Source: A Working Guide to Process Equipment

CHAPTER 38
Safety Concerns
Relief Valves,
Corrosion, and Safety Trips

Y
our process unit likely contains a wide variety of safety
features and equipment. These safety features fall under the
following three categories:

• Relief valves
• Corrosion monitoring
• Alarms and trips

It is certainly my experience that the most common and

catastrophic accidents in process units are related to corrosion-type
failures. I cannot bring to mind any process vessels that were
overpressured and failed because a relief valve did not open.
Relief valves were invented to prevent steam boilers from blowing
up as a result of excessive steam drum pressure. This was a distressingly
common occurrence in the nineteenth century. The relief valve is also
called a safety (or “pop”) valve. When the pressure in a vessel exceeds a
preset amount, the relief valve is supposed to pop or spring open. Gas
will then be vented from the vessel until the pressure in the vessel drops
by 10 to 20 psi below its relief-valve-set pressure. We usually operate
pressure vessels 25 psig or 10 percent below the relief-valve setting.
The relief-valve-set, or pop, pressure is adjusted in a machine
shop. A large threaded nut on top of the relief valve is used to make this
adjustment. Air pressure is applied to the valve, and the technician
adjusts this nut until the relief valve opens at the proper pressure.
The pressure at which relief valves are set to open is the vessel
design pressure. The vessel design pressure (maximum allowable
working pressure, MAWP) is stamped on the manufacturer’s nameplate.
It is illegal to set the relief valve at a higher pressure. The vessel is
probably hydrostatically tested at a pressure 50 percent greater than its
design pressure. This test pressure also is listed on the nameplate. You
cannot use the test pressure as a guide to set the relief valves.

459
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 459 3/20/08 12:56:05 PM

 Safety Concerns

460 A Working Guide to Process Equipment

Sometimes vessels are rerated. Often, we can increase the design

pressure of a vessel by taking credit for an excessive corrosion
allowance, which is an extra thickness of metal added to the vessel
wall. Designers do not include this extra thickness in their pressure-
rating calculations. However, if experience teaches that there is little
corrosion, then the vessel may be rerated and receive a new pressure-
rating stamp. There are rigorous, formal, legal procedures to follow
in rerating a vessel.
On the other hand, vessels must often be derated for age and
excessive corrosion. Either way, the relief valve needs to be reset.
There is only one way to know the actual relief-valve setting, and that
is to climb up to the relief valve and read the tag that was fixed to the
valve at the time of its last setting.
I used to reset relief valves on small-pressure vessels in the natural-
gas fields in southern Texas. After adjusting the nut on the valve with
a wrench, I would raise the vessel pressure to check whether the valve
would relieve at the proper pressure. If not, I would continue to adjust
the nut and, by trial and error, find the proper relief-valve setting. This
is not a particularly practical method to adjust relief valves on stream
on a process unit. It is also illegal to do so.
Many relief valves pop open above or below their set point. Such
valves have to be removed from the vessel and reset in a machine
shop. Often, there are isolating block valves, located beneath the relief
valve, that permit the relief valve to be pulled while the vessel is still
in service. These block valves are perfectly legal, provided they are
chain-locked open. It is unlawful to have an isolating valve below a
relief valve that is not chain-locked open or sealed open in some
positive manner.

38.1 Relief-Valve Plugging

Often, relief valves do not open because their piping connection to
the vessel is plugged with corrosion products, salts, or coke. Even if
this connection is only partially plugged, the effective capacity of the
relief valve is greatly diminished.
A rupture disk is a thin sheet of metal installed below the valve,
intended to protect the relief valve from plugging. The rupture disk
ruptures at the relief-valve-set pressure. A better approach to retard
this plugging problem is to maintain a steam purge, or inert-gas bleed,
below the relief valve to prevent the accumulation of solids below the
valve.
Often, relief valves fail to close once they have popped open. I
hate this! I hate the tension. Will it or won’t it reseat by itself? Or will
some operator have to climb up the 180-ft crude distillation column
and hammer on the side of the relief valve until it reseats? Especially
if the relief valve is venting to the atmosphere rather than the flare,
this can be a nasty and dangerous job.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 460 3/20/08 12:56:06 PM

