380 UNDERSTANDING PROCESS EQUIPMENT FOR OPERATORS AND ENGINEERS

Vac. off gas 150 psig

steam

90%
open
Spill-
back

“A”

“B”

“C”

FIG. 48.7 Spill-back pressure control. Too many jets in service.

excessive cooling air flow. However, there is a more energy-efficient way to reduce air flow.
That is, reduce the blade angle pitch. I’ve only done this once. At a Sasol plant in South
Africa. I reduced the fan blade pitch from 20–25 degrees, down to 10–15 degrees.
I could only make these adjustments to the nearest five degree increment. Two nuts
had to be loosened on each blade. The result was small. Twelve percent reduction in
the measured amp electric load on the AC motor. But it only required 1 h.

Centrifugal Compressors
As the molecular weight of gas flowing to a centrifugal compressor increases, the gas den-
sity will also increase. This will not affect the polytrophic head produced by the compres-
sor. However:
Delta P ¼ ðPolytrophic HeadÞ  ðDensityÞ
 Chapter 48 • Suppressing CO2 Emissions 381

Gas

PC

A
b
Wet s
gas o
r
FC
b
e
r

Wet-gas drum PC

P1

Motor

Compressor
FIG. 48.8 Suction throttling pressure control.

Thus, the compression ratio of the machine will increase, as the molecular weight of
the feed gas increases. As a consequence, either the compressor’s discharge pressure will
rise or the compressor’s suction pressure will decline. Either way, the operator will be
forced to slow down the compressor to avoid the unwanted increase in compression ratio.
Suppose the machine is driven by an ordinary fixed speed, 3-phase, AC motor.
Then what?
The usual operating response is to open the spill-back suction pressure control valve
(Fig. 48.8). Which works fine. But, the recycled gas has to be recompressed, which wastes
energy. A more energy-efficient response to excessive molecular weight is to suction throt-
tle. It may seem odd, but suction throttling is a more energy-efficient response to control-
ling the delta P developed by a constant speed centrifugal compressor, than is spilling back
gas to the compressor’s suction. Try it yourself. Throttle the back-pressure control valve by
a few psi on your constant speed compressor. Watch the spill-back valve close. You will see
that the amp load on your motor driver will decline. But, don’t cause the spill-back valve to
close too much. You may cause the compressor to surge and potentially damage the rotor.
382 UNDERSTANDING PROCESS EQUIPMENT FOR OPERATORS AND ENGINEERS

Discharge

Adjustable
unloading Piston
pocket

Cylinder
head

Suction
FIG. 48.9 An unloading pocket reduces engine load and volumetric capacity.

The relationship between the energy requirement to run a constant speed compressor,
the operation of the suction throttle valve, and spill-back valve, and surge, is a complex
subject. I’ve explained these relationships in detail in my book: Troubleshooting Process
Operations, 4th Edition – PennWell Publications.

Reciprocating Compressors
The most energy-efficient method to control the flow through a recip is by opening the
head-end unloader pocket. That’s the large wheel at the head-end of each compression
cylinder. Typically, this wheel is not operational through lack of use, so you will need a
large valve wrench. Opening the head-end unloader all the way will reduce the volume
of gas flow by a maximum of 25% (Fig. 48.9). Reducing net gas flow through your recip-
rocating compressor, by opening the discharge to suction spill-back valve, is a direct waste
of compression energy. Use of the valve unloaders (really disabling the compressor’s cyl-
inder valves) is not quite as bad as opening the spill-back valve, but not nearly as efficient
as using the head-end unloader pocket. Opening this pocket takes several minutes, and
can save up to 25% of compression energy, that was being wasted by the discharge-to-
suction gas recycle valve.

