Higman Chris Gasification (Газификация угля)

Подождите немного. Документ загружается.

310

Gasification

Molecular Sieves

The most common application of molecular sieves in connection with gasification

plants is the removal of water and CO

upstream of cryogenic units. Processes work-

ing at cryogenic temperatures, such as air separation or cryogenic gas separation,

require a feed gas completely free of these components, which would otherwise

freeze and deposit on the inlet heat exchangers and finally block them.

The classic cycle described above is usually employed. In air separation duty,

water and CO

are not the only considerations. The prepurification unit also prevents

the ingress of hydrocarbons into the cold box as a safety measure. Recently the

ingress of NO

into the cold box has also become an issue of concern. For air sepa-

ration. a combination of molecular sieve and silica gel is often used.

Pressure Swing Adsorption

Pressure swing adsorption (PSA) operates on an isothermal cycle, adsorbing at

high pressure and desorbing at low pressure. The principle application of PSA is for

hydrogen purification, although there are a number of others including air separation

(see Section 8-1).

The optimum pressure for hydrogen purification lies in the range 15–30 bar. At

higher pressures the hydrogen yield falls off, a point to be considered when integrating

a hydrogen off-take from a gasification plant optimized for a different application.

The hydrogen yield of a modern PSA unit usually lies between 80% and 92%.

Apart from the matter of pressure already mentioned, other influences are the quality

COMPONENT PARTIAL

PRESSURE, [bar]

LOADING,

[mol COMPONENT/mol SORBENT]

< T

Figure 8-10. Adsorption Loading at Different Temperatures and Pressures

Auxiliary Technologies

311

of the feed gas (the higher the quantity of impurities to be removed, the more hydrogen

is lost with them) and the tail gas pressure. Where the tail gas is burned in dedicated

burners, as for instance in a steam reformer hydrogen plant, the typical tail gas pres-

sure is 0.3–0.4 bar gauge. Where the tail gas pressure is higher (e.g., 3–5 bar gauge),

the drop in hydrogen yield can become very significant.

Additionally, the hydrogen purity can affect the yield, though only to a small

degree. Typical purities range from 99 to 99.999 mol%. An additional common

hydrogen specification is a limit on the amounts of carbon oxides (CO and CO

Levels of 0.1 to 10 ppmv are easily achieved. In the design of an overall gasification-

to-hydrogen system, it is useful to have an idea about the performance of likely

impurities in the PSA unit. A comparison of a number of components is shown in

Table 8-3. In this connection it is important to note that although water is strongly

adsorbed and so will not contaminate the product. It is disadvantageous to have large

quantities in the feed gas since this requires excessively large beds. Usually, cooling to

below 40°C with subsequent condensate separation is sufficient to provide an economic

design.

A further design consideration is the number of adsorber vessels. Early plants used

four beds, as is still the practice on smaller plants. Larger modern plants use as many

as twelve adsorbers. Sophisticated cycles have been developed to minimize the loss of

hydrogen on depressurization from the adsorption step to the desorption step, by using

this hydrogen to repressurize a bed that has just completed its desorption step. Thus

there can be a trade-off between a higher investment for an increased number of vessels

(and valves) and operating savings from an increased hydrogen yield.

Zinc Oxide/Copper Oxide

Adsorption of H

S onto zinc oxide is an effective method for removing trace quanti-

ties of sulfur from gas to achieve a purity of less than 0.1 ppmv, as is required by

copper or nickel catalysts. It is therefore the standard method of desulfurization

Table 8-3

Relative Strength of Adsorption of Typical Impurities

Non-adsorbed Light Intermediate Heavy

CO C

He N

Ar C

Source: Miller and Stoecker 1989

312

Gasification

upstream of natural gas steam reformers. The adsorption takes place via the reaction

of hydrogen sulfide with zinc oxide to form zinc sulfide. In situ regeneration is not

possible, and this places a limitation on the amount of sulfur that the process can

accept in the inlet gas.

There are two generally accepted designs for zinc oxide desulfurization units. In a

guard bed function or where the sulfur load is low, a single bed is provided, sized to

adsorb the total quantity of sulfur to be expected between planned turnarounds, say

one or two years. Where the sulfur load is higher and a single bed would become

unmanageably large, a two-vessel series arrangement is provided and provision is

made for exchanging the adsorbent online. With this arrangement, the individual

bed can be sized smaller, such as for a six-month interval between bed replacement.