 Safety Concerns

Chapter 38: Safety Concerns 461

Furnace tubes, process piping, and heat exchangers may also

have to be protected by relief valves. Incidentally, the small 1-in relief
valves you see on many tank field loading lines and on heat exchangers
are not process relief valves. They are there for thermal-expansion
protection only. This means that if you block in a liquid-filled
exchanger and the liquid is heated, the liquid must expand or the
exchanger will fail. That is what the thermal relief valve is there to
prevent.
Typically, heat exchangers, piping, and furnace tubes have a
certain flange rating. For example, 150 psi flanges have a pressure
rating of about 220 psig, and not 150 psig. A 300 psi flange has a rating
of roughly 430 psig. If the upstream process pump has a shut-in or
dead-head pressure exceeding this flange rating, then a downstream
protective relief valve is required.

38.2 Relieving to Atmosphere

In 2004, at the B.P. Refinery in Texas City, 15 contract workers were
killed and 170 seriously burned. The relief valve opened on the top of
a naphtha splitter. The relief valve did not vent to the flare, but to a
blowdown stack. The blowdown stack was open to the atmosphere
and drained to a sewer. This drain had a loop seal to prevent vapors
from blowing into the sewer.
The operators were starting the tower:

1. The tower was on total reflux.

2. The bottom level control valve was closed in the field.
3. The bottom flow meter was off-zero, and showed a substantial
flow.
4. The bottom liquid level indicator on the panel indicated a less
than full level, even when the bottoms level was above the
top level tap.

The operators, being unaware of the last three facts, filled the
tower almost to the top with naphtha and increased heat flow to the
reboiler. The relief valve, set for 50 psig, opened as the liquid in
the tower swelled up due to the extra heat. The boiling naphtha
partially flashed as it flowed from 50 psig in the tower to zero psig
in the blowdown stack.
As the boiling naphtha flowed through the loop seal of the
blowdown stack drain, the loop seal became vapor locked. Or perhaps
the volume of boiling naphtha exceeded the drain’s capacity.
Regardless, the naphtha backed up in the blowdown stack to the inlet
nozzle. The evolved vapors blew the liquid out of the top of the stack.
The naphtha ignited in the midst of the contractor trailer park built
around the blowdown stack. Two additional points are relevant:

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 461 3/20/08 12:56:06 PM

 Safety Concerns

462 A Working Guide to Process Equipment

• In the 1970s I assisted, in a minor capacity, Gary Elmer in the

design of this tower. At that time, the refinery was an Amoco
Oil plant, and Gary and I worked in the Chicago Engineering
Design Division. As process design engineers, we did not
worry about relief valve vents.
• In 1974, I was promoted to the position of operating supervisor
at the Amoco Refinery in Texas City. I had four tall distillation
towers on my alkylation unit. The towers had relief valves at
the top of the tower. These relief valves vented to the
atmosphere. Over-pressuring was common and I saw the
white plume blowing from the valves on many occasions.
Never did I stop and think, “What will happen if a relief valve
opens when a tower is in fully developed flood?”

Now I and everyone else in Texas City knows what happens.

How about your unit? Do you have relief valves on vessels in
hydrocarbon service that open to the atmosphere? What would
happen on your unit if the safety relief valve opened when the vessel
is full of liquid?
I have asked this question at many plants in the past years. The
majority still have safety reliefs venting to the atmosphere even
though everyone in the United States refinery industry knows what
happened in Texas City in 2004. To summarize, when considering the
safe operation of relief valve systems, the following assumptions
should be made:

• The valve will open some day.

• When the valve opens, the process vessel will be full of
liquid.
• The liquid will be at its boiling point at the pressure at which
the relief valve opens.
• There will be a source of ignition present downstream of the
relief valve.

As you evaluate the above criteria, recall that there were scores of
towers in Texas City operating for many decades. These towers relieved
to the atmosphere (some of which I designed), and nothing ever
happened. Nothing ever happened until that terrible incident in 2004.

