Nov-2021

Premature foam-flood in an amine absorber: Part 1

Troubleshooting and hydraulic analysis identified vapour channelling in fixed valve trays as the root cause of premature foam-flooding in a high pressure amine absorber.

HENRY Z. KISTER, Fluor
ROHIT KUMAR Bharat, Petroleum Corporation Limited

Article Summary

The lean amine flow rates in the high pressure (HP) amine absorber of the hydrocracker unit at Kochi refinery were limited by recurring foam-over events. To increase liquid handling capacity, the tower was retrayed by a major tray supplier (Fluor was not involved). The retray was quite standard, using high-capacity valve trays with small fixed valves and push valves and increasing downcomer top area. Surprisingly, higher lean amine flow rates could not be achieved without foam-overs. Even at the same amine and gas rates, the frequency of the foam-over events jumped from once or twice per month to once per day. Some 54 days after the new trays began operation, the bottom drain line from the pre-knockout drum plugged, causing precondensed lighter hydrocarbons entrainment into the amine system, and severely aggravating the foaming issues. This required a reduction in both the amine and gas flow rates. The problem was eventually resolved by replacing the tower with a random packed tower and a new Pre-KOD and unplugging the pre-KOD bottom drain line.¹

Although the problem was solved, the mystery remained. Why did the revamp trays perform worse than the original trays when they should have performed better? The refinery sought answers from the supplier and the supplier’s tray rating software, but neither of them were able to provide any clues.

Aware of Fluor’s distillation expertise, the refinery had Fluor join their task force analysing data and performing a ‘post mortem’. This joint investigation disclosed a phenomenon that to the best of our knowledge had not been previously reported: vapour channelling inducing premature tray foam-flood. This first part of the discussion presents our initial analysis and findings. Details of the follow-up investigation and its findings will appear in the second part of this article, to appear in the Q4 2021 issue of PTQ.

Process description
The vacuum gasoil mild hydrocracker (VGO MHC) unit of the Kochi refinery (1.7 million t/y) treats VGO from the crude vacuum distillation unit to produce low sulphur VGO feedstock for the FCC unit, together with conversion of excess VGO to low sulphur diesel. The design maximum feed sulphur in VGO was 1.6 wt%.

The effluent from the MHC unit, consisting of the cracked products, desulphurised VGO, hydrogen sulphide (H2S), ammonia, and excess hydrogen, are separated in separators/flash drums into liquid and gas streams. The gas streams are further purified in hot and cold separators to remove heavier ends and in a HP amine absorber to remove H2S. The unit process is described in detail with the aid of a flow diagram in a companion article.¹

Figure 1 shows the HP amine absorber. Feed gas containing less than 1.5 vol% H2S enters via a bare nozzle into a pre-knockout drum (Pre-KOD). A small amount of liquid knocked out in this drum drains via the bottom 2” line. Gas from the drum ascends into the absorber, entering it via another bare nozzle. The purified lean gas leaving the HP amine absorber goes to a downstream compressor knockout drum (D/S KOD) to remove any amine droplets before being routed to the suction of a recycle gas compressor. There is a gas bypass that can divert some of the feed gas to the absorber overhead, at the expense of not removing its H2S. The absorber pressure is 105 kg/cm2g.

Vessel inspection during a unit turnaround in November 2017 identified a number of internal cracks and hydrogen blisters in the HP amine absorber. Cracks were eliminated by grinding the blistered surfaces and weld filling the internal cracks followed by post-weld heat treatment. In January-March 2019, another unplanned shutdown was taken to partially replace the bottom portion of the absorber shell due to aggravation in the cracks.¹

The major reason identified for hydrogen-induced cracking and hydrogen blisters was high rich amine H2S loading of more than 0.50 mol/mol of methyl diethanol amine (MDEA), well above the 0.40 mol/mol maximum limit practised by the industry. The required lean MDEA flow rate to maintain the H2S loading below 0.40 mol/mol of MDEA was limited by the onset of foam-over events.

Retray details and preliminary hydraulic evaluation
The amine absorber is 2.0 m in diameter and had a total of nine single pass standard uncaged moving valve trays with sharp (not venturi) orifices at 600 mm spacing. The original trays were designed for 115 t/h of lean amine, but in practice the absorber operated at 70-80 t/h of lean amine, with occasional runs just above 100 t/h. To increase the lean amine flow rate, and with the outlook of processing high sulphur VGO, the tower was retrayed to increase its liquid handling capacity to 150 t/h.

The retray was designed and performed by a major tray supplier with mini-fixed valve trays, with peripheral push valves along the tower walls, pushing the liquid in the direction of flow. To increase the liquid handling capacity, the straight downcomers were replaced by sloped downcomers, with larger downcomer top areas and smaller bottom areas, while keeping the tray active areas unchanged. The downcomer top to bottom area ratio was 2:1, in accordance with good design practices.^2,3 The clearances under the downcomers were also raised to lower downcomer back-up. The weir heights were raised to keep the downcomers at a zero static seal. A comparison of the old to the revamp tray design is shown in Table 1.

The absorber lean amine inlet, rich amine outlet, gas entry, and bottom seal pan remained unchanged. The bottom seal pan overflow weir was 103 mm tall and 1500 mm long, 750 mm below the bottom tray. The straight bottom downcomer was replaced with a sloped one of the same top and bottom areas as the other revamp trays, but its clearance was increased from 50 mm to 90 mm. No changes were made to the pre-KOD vessel.

Operating trends during a foaming event
Figure 2 shows the operating trends during a foaming event, a short time after the absorber returned to service following the retray. The tower operated steadily at a gas flow rate of 43 t/h and a lean amine flow rate of 74 t/h, with a differential pressure (dP) of 11 mbar/tray. At about 16:19, the dP started to rise, slowly at first, then faster, peaking at 19 mbar/tray at 16:23. The absorber bottom level fell as the dP rose. Both the rise in dP and the reduction in bottom level are symptoms of flooding. Just before the dP peaked, the operators took the corrective action of greatly reducing the lean amine rate to 50 t/h. Amine carry-over to the D/S KOD started at 16:23, as seen by the rapid rise in level in the drum. The drum filled up with liquid within a minute or two. Upon reduction in the lean amine rate, the carry-over declined as seen by the more gradual fall in liquid level in the D/S drum, becoming low by 16:37. Simultaneous draining of the D/S KOD was carried out at field. Throughout the event, there was little change in the gas flow rate and in the Pre-KOD level.