Revamps 2024 Issue

Vane pack separator

31.926 27.365 22.804 18.243 13.682

Draw o

Gas outlet

Flash feed pipe

9.122 4.561 0 Velocity (m/s)

Tray

Figure 2 Vane pack arrangement in reabsorber C1

• Big vane spacing against plugging • Robust design.

Figure 3 Vapour flow trajectory

As the existing vane pack was inadequate for the new loads, the proposed solution was to retrofit the gas/liquid separator to reduce the methanol concentration in the tail gas from reabsorber C1 to meet the methanol emission cri- teria of less than 35 ppm in the tail gas from scrubber C2. Detailed evaluation of existing vane pack The plant owner requested an evaluation of the existing vane pack’s adequacy. To analyse the actual column perfor- mance, Sulzer collected operating data, including the tail gas analysis with methanol concentration. Table 1 summarises the hydraulic evaluation of Mellachevron H51Z in the design case and actual operation case. Note that Mellachevron ‘MCV’ is the trademark for Sulzer’s vane pack. The actual operating capacity was 116% of the design case, while the existing vane pack was designed with a maximum turn-up of 110%. As the operating loads were beyond the maximum design range, there was a chance of potential liquid droplet entrainment at the vane pack considering: • Gas velocity was exceeding maximum design velocity. • The complex vapour trajectory into the vane pack located at the column side wall due to lack of space. • The presence of the inlet pipe and the flash box, which were affecting the vapour flow trajectory. All these issues caused an increase in liquid droplet entrainment in the vane pack outlet, resulting in an increase in methanol content in the tail gas exiting reabsorber C1. In the design of vane packs, both vapour flow trajectory and vapour velocity are important to performance. Poor vapour distribution into the vane packs leads to vapour velocity being higher or lower than design guidelines. If the vapour velocity is too high, re-entrainment will happen. If the vapour velocity is too low, the inertial force will be insuf- ficient, resulting in an increase in droplet cut-off size. The MCV H51Z in reabsorber C1, with layout as seen in Figure 2 , was located at the side wall of the column due to lack of space. The housing stretches 900mm into the col- umn of diameter 3,800mm, affecting the vapour flow below the housing. The layout of the housing and column internals (tray deck, draw-off, and flash box) is not symmetrical, which may cause poor vapour distribution into the vane packs. A computational fluid dynamic (CFD) simulation was per - formed to simulate the fluid flow parameters, including gas

distribution quality, gas flow trajectory, gas velocity, and pressure drop at the vane pack entry. Figure 3 shows the complex vapour flow trajectory, while Figure 4 shows the gas load factor at the vane pack entry. The gas load factor ' l ’, also called the K-factor, is used to indicate vapour veloc- ity. K-factor is influenced by many parameters, including separator internal type, operation pressure, vapour, and liq- uid physical properties. As observed in the CFD, the vapour flow distribution at the vane pack entry is non-ideal and would lead to excessive entrainment, as follows: • The inlet pipe and the flash box not only obstruct the vapour flow but also re-entrain extra liquid droplets. • The vapour rising from the tray below, at the opposite end of the vane pack, will reach the top of the column and then be routed back to the vane pack entry, generating vapour stream circulation during the process, as shown in Figure 3. • Complex vapour flow patterns with multidirectional veloc - ity vectors, influencing the MCV capture of liquid droplets and liquid drain. • Obvious red zones of high velocity at the bottom of the vane pack, potentially leading to excessive entrainment. Revamp solution The main motivation of the revamp is to reduce the meth- anol concentration in the tail gas from scrubber C2 to less than 35 ppm, achieved through lower methanol content from the overhead of reabsorber C1. To achieve this strin- gent requirement, it was observed that the methanol con- centration after heat exchanger E1 must be less than 350 ppm. The revamp targets are as follows: • Capacity at 116% of original design. • Methanol concentration after heat exchanger E1 less than 350 ppm. CFD studies have shown that the vapour flow distribu - tion into the vane pack is not ideal. However, changing the location of the vane pack may not be feasible without loss of trays below it and extensive modification work at the site, which may lengthen the installation time. Hence, Sulzer proposed a retrofit solution with the vane pack in its current location, considering two key areas of focus to ensure good gas/liquid separation performance. • No hot works on column shell. • Turnaround within four days.

Revamps 2024