38.3 Corrosion Monitoring

On older process units, you may still encounter piping with sentry
holes. Let’s say I have a 0.5-in thick pipe. The corrosion allowance for
the pipe is 0.25 in. A number of small holes are drilled into the pipe to
a depth of 0.25 in. When we start leaking at these small holes, this
means that the pipe has corroded to its discard thickness in the area

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 462 3/20/08 12:56:07 PM

 Safety Concerns

Chapter 38: Safety Concerns 463

of the sentry hole. Incidentally, you can stop the resulting leak, at least
in carbon steel water lines, with a brass wood screw and a screwdriver.
I have also done this on hydrocarbon lines under an 80 psig pressure,
but perhaps that was not too smart.
A more modern method to check for loss of thickness in process
piping is by ultrasonic testing (UT) or Sonaray. These are portable
instruments used to check pipe thickness on-stream. Do not forget,
though, that the thin elbow, the one that is sure to fail, is always out
of reach unless a ladder can be found. And the inspector cannot find
the ladder. The outside radius of elbows are typically the thinnest
portions of a piping run. I became the world’s leading expert on not
finding ladders and not UT-checking thin elbows at the Amoco Oil
Refinery in Texas City, in 1976—just prior to the alkylation unit
explosion. The explosion that wrecked half of the refinery.

38.3.1 Corrosion Coupons

I always like corrosion coupons. They are a tangible piece of hardware
that you can hold in your hand. A corrosion coupon is just a piece of
metal, perhaps 0.25 in thick, 3 in long, and 0.5 in wide, that is inserted
into the flowing process stream through a packing gland in a valve. It
can be easily pulled for inspection every week or every month. The
corrosion engineer weighs the coupon for metal loss and inspects it
for pitting and other forms of corrosive attack. Often, a number of
corrosion coupons of various metallurgy are used. The corrosion
engineer then calculates, on the basis of the metal lost from the coupon
in one month, the mils per year corrosion rate.
For example, a steel pipe is 0.5 in thick, or 250 mils. Its discard
thickness is 0.125 in thick, or 125 mils. If the corrosion rate is measured
at 25 mils per year, then the expected life of the pipe is 5 years.
Corrosion rates in excess of 10 mils per year are normally considered
excessive and unacceptable, at least in petroleum refineries.

38.3.2 Corrosion Probes

This is an electronic method used to measure corrosion rate in mils
per year. The corrosion probe can be inserted through a packing
gland. It is read periodically with a portable instrument that measures
the change in electrical conductivity of the probe. It is simple, but
perhaps a little less reliable than the coupon.
There is another sort of probe that measures ionic hydrogen
penetration into and through steel walls. Such hydrogen activity is a
product of corrosion and/or high hydrogen partial pressures in a
vessel. This sort of hydrogen activity promotes hydrogen blistering
and stress corrosion cracking of vessel walls. A sketch of a hydrogen
probe is shown in Fig. 38.1. It consists simply of a small chamber
welded onto the exterior of a vessel. A pressure gauge is connected to
this chamber.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 463 3/20/08 12:56:07 PM

 Safety Concerns

464 A Working Guide to Process Equipment

H2

H+
Ions

FIGURE 38.1 A hydrogen probe.

The hydrogen ions (or protons) dissolve and pass through the
iron lattice structure of the vessel wall. When some of them emerge at
the outside of the vessel wall, they are trapped inside the hydrogen
probe chamber. There, the ionic hydrogen is converted into molecular
hydrogen. The rate of pressure increase inside the chamber is a direct
measure of the amount of ionic hydrogen infiltration through the
lattice structure of the steel wall of the process vessel.
Often massive corrosion failures occur suddenly, without the
warning of any small leaks. Lines part at welds, vessels burst apart as
a result of hydrogen-assisted stress corrosion cracking, and thin
elbows peel back like the top of a soup can. Process plants are
dangerous places, mainly because of corrosion; hence the importance
of monitoring corrosion. It is the responsibility of the unit chemical or
process engineers to monitor corrosion on their units.

