Liquefied Natural Gas properties and processing

Liquefied Natural Gas
(LNG)
By Eng./ Ramy Elzeiny

Introduction
At atmospheric pressure, and at the normal
boiling point of methane:
The liquid density is approximately 610
times greater than that of the gas at
ambient temperature and pressure.

Introduction
Consequently, a given volume of liquid
contains over 600 times the heating value
as the same volume of ambient gas.

Introduction
This density increase at ambient pressure
makes it attractive to:
 Liquefy,
 Transport,
 Store Natural Gas In Large Quantities

Introduction
Liquefaction and transport becomes
economically feasible when the size of the
reserves justify the capital investment of a
liquefied natural gas (LNG) plant.
Storage applications include storage at LNG
terminals and, just as important, storage for
peak shaving operations of gas utilities.

Introduction
After a general discussion of peak shaving and baseload
plants, this chapter focuses on eight topics:
1. Liquefaction cycles
2. Storage
3. Transport
4. Re-gasification and cold utilization
5. Plant efficiency
6. Safety and environmental considerations

Introduction
Figures 12.3, 12.6, and 12.7 clearly show the
large seasonal shifts in gas demand that
result in the need for gas storage facilities.

Introduction
Because natural gas fields are generally
located far from residential and industrial
consumers, storing large quantities of gas
near the point of consumption to
supplement the normal supply of pipeline
gas during periods of peak demand (peak
shaving) is essential.

Introduction
Gas storage techniques:
1. Underground storage in
 Depleted oil or gas fields,
 Salt caverns,
 Abandoned mines that can be effectively sealed.
2. Aboveground storage

Introduction
LNG Plants Categories
‘’Peak shaving plants’’ combine all three of
 Liquefaction,
 Storage,
 Re-gasification

Introduction
“Stranded” utilities” those not connected to
the national pipeline grid, rely upon LNG
received by truck to support their
customers.

Introduction
“satellite facilities” LNG facilities contain only
storage and re-gasification units

Introduction
Figure 13.1 shows a block diagram of the
common steps involved in a peak shaving
facility.
 Gas treating
 Compression
 Liquefaction,
 Liquid storage,
 Re-gasification.
 Odorant injection.

Baseload Plants And Stranded
Reserves
Baseload plants exist to provide the industrial
world with gas from stranded reserves in
remote places.

Reserves
Stranded gas reserves are located where:
 No economic use for the gas exists at the point of origin
 Pipeline transportation to the end user is not feasible.

Reserves
When compressed gas pipelines are impractical or impossible,
a limited number of conventional options are open such as:
 Compression and transport of the gas in specially built ships
 Conversion of the natural gas into a liquid through gas-to-liquid (GTL)
technology,
 Liquefaction and shipment of the gas in specially built LNG vessels.
 Conversion of the natural gas to hydrates for shipping.
Presently, LNG is the most viable option in almost all
situations involving stranded reserves , if the gas can be
pipelined to a seaport.

Reserves
As Figure 13.2 shows, bringing the gas from the field to the
customer involves four steps:
 Gas production, gathering, and processing
 LNG production, including gas treating, liquefaction, NGL
condensate removal, and LNG storage and loading.
 LNG shipping
 LNG receiving facilities, which include unloading, storage, re-
gasification, and distribution
Depending on the specific situation, not all plants will have all the
processes shown, and some plants may have additional processes.

Reserves

Reserves
To economically justify a traditional baseload LNG
plant requires reserves of approximately 3 Tcf (80
Bm3).
Newer designs have reduced the reserve volumes
down to around 1 Tcf (30 Bm3).

Gas Treating Before Liquefaction
Production of LNG requires temperatures as
low as −258°F (−161°C), the normal boiling
point of methane, and, consequently, the
allowable impurity levels in a gas to be
liquefied are much lower than that of a
pipeline-quality gas.

For example,
CO2 Content
Gas for pipelines contain a maximum of 3 to
4 mol% carbon dioxide ,
Gas for liquefaction should have a carbon
dioxide content of less than 50 ppmv.

Obviously, gas processed for LNG must have much
more aggressive removal of water, nitrogen, and
carbon dioxide than does gas destined for
pipelines.
The tight specifications on all the above
components, except for nitrogen and mercury, are
needed to avoid solids deposition that will plug the
heat exchangers.

