THE COLLIERY GUARDIAN

AND

JOURNAL OF THE COAL AND IRON TRADES.

Vol. CXIII.

FRIDAY, JUNE 15, 1917.

No. 2946.

Recovery of Benzol and its Homologues from Coal Gas.

By J. E. CHRISTOPHER.

The term ‘1 benzol ’ ’ is used in a rather indiscrimi-
nate manner. Certain definite chemical compounds,
each with decided characteristics,* enter into its com-
position, but in general we may look upon benzol as a
mixture of compounds of the aromatic series of the
general formula,- CMH2W_6, associated with certain
impurities. The compounds of importance in benzol
are: —

Boiling
point.

Main products:—	Degs.	C.

Benzene, CflH6 ................ 80*5

Toluene, C7H8 ................ 110*0

Ortho-xylene, <\H10 ...... .... 142‘0

Meta-xylene, C8hT10............ 139 0

Para-xylene, G8H 0............ 138*0

Propyl-benzene, C9H12 ........ 160*0

Impurities:—

Carbon disulphide, CSa.
Thiophene, C4H4S.
Paraffins, C%H2w + 2-
Phenols, CwH2w - 7 HO.
Pyridine, C5H5N.
Naphthalene, C10H8.
Cumarone.

Creosote oil, etc.

The general scheme of benzol recovery may be .
divided into the following headings : —

(a)	Extraction of a complex mixture of aromatic
hydrocarbons and associated impurities by treating the
gas with solvents of higher boiling point.

(b)	Expulsion of this mixture from the solvent by
regulated heat treatment.

(c)	Preliminary fractionation into groups of com-
pounds according to the respective boiling' points.

(d)	Chemical treatment of the various fractions.

(e)	Re-distillation and careful fractionation of the
chemically treated products.

Extraction.

With reference to the first heading, the crude pro-
ducts could be extracted by cooling the gas to a very

low temperature whilst subjecting it to a high pres-

D

G

T
♦
i

o o o o
o o o
o o o o
o o o o
o o o o
O O O O*

Fig. 1.—Diagram of Debenzolising Plant.

sure, but owing to the relatively small proportion of
benzol vapours in the gas, and consequently the large
volume of gas to be thus treated, this process is too
costly to work under ordinary circumstances, and the
method usually adopted is to scrub the gas with an
absorbent. The apparatus used may be of the tower
type, rotary type, centrifugal type, or intensive spray
type. For absorption purposes, blast furnace creosote
may be used, but whilst it has an advantage in being
free from naphthalene, there is a risk of the benzol
produced being contaminated with paraffin compounds
—an important point in benzol required for nitrating
purposes and aniline manufacture.

The great majority of plants use some form of coal
tar oil as a scrubbing agent, creosote being generally
preferred. The quality of the oil used is of the highest
importance, and periodical tests should be made to
ensure a requisite standard being maintained. A good
wash oil should be fluid, free from water, light oils, and
naphthalene as far as possible, and of satisfactory dis-
tillation test, and specific gravity. A good: specifica-
tion for a suitable oil would be: Specific gravity, 1-03
to 1-05; retort test with bulb of thermometer in the
liquid, no distillate below 200 degs. Cent., 70 to 90 per
cent, between 200 and 300 degs. Cent. The distillate
up to 300 degs., on cooling to 10 degs. Cent., should
not show more than 7 per cent, of solids.

The writer would prefer a wash oil with a distillate
up to 300 degs. Cent., approaching the lower limit
given, as this allows a certain proportion of anthra-
cene or “green” oil to be introduced. This tends to
bring about a cleaner separation of the crude benzol
at a later stage, whilst the viscosity of the oil, though

thereby increased, does not seriously affect the work-
ing of the plant. As to the amount of wash oil circu-
lated, figures of 50 to 120 gals, of wash oil per ton of
coal have been given, the writer’s experience being in
favour of 90 gals, of oil circulated per ton of coal, the
benzolised oil containing about 3 per cent, of crude
benzol.

A portion of the wash oil should be removed from
circulation from time to time and replaced by fresh
oil as the oil tends to thicken, with a decided deteriora-
tion in absorbing power. The wash oil is usually circu-
lated through two or three scrubbers in series, and in
the opposite direction to the flow of gas, the final gas
being treated with the debenzolised oil, which thus
absorbs the benzol in stages, and after circulating
through the system passes to the distillation plant as
benzolised oil.

Heat Treatment.

The expulsion of benzol from the enriched oil
depends mainly on the application of heat. Thus, it
could be readily driven off in a pot still by an open fire,
and, in fact, a tar dehydrating plant has been success-
fully used for this purpose. In dealing with the com-
paratively large volume of benzolised oil obtained at an
average coke plant, a continuous process of distilla-
tion is preferred, enclosed and open steam being the
heating agent in this case. The hot debenzolised oil,
after satisfactory cooling, is re-used in the scrubber,
thus completing the circuit. The hot benzol vapours,
excess of steam, and debenzolised oil carry away con-
siderable heat from the still proper, and a modern
plant aims at the recovery of as much of this “ spent ”
heat as possible, transferring it by means of pre-
heaters to the incoming “saturated” oil.

