plant. It
is known that, in general, a PDT reading does not indicate the
gas static head in a column, as the gas static head in the PDT
connecting piping would offset the one in the column. However,
in some situations, corrections must be made when the
PDT reading has to be rigorously interpreted. For instance, in
many applications, inert gas such as nitrogen (N2
) is used to
purge the piping to prevent condensation of process gas due
to lower temperature outside the column. The detailed correction
method can be found in literature.1
Since the pressure drop measured by the PDT in the ASU
was much higher than the predicted dynamic one, corrections
were conducted to verify whether the deviation can be explained.
A particularity of cryogenic distillation (as found in an
ASU) is that the temperature of the piping connecting to the
PDT is higher than the temperature inside columns, thus no
inert gas needs to be injected into the piping. Instead, process
gas enters the pressure piping directly.
To evaluate the impact of the gas static head on the PDT
reading, the space arrangement of PDT piping connected to
the HP column in the cold box was retrieved from installation
drawings. The piping in Unit 1 is shown in FIG. 2 and simplified
in FIG. 3. The representative piping with numerous bends
counters thermal expansion and contraction in a range of more
A
http://www.GasProcessingNews.com
Hydrocarbon Processing - April 2022

Table of Contents for the Digital Edition of Hydrocarbon Processing - April 2022
