Advances in Cryogenic Air Separation

5/23/2018 Advances in Cryogenic Air Separation

1/25

Acknowledgement: Paper presented by Prof. K. Chowdhury at the 21st National Seminar on

Industrial Gases, 29th-30th January 1999, Bangalore, India.

ADVANCES IN CRYOGENIC AIR SEPARATION

Kanchan Chowdhury

Associate Professor

Cryogenic Engineering Centre

Indian Institute of Technology, Kharagpur 721 302(WB)

ABSTRACT

The use of atmospheric gases like oxygen, nitrogen, argon and rare gases has steadilyincreased during this century in a variety of commercial, defence, nuclear, space and power

applications. Keeping in tune with the technological advancement in other areas, air separation

technology too has improved in efficiency through improvement of process equipment and

cycles as well as through the assimilation of advanced technology in other areas. Attempt hasbeen made in this paper to highlight some of the areas in air separation technology, particularly

in process cycles, where the advancement has taken place. After improving component

efficiency to a large extent, the air separation industry is trying hard to reduce the power cost by

decreasing the losses due to the Second Law irreversibility in heat and mass transfer. The

constructions of large air separation plants have been made possible by extensive instrumentation

and computer control. Some recent patents show that this industry is truly following the dictum"million drops make an ocean" and inching towards the reduction of energy at every level.

1. INTRODUCTION

Air comprises of the constituents whose volume percentages and boiling points are given

below[1]:

Table 1. Composition of air and boiling points of components

Gas Volume or mole % ppm Boiling point (K)

N2- Nitrogen 78.11 --- 77.36

O2-Oxygen 20.95 --- 90.18

Ar - Argon 0.934 --- 87.28

Ne -Neon --- 18.2 27.09

He -Helium --- 5.2 4.21

Kr -Krypton --- 1.14 119.83


2/25

Xe - Xenon --- 0.086 165.0

H2- Hydrogen --- 0.5 20.3

N2O - Nitrousoxide Oxide

--- 0.5 183.7

CO - Carbon

monoxide

--- ~0.1 81.6

SO2 - Sulphur

dioxide

--- ~0.04 263.2

O3- Ozone --- ~0.02 161.3

Although pressure swing adsorption process and membrane separation process are increasingly

being used for the production of atmospheric gases, cryogenic separation is likely to remain asthe "workhorse" of the air separation industry in the foreseeable future, due largely to its

flexibility in terms of purity, flow rate, pressure and state (liquid or gas) of the products. In the

field of bulk production of cryogenic fluids and atmospheric gases, the cost of production

remains the major factor in the spread of its use in a variety of application areas.

Although, air separation technology has developed into a mature technology by 1935, itwas around 1955 when the real need for the improvement of air separation technology was felt.

The reason behind this need lay in the increasing use of oxygen in steel making, particularly the

invention of L-D process in Austria. Even today, by the end of the century, after about 90 yearssince Prof. Carl von Linde introduced his concept of double column, technologists around the

world are working towards the overall improvement of the Air Separation Technology through

invention of new processes and assimilation of the developments in other related technologies.These efforts have lead to cheaper methods of production of atmospheric gases. As fibre-

reinforced plastics and aluminium are continuously replacing steel as manufacturing and

construction materials, the cost of oxygen, which is supplied across the fence to the steel plants,

need to be reduced continuously in order to enable steel to be produced at a cheaper price.Reduced power consumption enables the lower production costs and the resultant lower cost

products can be used to stimulate the market growth [2]. Improvement of the efficiency of an air

separation plant, which forms an important objective of the industrial gas companies, can be

achieved by development of more advanced and efficient equipment and optimisation of designof process cycles.

2. CRYOGENIC DISTILLATION OF AIR


3/25

Cryogenic separation of air is done in a distillation column following the same principle

as in case of petroleum distillation. Crude petroleum is available in liquid form and a small part

is vaporised by application of heat to create both vapour and liquid in the distillation column.The principle of separation is based on the differences of boiling points of different constituentsin crude petroleum. In case of cryogenic air separation, however, the input air is available in the

form of gas and one has to liquefy a small part of it to effect distillation. Table 1 shows the

differences of boiling points of different constituents of air. Therefore, while petroleum

distillation involves boiling of liquid, distillation of air involves the refrigeration andliquefaction. While petroleum distillation is accomplished at a temperature 330

OC higher than

that of ambient, cryogenic distillation involves a temperature of about 230OC below the ambient.