Zinc oxide can adsorb sulfur present as H

S almost completely. Performance with

other sulfur compounds (COS, mercaptans) is not as good. In cases where sulfur is

present other than as H

S, it is necessary to hydrogenate these components to H

upstream of the zinc-oxide bed. This is normally done over a cobalt-molybdenum

(CoMox) or nickel-molybdenum (NiMox) catalyst.

Zinc oxide adsorption is essentially a process for polishing or guard bed duty.

This becomes clear when considering a zinc-oxide bed for the carbon monoxide

plant described in Section 7.1.4. Operating in its optimum temperature range of 350

to 400°C, zinc oxide has a pick-up capacity of around 20% by weight. Assuming a

sulfur content of 100 ppmv in the natural gas, the total sulfur intake is about 10 tons/

year, requiring replacement of about 50 tons/year zinc oxide. Compare this with the

nearly 30 t/d sulfur intake of the 1000 t/d methanol plant of Section 7.1.2, and the

limitations become very apparent.

Given these numbers, zinc oxide in the gasification environment is limited either

to guard bed duty, for example, upstream of a low temperature shift or methanator

catalyst or to natural gas feeds. As discussed in Section 7.1.4, there are arguments

for desulfurizing either upstream or downstream of the partial oxidation reactor.

Where extreme sulfur cleanliness is required, copper oxide can be used for final

desulfurization down to 10 ppbv. Commercial adsorbents are available for this pur-

pose, either in a mixed ZnO/CuO formulation or as a separate polishing bed.

8.2.4 Membrane Systems

Permeable gas separation membranes in syngas service utilize differences in solu-

bility and diffusion of different gases in polymer membranes. The rate of transport

of a component through the membrane is approximately proportional to the differ-

ence in partial pressure of the component on the two sides of the membrane. Polymer

membranes have found increasing use in a number of applications, including natural

gas processing (CO

removal) and in the synthesis gas environment for hydrogen

separation out of the main syngas stream.

The design of a polymer membrane system exploits the different permeability

rates of the components in the feed gas. An idea of the relative rates through a typical

Auxiliary Technologies

313

hydrogen separation polymer can be gained from Table 8-4. Thus a good separation

can be achieved between, for example, hydrogen and CO or N

. Separation from

will be only moderately satisfactory, however.

Membrane units are usually supplied packed, typically as a bundle of hollow

tube fibers. The feed is supplied to the shell side of the bundle and the permeate

(hydrogen rich stream), which passes through the fiber-tube walls, is collected on

the tube side. Design variables are the pressure difference selected and the total

surface area of the polymer.

For the system designer, the integration of a membrane unit has two important

characteristics. First, permeable membranes provide the only system leaving the

carbon monoxide at essentially the same pressure level as at the gas inlet (less

hydraulic losses only) and the hydrogen on the low-pressure side. This is exactly the

reverse of the pressure swing adsorber.

Second, as mentioned above, it must be recognized that since all permeable

membranes work on the basis of different rates of diffusion, they can only have a

limited selectivity. This can be disadvantageous, since in a hydrogen extraction

application, the product hydrogen is not very pure, and the diffusion of CO through

the membrane can be considered as a loss of high pressure gas.

Nonetheless, skilled integration of membrane and PSA technologies can together

provide some extremely attractive solutions. Consider the following situation where

20,000 Nm

/h pure hydrogen is required from a main stream of syngas in an IGCC

(Figure 8-11 and Table 8-5). The membrane is used to produce a raw hydrogen at

reduced pressure (but still adequate for PSA feed) with only a small loss of other

syngas components for the gas turbine. The raw hydrogen has a purity of about 70–

90 mol%, depending on syngas composition and pressure, which allows the PSA to

have a significantly higher efficiency than would be the case with syngas feed. Fur-

thermore, the much smaller quantity of tail gas to be adsorbed allows the PSA unit

to be smaller too.

Care should be exercised with liquid carry over from an upstream AGR system.

In some cases these can damage the membrane. Proper separation at the AGR outlet

should however be sufficient to prevent problems (Collodi 2001).

