Troubleshooting crude vacuum tower overhead ejector systems

Use these guidelines to improve performance and product quality

J. R. LINES AND L. L. FRENS, GRAHAM MANUFACTURING CO. INC., BATAVIA, NEW YORK

R outinely surveying tower overhead vacuum systems can

improve performance and product quality. These vacuum sys-
tems normally provide reliable and consistent operation.
However, process conditions, supplied utilities, corrosion, erosion
and fouling all have an impact on ejector system performance.

Refinery vacuum distillation towers use ejector systems to main-

tain tower top pressure and remove overhead gases (Fig. 1).
However, as with virtually all refinery equipment, performance
may be affected by a number of variables. These variables may act
independently or concurrently. It is important to understand
basic operating principles of vacuum systems and how perform-
ance is affected by:

• Utilities Fig. 1. Twin-element, three-stage ejector system on crude vacuum

tower.
• Corrosion and erosion
• Fouling An ejector may operate unstably if it is not supplied with enough
energy to allow compression to its design discharge pressure. If
• Process conditions.
the actual motive steam pressure is below design or its tempera-
Reputable vacuum-system suppliers have service engineers that ture above design, then, within limits, an ejector’s nozzle can be
will come to a refinery to survey the system and troubleshoot per- rebored to a larger diameter. The larger nozzle diameter allows
formance or offer suggestions for improvement. A skilled more steam to flow through and expand across the nozzle. This
vacuum-system engineer may be needed to diagnose and remedy increases the energy available for compression.
system problems.
If motive steam supply pressure is more than 10% to 20% above
design, then too much steam expands across the nozzle. This
UTILITIES tends to choke the diffuser. When this occurs, less suction load is
handled by the ejector and vacuum tower top pressure tends to
When a vacuum system is initially designed, utilities are estab- rise. If an increase in tower top pressure is not desired, then ejec-
lished and the most extreme conditions are usually used for the tor nozzles must be replaced with ones with smaller throat
design basis. Once operating, actual utility supply conditions can diameters.
be different than those set at the design stage and vary occasional-
ly. Important utilities for ejector systems are motive steam and Steam quality is important. Wet steam can be damaging to an
cooling water. Motive steam pressure, quality and temperature are ejector system. Moisture droplets in motive steam lines are accel-
critical variables. Flowrate and inlet temperature are important for erated to supersonic velocities and become very erosive. Moisture
cooling water. in motive steam is noticeable when inspecting ejector nozzles.
Rapidly accelerated moisture droplets erode nozzle internals. They
Motive steam conditions. These are very important and have a etch a striated pattern on the nozzle’s diverging section and may
direct impact on an ejector’s operation. If motive steam supply actually wear out the nozzle mouth. Also, the inlet diffuser tapers
pressure falls below design, then the nozzle will pass less steam. and throat will have signs of erosion. The exhaust elbow at the
When this happens, the ejector is not provided with enough ener- ejector’s discharge can erode completely through. Severe tube
gy to compress the suction load to the design discharge pressure. impingement in the intercondenser can also occur depending
The same problem occurs when the supply motive steam temper- upon ejector orientation. To solve wet steam problems, all lines
ature rises above its design value. Result: Increased specific up to the ejector should be well insulated. Also, a steam separator
volume and, therefore, less steam passes through the nozzle.

