A gas flare,
alternatively known as a flare stack, is an elevated vertical
conveyance that is part of installations such as oil and gas
wells, oil rigs, refineries, chemical, petrochemical and natural
gas plants, and other facilities (Fig. 1). On oil-
production rigs and in refineries and petrochemical plants, the flare
stacks primarily serve to protect vessels or pipes from
over-pressuring during unexpected plant upsets.
Fig. 1. Flare stack.
Whenever plant equipment is over-pressured, the pressure,
relief valves on the equipment automatically release gases (and
sometimes liquids as well) that are routed through large piping
runs called flare headers after liquid hydrocarbons are
completely vaporized and then send to the flare stack. The
released gases are burned as they exit the flare stack. The
size and brightness of the resulting flame depends on how much
flammable material was released.
Steam can be injected near the end of the flare tip to
reduce formation of black smoke. The injected steam does,
however, increase the noise level of the burning gas. To keep
the flare system functional and instantly useable, a small
amount of purge gas is continuously burned. It thus resembles a
pilot light, maintaining the system ready for its primary
purpose as an over-pressure safety system. The continuous gas
source also helps to prevent oxygen ingress into the
As mentioned earlier, flare systems enhance plant safety by
dependably disposing of all hydrocarbons discharged during
plant upsets. All safety valve releases go to the flare system.
There are, however, two types of flare feeder systems in
A wet flare header is used to handle flare gases
that contain moisture but are not cold gases.
An intermediate flare header, which could
contain some moisture and normally handles some cold vapors (up
A dry flare header designed to handle dry flare
gases. These will also be cold, with normal temperatures below
A low-pressure acetylene flare header,
exclusively provided to handle acetylene-rich gases.
At Reliances Nagothane facility, a large flare stack
with a design load of 1,000 metric tph is located on the
north-east side of a gas-cracker plant. Flare headers from
individual plantspolypropylene, low-density polyethylene,
linear-low density polyethylene, gas cracker-OSBL and gas
cracked-ISBL) join the main flare header, which routes to the
flare stack (Fig. 2). The main flare header leads to a knockout
drum in the flare area.
Fig. 2. Flare system
schematic at the
Nagothane plant in India.
The purpose of the knockout drum is to separate entrained
liquid droplets carried with the gases passing through relief
valves. Liquid capture avoids the danger of burning droplets
falling from the top of the flare stack. Flare gas free of
liquid flows to a water-seal drum; its purpose is to provide
protection against pulling a vacuum and to prevent a flash back
in the flare header. Occasionally, the stack draft effect could
decrease pressures below atmospheric at minimum flare gas
loads. The water seal also eliminates the ingress of air into
the flare system and any attendant risk of explosion. The flare
stack at this facility is 100 m high and has a diameter of 1.52
m. The flare height of 100 m includes the flare tip and a
molecular seal installed just above the flare stack and below
the flare tip.
The molecular seal consists of a gas lead pipe
and an inverted cylinder over the pipe. Gas flows in an upward
direction, turns through 180° and flows downward for a
short length before being redirected again through 180° and
back to the original flow direction. In the static condition,
gas lighter than air will tend to collect in the upper bend and
heavier gases will tend to settle at the lower bend, sealing
off the stack against backflow of air. The flare tip is mounted
on top of the molecular seal and contains three pilot burners.
Damage to the flare tip due to flame burn back near the tip is
avoided through the use of refractory lining on all exposed
anchor and mesh surfaces.
Exploring the failure history.
The flare stack at Nagothane was commissioned in 1989. Since
then and at every plant shutdown, the flare tip was being
replaced because it experienced damage during flaring
operation. Until 2010, the flare stack structure had never been
repaired and neither had it been repainted after plant
commissioning because no time was available during annual or
major shutdowns. However, inspections of the flare stack
structure, ladders, grating, clamps and associated piping was
conducted before a major turnaround scheduled for early
The support structure of the flare stack was found damaged,
and loss of thickness was observed and measured at various
locations, mainly under the support plates (Figs. 3 and 4).
Fuel gas and steam piping were found damaged as well and some
grating had been totally eaten away by corrosion. Replacement
or repair of the entire structure during a projected 17- day maintenance shutdown was
contemplated but judged very difficult. It was also realized
that working on a flare stack structure during normal plant
operation involves high risks; needless to say, flaring can
occur at any time due to plant upsets.