Therefore, while heat has to be protected from being leaked out to atmosphere in case of

petroleum distillation, the same has to be prevented from leaking in case of cryogenicdistillation. Therefore, besides distillation columns, most of the equipment involved in air

separation are dedicated to liquefaction processes which involves compressors, equipment for

separation of moisture, CO2, hydrocarbons, heat exchangers, expansion devices, cryogenic

pumps etc. The block diagram in Fig 1 depicts the processes involved in cryogenic air

separation.

2.1 Process of compression

Compression process involves air filters, compressors, intercoolers and aftercoolers.There has been a lot of efforts by compressor manufacturers to improve the efficiency of

compression during the past half a century. Moreover, with the improvement of air separationprocesses and increase of the plant-size, it has been possible to shift from the high pressure low

capacity reciprocating compressors to the low pressure high capacity centrifugal compressors.

The latter is inherently more efficient. A lot of improvement has been achieved in theperformance of intercoolers and aftercoolers. Steep rise of cost of energy around the globe since

1970's oil crisis has motivated many companies to make more capital investment in intercoolers

and aftercoolers resulting in a steady decline in the operating cost of compression process.


4/25

2.2 Pre-purification processes

Reversing regenerators were used for separation of moisture and CO2till about 1960. The

process involved using two alternating regenerators filled with high heat capacity stones which,in the steady state, acquired a temperature gradient of 300 K to 80 K. During one cycle, one

regenerator is used to cool and purify the incoming stream while the other regenerator was usedto warm the outgoing stream and simultaneously re-evaporate and remove the frozen impurities

those were deposited during the cool down cycle.

First used around 1955, reversing heat exchangers, which are brazed aluminium heat exchangers,

were used to remove CO2and moisture from the air apart from the primary job of exchanging

heat between the warm incoming air and the cold separated products. In around 1960, a plant of300 TPD (tons per day) of oxygen used 20 cores in parallel of 0.43 m x 0.53 m x 2.7m long. By

1985, the size of brazed aluminium core reached around 1.2m x 1.3m x 6 m long [3]. A typical

brazed aluminium heat exchanger is shown in Figure 2.

With the appearance of molecular sieve in the market it has become possible to remove moistureand CO2 by adsorption near ambient temperature (5

OC). While the reversing heat exchangers

were subject to failure by fatigue, by water freezing and by corrosion, with the front end

purification, the operational reliability and down-time of the plant have considerably improved,particularly when the plants are located in heavily polluted areas. There have been continuous

efforts by industry to increase the adsorption temperature (so as to decrease the load on

refrigeration system), to increase its mechanical strength (so as to decrease its tendency to break-

down into dust) and to decrease the reactivation temperature (so as to decrease the energyconsumption in regeneration heating and later cooling). The recently used molecular sieves are

also capable of removing hydrocarbons, thus making it unnecessary to use silica gel adsorption

of hydrocarbon from rich liquid and liquid oxygen.

2.3 Refrigeration systems


5/25

Refrigeration in an air separation plant is done for two reasons: (1) reduction of air temperature

to facilitate front end purification and (2) to pre-cool air or nitrogen while they are being cooledto cryogenic temperature or are liquefied. With the advancement of the efficiency of

compression, refrigeration and heat exchanger technologies, the energy requirement of

refrigeration has decreased over the last few decades. Further, the recent air separation plantshave made all out efforts to utilise evaporative cooling in a packed-bed direct contact heat

exchanger utilizing bone-dry waste nitrogen and using the chilled water to cool the air before

front end purification. This is a big energy saving device.

2.4 Heat exchangers

The heat exchanger played the most crucial role in the establishment of commercial cryogenics

and bulk production of oxygen, when in 1902, Brin's Oxygen Company (the forerunner of theBritish Oxygen Company) acquired the sole rights to Dr. Hampson's patent on coiled tube heat

exchanger[4]. This heat exchanger served as a main heat exchanger in air liquefaction for morethan half a century. The shell and tube configuration used to be used as the condenser-reboiler.Both these applications of heat exchangers have now been replaced by plate-fin-heat exchangers

(PFHE), shown in figure 2. PFHEs have the advantages over other forms of heat exchangers,

such as: 1) Very close Temperature approaches and high thermal effectiveness. PFHE canachieve temperature approaches as low as 1

OC between single phase streams and 3

OC between

multiphase streams. 2) Large heat transfer surface per unit volume is available in PFHE . About

1000 m2

heat transfer surface area per cubic meter volume is available compared with 300 m2per

m3in a conventional shell and tube heat exchangers. 3) low weight per unit heat transfer. For a

given volume the PFHE weighs one-third compared to the conventional heat exchangers. 4) The

possibility of heat exchange between many process streams has made PFHE very welcome

addition as a vital equipment for air separation plant. However, PFHE is very expensive andbecause of its small flow passages blockage may arise from solids carried by process streams.The use of plate-fin heat exchangers has greatly improved the efficiency of air separation plants.

