Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 21

Planetary coolers, sometimes called satellite coolers can be grouped with

the more general class of integral coolers, the common feature of which is
that they are a rigidly-attached part of the structure of the rotary kiln, and
rotate with it.

Up to 1920, the rotary coolers pioneered by Hurry and Seaman were


standard on almost all rotary kiln installations, but they had a number of
disadvantages that were by then well-known:

 Because the clinker (at 1200°C) must fall from the kiln outlet into the
cooler, the kiln and its associated structures must be built high in the
air, on massive and expensive piers.
 In order to recuperate the heat from the clinker and at the same time
raise the temperature of the combustion air, the kiln must provide
enough suction to draw air through the cooler, and this necessitates
effective seals both between the kiln and the stationary kiln hood and
between the kiln hood and the cooler. In the hot, abrasive conditions
encountered, these moving seals are difficult to keep even
approximately leak-proof.
 The cooler represents a separate item of plant, with its own drive and
rollers requiring maintenance, but without which the kiln must stop,
and so the potential for mechanical failure of the kiln system is
increased.

The possibility of a cooler rigidly connected to the kiln evidently provides a


possible solution to these problems.

In-line cooling zones

An early, mechanically inexpensive approach to tackling the problem was


simply to extend the kiln tube and use this extension for cooling. This would,
of course, involve extending the kiln firing pipe so that fuel was injected
upstream of this cooling zone. This involved a more or less cumbersome
cantilever arrangement to support the pipe. Some cooling arrangement was
also needed, to prevent fuel from catching fire or coking before it reached the
end of the pipe. Firing pipes even of ordinary length are subject to
considerable danger of damage due to the intense radiant heat in the kiln
and the tendency of chunks of loose coating to fall on it from above. Such
extended pipes were therefore unreliable. The cooler could be simply a
refractory-lined extension of the kiln, transferring heat inefficiently by
radiation only, or lifters, as used in rotary coolers, could be used to improve
heat exchange, but necessitating a complicated cantilevered hood to protect
the firing pipe from the cascading clinker. A number of kilns were equipped
with this kind of cooler:

?Barton A1 1911
Masons A1 1913
Rhoose A1 1913
Rochester
1914
A2
Masons A2 1923
Rhoose A2 1923
Chinnor A1 1928
?South
1938
Ferriby A1

The coolers in practice were almost useless. Masons, using them up to the
1960s, recorded clinker outlet temperature around 1000°C, and the
“quenching” of the clinker, necessary for good quality, was delayed until it fell
out onto the conveyor.

A modest improvement in performance could be obtained by fitting a “van


der Werp” cooler. This involved perforating the shell of the kiln, and blowing
air into the area where the clinker bed lay, from a sealed-on pressurized
hood.

Early planetary coolers

The obvious demerits of the in-line cooler were tackled by allowing the clinker to run into tubes outside, and
parallel with, the kiln tube. Two designs immediately emerged – the ordinary, downhill-flowing tubes, and the
“reflex” tube, in which some sort of screw action caused the clinker to proceed uphill to the outlet.

Early planetary cooler designs

Both designs do away with the leaky "kiln hood" used with other coolers. Instead, the front of the kiln is
covered with a small "shield plate" pressed up against the kiln's front closure plate, allowing an effective and
easily maintained seal with little inleak.

The advantage of the reflex design was a lower and more compact design, with straightforward access to
the kiln for observation and maintenance. Disadvantages were the necessity of intact internals to maintain
the uphill movement of clinker, particularly difficult in the refractory-lined zone. More important, there was
the problem of supporting the outlet end at a point on the kiln shell where the temperature is high and
variable.

The downhill discharging design avoided these problems, but had the major disadvantage that access to the
front of the kiln could only be achieved by means of a cantilevered tubular steel tunnel. This feature was all
the more inconvenient if it was necessary to withdraw hot air from the kiln for coal drying.

Both designs had major mechanical problems at the point where clinker entered the tubes. White hot clinker
is falling in one direction, and high-temperature secondary air is flowing into the kiln, and it was difficult to
keep the steel cool enough to maintain its strength, particularly with the smaller early kilns, where the small
ducts had not much room for a thick layer of refractory to protect the steel. The connection was under
continual flexural stress as the kiln turned, so tubes cracked, leaked hot clinker or fell off. Clinker dropping
back out of the cooler over the firing pipe was a another nuisance. A further mechanical problem was in the
design of the kiln. The cooler assembly was very heavy, the mass being supported largely by the kiln's front
tyre, and unless many closely-spaced tyres were provided, the kiln was subject to large bending forces.

