Page 142 - Handbook Of Multiphase Flow Assurance

138 5. Flow restrictions and blockages in operations

the restriction existed in the uninsulated inlet line leading from the knock out drum to the
flare stack.
January 2001 Canada well
A plug developed in sour gas well tree during well workover to repair tubing to annu-
lus leak under cold weather conditions −15 °C. Pressure in the tubing/casing annulus was
2100 psi. Prior to running a wireline punch tool, attempt to pressure test well lubricator was
unsuccessful as operator found tree valves frozen. Steam was applied to thaw the valves.
After thawing the tree, the wireline lubricator was found to be plugged.
Well valves were then closed and the lubricator pressure bled off. The lubricator assembly
was then removed from the tree and laid down on the catwalk.
Steam was applied to the lubricator to thaw it out and remove the stuck tools when a loud
boom was heard. The bottom section of the lubricator was removed to inspect what had hap-
pened and it was observed that the tubing punch was missing. While applying steam to the
outside of the lubricator, trapped pressure inside the lubricator was released and the tubing
punch shot out of the lubricator and hit the engineer's shack 28 m away, embedded in the ver-
tical metal pole of the propane tank skid within 1.5 ft of two 400 lb. propane tanks. The tubing
punch fired on contact with the skid.
After this occurred, the supervisor returned and installed the drag cap on the remaining
sections of the lubricator string. All personnel were going to the doghouse for a meeting to
discuss what had just happened when a second bang was heard. The remaining tool string
shot through the drag cap 100 m into a field, passing within 2 ft of 3 workers walking to the
doghouse. The two remaining sections of the lubricator assembly recoiled across the catwalk
hitting the xmas tree and bending the upper master valve stem. The lower master valve re-
mained undamaged and the well remained secure.
June 2008 US onshore gas plant
A cryogenic service instrument valve failed to close when solids suspected of being hydrates
cut into the Teflon valve seat. Process equipment did have hydrate problems prior to this event.
Hydrates may have formed as a result of maintenance activities that used water to hydrotest
the cryogenic equipment. Although systems were purged with warm nitrogen after hydrostatic
testing, inadequate purging may have left water. A hydrocarbon leak in an instrument loop
caused an uncontrolled release of cryogenic fluid. After the plant emergency shutdown of the
train, the contents of train were manually depressurized to flare to reduce the amount of hydro-
carbon leaking to the atmosphere. As a result, the flare piping fell below its design temperature.
Train was injected with methanol and drained out of the exchangers prior to startup. Learning:
using water for hydrotesting in cryogenic service requires robust drying to prevent hydrate.
April 1991 US refinery
Hydrates of isobutane caused the blockage of a pump. After isolation by valve closure,
the pump was removed and hydrate melted by steam. A butane cloud was released from
a hydrocracker and ignited by a fired heater. After repairs, methanol is added to the run-
down lines. TDC alarms on propanizer flows alert the operators to the formation of hydrates.
Piping has been changed to balance the flow in the overhead lines.
February 2005 US onshore pipe
A hydrate plug was formed on purpose in a gas pipeline, with undulating profile, to val-
idate methods of hydrate plug location prediction, hydrate location detection, and hydrate