The improvement of efficiency of an air separation plant due solely to the improvement of heat

exchangers occurs primarily for three reasons[3]:

1. A closer temperature approach at the outlet of the PFHE(when used as the main heatexchanger) minimises the cold loss.

2. Lower pressure drop in the PFHE (when used as the main heat exchanger) results in alower pressure in the low pressure column, which meant a lower saturation pressure andconsequently a lower saturation (evaporation) temperature in the reboiler in the

condenser-reboiler. This results in the fall of required pressure at the high pressurecolumn and reduces the compressor discharge pressure.

3. Availability of higher surface with the PFHE (when used as condenser-reboiler) arealeads to a better heat exchange in the condenser-reboiler, which results in a lower T.

This results in the requirement of a lower saturation (condensation) temperature requiredin the condenser side of the condenser-reboiler, which means a lower pressure

requirement in the high pressure column and a lower discharge pressure from the

compressor.

Due to the above-mentioned advantages, the last few decades have seen the feed compressor


6/25

pressure decreasing from 7 bar(abs) to 5 bar(abs), which has resulted in a large improvement of

plant efficiency.

3. PROCESS CYCLES FOR CRYOGENIC AIR SEPARATION

At this point, we discuss some of the conventional air separation plants, which have been sold inthe market during the last few decades. Fig 3 shows a high pressure air separation plant

producing LN2, LOX and liquid argon.

These liquid plants, built around 1965-1970 used "high pressure" (140bar) reciprocating air

compressor, expansion engine and liquefied waste nitrogen recycling. WN2recycling helped in a

good argon recovery. There has been plants built around 1975, based on medium pressure (40 to50 bara) which used (either or both) recycling of liquefied low pressure pure nitrogen and

medium pressure pure nitrogen (Fig. 4). By this time, reliability of expansion turbines had

largely been improved. Compared with the high pressure plants, the medium pressure plantsbrought down the power consumption from about 1.7 units to about 1.5 units per Nm3of LOX.

With the development of high efficiency low pressure turbines, the compressor pressure has been

successfully brought down to the bottom column pressure at about 6 bara (1990 decade) (Fig. 5),

which are called "low pressure plants". These plants can produce liquid with a specific powerconsumption of about 1.3 units. During these 20 years (1970 to 1990) there have been large

improvements in the process equipment, which are primarily confined to compressor efficiency,

heat exchanger efficiency, quality of molecular sieve and turbine efficiency. But the most

impressive development of the 80's decade has been the development of structured packing bySulzer for the distillation columns[5, 6]. Fig 6 shows a primarily gaseous O2plant, which uses

structured packing for the columns. The reduction of price, easier availability and enhancement

of performance-related knowledge of the designers have lead to the widespread use of structuredpacking in this decade (1990), which has resulted in the elimination of H2dosification for theproduction of pure argon. Thanks to the use of structured packing, argon can now be produced at

5 ppm impurity without going through the expenditure and hazards of using H2.


7/25


8/25

3.1 Structured Packing


9/25

The use of structured packing instead oftrays for the distillation of air provides a

substantial reduction in the powerrequired to drive the air compressor.Packing has a pressure drop which is

about 10% to 20% that of trays. Fig 7

shows how the use of structured packingcan reduce the feed compressor discharge

pressure to 5.65 bara from 6.33 bara [7].

The resultant saving in power can be

substantial.

The distillation equipment is a masstransfer equipment required to produce a

degree of mass transfer from liquid to

vapor and reverse, which is achieved in aplate column by bubbling the vapor

through a pool of liquid a number of

times. Packed columns achieve this masstransfer by countercurrent flow of the

liquid and vapor stress. Packing is a

material of high surface area per unit

volume and of high voidage that is used to

fill the distillation column.