The seriousness of these problems diminished with the arrival of larger kilns and coolers, where more
refractory insulation could be used, and more elaborate expansion joints could be accommodated. The
relatively modern forms of the planetary cooler were all of the downhill-discharging design.

Early FLS "Unax" downhill discharging planetary cooler at Hope, 1929.


However, all planetary coolers have the problem that, in order to protect the kiln outlet ports, the flame must
be projected well beyond them. This means that a length of the kiln - typically one to two diameters beyond
the ports - is not available for sintering, and the kiln's "effective length" is diminished by this amount.
Furthermore, rather sluggish cooling of the clinker occurs in this zone, to the detriment of clinker quality.
From 1928 until WWII the majority of British kiln installations had "early" planetary coolers. Grate coolers
started to be used from the late 1930s, and these became popular because of the known problems with
planetary coolers and the rapid cooling effect of grates, and in a number of instances, kilns with planetary
coolers were converted to grate cooling.

Modern planetary coolers

Difficulty in controlling the earlier grate coolers prompted a return to the use of planetary coolers on larger
kilns in the post-war period. The problem of supporting the much larger cooling tubes was solved by
extending the kiln tube and providing another tyre downhill from the cooler outlet, but at the expense of an
extremely long access tunnel.

Later planetary cooler design


Picture: Rugby archive RC10. Rochester kiln 6 under construction, with the cooler tubes yet to be added,
showing the substantial length of kiln shell used to support the cooler. The access tunnel, to the right,
extends into the shell nearly as far as the cooler ports.

Picture: Rugby archive RC10. Rochester kiln 6 completed, showing the hood surrounding the clinker
discharge and the end of the access tunnel in the foreground.
Picture: Rugby archive cat. no. RC10-5-1. Rochester kiln 6 showing the massive second pier rollers
required for these coolers.

The new design began to be used in the mid 1960s. Here is a complete list.

Date Kiln Process Supplier


1963 Rhoose 3 Wet FLS
1963 Pitstone 4 Wet FLS
1964 Barrington 4 Wet FLS
1964 Limerick 4 Wet FLS
1965 Limerick 5 Wet FLS
1966 Padeswood 3 Long Dry FLS
1967 Ribblesdale 5 Wet FLS
1968 Rugby 6 Wet FLS
1970 Pitstone 5 Long Dry FLS
1972 Platin 1 Long Dry FLS
1975 Aberthaw 6 SP KHD
1976 Ribblesdale 6 Wet FLS
1977 Platin 2 SP FLS
1977 Ketton 7 SP FLS
South Ferriby
1978 Lepol FLS
3
1980 Rochester 6 Lepol FLS
1983 Limerick 6 SP FLS
1989 Derrylin 1 SP FCB

Platin 2 and South Ferriby 3 were subsequently converted to grate coolers. The revival of planetary coolers
in this form occurred because, despite the poor quenching and higher clinker temperatures produced, they
were much simpler and cheaper than grate coolers, and could produce high secondary air temperatures.
Because, unlike grate coolers, they produce no waste air, the heat loss from this is avoided, as is the need
to clean the dusty air.

With the advent of precalciners, excess hot air from a grate cooler, previously wasted, could now be
profitably used as tertiary air for the calciner, whereas the planetary cooler could produce no tertiary air.
Because of this, planetary coolers became obsolete.

© Dylan Moore 2011: last edit 15/01/2015.