Hydrocarbon Processing®, March 1995 1

 EJECTOR FUNDAMENTALS
The basic operating principle of an ejector is to convert pressure
energy into velocity. This occurs with adiabatic expansion of
motive steam across a converging/diverging nozzle from motive
pressure to suction load operating pressure. Supersonic velocity
from the nozzle mouth results. Typically, velocities of mach 3 to 4
are achieved.
In operation, motive steam expands to a pressure below the suction
pressure. This creates a driving force to bring the suction load into
the ejector. High-velocity motive steam entrains and mixes with
the suction load gas. The resulting mixture is still supersonic. As
this mixture enters the converging/diverging diffuser, high velocity
is reconverted into pressure. A diffuser’s converging section reduces
velocity as crossflow area is reduced. The diffuser’s throat is
designed to create a normal shock wave. A dramatic increase in
pressure occurs as the flow across the shock wave goes from super-
sonic to sonic to subsonic after the shock wave. In the diffuser’s
diverging section, cross-sectional flow area is increased and velocity
is further converted to pressure. Fig. 2 details ejector components
Fig. 2. Ejector components and pressure profile. and a pressure profile for an ejector having a compression ratio in
excess of 2:1.
Ejector systems are required to operate over a wide range of condi-
with a trap should be installed immediately before an ejector’s tions—from very light loads to loads above design. An ejector
motive steam inlet connection. In some cases, a steam superheater system must stably adapt to all anticipated operating conditions.
may be required. Determining the design non-condensable and light-end hydrocar-
bon loading is essential for stable operation. Furthermore, an
Wet steam can also cause performance problems. When water accurate understanding of system back pressure is important.
droplets pass through an ejector nozzle, they decrease the energy Ejector systems may be configured a number of different ways to
available for compression. The effect is a decrease in load han- offer flexibility in handling various feedstocks and differing refin-
dling ability. With extremely wet steam, the ejector may even ery operations. A single vacuum train with one set of ejectors and
condensers has the lowest initial capital cost, but flexibility is limit-
break operation. ed. Often, parallel ejector trains are installed for each stage. Each
parallel ejector will handle a percentage of the total loading. For
Cooling water. Ejector system intercondensers and intercon- example:
densers are designed to condense steam and condensible • Twin element ejectors, each designed for 50% of total load
hydrocarbons, and cool non-condensible gases. This occurs at a
pressure corresponding to the preceding ejector’s design discharge • Triple element ejectors, each designed for 40% of total load-
ing for 120% capacity
pressure and the following ejector’s design suction pressure. When
the cooling water supply temperature rises above its design value, • Twin element, 1⁄3:2⁄3 ejector trains
ejector system performance is penalized. A rise in cooling water • Other configurations.
temperature drives down a condenser’s available log-mean temper-
Parallel ejector trains allow one train to be shut down for mainte-
ature difference (LMTD). The condenser does not condense nance while the column operates at reduced conditions. Also, at
enough and more vapors are carried out with the non-condensible light loadings, a train may be shut down to conserve refinery oper-
gases as saturation components. A pressure drop increase across ating costs. Fig. 3 shows a typical vacuum tower ejector system
the condenser is noticeable. The ejector following this condenser with a triple element ejector and first intercondenser. The second
cannot handle the increased load at this pressure. Pressure rises intercondenser and aftercondenser are a single element.
and the preceding ejector does not have enough energy to dis-
charge to the higher pressure. Result: The preceding ejector
breaks operation and the system may become unstable. CORROSION AND EROSION

This also occurs if the cooling water flowrate falls below design. Corrosion may occur in ejectors, condensers or vacuum piping.
At lower-than-design cooling water flowrates, there is a greater Extreme corrosion can cause holes and air leaks into the system.
water temperature rise across a condenser. This also lowers This destroys vacuum system performance.
LMTD and the above situation occurs.
Erosion may occur within the ejectors. Poor steam quality and
Problems with cooling water normally occur during summer high velocities erode diffuser and motive nozzle internals. An
months. This is when the water is at its warmest and demands on ejector manufacturer will provide certified information that gives
refinery equipment are highest. If the cooling water flowrate or the motive nozzle and diffuser throat design diameters. If a rou-
temperature is off design then new ejectors or condensers may be tine inspection of these parts indicates an increase in
required to provide satisfactory operation. cross-sectional area over 7%, then performance may be compro-
mised and replacement parts will be necessary.

Hydrocarbon Processing®, March 1995 2

Corrosion is a result of improperly selected metallurgy. Ensure that
the most appropriate materials are used before replacing parts. A
common corrosion problem occurs when carbon steel tubing is
used in condensers. Although carbon steel may be suitable for the
crude feedstock handled, it is not always the best practical choice.
It does offer the initial advantage of lower capital cost. However,
operating problems far outweigh modest up-front savings.

Vacuum towers undergo periods of extended shutdown for routine

maintenance, revamp or other reasons. During this period, a con-
denser with carbon steel tubing will be exposed to air and will rust
and develop a scale buildup. When the system starts up, the con-
densers are severely fouled. They will not operate as designed and
vacuum system operation is compromised. Modest savings in ini-
tial investment for steel tubing is quickly lost with less-than-
optimal tower operations due to rusted and scaled tubing. Vacuum
system manufacturers often caution against using carbon steel tub-
ing.