Liquid runs the down the surface and vapor flows through the voids and mass transfer occurs atthe liquid interface. Structured packing has a regular structure, as opposed to the randomness of

dumped packing and it is supplied in section that fit snugly, directly into the column. In spite of

its recent cost reduction, structured packing still costs more than trays. The additional cost of

packing over trays has to be offset against the power savings and additional argon recovery. Forstructured packing, HETP or height equivalent to a theoretical plate is defined as the height

difference between the points in a packed column where the gas composition is in equilibrium

with the liquid composition.

The salient points highlighting the advantage of structured packing over the plate is given below:

1. The pressure at the air compressorThe pressure at the air compressor outlet is dictated by the pressure required to condense

nitrogen in the condenser-reboiler within the main rectification columns. For every 0.1bar drop in pressure on the reboiler side (upper column), there is approximately 0.3 bar

advantage (reduction) in the pressure on the condenser side (lower column). Therefore, it

makes sense to use structured packing in the upper column.

2. Turndown flexibilityTwo factors which determine the flexibility of load variation or the operating range in aplate column, has been dealt in detailed by Latimer [1]. These are: entrainment-cum-


10/25

flooding and weeping. At higher vapor rate, liquid droplets are entrained in the vapor

stream and are returned to the plate above. This phenomenon is called liquid entrainment.It has a completely negative effect on the separation efficiency, because the enriched

liquid is simply mixed back again with the leaner liquid. The reduction in plate efficiency

is a function of the liquid entrained and beyond a few percent entrainment, there is anatural fall in efficiency. Further entrainment causes flooding. The operating range of the

column is generally 80% of the physical flooding point because of the reduced plate

efficiency.

At low vapor rates, with most tray designs, the liquid cannot be supported on the platesand "weeping" occurs, whereby some of the liquid passes through the plates, bypassingvapor liquid contact and causing vapor instabilities. This causes a reduction in the overall

plate efficiency, giving a minimum useful operating point for a plate [6].

Structured packing suffer from problems like flow maldistribution and inadequatewetting. Maldistribution occurs when the packing is large compared to the columndiameter. At low liquid rates, packing may suffer from inadequate wetting. Nevertheless,

it is claimed that structural packing, with the flexibility of operating at 30% to 110% of

design, can be used in an ASU to run it at a very low partial load [5].

3. Rapid Response to changeUse of structured packing contributes to another factor of significant importance to theoperation of the plant. The small hold-up in the packed column helps a plant with packed

columns to have a better load-following characteristics and responds quickly to a change

in oxygen and nitrogen demand.

4. Better recovery of argonBecause of low absolute pressures which are obtained in both the upper and lower columns, the

vapor-liquid equilibrium relationships become more favourable and the recovery of argon is

increased.

3.2 Improvements in Process Cycles

The key to improvement in the efficiency of process cycles lies in the reduction of losses. These

losses are of two kinds:

1) The loss due to First Law Of Thermodynamics and

2) The loss due to Second Law of Thermodynamics.

1) The first law of thermodynamics refers to the loss due to transfer of heat. For example, heat

inleak into the system is a loss of First Law of Thermodynamics. Similarly, the exit temperature

loss in the main heat exchanger causes a loss of cold (which had been produced after spendingenergy) and results in a loss of plant efficiency, which is of First Law of Thermodynamics in

nature. Attempts have been made during the last decade to reduce these losses through the

application of high efficiency heat exchangers and improved insulating materials. It is, however,

very natural to reach a point of diminishing returns, when large capital investments in heat


11/25

exchange and insulation systems are not likely to achieve a corresponding reduction in the

operating costs. It is very important to mention here that this decision of diminishing return is afunction of the prevailing cost of energy and, therefore, changes with time. For example, the

pressure drop in a pipeline carrying liquid or gas, which were acceptable in 1960, should no

more be acceptable today and there is nothing called a permanent design formula. Therefore,standing at the end of twentieth century, it is very easy to conclude that efficiency improvements

in process equipment and the resultant reduction of First Law losses have almost reached a

plateau.

2) During the last two decades there has been a continuous quest to reduce the losses due toSecond Law of thermodynamics. Thermodynamic losses occur throughout the air separationplants. By understanding the nature of these losses, they can be minimised, resulting in increased

overall efficiency of the process cycle[2]. Losses can be divided into 3 categories described

below:

a. Irreversible Expansion or Compression:When a gas flows though a pipe, there will be some pressure drop due to friction.