The essential characteristic
of a grate cooler is a layer
of clinker spread on a more-
or-less horizontal perforated
grate, through which cold air
is blown. The grate is made
of steel, and the cold air
keeps it sufficiently cool to
avoid melting or burning. The
clinker progresses through
the cooler by moving more-
or-less horizontally along the
grate, and so the direction of
the cooling air is roughly at
right angles to the direction
of movement of the clinker,
and the cooler functions as a
cross-current heat
exchanger. This mechanism
is inherently less efficient
than a counter-current heat
exchanger such as a rotary
or planetary cooler, but there
are distinct advantages:

 the hot clinker at the


inlet is treated with
cold air rather than
partially heated air,
so that a rapid
quenching effect is
possible, improving
clinker quality by
preserving reactive
high-temperature
silicate polymorphs.
 by using air in excess
of that needed for
combustion, the
clinker can readily be
cooled below 200°C.
 it is easy to tap off
exhaust hot air
streams in desired
temperature ranges
for use in other
processes - e.g.
calciner tertiary air or
drying air for fuel or
raw materials.

It is fascinating to read the


Fuller cooler patent of 1937:
in distinct contrast to
discussions of modern
coolers, the preamble of the
patent focuses almost
exclusively on the benefits of
the cooler to clinker quality.

The types of grate cooler are


differentiated according to
the method whereby the
clinker is moved along the
grate, and are treated in
chronological order of their
introduction.

Grate coolers exchange heat between clinker and air, and because heat
losses are comparatively slight, the thermodynamics of the process is
simple enough. The specific heat of air is often assumed as an
approximation to be equal to that of clinker, although this is not quite true, as
shown.
This relates to a typical modern clinker, including the effect of the melting
endotherm, and to moist air. For a narrow cross-current heat exchanger with
no heat losses fed with clinker at 1300°C, the temperatures of clinker and
secondary air are:

Reference to the data sheets on fuels shows that the air needed for
combustion is usually in the range 335-345 kg per GJ of energy, or say 380
kg allowing a 10% excess. This means that an efficient kiln consuming 3.2
GJ per tonne of clinker will only take in 1216 kg of air per tonne of clinker -
i.e. the air/clinker ratio is 1.216. Using this air alone in the cooler would
result in a clinker temperature of 399°C, so the extra air needed to obtain a
reasonable temperature is surplus to requirements. On the other hand, a
typical Thames-side wet process kiln of the 1950s, consuming 8.0 GJ/t,
would use 3.04 kg of air per kg clinker, and could therefore obtain very low
clinker temperatures without (theoretically) any waste air production. Grate
coolers were developed in the historical context of these less efficient kilns,
and so the issue of waste air was not at that time serious.

The waste air stream is another distinct disadvantage of the grate cooler
design, and limited its popularity until the 1980s. Rotary and planetary
coolers operate under suction and although noisy, produce little atmospheric
pollution. The waste air of a grate cooler naturally contains a large amount
of gritty clinker dust. The early coolers, like the kilns, made little effort to trap
this, and produced a lot of fall-out in the immediate vicinity. Furthermore,
clinker dust is much more objectionable that kiln dust (which is largely
unchanged raw material) because of its ability to permanently coat surfaces
with hard hydration products. As environmental pressure grew, increasingly
elaborate dust filters had to be fitted to grate cooler exhausts, greatly
increasing the installation costs.

Bucket Grate Coolers

The first form of grate cooler used on British kilns was the Bucket Grate.
The cooler was built into a greatly-enlarged kiln hood, with length about 1.9
times the kiln diameter, and width 15% greater than the length. A concave
grate-plate occupied the entire base, with a pressurisation chamber below.
The nose of the kiln was surrounded with a steel cylinder about 1.3 times
the diameter of the kiln, on the outside of which were attached various
implements, rotating with the kiln, designed to distribute and agitate the
clinker on the grate, and to move it uphill towards the outlet of the cooler. A
more-or-less thin layer of undisturbed clinker protected the grate from direct
contact with hot material. The air-gap between the cooler shell and the kiln
nose, open to ambient air, prevented these from overheating.
Nine of these coolers were installed during the flurry of kiln construction in
the late 1930s, patents having been obtained in 1934. By the time kiln
construction resumed after WWII, the superiority of reciprocating grate
coolers had become obvious, and no more bucket grates were installed. As
far as I know, the chronological order of installations was:

1935 Rodmell A1 (retrofit) removed 1957


1936 Ribblesdale A1 (new kiln) shut down 1982
1937 Southam A4 (new kiln) shut down 1979
1937 Metropolitan A3 (new kiln) shut down 1970
1937 Ribblesdale A2 (new kiln) shut down 1983
1937 Pitstone A1 (new kiln) shut down 1979
1938 Southam A5 (new kiln) shut down 1979
1938 Rochester A3 (new kiln) shut down 1980
1938 Cliffe B1 (new kiln) shut down 1969

The cooler at Rodmell replaced the original Edgar Allen reflex planetary
cooler, and was in turn replaced in 1957 with a Fuller 633S reciprocating
grate.