Nitrogen is a volatile diluents which, at
higher concentrations, can raise the
potential for stratification and rollover
(discussed in section 8.4)

Elemental mercury presents serious problems in
cryogenic operations.
trace quantities of mercury condense in the
cryogenic heat exchangers and form an amalgam
with aluminum that can lead to exchanger failure.
Consequently, mercury must be removed to a level
of 0.01 mg/Nm3.

Table 13.4 compares compositional specifications for the two cases.

Table 13.5 shows the range of compositions
and properties for 17 LNG samples.
These compositions are for the LNG produced
and not the feed gas to the plant.

Liquefaction Cycles
The two most common methods that have
been used in engineering practice to
produce low temperatures are:
 Joule-Thomson Expansion.
 Expansion in an Engine doing External Work.

Joule-Thomson Cycles
The Joule-Thomson coefficient:
is the change in temperature that results when a
gas is expanded adiabatically from one constant
pressure to another in such a way that:
 No external work is done
 No net conversion of internal energy to kinetic energy of
mass motion occurs.

The Joule-Thomson coefficient:
Thermodynamically, it is an irreversible process that
wastes the potential for doing useful work with the
pressure drop.
However, it is as simple as a valve or orifice and
finds wide use in refrigeration cycles.

The thermodynamic definition of the Joule-
Thomson coefficient is:

One of the more important thermodynamic
relations that involves the Joule- Thomson
coefficient is

Combination of the above relation with the
ideal gas law (PV = RT) show that no
temperature change occurs when an ideal
gas undergoes a Joule-Thomson expansion.

For a real gas, the Joule-Thomson coefficient
may be:
1. Positive (the gas cools upon expansion),
2. Negative (the gas warms upon expansion),
3. Zero (no temperature change occurs)

The locus of all points on a pressure -
temperature plot where the Joule-Thomson
coefficient is zero is known as the inversion
curve.

Figure 13.6 shows that the Joule-Thomson
inversion curve for methane expansions
must take place below the curve to produce
refrigeration.

The behavior of several gases upon expansion
from 101 bar (1,470 psia) to 1 bar (14.5
psia) is shown in Table 13.6.

Two items should be noted
First, for both methane and nitrogen, the
cooling effect upon expansion when started
at ambient temperature (80° F, 27°C) is
relatively small.
Second, the cooling effect increases
significantly as the initial temperature is
lowered.

For helium, the expansion results in heating
the gas rather than cooling.

The temperature increase remains constant
because the Joule- Thomson coefficient
remains nearly constant over the
temperature range considered.

Because methane, the principal constituent of
natural gas, must be cooled to −258°F
(−161°C) before it becomes a liquid at 1
atmosphere pressure,
a liquefier that uses only a Joule-Thomson
expansion requires more than a compressor
and an expansion valve if it is to function at
reasonable initial pressures.

A counter-flow heat exchanger needs to be added to
make a complete system. A simple Joule-Thomson
system suitable for natural gas liquefaction is
shown in Figure 13.7.

The liquefaction cycle begins with natural gas
being compressed and sent through the
heat exchanger and expansion valve.

Upon expansion, the gas cools (approximately
84°F [47°C] if the gas is principally methane
and the expansion is from 1,500 to 14.7
psia [101 to 1 bar]),
but none liquefies because a temperature
drop of approximately 338°F (188°C) is
required to convert the gas to a liquid.

Thus, all of the chilled low-pressure gas is
recycled through the heat exchanger for
recompression.
This cold low-pressure gas lowers the
temperature of the high pressure gas
stream ahead of the expansion valve, which
results in a lower temperature upon
expansion.

As long as all of the gas being expanded is
recycled through the counter-flow heat
exchanger to cool the high-pressure gas
stream, temperatures will be progressively
lower upon expansion.
The process continues until liquid is formed
during the expansion from high to low
pressure.

The liquid formed is separated from the low-
pressure gas stream in the liquid receiver
and is ultimately withdrawn as the product.

The amount of low-pressure gas recycled to
the compressor is now significantly reduced,
which cuts back on the cooling effect in the
heat exchanger.

With the addition of makeup gas to the low-
pressure side of the compressor to
compensate for the liquid product being
withdrawn,
a steady-state is reached in the liquefaction
system and no further cooling can be
achieved.

The first law of thermodynamics for a steady-state flow
system is
Where:
 PE The potential energy per unit mass
 KE The kinetic energy per unit mass.
 h The enthalpy per unit mass.
 q Heat term per unit mass.
 Ws work term per unit mass.
 m The mass flow rate.