Table 8-4

Relative Permeability Rates of Typical Syngas Components

Quick Intermediate Slow

He CH

S N

Source: Kubek, Polla, and Wilcher 1997

314

Gasification

Hot Gas Cleanup

For power applications the energy loss involved in cooling synthesis gas down to

ambient or lower temperatures as required by current acid gas removal systems is

reason enough for the interest in so-called “hot gas cleanup.” Actually, hot gas

cleanup is a misnomer, and these technological developments should rightly be

called “warm gas cleanup,” since the target operating temperature range is between

250 and 500°C.

Impurities that need to be considered in a warm gas cleanup system include

particulates (fly ash and char) as well as gaseous compounds such as H

S, COS,

, HCN, HCl, and alkali species. At temperatures above about 500°C, alkaline

species will pass through a particulate filter, and this together with materials issues

is the principle reason why no attempts at hotter cleanup have been made.

MEMBRANE PSA

SYNGAS

DEPLETED

SYNGAS

RAW H

PURE

HYDROGEN

TAIL GAS

Figure 8-11. Membrane and PSA Combination

Table 8-5

Mass Balance for Membrane/PSA Combination

Syngas In Syngas Out Raw H

Pure H

Tail Gas

mol% mol% mol% mol% mol%

8.6 8.4 9.4 31.4

CO 43.8 52.4 7.9 26.3

45.3 36.5 82.4 100.0 41.2

2.3 2.7 0.3 1.1

Total (kmol/h) 6635 5360 1275 893 382

Pressure (bar) 50 49 25 24 1.3

Auxiliary Technologies

315

Technologies for warm gas cleanup using zinc-based sorbents have been built at

demonstration scale in Polk County and Piñon Pine without great success (Simbeck,

2002; U.S. Department of Energy 2002). In fact, neither of these units was ever oper-

ated. Both of these were designed essentially as desulfurization units with removal

efficiencies of up to 98%, which at the time of design conformed to existing power

station emission regulations. They did not address some of the other species, such as

nitrogen compounds and halides, nor for that matter mercury. Furthermore, the sulfur

removal efficiencies made them unsuitable for most chemical applications.

Nonetheless, the potential in terms of efficiency improvement remains and continues

to provide an incentive for research and development to find appropriate systems.

8.2.5 Further Developments

One current program is addressing some of these issues. It includes the use of

membrane technology for bulk desulfurization, zinc oxide or similar chemisorption

process for fine sulfur removal, a sodium carbonate-based sorbent for HCl removal,

and a high-temperature molecular sieve for ammonia removal (Gupta 2001). How-

ever, the use of polymer membrane technology is likely to limit its ability to operate

at high temperature and thus any efficiency gains on this basis. Mercury capture is

not specifically mentioned as being part of this program, although on the basis of

currently published flowsheets it could possibly be incorporated as a separate stage.

8.2.6 Biomass Syngas Treating

Treating the syngas generated from biomass has special problems—particularly those

associated with the presence of tars in the gas. Attempts have been made to reduce the tar

content by cracking (Morris and Waldheim 2002; Bajohr etal. 2002). Other attempts have

been made at using an oil wash (Boerrigter, den Uil, and Calis 2002; Hofbauer 2002). To

date, success has been limited to achieving a quality suitable for power applications.

Considerable work is still required to achieve a chemicals application syngas quality.

8.3 CO SHIFT

Besides having an important influence on the composition of the raw syngas from

the gasifier itself, the CO shift reaction

CO + H

OCO

−41 MJ/kmol (2-7)

can be and is operated as an additional and separate process from the gasifier at

much lower temperatures in order to modify the H

/CO ratio of the syngas or maxim-

ize the total hydrogen production from the unit. As can be seen from the reaction

2-7, one mole of hydrogen can be produced from every mole of CO. The reaction

←

→

316

Gasification

itself is equimolar and is therefore largely independent of pressure. The equilibrium

for hydrogen production is favored by low temperature.

The CO shift reaction will operate with a variety of catalysts between 200°C to

500°C. The types of catalyst are distinguished by their temperature range of operation

and the quality (sulfur content) of the syngas to be treated.

