Cryogenic Process
Cryogenic Process
Cryogenic Process
3
CRYOGENIC PROCESS
(EXPANSION &TURBOEXPANER)
2009
1. INTRODUCTION
The first turboexpander application for natural gas processing was accomplished
using Rotoflow technology in Texas in the early 1960s. It dramatically
demonstrated how efficiently the expansion turbine could condense heavier
components of the gas stream, while at the same time providing power to re-
compress the leaner gas. The current range of Rotoflow Turboexpander-
Generators grew from that original application. Forty years later, close to 4,000
units are in operation around the world, and Rotoflow has proven, many times
over, its ability to engineer machinery that delivers higher power levels, functions
at extreme operating temperatures, and achieves greater pressure ratios. Our
turboexpander experience and technology have become invaluable resources for
every segment of the natural gas and hydrocarbon industries. The rapid growth of
Rotoflow turboexpander technology has been a story of continuous improvement
in expander design, rotor and bearing design, efficiency optimization, and control
systems. This growth has been driven by the needs of the industry to increase
capacity, reduce costs, and maximize reliability. By adopting an internal business
structure that parallels the structure of the oil and gas industry, GE Energy is able
to address those needs directly, as an active partner in the search for more
effective solutions. Four of GE Energys Oil & Gas Business Units are closely
involved in the application of Turboexpander-Generator solutions for natural gas
and related processes.
The most efficient lean oil absorption plants recover only about 40 percent of the
ethane, 90 percent of the propane, and 100 percent of the butane and heavier
hydrocarbons from the gas. Additional heat is required to separate the products
from the lean oil, and additional cooling is required in order to re-liquefy.
the raw products before fractionation. Lean oil absorption plants usually have
higher operating costs than refrigeration plants or turbo expander plants.
Therefore, this old type process was dropped from selection.
Refrigeration Plants
If the major purpose of a plant is to condition rich gas to meet certain pipeline
specifications, the mechanical refrigeration plant may be the proper selection.
Refrigeration is used to condition produced gas to meet pipeline hydrocarbon dew
point specification, Btu specification, limited liquid recovery of heavier
hydrocarbons such as C5+ or a combination of these objectives. The straight
refrigeration plant is limited to chilling the gas stream to the range of -34C to
-40C. This limits product recovery to about 60 percent of the propane and much
less ethane at typical plant operating pressure. As the intention of the project is to
recover heavier components starting from Ethane, this technology was skipped.
First of all , NGL recovery is also called ethane recovery ,C2 + recovery or
cryogenic gas plants.
The word, cryogenics, defined as being "the science that deals with the production
of very low temperatures and their effects on the properties of matter." The
cryogenic gas plant is built around a process that produces very low temperatures
in order to affect the properties of natural gas, namely to cause certain of the gas's
components to condense to liquid. The gas components which condense form a
liquid hydrocarbon mixture known as Natural Gas Liquids or NGL. The NGL
mixture contains a number of valuable hydrocarbons that can subsequently be
separated into individual products, such as propane, gasoline and petrochemical
feedstock.
The cryogenic process has application as the most economical means for
recovering a high percentage of all hydrocarbons heavier than methane.
Specifically, the cryogenic process aims to recover ethane and heavier
hydrocarbons, such as propane, butane and gasoline components. Different
variations of this process are capable of removing more than 85% of the ethane
and essentially all of the heavier hydrocarbons found in produced natural gas. By
contrast, other processes may be more appropriate when the goal is to recover just
propane and heavier components. The main advantage of the cryogenic plant, in
terms of recovered product, is therefore its ability to recover ethane or high
propane recoveries.
There are three general methods which can be used to achieve the conditions
necessary to attain high ethane recovery levels.
1. J-T Expansion
2. Turboexpander
3. Mechanical refrigeration
Each of these processes has been used successfully, with the turboexpander being
the predominant process of choice for ethane recovery facilities. In the following
context both J.T. expansion & Turboexpander will be fully discussed in details.