The principle of one type of modern debenzolising
and absorption plant is shown in diagram form in
fig. 1. Wash oil is pumped from the scrubbers to

G

9

M

water

STEAM

ACID ere

T
t

Fig. 2.—Scheme of Benzol Rectification Plant.

elevated store tanks, from whence it flows by gravity
through the debenzolising system. In this type, the
oil is pre-heated in three stages: first, by means of hot
debenzolised oil in the heater A; secondly, by hot
benzol vapours and excess steam in B; and, thirdly,
by enclosed steam coils in the heater C. The oil,
heated to about 110 degs. Cent., passes into the main
still D, constructed on the principle of an ammonia
still, and passing over successive trays, is agitated
thoroughly by steam admitted in the lower part, and
the benzol is expelled. This agitation with open steam
is necessary if we wish to strip the oil satisfactorily,
but at the same time, the tendency for the more vola-
tile products to carry forward a portion of the heavier
oil is increased, and, in addition to the lighter pro-
ducts, an appreciable proportion of creosote oil and
naphthalene pass forward. Accordingly, a fraction-
ating column E is attached, to restrict the passage of
these mixed vapours, and, by atmospheric cooling, to
condense selectively some of the higher boiling pro-
ducts. The ascending vapours are made to bubble
through the descending condensate, and the separation
of the products is rendered more complete. The
vapours leaving this column pass through the condenser
heater B, and are finally condensed in the water-cooled
condenser F. The mixture of water and crude benzol
passes to the dividing box G, in which the water is
siphoned off from the bottom, whilst the crude benzol
is run off from the top to the store tank. The hot
debenzolised oil from D passes through the pre-heater
A, and is finally cooled by a water-cooled system of
piping at H, passing thence to the absorption system.
The above diagram is given to show the general prin-

ciple, but the types of crude benzol plants in vogue
vary considerably with regard to the relative disposi-
tion of the several units. Thus, in some cases, instead

of an atmospherically cooled column E, a multi-tubular
cooler is superimposed, cold benzolised oil flowing round
the outside of the tubes, and regulating the tempera-
ture of the vapours passing through. In some cases
the incoming oil is pre-heated by benzol vapours only,
whilst sometimes a separate steam heater is dispensed
with; in which event a somewhat larger main still is
used, with enclosed steam coils in some or all of the
tray sections. Again, the pre-heater, condenser, and
separator may be superimposed.

It is important in the working of a crude benzol plant
that ammonia should be satisfactorily removed before
the entry of the gas into the benzol scrubbers. If not
completely removed, the ammonia compounds are
absorbed by any water in the wash oil, and are decom-
posed in the steam pre-heater, the liberated acids
attacking the heating coils — more especially if the
latter are of wrought iron or steel. The writer has
known cases in which cast iron coils have been
destroyed in a few weeks, a solid deposit in the heater
indicating a corrosion due mainly to the formation of
iron sulphide. The evil may be mitigated somewhat
by keeping the water content of the oil at its lowest
limit, a final heating of the oil by a closed steam coil
in close proximity to the open steam jets being some-
times advantageous in this respect. In all cases, heat-
ing coils should be arranged to permit of easy access
for examination or removal. A perforated heating
coil gives rise to a sudden increase in the amount of
open steam, with a consequent priming of the still or
a sudden reduction in the distillation test—indicating
a higher proportion of heavier products carried over
mechanically. If not remedied immediately, there is
a risk of water passing into the oil, and the corrosion
would be intensified. At some works crude benzol
only is produced, and, to avoid carriage by rail of an
undue proportion of washed oil in the benzol, a distil-
lation test is usually stipulated, a very common specifi-
cation requiring a distillate of 65 per cent, when
heated in a retort to 120 degs. Cent. If the crude pro-



duct is to be worked up at the coke plant, this test is
not necessary, and it is usually an advantage to work
the crude benzol plant to give a benzol of lower distil-
lation test, often as low as 50 per cent, at 120 degs.
Cent. The after-treatment of the crude benzol is
dependent on the class of products sought. In a few
cases it is split up into products, each of a certain
range of distillation, without chemical treatment.
More commonly it is rectified by distillation and
chemical treatment, and split up into products, in
which benzene, toluene, and xylene predominate
respectively. In other cases, these semi-pure products
are further chemically treated and fractionally dis-
tilled to form products more or less approximating to
the definite chemical compounds, benzene C6H6,
toluene C7H8, xylene C8H10, etc. The majority of
benzol plants aim at turning out products of the follow-
ing designation and distillation test: —

90% benzol ......90%	distilling	over	between	80°	and	100°	C.

90% toluol ......90%	„	„	„	100°	„	120°	C.

Solvent naphtha. 90%	„	„	,,	135°	,,	160°	C.

Heavy naphtha... 90%	„	,,	,,	160°	„	190°	0.

The above may be “ washed ” or “ unwashed ” pro-
ducts, but usually the former. Before the war, the
following specification was also in vogue : —

50—90% benzol ....50% at 100° C., 90% at 120° C.

And occasionally pure products were sought, more or
less approximating to : —

Purebenzene... 95% } distilling C0*8°C.(nowwithin0*5°C.)
Pure toluene..95% [-overwithin■< 0*8°C.( ,,	„	0*5°C.)

Pure xylene ... 95% j a range of (.3 to 4*5° C.

The above tests are made on a retort with the ther-