Application of the equation to the components inside
the thermodynamic boundary of Figure 13.7 (heat
exchanger, Joule-Thomson valve, and liquid
receiver) gives the relation
Where the overall enthalpy change of the gas, h, on
a mass basis equals the heat leak ,qL, per unit
mass of gas.

On a per unit of mass flow of entering gas, and
defining f= m1/m2 , the fraction of entering gas
withdrawn as a liquid, the equation becomes:

For a given system, h2, h3, and qL are
essentially fixed, so the only way to increase
liquefaction is to decrease the inlet gas
enthalpy, h1, which is done by increasing
the inlet pressure,
assuming that the compressor outlet gas
temperature remains constant. Thus, more
compressor work should lead to more liquid
production.

Example 13.1 Methane is to be liquefied in a Joule-Thomson
cycle as shown in Figure 13.7. The methane enters the
heat exchanger at 80°F and 1,500 psia and expands to
14.7 psia.
1. Calculate the fraction of methane entering the system
that is liquefied.
2. Estimate the % decrease in production if a heat leak ql of
15 Btu/lb of methane entering is present and if a
temperature approaches of 5°C is obtained at the warm
end of the exchanger.
3. Calculate the fraction liquefied if the pressure is 2,000
psia.

Calculate the fraction liquefied—An ideal heat
exchanger is assumed (no warm end ∆T and no
pressure drop). From the methane pressure -
enthalpy diagram and saturation table (Appendix
B) the following values are obtained:
h1 = 350 Btu/lb (80°F, 1500 psia)
h3 = 392 Btu/lb (80°F, 14.7 psia)
h2 = 0 Btu/lb = (−259°F, 14.7 psia, liquid)
Then by use of Equation 13.6

Effect of heat leak on production—Use the same cycle
but now have a 5°F temperature difference at the warm
end of the heat exchanger (t1, t3) and a heat leak, = 15
Btu/lb. This change lowers the recycle gas outlet
temperature to 75°F and h3 = 390 Btu/lb (75°F, 14.7 psia)
The liquid fraction generated now becomes:

Effect of pressure on production—Determine how liquid
production is affected by increasing the pressure on the
inlet gas to 2,000 psia from 1,500 psia.
h1 = 337 Btu/lb = (80°F, 2,000 psia)

This example illustrates the effect of pressure
and heat exchanger performance on liquid
yield.
For example, if the warm end ΔT for the heat
exchanger is approximately 77F (43C), a
very unlikely value, the liquid yield is
reduced to zero, even if no external heat
leaks are present.

This outcome raises the question of whether an optimum pressure exists.
In Equation 13.6, f will be a maximum when (h3-h1-qL) is a maximum
because the other terms are independent of inlet pressure.
The enthalpy of the liquid, h2, depends only on the liquid receiver
pressure, which we hold constant at the lowest pressure
(approximately 14.7 psia [1 bar]).
Also ql is independent of pressure and h3 is fixed at the lowest pressure
and the highest temperature (the inlet temperature for zero ∆T at the
warm end of the heat exchanger). Thus we maximize f when h1 is a
minimum.
The mathematical criterion is

Because thermodynamic optimum
pressure will occur when μ=0 or when the
inlet conditions are on the inversion curve.
However, many other factors must be
considered in selecting the economically
optimum inlet conditions. Considerable
improvement can be achieved in this simple
Joule-Thomson cycle, but at the expense of
added equipment and complexity of
operation.

The addition of an external source of
refrigeration markedly improves efficiencies,
as does the use of a double expansion of
the high-pressure gas instead of a single
expansion.

Although both of these techniques are
extensively used in air liquefaction plants,
only the dual-expansion process has found
favor in LNG processing.

Figure 13.8 shows the schematic of a
commercial facility that used the Joule-
Thomson cycle ,
This plant served a stranded utility, and its
total production was transported overland
by truck. It was designed and built to allow
easy movement to a new location.

Feed to the plant is obtained from a natural gas pipeline at 40°F (4°C) and pressures in
excess of 300 psig (20 barg).
The inlet gas is regulated to 300 psig (20 barg) and passed through a molecular sieve dryer
to remove both water vapor and carbon dioxide.
The gas then is compressed to 3,000 psig (210 barg) in an electrically-driven, two-stage
reciprocating compressor.
After passing through the three-stream heat exchanger, the gas undergoes a double Joule-
Thomson expansion, first to 300 psig (21 barg), and then to 10 psig (0.7 barg) to liquefy
the stream.
The LNG is transferred to one of the two storage tanks at the facility, either a 21,000 gallon
(Imperial) horizontal cylindrical tank that uses vacuum perlite insulation or a 35,000
gallon (Imperial) aluminum tank embedded in the ground.
the LNG is transferred to a 21,000-gallon (Imperial) storage tank before re-gasification and
distribution in the town’s natural gas system.