Fouling. Intercondensers and aftercondensers are subject to foul-

Fig 3. Typical multi-stage system.
ing like all other refinery heat exchangers. This may occur on the
tubeside, shellside or both. Fouling deters heat transfer and, at
some point, may compromise system performance. Process conditions. These are very important for reliable vacuum
system operation. Process conditions used in the design stage are
Cooling tower water is most often used as the cooling fluid for vac- rarely experienced during operation. Vacuum system performance
uum condensers. This water is normally on the tubeside. Typical may be affected by the following process condition variables that
fouling deposits on tubing internals cause a resistance to heat trans- may act independently or concurrently:
fer. Over a prolonged period of time, actual fouling may exceed the
design value and condenser performance falls short of design. • Non-condensible gas loading, either air leaks or light-end
hydrocarbons
Vacuum tower overhead gases, vapors and motive steam are nor-
• Condensible hydrocarbons
mally on the condenser’s shellside. Depending on tower
fractionation and the type of crude processed, a hydrocarbon film • Vacuum tower loading
may develop on the tube’s outside surface. This film is a resistance
• Vacuum system back pressure
to heat transfer, and over time, this fouling will exceed design.
Once this occurs, condenser performance falls short. • Condenser condensate barometric leg.

Routine refinery procedures should include periodic cleaning of

condenser bundles. Cleaning procedures must be for the con-
denser’s tubeside and shellside. To facilitate shellside cleaning, a
common practice is to specify removable tube bundles (i.e.,
TEMA designation BXU, AXS or AXT).

Fig 4. Typical first-stage ejector operating curve. Fig. 5. Preferred recycle control scheme to maintain tower pres-
sure at design when handling overheads below design.

Hydrocarbon Processing®, March 1995 3

Non-condensible loading. Vacuum systems are susceptible to Non-condensible loadings must be accurately stated. If not, any
poor performance when non-condensible loading increases above vacuum system is subject to performance shortcomings. If non-
design. Non-condensible loading to a vacuum system consists of condensible loadings are consistently above design, then new
air leaking into the system, light-end hydrocarbons and cracked ejectors are required. New condensers may be required depending
gases from the fired heater. The impact of higher-than-design on severity.
non-condensible loading is severe. As non-condensible loading
increases, the amount of saturated vapors discharging from the Condensible hydrocarbons. Tower overhead loading consists of
condenser increases. The ejector following a condenser may not steam, condensible hydrocarbons and noncondensibles. As differ-
handle increased loading at the condenser’s design operating pres- ent crude oils are processed or refinery operations change, the
sure. The ejector before the condenser is not designed for a higher composition and amount of condensible hydrocarbons handled
discharge pressure. This discontinuity in pressure causes the first by the vacuum system vary. A situation may occur where the con-
ejector to break operation. When this occurs, the system will densible hydrocarbon loading is so different from design that
operate unstably and tower pressure may rapidly rise above design condenser or ejector performance is adversely affected.
values.

Hydrocarbon Processing®, March 1995 4

This may occur a couple of different ways. If the
condensing profile is such that condensible hydro-
carbons are not condensed as they were designed
to, then the amount vapor leaving the condenser
increases. Ejectors may not tolerate this situation,
resulting in unstable operation.

Another possible effect of increased condensible

hydrocarbon loading is an increased oil film on the
tubes. This reduces the heat-transfer coefficient.
Again, this situation may result in increased vapor
and gas discharge from the condenser. Unstable
operation of the entire system may also result.