Extending Mankind’s Future on Planet Earth

I imagine all my readers now believe in Global Warming. The problem is rapidly spinning
out of control. In summary, I’ve listed ideas that you can use to reduce CO2 emissions:
1. Tray Deck Levelness in Fractionators
2. Use of Picket Weirs
3. Optimizing Pressure in Distillation Towers
 Chapter 48 • Suppressing CO2 Emissions 383

4. Heat Exchangers—Shell Side

5. Heat Exchangers—Tube Side
6. Fired Heaters
7. Steam Turbines—Topping
8. Steam Turbines—Condensing
9. Motor-Driven Centrifugal Pumps
10. Fixed Speed Centrifugal Compressors
11. Fixed Speed Reciprocating Compressors
12. Steam Condensate Recovery

Tray Deck Levelness

Ordinary valve, grid, and sieve trays will leak at the low points of the tray decks. The vapor, on
the other hand, will take the path of least resistance, and flow to the higher points on the tray
decks. The result is vapor-liquid channeling, and therefore low tray fractionation efficiency.
To improve fractionation, the operator increases reflux rate, which then requires more
reboiler heat. The resulting incremental heat comes from plant steam, which wastes energy.
Leveling tray decks makes the trays work more efficiently and saves energy, especially
when the tower is not heavily loaded and/or the tray deck flow path length is long.

Use of Picket Weirs

Picket or notched weirs are not needed if:
• The weirs are level.
• If the weir loading is greater than 4–6 GPM per in. of weir length.
If not, use of picket weirs to increase weir loadings to 4–10 GPM per in. is going to increase
tray fractionation efficiency.
But why?
If the crest height (i.e., the height of liquid over the top of the weir) is zero or very small,
on the left side of a tray, then all of the liquid will flow across the right side of the tray. The
liquid on the left side of the tray deck will be stagnant.
Liquid that is stagnant cannot alter the composition of the upflowing vapors. Hence
tray efficiency is lost and can only be regained by increasing both the reflux and reboiler
rates. Which wastes energy.

Optimizing Pressure in Distillation Towers

Lowering the tower pressure will increase the difference between the vapor pressures of
hydrocarbons. That is, lower pressure increases relative volatility. The greater the relative
volatility, the less reflux and reboiler heat is needed to make the required split. However,
lowering the tower pressure too much may cause excessive entrainment, which reduces
fractionation efficiency.
384 UNDERSTANDING PROCESS EQUIPMENT FOR OPERATORS AND ENGINEERS

Optimum tower pressure can be determined by running plant tests at various pres-
sures. There is no way to calculate the optimum tower pressure for an existing unit. It’s
dependent on:

• Tray deck levelness

• Weir levelness
• Nonideality of the key components
Typically, at reduced rates, a lower operating pressure is more optimum.

Heat Exchangers—Shell Side

The best heat transfer is always obtained by placing the high viscosity fluid (i.e., cold
crude; tar) on the shell side. To enable cleaning of the shell side in tar, or asphalt service,
use:

• 100 tubes
• 1½00 tube spacing
• Rotated square pitch
• 316 ss tubes
This will provide straight, ½00 gaps, to hydroblast the shell sides during a turnaround. Fill
the space between the tubes caused by the channel side pass partition baffles with dummy
tubes. This is only necessary if the shell side tube support baffle cut is perpendicular to the
pass partition baffles.
Make sure there are seal strips that prevent shell side bypassing across the gap created
on the top of the shell, by the shell side inlet flow impingement plate. Too often, these seal
strips are lost in maintenance activities when the bundle is pulled.

Heat Exchanger—Tube Side

Tube velocities should be kept between 6 and 12 ft/s. Above 12 ft per second, tube erosion
may occur.
During a turnaround, make sure the channel head pass partition baffle is tightly
secured to the channel head tube sheet gasket. Leave off the channel head cover to check
this closure. A small gap will likely erode out into a major bypass stream.
High alloy tubes, which remain smooth, will resist fouling far better than carbon steel
tubes. However, do not use 316 ss with carbon steel tube support baffles. This will result in
galvanic corrosion of the baffles.
I never use tube inserts. They’re great at first, but interfere with tube cleaning during a
turnaround. Twisted tubes create shell side cleaning problems. Helical baffles, on the
other hand, usually work fine, but are expensive and only effective, if controlling resis-
tance to heat transfer is on the shell side.