Although the pressure falls, there is no recovery of energy and so the work, which was

spent in increasing the pressure, is irreversibly lost. Similar losses also occur inmachinery such as compressors and expanders. The minimum work required to compress

a fluid from P1to P2isothermally is:

WORK = R TOln P2 / P1 . . . ( eqn. 1)

b. Irreversible Heat Transfer:In heat exchangers, heat flows from a warm stream to a colder stream. This exchange

takes place across a temperature difference of T = Tw Tc Since energy would be

required to transfer the heat back to the warm stream, this heat transfer represents a loss.Similar losses also occur in distillation where warm streams mix with colder streams on a

tray. A reversible heat pump transfers heat Q from a temperature Tcto Tw, using:

WORK = Q (Tw-Tc) / Tc . . . ( eqn. 2)

c. Irreversible Mass Transfer:In air separation plants, a waste N2stream is produced and vented to atmosphere. The mixing of

N2with the air is a loss, since energy is required to recover the N2stream from air. In addition toprocess mixtures such as this loss, losses can also occur when streams, which are not in

thermodynamic equilibrium, are mixed in distillation trays. In a reversible separation device,

using the perfectly selective properties of an ideal membrane to separate a gas into its

components, the work required is given by the expression:

WORK = R TOxiln xi, . . . ( eqn. 3)

where xi's are the mole fraction of the separated components.

Timmerhaus[8] has given the work involved in reversibly separating a specified gaseouscomponent from other components in air at 300K (Table 2 )


12/25

Table 2 : Work needed for reversible separation

Gas Mole % Work per mole of

mixture (KJ/kgmol)

Work per unit mass

of separated gas(KJ/kg)

Nitrogen 78.084 1311.6 60.0

Oxygen 20.946 1280.2 191.0

Argon 0.934 132.1 353.9

Carbon dioxide 0.033 7.42 511.1

Neon 1.818x10

-3

0.54 1472.5

Krypton 1.14x10-4

0.0404 438.1

Xenon 8.6x10-6

0.0037 328.1

Hydrogen deuteride 3.12x10-2

7.06 7485

Deuterium 1.56x10-2

3.80 6090

Helium 1.0x10-5

0.00427 14,220

For a modern plant, the specific power needed to produce the gaseous O2is of the order of 0.6KWh/Nm

3, while for liquid O2 it is of the order of 1.3 KWh/Nm

3. It may be worthwhile to

explain at this stage where this specific power is exactly utilized or lost in order to understand

the relevance of present improvements in the process cycles that are taking place today. Figure 8

shows the energy consumed in producing gas at atmospheric pressure, pressurised gas andliquids.

When air is separated into gaseous O2 and other constituents, it has to undergo the followingprocesses with the corresponding energy consumption. 1) Work of separation and 2) Make-up

work for heat inleak and heat exchanger losses. The latter can be accounted as de facto

liquefaction. (3) For producing pressurised gases, work is to be done in compressor or in liquid

pump. (4) Work is required for the production of liquid in an air separation plant.


13/25

By working out from the chart given in Table 2 for ideal separation work, the ideal powerrequirement for separating O2works out to be 0.08 KWh/Nm

3. However, column pressure drop,

finite temperature difference and mixing of fluids with different compositions cause the

deviation from ideality and thermodynamic efficiency of separation may be close to 0.25 to 0.30.

Therefore the energy requirement for separation alone works out to be 0.08/0.25 = 0.32

KWh/Nm3.

If insulation loses and heat exchanger losses are taken equivalent to about 6 % liquefaction of thefeed air, it means a refrigeration load of 0.027 KWh/Nm

3. The ideal work requirement to get this

refrigeration, according to equation (2), is 0.027 X (310K-77K) / 77K 0.081 units. If thethermodynamic efficiency of the actual refrigeration cycle is 30% of the ideal (Carnot Cycle),the actual power requirement is 0.081 / 0.3 = 0.27 units. So together, the specific power required

for producing gaseous O2works out to be 0.32 + 0.27 = 0.59 KWh/Nm3. In order to get a liquid

product we have to add to it the power required for liquefaction. The heat exchanger does not

have to handle oxygen and therefore the heat exchanger losses will decrease. Assuming that 18%O2 is liquefied, the power required will be 0.32 + 0.18 + 0.8 =1.3 KWh/Nm

3 .The recent

improvements in process cycle attempt to improve the work of separation, which has been

discussed in this paper with some detail.

3.2.1 Reduction of Separation Work

In order to reduce the net work consumption in a binary distillation, it is necessary toreduce the driving forces for heat and mass transfer within the individual stages. This reduces to

a problem of making the operating and equilibrium curves nearly coincident. The point is

illustrated in Fig.9. Fig 9(a) represents an ordinary distillation when the reflux is substantiallygreater than minimum. The driving forces for heat and mass transfer between the streams

entering a stage can be reduced by moving the operating lines closer to the equilibrium curve.

