qUlpmen

Chlorine Cylinders

Loss Prevention Engineers and Safely ProfessionaJs have repealed stories in Louisiana
about the misapplication of steam on one-ton chlorine cylinders. Two similar stories were
discussed at inter-plant loss prevention meetings a couple decades ago. Naturally, the de-
tails arc now a little vague but the stories were from companies located over 100 miles
apart in two separate petro-chemical sections of the state.
Process engineers who have requirements for low flow rates of chlorine can purchase
chlorine in J 50-pound and one-ton cylinders. These containers are very robust and ill-
dustries have safely used them for decades. The one-ton cylinders are 2 1/2 ft. (0.75 m)
in diameter and just under 7-ft. (2.1 m) long with extended cylinder walls to protect (he
valve". Chlorine ton cylinders are hydrotcsted at 500 psig (3,450 kPa gauge).
Chlorine can be provided from the ton cylinders as a liquid or a gas. The discharge
rate depends on the pressure within the cylinder and this is a function of the temperature.
In order to withdraw ga'\. you must vaporize liquid chlorine. Withdrawal tends to reduce
the temperature and hence the vapor pressure and the surrounding air must supply suf-
ficienl heal to boost the vapor pressure. Typical maximum discharge rate for a one-ton
chlorine cylinder in a 70" F environment is about 15 pounds per hour.
When chlorine gas is withdrawn from a chlorine cylinder at a mpid rate, the cylinder
will cool down. Some companies who need to boost the gaseous chlorine flow rate have
placed warm water spray or utility water on Ihe cylinder to add a well-controlled heat

input on cool days. However, some individuals were reported to have become more cre-
ative and blew live steam via a hose to even further increase the rate of vaporization.
That would :.eem like a good idea-unless you took the time to read the precautions.
Each chlorine ton cylinder has three fusible plugs on each end of the tank to prall'ct
it from over-temperature. These plugs are designed to melt between 158 0 and 165 0 F
(70 to 74" C). They are designed to protect the tank from overpressure if the cylinder
is accidentally exposed to fire or other sources of excessive heat. Ln fact. if yOll acci~
dentally heat steel containing chlorine to over 483 0 F (250" C). a steel-in-chlorine fire
can be ignited.
The protection philosophy seems to be to protect the cylinder from overpressure by
protecting it from over-temperature that would vaporize the liquid chlorine. Each of the
fusible plugs is in a 3/4-inch threaded plug. If the fuse mell~, the resulting hole is about
11/32 inch in diameter.
Stories from plant safety professionals indicate that live steam wa" exhausted on the
cylinders to make a crude vaporizer to increase chlorine flow. The frightful result wa"
more chlorine. but it was to the atmosphere, not to the process. There were no reports of
injuries when these memorable stories were told.
Repairs to a Pipeline Result in Another Iron-in-Chlorine Fire
When a welder finished welding a short section of 2-inch (5 em) piping in a rather
long insulated gac;eous chlorine equalizing line, it was late on the day shift. The weld that
joined the new section of piping 10 existing piping was about 6 inches (15 em) from ad-
jacent thermal insulation thai encircled the original chlorine line.
Tht: evening shift loader's first task was to return the repaired chlorine line to service.
The loader quickly pressured the section of piping with 100 psig (690 kPa gauge) com-
pressed air and determined the repairs were leak-free. He next opened a valve connecting
the 2-inch (5 em) pipeline to 160 psig (1100 kPa gauge) compressed chlorine gas.
Within seconds, the iron caught fire and the escaping gases roared like a jet as a
brownish-orange plume of ferric chloride drifted towards the fence line from the blazing
piping. The shift loader quickly isolated the piping system by closing the block vaJve he
had just opened. Once the oxidant was consumed. the fire self-extinguished. Eyewit-
nesses said that sparking and fire resembled a traditional roman candle. shooting sparks
of burning steel. Although the fire bwnt out quickly. it was remembered for years.
Many of the 1!>upervisors and loaders were aware that if chlorine comes in contact with
iron that has been heated above 483 0 F (250 0 C). a spontaneous iron-in-chlorine fire can
occur. Bul. the loader probably assumed the new weld on the repaired pipeline had
cooled to well below the auto-ignition temperature of iron in chlorine. The investigation
determined that the fire did nol originate at the new weld, but started under the adj~Kent
insulation, which trapped heal and did not allow the piping to cool quickly.

Just a lillie of the Wrong Lubricant

While auempting to free up a 2-inch (5 em) plug valve in chlorine service, which wa~
ht..'Coming difficult to use, a chemical process operator located a lubricant gun that was
marked "for chlorine service." The operator checked the cartridge, and it was properly
labeled as the specified chlorine-compatible grease.
TIle operator connected the grea"e gun to the valve fitting and hand pumped in a lllOall
amount of grease. Before the opermor could disconnect the grease gun, a mild explo~ion
occurred. Evidence showed lhat someone had contaminaled the gun with hydrocarbon
grease. The rapid chlorination of the hydrocarbon grease created destructive prcs'iures
high l.:llough 10 blow a 3/4-inch (2 em) hole illlhc bottom of the plug cock body. Some
chlorine c\Caped from the system via the damaged valve, but no one was injured.
The chemical process operator did everything correctly. However. someone else in
the organization apparently did not understand the dangers when he previously violated
good pr,lctice and used the grease gun for hydrocarbon grease service.
Normal Maintenance on a Brine Pump

The impeller within a large sodium chloride brine pump in a closed loop system of a
chlorine plant needed to be replaced. The job was not assigned to the regular mainte-
nance crew, but was given to a crew who worked throughout a major chemical plant in
areas that had a significant maintenance backlog. Pump specifications required that all
wetted surfaces would be fabricated of titanium. However, the job was being manned
with "outsiders" on an evening shift.
After the maintenance foreman searched for the appropriate titanium impeller and
could not find one, he decided to improvise. The maintenance foreman located a stain
less steel impeller that was available for other pumps of the same size and manufacturer.
He thought the stainless steel would be satisfactory, but would have a shorter life. The
mechanics completed the job and the maintenance foreman wem to eat an overtime sup-
per al the plant cafeleria where he joined the operaling foreman. The maintenance fore-
man just happened to mention the problems he faced securing the specified impeller and
his clever temporary approach to restore the pump.
After listening intently. the operating foreman explained that mere traces of chrome
salts in the brine system could create an explosive situation within the electrolytic chlo-
rine cells. Traces of chrome salts in the feed brine to the chlorine cells liberate hydrogen
gas in the chlorine cell gas. Hydrogen in the chlorine cell gas has a very wide explosive
range. Installation of stainless steel equipment in sodium chloride brine systems has dev-
astated chlorine processing equipment within other similar chlorine manufacturing
plants. The maintenance foreman had the improper pump impeller removed immediately
before any problems occurred.