38.4 Alarms and Trips

38.4.1 Safety Trips
One of the most common safety trips is the automatic fuel-gas shutoff.
We have this at home on our furnaces. We ignite the pilot light

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 464 3/20/08 12:56:08 PM

 Safety Concerns

Chapter 38: Safety Concerns 465

manually. The pilot light heats a thermocouple. The milliamp output

from this thermocouple opens the fuel-gas valve to the main burner.
The gas in the main burner is ignited from the pilot-light flame. Should
the heat from the pilot light diminish below a certain point, the fuel
gas to the main burner, as well as to the pilot light, will be shut off.
A less common type of fuel-gas trip to a heater is a low-pressure
trip. A pressure transducer generates a milliamp output from a boiler
feedwater pump. Should this milliamp output fall below a certain
level, the instrument air signal to the fuel-gas regulator actuator will
be shut off. These fuel-gas valves are air to open, meaning that loss of
instrument airflow causes the valve to close.
Some fired heaters, especially boilers, have a device called a “purple
peeper,” which is simply an optical device that looks at a flame. If it
does not detect light with a wavelength in the high-frequency (i.e.,
purple) end of the optical scale, it interprets this as a flame-out. The
fuel-gas regulator is automatically shut.
High process heater outlet temperatures are another trip point
for many heaters. These sorts of trips are subject to a rather deadly
malfunction. If the process flow is totally lost, the heater tubes may
become extraordinarily hot. But the high-temperature trip may be
located too far from the heater outlet to be affected by the high
temperature in the heater. This particular malfunction occurred in
a southern Louisiana refinery, and indirectly led to the deaths of a
score of workers. This is why we have P&ID (piping and
instrumentation diagram) reviews when our unit is being designed.
Fuel to a boiler may also be shut down by a low water level in the
boiler’s steam drum. The low-level trip is simply a float chamber. If
the float drops to the bottom of the chamber, it flips a mercuroid
switch, which shuts off the fuel-gas supply to the boiler. These
mercuroid switches look quite similar to the thermostat switch we
have inside our house, which switches the air conditioner on or off.
Note that these level trip float chambers, even though they are
quite short, still require two level taps on the vessel.
Some heaters also have a low fuel-gas pressure trip on the fuel
gas itself. The idea here is that if fuel-gas flow is lost, we do not want
it to surge back into the heater too quickly if the fuel gas pressure is
suddenly restored.

38.4.2 Compressor Trips

Our home circuit breakers, or fuses, are, of course, trips to prevent
overheating electric circuits or electric motors. The only difference
is that at work our electric circuit breakers have a built-in time delay.
This is needed to allow the motor driver to overcome the starting
torque inherent in most large pieces of rotating process equipment.
Compressors also have vibration trips. These trips measure the
amplitude of the vibrations—which, if they become excessive, will
shut off the fuel, steam, or electricity to the compressor’s driver.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 465 3/20/08 12:56:08 PM

 Safety Concerns

466 A Working Guide to Process Equipment

Turbines, both gas and steam, also have overspeed trips. These
consist of a little flywheel constructed from three balls. The little balls
are spread apart by centrifugal force. The greater the rpm, the greater
the centrifugal force. If the balls spread apart too far, they activate the
trip. We have James Watt to thank for this neat innovation, still used
in its original form.
Most large compressors also have a low-lube-oil-pressure trip. This
again would shut off the fuel or steam to a turbine if the lube-oil pressure
gets too low. It is interesting to note that the low-lube-oil pressure that
activates the trip is not the lube-oil pressure to the bearing that would be
damaged because of a lack of lube-oil flow. Rather, it is the low-lube-oil
pressure to the trip switch itself that directly shuts down the turbine or
motor. Thus, with a low-lube-oil-pressure, one can trip off a compressor
without actually losing lubrication flow to the bearings at all.
Compressors are also served by high-liquid-level trips in their
upstream knockout drums. These high-liquid-level trips work in the
same way as the low-level boiler trips discussed above, except that
the mercuroid switch is activated by a rising, rather than a falling,
liquid level so as to protect the compressor from a slug of liquid.
A final word about trips. Any trip that is not tested on some
routine basis can never work in an emergency. I will guarantee you
that when the coupling shears on that steam turbine driving that
4000-bhp compressor, the overspeed trip that should shut off the
steam flow to the turbine will not in fact trip if you have not tested
that same trip recently. I promise you that the trip valve mechanism
will be encrusted with hardness deposits from the steam. While the
trip lever may be unlatched, the turbine will continue to spin merrily
along, until it self-destructs as a result of uncontrolled overspeed.
And, ladies and gentlemen, you may imagine how I have become so
knowledgeable about this particular subject.