The buckets and paddles were very susceptible to wear, and the dead
material lying on the grate surface tended to blind it. This at least meant that
the hood could not be pressurised, and little dirty air passed out to the stack,
but inevitably heat transfer efficiency was poor, and even with much heat-
loss by radiation, clinker outlet temperature was typically 300-350°C,
although the grate loading for the listed kilns was only 10-15 t/d per m2.

Reciprocating Grate Coolers


Following the bucket grate of 1934, the first Reciprocating Grate was patented by Fuller in
1937, and early examples were installed in the USA. They started to be installed in the UK
after the delay of WWII, and became the most common type of cooler on large kilns.

Classic horizontal grate design. Plates had a working surface 1 ft square, plus 5
inches of overlap, a 2 inch lip and 0.25 inch gap between plates. The moving
plates oscillated with an amplitude of around 5 inches.

The cooler grate is composed of overlapping rows of perforated grate plates. Half of the rows
are static, fixed to the casing of the cooler. The alternate rows are carried on a movable frame
to which a reciprocating movement is imparted by an eccentric drive or hydraulic rams. The
overlying bed of clinker is pushed forward on the forward stroke, and the plates slide beneath
the bed on the return stroke. Fine clinker can fall though the grate holes, and so the under-
grate chamber contains drag-chain conveyor(s) to move the spillage to the outlet end of the
cooler. For most of its history, a crusher, usually in the form of a hammer mill, has been
placed at the end of the cooler. Larger clinker lumps, with a low surface area, are less
effectively cooled, and having been crushed, the rotary action of the hammer mill hurls the
fragments back up the cooler for further cooling.

The under-grate chamber is generally divided into a number of compartments, each with its
own fan, which can be separately pressurised. The chamber above the grate is refractory
lined. Areas of cold, "dead" clinker are provided at the sides of the grate to protect those areas
from over-heating. The hot air emerging from the bed passes out to the kiln, and the hot-end
grate pressure is controlled to provide a small negative pressure in the kiln hood. Because, in
general, more air passes through the grate than can be used by the kiln, outlet ducts are
provided on the side of the cooler parts of the over-grate chamber. The hot air passing out
through these may be used productively for process operations - e.g. for fuel or raw material
drying - or may simply be run to waste through an exhaust stack. The hot air is inevitably
heavily loaded with fine clinker grit, and so some sort of gas cleaning is provided prior to the
stack.

Both sloping and horizontal grates may be used. The original design had a 12° slope. Many
coolers during the 1950s had 10° slopes. In the 1960s, grates at 5° or horizontal were
employed. Grates for over 1000 tonnes per day clinker output are commonly constructed in
two separately-driven sections, the first (hot) part being sloped at 5° and the second
horizontal. Grates over 3000 t/d commonly have one sloping and two horizontal sections.
Fuller cooler sizes are designated by abbreviations, for example:

 748S means a sloping grate 7 plates wide and 48 plates long (i.e. roughly 7 ft by 48 ft)
 850H means a horizontal grate 8 plates wide and 50 plates long
 825S1050H means a sloping grate 8 plates wide and 25 plates long, followed by a
horizontal grate 10 plates wide and 50 plates long

Since all kilns, irrespective of process, deliver clinker at around 1300°, the size of cooler
required is solely related to the expected kiln output, and grates are typically designed for a
loading of 30 t/d per m2 - i.e. a kiln making 1800 t/d would require a grate area of 60 m2.

Classic horizontal grate cooler layout.

About 40 reciprocating grate coolers were installed, commencing with that on Norman C1 in
1949, which was a Fuller 620S cooler supplied under license by Vickers Armstrong. Through
the 1950s and 60s both grates and planetary coolers were installed, but with the arrival of
precalciners in the 1970s, needing a hot air supply that by-passes the kiln ("tertiary air"),
reciprocating grate coolers became the only viable option, and subsequent developments (see
below) were based upon the reciprocating grate design.