8.3.1 Clean Gas Shift

High Temperature (HT) Shift

Conventional (high temperature) shift uses an iron oxide–based catalyst promoted

typically with chromium and more recently with copper. The operating range of

these catalysts is between 300 and 500°C. Much above 500°C sintering of the

catalyst sets in and it is deactivated. HT shift catalyst is tolerant of sulfur up to a

practical limit of about 100 ppmv, but it is likely to loose mechanical strength, par-

ticularly if subjected to changing amounts of sulfur.

An important aspect in the design of CO shift in the gasification environment, where

inlet CO contents of 45% (petroleum residue fed) to 65% (coal) are common, is the

handling of the heat of the reaction, particularly under end-of-run conditions where an

inlet temperature of 350°C or more may be necessary. On the one hand, the reaction

must be performed in several stages to avoid excessive catalyst temperatures and to have

an advantageous equilibrium. On the other hand, optimum use must be made of the heat.

One such arrangement is shown in Figure 8-12. Desulfurized syngas containing

about 45 mol% CO, which leaves the AGR at about 54 bar and ambient temperature,

is heated and water saturated at a temperature of about 215°C by water that has been

preheated with hot reactor effluent gas. The saturated gas is further preheated to the

catalyst inlet temperature of between 300°C and 360°C. The steam loading from

the saturator is such that only the stoichiometric steam demand for the reaction is

required to be added from external sources. In the first stage, the CO is reduced to a

DESULFURIZED

GAS

SATURATOR REACTOR I

PROCESS

STEAM

REACTOR II

SHIFT GAS

DESATURATOR

HEAT

RECOVERY

Figure 8-12. CO Shift with Saturator-Desaturator Circuit (Source: Higman 1994)

Auxiliary Technologies

317

level of about 7–8 mol% at an outlet temperature of about 500°C. The outlet gas is

cooled to a temperature of about 380°C in the gas and water preheaters before entering

the second catalyst bed. Here the residual CO is reduced to about 3.2 mol%. The gas

is then cooled in a direct-contact desaturator tower. There are a number of different

designs, particularly for the first reactor, that incorporate the gas-gas heat exchanger

as an internal. In such reactors, the exchanger is arranged centrally inside an annular

catalyst bed with an axial (Lurgi) or axial-radial (Casale) gas-flow pattern. Alternative

methods of controlling the catalyst outlet temperature include interbed condensate

injection (e.g., Toyo). The use of an isothermal steam raising reactor has been pro-

posed, and although such a solution has been employed in a steam reformer plant,

none is recorded at the high CO inlet concentrations involved in a gasification plant.

Typical catalyst lifetime for the first bed in a gasification situation is two to three

years, which is considerably shorter than for a steam-reforming situation. This is gen-

erally attributed to the high operating temperatures associated with high CO concen-

trations in the inlet gas. On a moles-converted basis over the lifetime of the catalyst,

the performance in the gasification context is comparable with that of steam reforming.

Low Temperature (LT) Shift

Low temperature shift operates in the temperature range 200°C to 270°C and uses

a copper-zinc-aluminum catalyst. It is used in most steam reforming-based ammonia

plants to reduce residual CO to about 0.3 mol%, a requirement for a downstream

methanator, but has generally not been applied in gasification-based units. On the one

hand, it is highly sulfur-sensitive, and even with 0.1 ppmv H

S in the inlet gas, will

over time become poisoned. A second reason for its lack of use particularly in oil-

gasification plants, is the effect of the higher pressure on the water dewpoint in the

gas. Operation near the dewpoint will cause capillary condensation and consequent

damage to the catalyst. With a dewpoint of about 215°C and a temperature rise of

25–30°C, there is not much margin for error below the upper temperature limit of 270°C

when recrystallization of the copper catalyst begins. The first application of low

temperature shift at high pressure was in Shell’s Pernis gasification facility, which

has now performed successfully for several years (de Graaf et al. 2000).

Medium Temperature (MT) Shift

An improved copper-zinc-aluminum catalyst able to operate at higher exit

temperatures (300°C) than conventional LT shift has been developed, particularly

for use in isothermal reactors. No application in gasification plants is known.