Fig. 3. Damaged
Fig. 4. Extreme
corrosion on a flare stack in
The conventional mode of replacing or repairing a flare
structure, piping and subsequent painting would take more than
100 days. It was, therefore, judged impossible to do the entire
job during a planned shutdown of only a 17-day duration. With
that in mind, initial discussions were aimed at completing the
job in discrete phases; specifically up to 44 m elevation, in
steps dubbed non-shutdown or pre-shutdown tasks. The remaining
work from 44 m to 100 m elevation was to be done during the
scheduled major shutdown.
Scaffolding and crane arrangements were implemented as
pre-shutdown work (Fig. 5). That left about 24 days as the time
required for work conducted while the facility was shut down.
Therefore, and after further deliberations, it was agreed to
plan additional pre-shutdown work to a height of 65 m during
non-shutdown and carry out the remaining jobs from 65 m to 100
m elevation with the facility shut down.
Pre-shutdown scaffolding work.
Risk assessment and safety.
Listed among the special risks and risk mitigation
Heat radiation due to flaring
Exposure to work at heights above grade
Descending fire balls during heavy
Stinging insect attack or bites.
Among the major work items were protective metal shields of
1 mm thickness. These were installed at elevations 26 m, 44 m
and 65 m (Fig. 6). The sheet-metal guards were affixed to the
grating of all scaffolding grating. In addition, ceramic
blankets were fastened to the sheet metal to substantially
reduce the intensity of the radiant heat and avoid burning
Fig. 6. Lift
arrangement and protective
Flood lights were provided for job execution at night, and
water shields were installed at the 66-m elevation to effect
cooling during flare events. As part of the water-shield
system, two water-curtain nozzles were affixed horizontally to
the structure. Their effectiveness was demonstrated before they
were mounted at the jobsite.
Two lifts and a crane (hoist) were deemed appropriate for
worker rescue and to facilitate the lifting of both personnel
and materials. One crane was designated for emergency rescue of
workers; and a suitable cage was fabricated and load tested
before usage. Of course, the crane was also used for lifting
and lowering of materials. The rack and pinion lifts were rated
at 1 ton and 0.4 ton capacities, respectively. They too could
be used for rescue purposes and up to 22 persons could be
evacuated in case of an emergency.
Nomex coveralls were mandatory for all workforce members and
their supervisors. Whenever heavy flaring was to take place,
warnings would be issued to the workforce in the flare area
through redundant means, including mobile handsets and a
plant-wide loudspeaker (audio) system activated from the
control room. The water-spray curtain would commence
immediately so as to proactively cool the working area. All
persons could immediately retreat safely to the protective area
below the metallic shield (Fig. 7).
Fig. 7. Water
shield at 66 m.
In essence, the non-shutdown and shutdown work
Erection and dismantling of crane and lifts
Scaffolding erection and removal up to 100-m
Affixing of metallic shields on the scaffolding
gratings at elevations of 26 m, 44 m and 65 m
Water-shield system installation at the 66 m
Replacement or repair of 20 metric tons of
structure and replacement of 2.3 metric tons of grating (Figs.
8 and 9)
High-pressure water blasting of structure and flare
stack; power tool cleaning instead of manual wire brush; all
followed by painting
Insulation and cladding replacement of 3-in. and
8-in. steam lines up to 100 m elevation
Damaged fuel gas and steam line replacement.
Fig. 8. Platform
and clamps after
Fig. 9. Piping,
insulation and cladding
Initially, the non-shutdown tasks were planned to be done
during daylight hours. With unscheduled flaring on some days,
the daytime work had to be suspended. Lost time was recovered
and work execution scheduled on a round-the-clock basis using
floodlights at night to make up for lost time and to complete
the non-shutdown part of the repair job in time.
All repair work on the flare stack structure was
successfully done without incident within the scheduled
period-85 days for non-shutdown work and 16 days for
shutdown work. This was the first time in the history of
Reliance that repair work on a rather massive flare stack
structure (Fig. 10) has been done online without any safety
Massive flare stack at full usage.
Fig. 10B. Flare
stack structure after repair
NMD- Nagothane Manufacturing Division
PP- Poly propylene
LDPE- Linear density polyethylene
LLDPE- Linear-low density polyethylene
GC- Gas cracker
OSBL- Outside battery limit
ISBL- Inside battery limit
Surinder Singh is vice president
(mechanical) with Reliance Industries Ltd. at the
Nagothane Manufacturing Division in Maharashtra, India.
He has over 30 years of petrochemical industry
experience. At present, he is assigned as head of plant
mechanical maintenance for the entire
complex. He has had wide experience in plant downtime
reduction and major turnaround planning. He is credited
with filling lead roles and involvement in the
development of various safety procedures. Mr. Singh
graduated with a BSc degree in mechanical engineering
from Regional Engineering College, Kurukshetra, India.