. . . in development . . .

Chain Grate Coolers

The Chain Grate Cooler was developed in parallel with the Lepol
preheater, and was installed on a few Lepol kiln installations. The structure
of the grate is essentially that of the Lepol preheater. Secondary air and
waste air are taken off in the same manner as the reciprocating grate. Unlike
the reciprocating grate, the clinker remains more or less undisturbed once it
has been deposited on the grate, and so efficient operation relies critically
on uniform spreading of the clinker at the inlet end. The standard method, at
least on early coolers, was to provide a separate, high pressure, pulsed air
supply to fluidise the clinker at the inlet. The grate can't be protected by a
"dead" layer and so at the inlet end it can be subject to much higher
temperatures. To alleviate this, and to protect the grate from large falling
lumps, on the older grates the clinker was directed onto the grate via a
water-cooled steel chute. This represented a built-in heat wastage of
typically 50 kJ/kg, and presented the problem of disposal of a quantity of hot
water.

The chronological order of installations was:

1960 Cauldon A2 shut down 1985


1965 Weardale A1 shut down 2002
1965 Weardale A2 shut down 2002
1937 South Ferriby A2 extant (2012)

. . . in development . . .

Later Developments

The reciprocating grate cooler was standard for installation on modern


precalciner kilns. The increasing size of the kilns and the push for minimum
energy consumption led to a major reassessment of cooler design. The
"modern" grate begins with the IKN patents of 1983. The essential
characteristics progressively incorporated into new coolers included:

 plate air nozzles with improved aerodynamics and higher pressure


drop, producing an cooling airflow less sensitive to unevenness of
bed loading.
 plate designs that reduce or eliminate "fall-through" of fine clinker.
 air supply coupled directly to the plates, preventing air from
bypassing parts of the grate, and allowing a more targeted
application of pressure where needed.

All these are modifications of the basic reciprocating cooler design. In


addition, the crossbar cooler has become available since 2000. This has a
fixed grate with zero fall-through, the clinker being moved by reciprocating
cross-bars above the grate surface.

. . . in development . . .

© Dylan Moore 2012: last edit 15/01/15.


Stokes' Cooler

A rotary cooler consists of a slightly inclined tube of similar construction to a


rotary kiln, placed at a lower level than the kiln so that the clinker leaving the
kiln can fall directly into the upper end of the cooler. Ambient air enters at the
other end. The upper part of the cooler, because of the high temperature of
the clinker, is brick lined, while the lower part is usually provided with scoops
or "lifters" that pick up the clinker and cascade it through the air-stream.

The earliest rotary cooler described is in Stokes' patent of 1888, and a


prototype may well have been installed on the kiln at Arlesey. He says:

In order to produce the required temperature in the furnace I prefer to use


what is known as the "regenerative system" in which the air is heated
previous to combustion, either by contact with the issueing (sic) hot clinker,
the passage of the air being in the contrary direction to that of the hot clinker;
or by the passage through regenerative chambers, or parallel flues, of well
known construction, instead of the cooling drum aforesaid.

Stokes' drawings show a cooler 38 ft long and 2'3" diameter (the kiln was
28×5), lined for two-thirds of its length with 4" brickwork, while the rest has
"projecting shelves" for lifters. If the kiln had worked (which it most certainly
didn't) the cooler would have been very effective. In the ensuing 80 years in
which they were used, rotary coolers evolved little beyond this first design.

Hurry & Seaman Coolers

Hurry & Seaman always used rotary coolers in their kiln designs, but there is
a subtle change of emphasis. The early Hurry & Seaman kilns were oil fired,
and getting the kiln hot enough was not a major preoccupation. On the other
hand, the piles of white hot clinker that they produced needed more efficient
handling, and their coolers were designed to cool the clinker. Early designs
had two cooling tubes on successively lower levels. In the first, ambient air
cooled the clinker to below red heat before entering the kiln as secondary air.
The clinker then passed through a crusher into the second cooler, with air
drawn through by a fan and wasted. Water was sprayed into the upper end,
to increase the cooling effect, and also probably to hydrate free lime, which
was a major problem in early rotary kiln clinker.