To remedy performance shortcomings, new con-

densers or ejectors may be necessary.
Readings that should be taken include:
Tower overhead loading. In general, a vacuum system will track
tower overhead loading as long as noncondensible loading does • Motive steam pressure and temperature measured as close as
not increase above design. Tower top pressure follows the first- possible to each ejector’s inlet
stage ejector’s performance curve. Fig. 4 shows a typical
• Suction and discharge pressure and temperature of each ejector
performance curve. At light tower overhead loads, the vacuum
system will pull tower top operating pressure down below design. • Cooling water inlet and outlet temperature of each condenser
This may adversely affect tower operating dynamics and pressure
• Tubeside pressure drop across each condenser
control may be necessary.
• Condensate temperature from each condenser, if available.
Tower pressure control is possible with multiple element trains.
At reduced overhead loading, one or more parallel elements may With these readings, a step-by-step comparison against the origi-
be shut off. This reduces handling capacity, permitting tower nal design criteria should allow a determination of the cause for a
pressure to rise to a satisfactory level. If multiple trains are not deficiency in the system.
used, recycle control is another possible solution. Here, the dis-
charge of an ejector is recycled to the system suction. This acts as Tables 1 and 2 will help in analyzing the system to determine any
an artificial load, driving the suction pressure up. With a multi- problems. Table 1 pertains to ejector performance and the areas
ple-stage ejector system, recycle control should be configured to that effect an ejector, namely, motive steam pressure and quality,
recycle the load from before the first condenser back to system suction load and discharge pressure. Table 2 relates to condenser
suction (Fig. 5). This way, noncondensible loading is not allowed performance.
to accumulate and negatively impact downstream ejectors.
If there are still questions as to the exact problem after complet-
System back pressure. Vacuum system back pressure may have ing the above evaluation, it is advisable to contact the original
an overwhelming influence on satisfactory performance. Ejectors manufacturer or have a qualified vacuum service engineer visit the
are designed to compress to a design discharge pressure. If the site to help analyze the system.
actual discharge pressure rises above design, the ejector will not
have enough energy to reach the higher pressure. When this
occurs, the ejector breaks operation and there is a sharp increase CASE HISTORY
in suction pressure.
Operating survey of a vacuum system on a crude tower in a
When back pressure is above design, possible corrective actions are South American refinery. Problem: The refiner was dissatisfied
to lower the system back pressure, rebore the steam nozzle to permit with the vacuum tower operating pressure. Tower top pressure
the use of more motive steam or install a completely new ejector. exceeded its design value. Table 3 shows the results of the system
operating survey that was conducted.
Operating survey. The most practical way to troubleshoot a vacu-
um system is to perform a pressure and temperature profile while The system was shut down and equipment inspected.
the system is in normal operation. Then compare the readings Dimensionally, the ejectors were in satisfactory condition, but the
with the origin design criteria. Before performing a survey, it is second-stage ejector showed signs of erosion. The second-stage
critical to have accurate instrumentation like calibrated ther- ejector diffuser throat was 4% to 5% larger in diameter than
mometers, pressure gauges and an absolute pressure gauge for design. Each ejector had a heavy hydrocarbon film on the motive
vacuum readings. steam nozzle exterior and air chamber.

Hydrocarbon Processing®, March 1995 5

The first intercondenser was heavily fouled on the tube and shell- THE AUTHORS
side. The tubeside had an excessive white scale that caused the
high tubeside pressure drop. The shellside had a heavy black film James R. Lines is vice president of engineering
coating the tubes. This was similar to what was noticed in the for Graham Manufacturing Co., Inc., Batavia,
ejectors. The secondary condensers had similar fouling deposits. New York. Since joining Graham in 1984, he
has held positions as an application engineer,
Solutions: The higher-than-design cooling water temperature product supervisor and sales engineer focusing
and excessive fouling of the condensers affected condenser per- on vacuum and heat transfer processes. Mr
formance to the extent that inadequate condensation was taking Lines holds a BS degree in aerospace engineer-
place. The impact of this was with the ejectors. Higher vapor ing from the University of Buffalo.
loads exiting the first intercondenser could not be handled by the
second-stage ejector at 86 mmHg Abs. This forced the first-stage Lance L. Frens is a senior contract engineer/
ejector to break operation resulting in a substantial increase in technical services supervisor for Graham
tower operating pressure. Manufacturing Co., Inc., Batavia, New York.
Since joining Graham in 1967, he has held posi-
Quotes were made to replace existing condensers with new ones tions as a test technician and application engineer.
having higher design fouling factors and based on higher cooling Mr Frens has 27 years of hands-on experience
water temperature. Once installed, tower top pressure was at surveying, revamping and troubleshooting vacu-
design and operation was stable. um and heat transfer systems worldwide.

Hydrocarbon Processing®, March 1995 6

Reprinted from March 1995 Hydrocarbon Processing® 7