8.3.2 Raw Gas Shift

For applications where it is desired to perform CO shift on raw syngas, a cobalt-

molybdenum catalyst, variously described as a “sour shift” or “dirty shift” catalyst,

318

Gasification

can be used. In some parts of the literature this catalyst is described as sulfur tolerant.

This is actually a misnomer, since the catalyst requires sulfur in the feed gas to

maintain it in the active sulfided state. It is generally applied after a water quench of

the raw syngas, which typically will provide a gas at about 250°C saturated with

sufficient water to conduct the shift reaction without any further steam addition.

For an ammonia application the raw gas shift is typically configured as two or three

adiabatic beds with intermediate cooling resulting in a residual CO of about 1.6 or

0.8 mol%, respectively.

An important side-effect of the raw gas shift catalyst is its ability to handle a

number of other impurities characteristic of gasification. COS and other organic

sulfur compounds are largely converted to H

S, which eases the task of the down-

stream AGR. HCN and any unsaturated hydrocarbons are hydrogenated.

Carbonyls are decomposed and deposited as sulfides, which increases the pres-

sure drop over the bed. Selective removal of arsenic in the feed is also claimed

(BASF undated).

8.4 SULFUR RECOVERY

The sulfur compounds from the feedstock of a gasification-based process are

generally removed from the synthesis gas as a concentrated stream of hydrogen

sulfide and carbon dioxide known as acid gas. Depending on the design of the

upstream AGR unit, the acid gas may contain other sulfur species such as COS as

well as ammonia and hydrogen cyanide. It is unacceptable to emit H

S, a highly

toxic, foul-smelling gas, to the atmosphere, so it is necessary to fix it in one form or

other. There are essentially two alternative products in which one can fix the sulfur,

either as liquid or solid elemental sulfur, or as sulfuric acid. The choice of product

will depend on the local market. Where there is a strong local phosphate industry,

then there will be a good local market for sulfuric acid. If this is not the case, then

elemental sulfur will probably be the better choice, since bulk transport of this mater-

ial is much easier than of the concentrated acid.

8.4.1 The Claus Process

The basic Claus process for substoichiometric combustion of H

S to elemental

sulfur was developed as a single-stage process on the basis of reaction 8-5 at the end

of the nineteenth century. During the 1930s it was modified into a two-stage process

in which initially one third of the H

S was combusted to SO

and water, and in a

second low-temperature catalytic stage, the SO

was reacted with the remaining

S to sulfur. Operating the second stage at a comparatively low temperature

(200–300°C) used the more favorable equilibrium to achieve much higher sulfur

yields than had been possible with the original process.

Auxiliary Technologies

319

S + 1½ O

←→ SO

O (8-3)

2 H

S + SO

←→ 2 H

O + 3/8 S

(8-4)

3 H

S + 1½ O

←→ 3 H

O +3/8 S

(8-5)

Today there are innumerable Claus processes available, all of them ultimately vari-

ants of the modified Claus process. A typical standard Claus process is shown in

Figure 8-13. In the first combustion stage all the H

S is combusted with an amount

of air corresponding to the stoichiometry of reaction 8-5 at a temperature in the

range 1000–1200°C. The thermodynamics of these three main reactions is such that

about half the total sulfur is present in the outlet gas as elemental sulfur vapor, the

rest as an equal mix of H

S and SO

. The hot gas is cooled by raising steam and the

sulfur already formed is condensed out. The removal of sulfur at this point assists in

driving reaction 8-4 further to the right in the subsequent catalytic stage. The gas is

reheated and passed over an alumina catalyst at a temperature of about 200–300°C,

and cooled again to condense the sulfur formed. This may be performed a number of

times to remove further amounts of sulfur. Typically, two (as shown in Figure 8-13)

or three catalytic stages are used.

Oxygen Claus Processes

A standard air-blown Claus plant is limited in the dilution of H

S possible in the acid

gas. At concentrations less than about 25–30 mol% H

S, the temperature in the

COOLER I

COOLER II

PRE-

HEATER

ACID GAS

CLAUS

REACTOR I

CLAUS

REACTOR II

WHB

CLAUS

FURNACE

AIR

SWS GAS

MP STEAM

LP STEAM

BFW

TAIL GAS

PRODUCT SULFUR

SULFUR TANK

Figure 8-13. Typical Two-Stage Claus Unit (Source: Weiss 1997)