By the time Hurry & Seaman kilns began to be installed in Britain, single
cooler tubes were the norm. Because at that time it was recommended that
most or all of the combustion air should pass up the firing pipe, most cooler
air was diverted into coal drying or other duties. Water sprayed into the
cooler would have no detrimental effect on these.
Design features

The design of the cooler has to allow for the temperature of the clinker and
air. Because clinker enters at 1200°C or more, the upper part of the cooler is
lined with refractory brick. The brick lining generally prevents attachment of
lifters, so that little in the way of agitation of the bed of clinker is possible in
the hot zone, although "ripple" linings or cast refractory lifter blocks have
been tried, with little success.

The rest of the cooler, in which the clinker is usually below 800°C, is
provided with steel liner plates (mainly to protect the structural shell from
abrasion), and lifters. In the hotter part, the lifters are in the form of buckets
made from cast high temperature steel. In the cooler parts, mild steel
"channel" is generally used. At the outlet, a grid is usually provided, allowing
normal-sized clinker (<100 mm) to fall through to the clinker conveyor, while
larger lumps are separated out for crushing.

The cooling effect obtainable obviously depends upon the amount of


cascading of clinker that can take place. Provided that enough lifter capacity
is present to lift the entire clinker bed, the best results are obtained with a low
slope and high rotation speed.

Coolers on British kilns

Rotary coolers were installed on the vast majority of kilns in the first quarter of the twentieth century. They
were installed either beneath the firing floor, as in the original Hurry & Seaman designs installed by Fellner &
Ziegler, or beneath the kiln. The latter design had the advantage that floor space required for the kiln system
could be reduced, but meant that the cooler had to pass through one or more of the kiln piers, requiring a more
elaborate design of the latter.
The largest British rotary cooler, by Vickers Armstrong as installed at Kent and Shoreham, 92' long and 9' in
diameter. The design necessitated tall and massive kiln piers, but was easy to maintain and run.

One way of shortening the cooler so that the second kiln pier need not be pierced was the "double-back" or
"concentric" cooler, produced by several manufacturers in the period 1905-1920. The later versions were
frequently supplied with pressurisation air. Being very difficult to maintain, most of them were soon abandoned
in favour of more conventional designs. The standard design coolers shown above were installed at Kent in
1949 to replace the original (1922) concentric coolers. On some smaller kilns, such as Wouldham 7 & 8, the
concentric coolers were simply removed, and the kilns run with no cooling at all.

Concentric rotary cooler layout based on those at Johnsons.


A tentative list of the concentric cooler installations is as follows:

Kiln Date Supplier Replaced


Johnsons 1-3 1907 FLS no
Newhaven 1907 FLS no
Premier 1908 FLS no
Harefield 1909 FLS no
Wouldham 7-9 1910-12 FLS ?1929
Peters 1&2 1911-13 FLS no
West Kent 4 1911 FLS no
Rugby 1&2 1911-13 Newell ?
West Thurrock 1&2 1912-14 FLS ?
Wilmington 3&4 1913-20 FLS 1933?
Penarth 1914 FLS 1933
Coltness 1 1914 Pfeiffer no
Magheramorne 1&2 1914-21 FLS 1949
?Kirton Lindsey 1&2 1920-22 FLS ?1949
Kent 1&2 1922 FLS 1949

From 1923, rotary coolers started to lose ground to planetary coolers, and from 1937 to grate coolers. The
number of installations declined, although this to some extent reflects reduction in kiln installations, while size
of coolers increased. The largest installed (90'×9'0¾") were among the last, on Kent A1 and A2 and Shoreham
C1 and C2 in 1949-1950. The last installed was on the anachronistic Kirton Lindsey A5 in 1968.

Date Installed
1900-1909 93
1910-1919 39
1920-1929 37
1930-1939 14
1940-1949 1
1950-1959 7
1960-1969 3
Picture: ©NERC : British Geological Survey Cat. No. P538713. This is a view
(early 1960s?) into the outlet of Barrington A1 rotary cooler. Clinker cascades
through the stream of cold air being drawn in by the kiln overhead. Because
the cooler "lifters" are fairly worn, the hot clinker can be seen falling into the
cooler at the far end. Water is being squirted in, to aid the cooling action, at
the expense of cooling and diluting the secondary air.
© Dylan Moore 2011: last edit 15/01/15.

You might also like