The minimum reflux condition shown in Fig 9(b) corresponds to the upper and lower operating

lines having been moved as close as possible to the equilibrium curve. It can be seen that the


14/25

number of required theoretical plates has increased, which will increase the pressure drop in the

distillation column. Therefore, the advantage in the decrease of the separation work requiredbecause of the decrease in driving forces for heat and mass transfer will somewhat will be offset

by the pressure drop in the column. (It is exactly for this reason that structured packing gives

higher efficiency, because increase of number of plates does not necessarily increase the columnpressure drop). Even at minimum reflux, there are still substantial driving forces for heat and

mass transfer at compositions in the tower removed from the feed stage in a binary distillation.

Using different operating lines in those portions of the column, where the irreversibilities withthe original operating lines were more severe, can reduce these irreversibilities. Such a situation

is shown in Fig. 9(c), where there are two operating lines applied to different parts of stripping

section and two operating lines belonging to the rectifying section [9]. The situation shown in

Fig.9(c) is shown schematically in Fig.10. This knowledge of process improvement using


15/25

intermediate condenser and intermediate reboiler has been implemented as shown in Fig.11(a)

and Fig. 11(b) [10].

In Fig. 11(a) and 11(b), the improvements in the high efficiency process cycles are obtained byproviding intermediate height reflux to all three rectifying section of the high purity oxygen

plant: the high-pressure column, the low pressure column and the argon rectifying section. An

intermediate height reflux is defined as a reflux , which has a zone of counter-current vapour-

liquid rectification both above and below the point of introduction. In Fig 11(a), the argonsidearm intermediate height reflux is provided by evaporating partially depressurized liquid


16/25

nitrogen from the high-pressure column. The argon recovery increases to 80% and pressurised

nitrogen pressure increases from 3.4 bara to 5.3 bara and the quantity increased from 5% to 11%.

The primary difference between Fig

11(b) from Fig 11(a) is that the oxygen product is evaporated from the condensing nitrogen

versus air i.e. at a lower pressure. Argon recovery is also marginally higher owing to the two

sequential refluxes from boiling kettle liquid [10]

The ultimate of reducing thermodynamic irreversibilities within a distillation column would be to

introduce reflux to all the stages above the feed and to reboil at all the stages below the feed. Inthis way, the operating line at each stage will almost be coincident with the equilibrium curve as

shown in Fig.9(d). A schematic diagram of a device for carrying out such a process is shown inFig 12 [9]. There is no doubt that a considerable capital outlay is involved in altering an ordinary

distillation to bring the process closer to reversible distillation. The number of stages required for

a given separation becomes greater and the required heat duty will be split up between the

terminal reboiler, terminal condenser and those reboilers and condensers that are necessary togenerate intermediate boil up and intermediate reflux. Offsetting this need for considerable

additional capital investment are two factors:

1) The heat energy used in the distillation is degraded to a lesser extent. Much of the


17/25

reboiled heat are added at temperatures lower than the bottom temperature and much of the heat

removal can be effected at temperatures warmer than the overhead temperature. These decrease

the Second Law loss and improve the thermodynamic efficiency of the distillation column.

2) The reduced vapor and liquid flows towards the product ends may make it possible to use

towers of different diameters at different stages. This may improve the capacity control of the

distillation column.

3.2.2 Ultra High Purity Oxygen Production

One of the recent trends of air separation plant cycle is dedicated to the production of ultra highpurity oxygen. Although ppm impurity concentration level is very common now a day, it hasbecome a thing of the past for the semiconductor industry [11]. The acceptable impurity level has

already dropped below 10 ppb (parts per billion) and likely to touch ppt (parts per trillion) level

in the near future. This calls for removal of heavier hydrocarbons in distillation column. Air

contains acetylene (0.1 to 1 ppm), ethylene (0.01 to 2 ppm), propylene (0 to 0.2 ppm) and otherheavier hydrocarbons. Although these are adsorbed on molecular sieve bed, some ethane (0.02 to

0.1 ppm), propane(0 to 0.1 ppm) and methane (2.0 to 10 ppm) flow with air to the cryogenic feed

process.