We have seen how turboexpander plants remove both sensible and latent
heat from a feed gas stream to condense hydrocarbon liquids. Any gas will
possess both sensible heat and its latent heat of vaporization. Heat is a form
of energy, so a gas contains heat energy. A gas may contain other forms of
energy as well, such as pressure energy and velocity. The total energy, of all
kinds, possessed by a gas is called its "internal energy". Another name for
this internal energy is enthalpy.
The effect is named for James Prescott Joule and William Thomson (fig.1), 1st
Baron Kelvin who discovered it in 1852 following earlier work by Joule on Joule
expansion, in which a gas undergoes free expansion in a vacuum.
Temperature change of either sign can occur during the JouleThomson process.
Each real gas has a JouleThomson (Kelvin) inversion temperature above which
expansion at constant enthalpy causes the temperature to rise, and below which
such expansion causes cooling. This inversion temperature depends on pressure;
for most gases at atmospheric pressure, the inversion temperature is above room
temperature, so most gases can be cooled from room temperature by isenthalpic
expansion.
As a gas expands, the average distance between molecules grows. Because of
intermolecular attractive forces (see Van der Waals force), expansion causes an
increase in the potential energy of the gas. If no external work is extracted in the
process and no heat is transferred, the total energy of the gas remains the same
because of the conservation of energy. The increase in potential energy thus
implies a decrease in kinetic energy and therefore in temperature.
3. For continued operation during those times when the expander is down for
maintenance.
6.1. History
The possible use of an expansion machine for isentropically creating low temperatures
was suggested by Carl Wilhelm Siemens (Siemens cycle), a German engineer in 1857.
About three decades later, in 1885, Ernest Solvay of Belgium attempted to use a
reciprocating expander machine but could not attain any temperatures lower than 98 C
because of problems with lubrication of the machine at such temperatures.
The first turboexpanders seem to have been designed in about 1934 or 1935 by Guido
Zerkowitz, an Italian engineer working for the German firm of Linde AG.
In 1939, the Russian physicist Pyotr Kapitsa (fig.4) perfected the design of centrifugal
turboexpanders. His first practical prototype was made of Monel metal, had an outside
diameter of only 8 cm , operated at 40,000 revolutions per minute and expanded 1,000
cubic meters of air per hour. It used a water pump as a brake and had an efficiency of 79
to 83 percent. Most turboexpanders in industrial use since then have been based on
Kapitsa's design and centrifugal turboexpanders have taken over almost 100 percent of
the industrial gas liquefaction and low temperature process requirements. In 1978, Pyotr
Kapitsa was awarded a Nobel physics prize for his body of work in the area of low-
temperature physics
Many hydrocarbon turboexpanders, or expanders for short, are employed in processes for
the purpose of extracting heavier hydrocarbons from the gas stream. Fig.(5) is a
simplified schematic for such a process. The high-pressure gas mixture enters the process
at condition (1)and is cooled in a gas-to-gas heat exchanger. Usually some liquids are
condensed at (2) and are separated out in the high-pressure separator. The overheads from
this separator, (3) , are then expanded nearly isentropically in the expander and additional
liquids are condensed at (4) . These liquids are removed in the low pressure separator and
the overheads, (5) , are used for precooling the incoming gas in the gas-to-gas exchanger.
The warmed gas at (6) is sent to the booster compressor driven by the expander where the
gas pressure is increased to conditions at (7) .
For a natural gas dew point control process, the gas at 1 is to be conditioned for pipeline
transmission. The pressure drop between (3) and (4) is usually relatively small because
the required temperature drop between (3) and (4) is low. This is because only limited
quantities of hydrocarbon liquids need to be removed to make the hydrocarbon dew point
acceptable. For a natural gas ethane extraction process, the gas at 1 is being processed to
remove essentially all of the C2+ hydrocarbons from the stream. In this case, the pressure
drop between (3) and (4) is relatively large and the resulting low temperature at (4)
facilitates proper condensation in the demethanizer tower. In this process, relatively large
quantities of liquid are often extracted.