Expander Cycles
The point was made during the discussion of
the Joule-Thomson expansion that it was a
thermodynamically irreversible process.
Expansion of high-pressure gas to the lower
pressure in a reversible or nearly reversible
manner provides two distinct improvements
over the Joule-Thomson expansion.

Expander Cycles
First, in the reversible expansion, a large
fraction of the work required to compress
the gas can be recovered and used
elsewhere in the cycle. This property
provides an increase in cycle efficiency.
Second, the reversible process will result in a
much larger cooling effect.

Expander Cycles
For example
reversibly and adiabatically expanding
methane gas from 75 psia and 80°F (5.1
bar, 27°C) to 14.7 psia (1.01 bar) cools the
gas to −94°F (−70°C), a temperature drop
of 174°F (97°C).

Expander Cycles
A Joule-Thomson expansion between the
same pressure limits cools the gas
approximately 4°F (2.2°C).
Adiabatic reversible turbo-expansion provides
the most cooling possible over a given
pressure range.

Expander Cycles
Several options are available for selection of expanders for
LNG use, both:
 In the type of expander
 In the basic cycle itself.
Expanders are basically compressors with the flow reversed.
Expander types:
 Positive displacement
 Dynamic expanders are available.

Expander Cycles
Expander History:
►In 1902, Georges Claude pioneered
expander use in air liquefaction.
►Claude’s expander was a reciprocating
machine, as were most early machines used
in cryogenic processes, such as those
developed by Heylandt in 1912 and later by
Collins (1947).

Expander Cycles
► Barron (1966) reports reciprocating machine adiabatic
efficiencies of 70 to 80%.
He attributes reciprocating expander inefficiencies to four
causes:
 Inlet and outlet valve losses
 Incomplete expansion
 Heat transfer
 Piston friction
Reciprocating machines are rarely used in LNG facilities.

Expander Cycles
Similar to dynamic compressors, dynamic
expanders can be:
1. Centripetal flow.
2. Axial flow.

Expander Cycles
In centripetal turbo-expanders, the gas enters
through nozzles around the periphery of
the wheel, expands, and transmits work to
the wheel, which causes it to rotate, and
finally exhausts at low pressure at the axis
of the machine.

Expander Cycles
Axial-flow expanders have as their counterparts
steam turbines.
Axial-flow expanders are about 80% efficient
(Swearingen, 1968).
Centripetal machines have isentropic efficiencies on
the order of 85 to 90%.

Expander Cycles
Turbo-expanders are high-speed machines,
generally designed to operate from 10,000
to 100,000 rpm, depending on the
throughput.

Expander Cycles
For design purposes, several techniques may
be used to compute the expected enthalpy
change, but the simplest and apparently
satisfactory method is to use the ideal value
from a P-H or T-S diagram, and correct this
value with the anticipated turbo-expander
efficiency (Swearingen, 1968; Williams,
1970).

Expander Cycles
The work generated in the expander must be
removed from the system if the full
thermodynamic efficiency of the cycle is to
be realized.

Expander Cycles
The general practice in large-scale operations
is to couple the turbo-expander to a gas
compressor.
Expander-compressor combinations require
considerable care in their selection and
operation.

Expander Cycles
Reciprocating expanders would naturally be
coupled with reciprocating compressors, and
turbo-expanders coupled with centrifugal
compressors. The available expander work
can be very large. Swearingen (1968) states
that a turbo-expander handling 500 MMcfd
(14 Sm3/d) at pipeline pressure would
develop 10,000 hp (7,500 kW).

Expander Cycles
Surprisingly, the turbine rotor would only be
18 inches in diameter. In small-scale
operations, recovery of the expander work
is often not economically feasible. In this
case, the turbo-expander is simply coupled
to a braking device that dissipates the work.

Expander Cycles
Swearingen (1970) and the Engineering Data
Book (2005b) discuss what must be
considered in:
 The selection,
 Operation,
 Maintenance of turbo-expanders.

Expander Cycles
All expander cycles fall into two groups:
 Closed cycles
 Open cycles.
Note that most expander cycles have J-T
valves as well as turbo-expanders.

Expander Closed Cycles
In a closed expander cycle, the fluid being
expanded is not the fluid to be liquefied; the
expander simply acts as an external source
of refrigeration, similar to the propane
refrigeration.