Other heavier contaminants are krypton

( 1.1ppm), xenon ( 0.08ppm) andnitrous oxide(0.01 to 0.05 ppm). The

conventional standard grade oxygen is

composed of about 99.7% O2, 0.3% Ar,10 ppm methane, 0.5ppm other

hydrocarbons, 5 ppm krypton, 0.4 ppmXenon, 0.1 ppm nitrous oxide and no

nitrogen.

Fig 13 is a schematic diagram for

conventional process for gaseous O2

production. Fig 14 shows theconcentration profile for O2 and

methane in the liquid phase in the upper

column for a conventional oxygen plant

of figure 13. It can be easily seen from

Fig.14 that methane concentrationreduces to almost nil at a few plates

above crude O2 feed location. Fig 15and Fig 16 are two suggested air

separation cycles from where ultra high

purity O2can be produced.


18/25

3.2.3 Some Recent Patents

(i) Improving Condenser-Reboiler efficiency by Removing Hydrocarbons

A recent patent by Honda and Kishida[12]describes a process (Fig. 17) where a

reduction power cost is effected by

removing hydrocarbon in a cleaning section

at a lower position in the low pressurerectification column. The development

relates to an air liquefaction separation

apparatus where a low pressure column of a

double rectification column has a cleaningsection at a lower position and a liquid

oxygen is partly withdrawn from the spaceabove the cleaning section so as to supply itto a main condenser-evaporator. The liquid

oxygen supplied to the main condenser-

evaporator is subjected to heat exchangewith a nitrogen gas separated at the head of

a high-pressure column and gasified into an

oxygen gas. This oxygen gas is introducedto the space under the cleaning section.

Hydrocarbons contained in the oxygen gas

ascending through the cleaning section are

washed down by the reflux liquid oxygendescending though the cleaning section to

provide a clean oxygen gas.

The liquid oxygen passed through the cleaning section is withdrawn from the low-pressurecolumn. Thus, since the hydrocarbons are prevented from being concentrated highly to or over

critical levels in the liquid oxygen in the main condenser-evaporator, submergence in the main

condenser-evaporator can be reduced to minimize the influence of the depth of the liquid, thusimproving the efficiency of the condenser-reboiler. This will lower the pressure of the nitrogen

gas and reduce the operating cost by reducing power of the compressor for compressing the feed

air.


19/25

(ii) Variable Production Rate of LOX And LIN from

the Same Plant

A very useful patent has been taken by R. Agarwal

et al.[13] where variable rate of production of liquid O2andliquid N2can be obtained from a single plant. This is likely

to be a useful concept in case of fluctuating demand at a

particular location (Fig. 18). The air separation plant uses a

liquefier and a two stage distillation column capable ofoperating in two modes, namely a first mode of operation

during which only liquid nitrogen is produced and a second

mode of operation during which liquid nitrogen and liquid

oxygen are produced. By adjusting the duration ofoperation in each mode, any ratio of liquid nitrogen to

liquid oxygen greater than the ratio achieved during the

second mode of operation can be achieved. In the firstmode of operation, a condenser is used to condense the

lower-pressure-stage gaseous nitrogen into lower-pressure-

stage nitrogen condensate. To condense the lower-pressure-stage nitrogen, either at least a portion of the crude oxygen

liquid from the higher-pressure-stage, a least a portion of

the oxygen-enriched liquid from the lower-pressure-stage,

at least a portion of the liquefied air, or mixture thereof, areintroduced to the condenser. In the second mode of

operation, the top condenser is not used. Instead, all of the

crude oxygen liquid is introduced into the lower pressure

stage, which produces a bottom liquid oxygen stream and alow pressure overhead waste stream containing nitrogen.The system includes fluid flow lines and valves for

directing the flow of certain fluids, particularly the crude

oxygen liquid and the oxygen enriched liquid, during the

two modes of operation.

iii. Decreasing Separation WorkA double column cryogenic rectification system for producing lower purity oxygen wherein aminor portion of the feed air is successively condensed in two vertically oriented stages within

the lower pressure column before undergoing rectification. This concept is primarily aimed at

reducing the work of separation (Fig. 19)[14].


20/25


21/25


22/25


23/25


24/25

4. IMPROVEMENT IN RELATED TECHNOLOGICAL AREAS

No industry can be insulated from the technical changes which take place in other related areas

and air separation industry is no exception. It has gained from the technical advancement ofmaterials, production technology, transportation and, more importantly, from the advent of high-

speed computers and sophisticated instrumentation.