The radial inflow expander, utilizing variable inlet guide vanes, produces high
efficiencies over a broad operating range. Careful design of the wheel blade angles
and contour optimizes aerodynamic performance without compromising the
mechanical design integrity of the wheel. A typical expander wheel is shown
below. (fig.8)
The expander-compressor has a bearing on each end of the shaft. These bearings
must be continuously lubricated with clean lubricating oil(fig.10) of the approved
type at the proper temperature. Lubrication failure for a short-period of time may
result in a bearing failure which may seriously damage the machine .
The lube oil system is shown in the drawing below. Pressures shown are typical
for a plant in which the expander outlet pressure is 2070 kPa [300 psi]. Flow is as
follows:
Oil from the reservoir enters one of the pumps which is able to raise its pressure
several thousand kPa [several hundred psi]. A control valve in the pump discharge
line releases excess pressure through a spillover line that returns to the reservoir.
Oil from the pump enters a temperature control valve, which is positioned by a
temperature controller that allows some of the oil to by-pass the cooler to maintain
a constant temperature. The oil then flows through one of two filters to remove
solid particles and enters the bearings on each end of the shaft. The oil flows out of
the bearings(fig. 11) and drops to the bottom of the housing and flows by gravity
into the reservoir, and the cycle is repeated.
are now considered standard practice for certain applications. The application
of magnetic bearing technology to hydrocarbon turboexpanders has
important advantages over traditional oil bearing designs.
Some of the advantages of magnetic bearing technology are:
Eliminates Risk of Process Contamination: Since the lube oil is
completely removed from the system, there is no risk of oil migrating into the
process.
Bearing Losses: The frictional losses of magnetic bearings are usually lower
than conventional oil bearings. This additional power is not lost but recovered
as useful energy for compression or power generation.
Eliminates the Risk of Oil Dilution: In oil bearing systems, the mixing of
heavier seal gas and oil can result in a measurable dilution of the lube oil. If
not properly controlled,
this may cause insufficient oil viscosity, reduced load capacity, and increased
vibration. This potential problem is eliminated with magnetic bearings.
Lower Utility And Maintenance Costs: By eliminating the lube oil
system, the electric power and other utility requirements are reduced.
Maintenance requirements on pumps,
filters, coolers, and other maintained components are also reduced or
eliminated.
Reduced Weight And Space: Elimination of the lube oil system reduces
the overall weight and dimensions of the support skid. This makes magnetic
bearings ideally suited
for offshore applications, or wherever space is limited.
6.4.3 Seal Gas(fig.13) : Use a suitable gas stream with filtering and pressure
control to maintain proper gas pressure at the shaft seals. The seal gas
should be introduced before the lube oil system is started because there
might be a pressure upset which
would put enough oil into the process to cause a problem.
Each of the main rotating components (radial bearings, thrust bearings, and
shaft seals) can be damaged or eroded by improper oil filtration, lack of oil
flow, improper gas dehydration, and improper seal gas filtration.
Thrust Control
In any type of centrifugal device, thrust forces develop which tend to move the
shaft toward one end or the other. If it were to move laterally along its axis, the
impellers would touch the casing and quickly wear out.
In an expander-compressor, thrust bearings on each end of the shaft prevent
lateral movement. However, the thrust forces against the bearings must be
controlled at a moderate level to prevent bearing failure and serious damage to the
machine.
The thrust force is due to a difference in suction and discharge pressure acting on
the front and rear face of an impeller. Look at the expander impeller in the
following drawing. High pressure inlet gas enters at the tip of the impeller, and
leaks around the labyrinth impeller seal to the rear face and exerts a force to the
left. Low pressure outlet gas pressure is imposed on the front or left side of the
impeller. In order to neutralize the thrust in the expander impeller, gas which leaks
around the labyrinth seal on the rear face is slightly above outlet pressure.(fig.14)
C02 can be expected to freeze (form a solid phase) if the temperature of the process is
less than -80F, the triple point of C02. In the typical turboexpander process,
temperatures lower than the triple point are usually necessary to achieve the desired
levels of ethane and propane recovery. Carbon dioxide has a limited solubility in both the
liquid and vapor hydrocarbon phases at typical turboexpander plant process conditions.