38.5 Autoignition of Hydrocarbons

Heavy hydrocarbons, such as vacuum tower bottom products, tar,
and pitch, are frequently pumped and stored in tanks above their
autoignition temperatures. The process reason for this is to prevent
these fluids from “setting up” or solidifying in the pipelines or tanks.
At ambient temperatures, even in hot climates, these fluids would
normally be either solid or extremely viscous.
A hydrocarbon above its autoignition temperature will
spontaneously ignite upon exposure to air without any source of
ignition. This is why there are often fires on pumps that pump heavy
hydrocarbons. If the seal on a pump in such a service begins to leak,
the heavy material then leaks out and spontaneously catches fire. Be
careful when taking pressure surveys or sampling from these hot,
heavy hydrocarbon lines to avoid spillage. The risk of autoignition
increases for fluids above their autoignition temperature as the
surface area of spillage increases. As the surface area increases, the

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 466 3/20/08 12:56:09 PM

 Safety Concerns

Chapter 38: Safety Concerns 467

surface area of the hot, heavy hydrocarbon exposed to the oxygen in

the surrounding air also increases. Spillage on the ground, leaks into
insulation material, and spillage onto rags are all increasing the risk
of autoignition by increasing the surface area exposed to oxygen.

38.5.1 A Hazardous Piping Configuration

Sections of piping that contain a pair of alternate 90-degree elbows or
alternate right-angled bends are hazardous. These should be avoided
when designing pipe work because:

• When fluid flows through such pipe work, it changes

momentum at the opposing angles of the turns that the pipe
work requires it to make. These changes in momentum act
upon the pipe work configuration to make it rotate or spin,
just like a two-sailed windmill where the right-angled ends
are the “sails,” or just like one of those fireworks that come as
a pair of identical fireworks mounted in opposite directions at
either end of a piece of wood—we pin the wooden strip to a
fence, light the pair of fireworks, and watch the assembly
spin. Piping sections are anchored, but considerable stresses
are set up within such a system.
• The changes of momentum occurring in the fluid as it flows
around the bends make the inner wall at the outer edge of the
elbows particularly susceptible to erosion.

Outages and some serious incidents have occurred as a result of

the use of this type of piping configuration (see Fig. 38.2).

“ Impelled
” dir
ec
ti
on
of ro
tatio
n

Fluid flowing
in the pipe section
tion
rota
of
on

cti
dire ”
“ Impelled

FIGURE 38.2 A hazardous piping configuration.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 467 3/20/08 12:56:09 PM

 Safety Concerns

468 A Working Guide to Process Equipment

38.6 Paper Gaskets

I was talking to my friend Dan Rogers at a Texaco refinery. Dan said to
remind my readers to beware of “paper gaskets.” Gaskets seal the ends
of flanges used to assemble piping spool pieces. High temperature and
high pressure gaskets are made of metal. In the past few years, the use
of so-called “paper gaskets” has become common. Theoretically, these
gaskets have a temperature rating of 500°F. In practice, they should
only be used for:

• Blinding equipment during a unit turnaround when the blind

flange will not be exposed to much more than ambient
temperatures.
• Cooling water, plant air, fire water, low pressure steam.
• Services below 250°F or where a leak will be an inconvenience
and not a safety issue.

These paper gaskets will look pale green-blue when viewed

edge-on between the flanges. Dan says that their use has caused
several fires at his plant for services below their rated 500°F
temperature limit.