One of the reasons of reduction of energy requirement in an air separation plant is in increase of

its sizes from a few hundred Nm3 capacity during WWI, to a few tens of tons per day during

WWII, to a few hundred tons per day during 1960, to a few thousand tons per day now. This

increase of sizes of ASUs have been made possible by making large sizes of distillation columns,

liquid pumps, apart from construction of large heat exchangers and compressors, whichdepended on the development of these industries and improvement of production processes.

Large size also leads to reduction of losses, particularly insulation losses per unit product. The

advancement of instrumentation in process industry has seen the changes from mechanicalinstrumentation, to pneumatic instrumentation, to Programmable Logic Controls (PLC) andfinally to Distributed Control System (DCS). Accurate feed-forward and feed-back controls have

now been possible with the availability of on-line data. Steady state and dynamic simulation

programs, with the advancement of computers, now produce results that are far more accurate.Computer networking has now made it possible for someone to monitor and control the

operation from a far-off place. The instruments, for the purpose of measurement, control and

analysis have become more reliable with great improvement of accuracy, repeatability,sensitivity and response time. All these have made a large plant to be safe enough to be built and

operate. The technology of building large-sized (5000 Tons) storage tanker and reliable transport

tankers have made the large plants a commercial viability. The improvement in superinsulation

has resulted in reduction of losses to a great extent. All these above advancement of technologieshave contributed positively to the advancement of air separation technology.

1. CONCLUSIONSUnless there is a major conceptual breakthrough, the trend in the improvement of the

efficiency of cryogenic air separation is likely to follow the same trend in the foreseeable

future. As we have seen in this paper, the process improvement means a lower energyconsumption with the associated penalty of higher capital investment. The new processes

envisage structured packing, more column height and more number of heat exchangers

for the plant. It is also important to note that the optimum level of capital outlay will

depend on cost of energy prevailing at any instant of time.

2. ACKNOWLEDGEMENTThe author is grateful to A. R. Singh and Pankaj Gupta, Consultant and Secretary of All IndiaIndustrial Gases Manufacturers' Association, New Delhi respectively, for suggesting the subject

of the present paper.

7. REFERENCES

1. R. E. Latimer, Distillation of Air, Chemical Engineering Progress, Vol. 63, no. 2,February, 1967.


25/25

2. T.D.Atkinson and T.Rathbone, Advances in Cryogenic Air Separation, in Separation ofGases, Proceedings of the fifth BOC Priestley Conference, Page 35-52, 1989.

3. M.A.Taylor, Plate-Fin Heat Exchangers: Guide to their Specification and Use, AmendedOctober 1990, Published by HTFS, Harwell Laboratory, UK.

4. R.G.Scurlock, A Brief History of Cryogenics, Advances in Cryogenic Engineering, Vol.37, part A, 1992.

5. E. Schoenpflug, W.Deinart and G.Rueckborn, Recent Advances in Air SeparationTechnology, Cryogenics, Vol. 30, Page 17-22, September Supplement, 1990.

6. L.M.Rose, Distillation Design in Practice, Elsevier, Chapter 7, Column Fundamentals,page 137-167

7. M.J.Lockett, Gas Separation by Distillation, in Separation of Gases, Proceedings of thefifth BOC Priestley Conference, Page 19-34, 1989.8. K.D.Timmerhaus and T.M.Flynn, Cryogenic Process Engineering, Plenum Press, 1989.9. C. J. King, Separation Processes, McGrawHill Book Company, 197110.R. Agarwal, D.C.Erickson and D.W.Woodward, High efficiency Processes for Cryogenic

Air Separation, in Process and Equipment 1989, The Seventh Intersociety Cryogenic

Symposium.

11.R. Agarwal, Production of Ultra-high Purity Oxygen: A Distillation Method for the Co-production of the Heavy Key Component Stream Free of Heavier Impurities, I and EC,1994.

12.H. Honda and Y.Kishida, Air Liquefaction Separation Process and Apparatus Therefor,Patent no. 763301 (1996)

13.R. Agarwal et al., Method and Apparatus for Producing Liquid Products from Air inVarious Proportions, Patent no.660311 (1996)

14. D.P.Bonaquist, Cryogenic Rectification System with Staged Feed Air Condensation,Patent no.617591 (1996)

This page has been developed by Mr. Uttam Bhunia & Prof. Kanchan Chowdhury 1999, Cryogenic

Engineering Centre, Indian Institute of Technology, Kharagpur 721302, India. Suggestions may please be

forwarded to: [email protected]

Documents

Advances in Cryogenic Air Separation