Thus, the concentration of C02 in both the liquid and vapor phase must be considered
along with the temperature-pressure levels in the turboexpander plant demethanizer.
White et al. have presented a generalized correlation for predicting the conditions under
which CO2 freezeup can occur. Bergman and Yarborough have performed a series of
CO2 freeze out experiments on light hydrocarbon systems. Their work resulted in
correlations similar to the one given by White. The liquid freezeup curves from these two
correlations are essentially identical except at the high temperature end (-100F).
The possibility of CO2 freezeup should be checked at the outlet of the turboexpander and
in the top section of the demethanizer column. C02 freezeup can be avoided by: (1)
If the expander liquid is fed to the top tray of a demethanizer, the CO2 will concentrate in
the top equilibrium stages. This means that the most probable condition for solid CO2
formation may be several trays below the top of the tower rather than at expander outlet
conditions.
Turboexpander efficiencies typically range from about 60% to about 85%. Lower
efficiencies have been observed in some cases. A best first guess for turboexpander
efficiency would be in the range of 70-75%. After preliminary calculations have been
performed, this initial assumption can be adjusted using a procedure to be outlined later.
1. Any errors in the efficiency are directly reflected in the recompression power
requirements. Admittedly, these effects may be small, but on very large plants, they can
become more significant.
2. The errors in the predicted liquid formation directly affect the C2+ liquid recovery
and the heat balance on the demethanizer. Errors here are probably more significant than
the errors in the recompression load.
3. False C02 freezeup problems may be predicted if too high efficiency is used.
Conversely, a potential CO2 freezeup may not be properly predicted if the efficiency is
too low. This issue will be discussed in a later section.
The efficiency of a compressor driven by a turboexpander appears to be about 5% lower
than the turboexpander efficiency. This observation is completely empirical but seems to
match the performance of several operating plants.
Some feel that about 20 mass percent liquid in the expander outlet is the maximum liquid
formation that can be tolerated in the expander. is not formed in the area of the wheel, but
that liquid is formed in the outlet nozzle of the expander. Residence times in the area of
the wheel are on the order of nanoseconds (10" seconds). This is too short a time for big
(micron size) drops to form. Further, the ability of some specially adapted
turboexpanders to handle liquids tends to refute this limitation.
In compression systems, maximum compression ratios of three to four are routinely used.
The major reason for this limitation is the discharge temperature of the gas; a secondary
reason is the loading on the thrust bearings in the machine. Thrust bearing loading
A Joule Thomson valve is normally installed in parallel with the turboexpander. This
installation is made for several reasons: (1) startup; (2) continued operation if the
turboexpander fails; (3) permit operation at flow rates above the turboexpander limits.
So-called constant entropy nozzles have been used in. the space industry for years. The
Joule Thomson control system might be replaced with one of these constant entropy
nozzles to maintain turboexpander performance of the plant at high through puts or if the
turboexpander is out of service.
Among the factors that must be considered in the purchase of a turboexpander plant is the
high speed of the turboexpander (rpm >8000). Because of this high speed,
maintenance/part replacement problems can crop up particularly in remote installations.
Unit noise levels are high, special precautions must be taken to protect the operator from
this noise.
How much, if any, difference in the liquids recovery and temperature level should
be expected between the turboexpander and Joule Thomson processes? This
question cannot be answered specifically for every system, the results will be
contingent on the temperature-pressure-compositions encountered in the specific
case. However, one can say that if the operating pressure levels are identical, the
turboexpander process will generally produce lower temperatures and more liquids
than the Joule Thomson process. For these conditions, the turboexpander process
will require less recompression horsepower to recompress to the residue gas to
pipeline pressure.
Clearly the turboexpander process has a far better capacity for recovering liquids;
essentially 100% of the C4+ are recovered in the liquid stream, while only 61% of
the C4+ are recovered in the liquid for the Joule Thomson process. This
incremental liquid recovery is due simply to the much lower temperature achieved
in the turboexpander process (about 75F lower).