38.7 Calculating Heats of Reaction

My client had burned a hole in a hydrotreating reactor. The reactor was
not in service, it was on standby with once-through hydrogen flowing
through the reactor. The hydrogen was coming from a hydrogen
production plant, which was in a start-up mode of operation. The failure
in the side of the reactor looked like someone had burned a hole through
the 4-in steel wall from inside the vessel. This was consistent with plant
data that indicated reactor temperature had exceeded 1800°F.
The failure was due to combustion of oxygen plus hydrocarbons.
The combustion reaction involved was:

CO + 3H2 = CH4 + H2O (38.1)

This is a methanation reaction. The hydrogen reacts with the

carbon monoxide to produce water, methane, and heat. My job was to
calculate how hot the reactor could get with 4 percent carbon
monoxide in the hydrogen supply gas. To do this I looked up in
Perry’s Engineering Handbook the heats of formation of:

• CO = 24.21 kcal per gram-mole

• H2 = 0 (as are all elements)
• CH4 = 17.88 kcal per gram-mole
• H2O = 15.80 kcal per gram-mole

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 468 3/20/08 12:56:10 PM

 Safety Concerns

Chapter 38: Safety Concerns 469

Adding up the heats of formation on both sides of Eq. (38.1)

and taking their difference gives a total heat of reaction of 49.2 kcal
(49,200 calories). This would be the heat of reaction if there were
25 percent CO and 75 percent H 2 in the feed gas. However, the feed
gas contained 4 percent CO, not 25 percent CO. Therefore, the
heat of reaction was:
● (4 percent divided by 25 percent) × 49,200 = 7900 calories

The 96 percent hydrogen and 4 percent CO in the feed gas would

be converted to:
● 92 mole percent hydrogen
● 0 carbon monoxide
● 4 mole percent methane
● 4 mole percent water

The combined specific heat of this mixture is:

● 7.0 calories per mole per 1°C
Dividing the heat of reaction (7900 calories) by the specific heat of
the products of reaction (7.0 calories per mole per 1°C) results in a
temperature rise of 1130°C or 2040°F.
This calculation shows how 100°F hydrogen flowing to the reactor
could have reacted, when exposed to a catalyst bed, to produce
temperatures of 2000°F.

38.8 Hot Water Explodes Out of Manway

I just put down the phone. Two pipefitters were burned opening
the demineralizer drum shown in Fig. 38.3. My client wanted to
understand what happened. It’s a common incident.
The water-filled drum had been depressured by venting to the
atmosphere. The drum had been drained to a level just below the
manway. The manway was carefully and slowly unbolted and
removed. Nothing happened for a few minutes. Suddenly the hot
water boiled out of the open manway.
This happens because the drum is not at a uniform temperature.
The water in the drum is at its bubble point or saturated liquid
temperature throughout the drum. But because the water in the lower
portion of the drum is at a higher pressure due to the static head of
water, its saturation temperature is also higher than the water at the
top of the drum. Also, the cooler water at the top of the drum is a little
denser than the hotter water at the bottom.
If something happens to disturb this delicate balance, the hot
water at the bottom will move toward the top. As the hotter water

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 469 3/20/08 12:56:10 PM

 Safety Concerns

470 A Working Guide to Process Equipment

FIGURE 38.3 Open steam vent

Boiling water can
explode out of
manway.

Zero
psig
Open
manway

212°F

Hot
water

220°F

Closed
drain

moves up, it loses a bit of head pressure and starts to vaporize. The
bubbles of steam stir up the vessel’s contents and accelerate boiling
and steam evolution. The sudden generation of steam pushes the hot
water out the open manway.
Honestly, this explanation is my best guess. But what I’m not
guessing at is that this incident is not unusual. It’s safer to drain
saturated water levels well below manways before they are opened. I
have accidentally drained 180°F heavy naphtha on my gloved hand
without injury. A similar careless incident with 180°F water resulted
in a trip to the infirmary. Excluding fire, hot water is more hazardous
to personnel than hydrocarbon liquids of the same temperature. Treat
hot water with care and respect.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2009 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.

Lieberman_Ch38_459-470.indd 470 3/20/08 12:56:11 PM