For example, in LNG production, nitrogen may
be used in a closed expander system to
liquefy natural gas. A very simple schematic
of a closed cycle is shown in Figure 13.9.

The compressed nitrogen is expanded, and
the cold gas is then used to cool and liquefy
the natural gas stream. Actual cycles for
producing LNG are far more complex.

The closed cycle has several advantages over
the open cycle, in which the natural gas
itself is expanded.
First, if nitrogen is used, safety is enhanced,
because the closed cycle reduces the
number of processing steps in which
flammable natural gas is used.

Second, the closed nitrogen cycle has been
reported (Anonymous, 1970) to require
simpler and less expensive shutdown
procedures than its open cycle counterpart
and appears to be the most economical
process under many conditions.

Finally, because the natural gas is not passing
through the expander, the process purification
system is not so critical.
Gas passing through the high speed expander must
be free of condensed phases and any components
that solidify at the expander exhaust temperature,
because deposition on the rotor will destroy it.

Expander Open Cycles
An open expander cycle uses the gas being
liquefied as the expanding fluid and has the
advantage over the closed cycle of being
less complex.

A basic expander cycle is shown in Figure
13.10. In this example, the cold exhaust
stream from the expander is simply used as
a source of refrigeration, and the high-
pressure gas is liquefied as it expands
through the Joule-Thomson valve.

The first law of thermodynamics for a steady-
state flow system applied to the two heat
exchangers, the expander, and the liquid
receiver gives the following equation:

Where and (h4 − h6) is the work done by the
expander.
The quantities in Figure 13.10 represent
 m the mass flow rate into the liquefier,
 e the fraction of the gas to the expander
 mf the mass flow rate of liquefied product,
 me the mass flow rate to the expander,

An industrial LNG facility that uses an open expander cycle
(Figure 13.11) is the Chula Vista plant of the San Diego
Gas and Electric Company (Hale, 1966). The plant receives
25 MMscfd (0.71 Mm3/d) from a natural gas pipeline at
300 psia (21 bar) and 90°F (32°C). The gas is first
prepared for liquefaction by removal of CO2, H2S, and
water by physical adsorption on a molecular sieve. The
stream then splits, with about 21 MMscfd (0.59 Mm3/d)
going to the expander to provide refrigeration. The
refrigerant is initially cooled in the first heat exchanger
before going to a separator. Liquid from the separator
expands through a Joule-Thomson valve.

It recombines with the vapor from the separator,
which has been through a turbo-expander where
the pressure drops to 60 psia (4.1 bar) and the
temperature drops to −175°F (−115°C), and
through the second heat exchanger. This stream
provides cooling to both the incoming refrigerant
stream and the fraction to be liquefied in the first
exchanger. The gas then is compressed to 82 psia
(5.6 bar) before being odorized and sent to the
local power plant. Compression comes from work
done by the turbo-expander.

The 4 MMscfd (113 MSm3/d) of gas in the liquefier stream
passes through all three heat exchangers and a Joule-
Thomson expansion valve. Liquid and vapor are then
separated; the vapor stream passes through the heat
exchangers and then goes to fuel for the power plant.
Three-fourths of the gas that enters the liquefier becomes
liquid. When desired, the LNG is gasified by pumping the
liquid to 460 psia (32 bar) and vaporizing it in a hot water
heat exchanger. The gas, at 400 psia (27.5 bar) and 60°F
(16°C), is then ready for distribution. The vaporizer
capacity is 60 MMscfd (1.7 MMSm3/d).

The LNG is stored in a single 175,000-barrel (27,800 m3)
aboveground storage tank but has 1 MMscfd (28 MSm3/d)
of boil-off. The boil-off provides some refrigeration and is
compressed and combined with the vapor from the
separator before going to the power plant. With a net
liquefaction rate of 2 MMscfd (57 MSm3/d), 315 days are
required to fill the storage tank, but only 10.3 days are
required to empty the tank if vaporization is at the
maximum rate. This outcome matches the gas demand, as
the company typically has surplus gas available about 300
days a year. During this period, the storage tank is filled.
During the much shorter periods of peak demand, the LNG
is vaporized and placed in the distribution system.

Liquefied Natural Gas properties and processing

Recommended

Recommended

More Related Content

Similar to Liquefied Natural Gas properties and processing

Similar to Liquefied Natural Gas properties and processing (20)

Recently uploaded

Recently uploaded (20)

Liquefied Natural Gas properties and processing