You are currently viewing: Articles


Replacing a corroded column with packing internals

A corroded column was replaced with a new column equipped with high capacity packing. The cross sectional area is less than half that of the original.

Sulzer Chemtech
Viewed : 482
Article Summary
A direct contact after cooler (DCAC) column equipped with sieve trays in an air separation unit (ASU) was severely corroded and had to be replaced. Rather than go for a one to one replacement, a modern packing solution was selected based on a lower pressure drop and much higher capacity. During the process design stage, the packing height specified was confirmed to be adequate to meet the heat transfer requirement. During the detailed engineering phase, computational fluid dynamics (CFD) simulations were utilised to assess the acceptability of the existing inlet arrangement. Trial runs were conducted shortly after the new column was erected. The measured temperature of the air from the top of the new DCAC column matched the design value very well.

In an ASU, compressed air is brought into contact with chilled and cooling water in the DCAC. The primary function of the DCAC column is to cool the hot air and reduce moisture.1 As the water vapour content in compressed air is linked directly to temperature, the compressed air must be cooled to 8°C~15°C. Otherwise, the downstream molecular sieve adsorbers may be laden with water in preference to CO2. If CO2 is not removed from compressed air, it can preferentially freeze and cause plugging of downstream equipment. A simple sketch of a DCAC column is shown in Figure 1.

An operator in the Asia Pacific region approached Sulzer to replace a severely corroded DCAC column. The existing column was 
4 m in diameter and 14 m T/T, and was equipped with 10 sieve trays. The dimensions of the new column were specified as in Table 1.

Besides, a detailed design of the air inlet was provided in the specification sheet: an open pipe type with a disk beneath the downward opening to direct the incoming vapour upwards.

The proposed operating conditions and process flows for the new column were unchanged. Based on the stream data and packing type, the capacity and the pressure drop were evaluated, with the results listed in Table 2. It can be seen from the packing hydraulics that the new column was not highly loaded.

The key function of the DCAC column is to achieve the required temperature specification of the cooled air stream. A temperature outside this range will negatively impact the vital downstream molecular sieve adsorbers’ ability to condition the incoming air stream. Table 3 summarises the top pressure of the DCAC column and the temperatures of various feeds. Based on past operation of the old column, the best ever achieved temperature of cooled air from the top of the DCAC was 15.5°C. Sulzer was asked to evaluate the specifications and guarantee an approach temperature (the temperature difference between the cooled air and the chilled water) of less than 1.5 °C.

To guarantee the temperature of the cooled air, heat transfer calculations must be carried out to evaluate the process risks.

Heat transfer calculations
In the DCAC column, as the temperature of the compressed air is higher than that of the water, sensible heat is transferred from the hot air to the water. Meanwhile, due to condensation of water vapour, the latent heat of the hot air is also transferred. It should be highlighted that condensation is essentially a mass transfer matter. Figure 2 illustrates the concentration and temperature profiles of water in the air and water phases for a DCAC column. From this figure, it can be easily understood that resistance to sensible heat transfer exists dominantly in the gas phase while for latent heat transfer or mass transfer there is no resistance in the liquid phase due to a nearly zero concentration difference of water between the water phase and a water laminar film. Therefore, for sensible 
heat transfer and latent heat transfer in the DCAC column, only the gas phase needs to be looked at closely.  

Process simulations provide useful insights, such as the driving force for respective heat transfers as well as the thermal properties of the two phases. However, they can not predict how fast heat will be transferred over a specific type of packing.

Specifically, for sensible heat transfer, the heat transfer coefficient in the gas phase should be known. This, apart from the thermal properties of the gas, is also related to gas turbulence inside the packing. Subsequently, the required packing area2 can be calculated as follows:


where QS (W) is sensible heat, U (W · m-2 · K-1) is gas heat transfer coefficient, and TG and Ti (K) are temperature of bulk gas and temperature of the interface respectively.

As for latent heat, the required packing area3 can be determined by:

where QL (W) is latent heat, kG (kmol· m-2 · s-1) is mass transfer coefficient, hG and hi (mol · mol-1) are humidity in gas bulk and the interface respectively, Mair (g· mol-1) is molecular weight of air and ∆HLV (J/kg) is latent heat of water.

Similarly, the mass transfer coefficient is also related to gas turbulence inside the packing.

Depending on the application and specific customer requirements, the packing surface area can be adapted to fit the purposes best. It should be noted that the packing area described above refers to the effective area available for heat transfer, which may or may not be the same as the geometrical surface area of the packing.4 Nevertheless, Sulzer’s proprietary packing correlations predict packing effective area as well as mass transfer and heat transfer coefficients at specific operating conditions.

Based on the specified packing type and height, and the flow rates and conditions of feeds, we calculated the cooled air outlet temperature to be 14.5°C with an approach temperature of 0.5°C. The impact of possible variations in water flow rate (±10%) was also investigated. Eventually, Sulzer guaranteed that 1.5°C of approach temperature could be achieved.
Current Rating :  4

Add your rating:

Your rate: 1 2 3 4 5