Examining the cornish engine

was almost exclusively limited in application to water works and
mining installations. It was also big; 9′ and 10′ strokes
with bores up to 90′ or 100′ were common by the middle of
the 19th century.

We can examine its decendency from the atmospheric engine in
order to better understand its development and operation.

Although the shaft is no longer used for pumping, it served
until recently as a ventilating shaft, and air ducts are located
behind the concrete walls.

This engine is now preserved and open to the public. Notice the
resemblence to the engine house model shown in the July/August
issue of IMA.

The windows indicate the three floor levels of these engine
houses. The lower one is level with the cylinder bottom; the middle
several feet below the cylinder top, and the upper level with the
walking beam.

The beam bearings rested on the wall of the engine house itself,
and as a result the shaft side wall was heavily built, often being
three feet thick.

The first engines built by Newcomen were single cylinder
machines set up next to mine shafts for the purpose of removing
ground water. The cylinder was vertical, often mounted directly on
top of a round boiler, in appearance much like a brewery still.

The piston was connected to one end of a wooden beam, the other
end of which was over the mine shaft and connected to the pumps
down in the mine. The beam was usually mounted on a solid mass of
cut stone which supported the middle pivot bearing.

The engine was so balanced that when it was at rest the weight
of the pump gear pulled the piston to the top of the cylinder.

To start the engine steam at low pressure, only 1 or 2 p.s.i.
was admitted into the cylinder till all the air was blown out. The
admission valve was then closed and a jet of cold water sprayed
into the cylinder which condensed the steam forming a vacuum, and
also cooling the cylinder at the same time. Atmospheric pressure
then furled the piston down raising the pump gear through the
medium of the walking beam. At the bottom of the piston stroke the
steam valve opened to repeat the cycle.

Watt’s first and major contribution was to relocate the
condensing part of the cycle to a separate con-densor thus
materially improving the engine’s thermal efficiency.

He also went a step further and covered the open top of the
cylinder to keep cold air out. By substituting a piston rod and
packing gland for the chains of Newcomen he gave the steam engine
its potential for becoming double acting.

Initially, however, Newcomen’s cycle was followed except
that steam was allowed to flow into the top of the cylinder, at
first merely for the purpose of keeping cold air out.

When the piston reached the bottom of its stroke the steam above
the piston was admitted below the piston, breaking the vacuum and
equalizing the pressure in both ends of the cylinder so that the
piston could be drawn up by the pump rods. At the top of its stroke
an exhaust valve opened in the bottom of the cylinder and an
admission valve let steam into the top and the cycle began
anew.

At this point it was obvious that by increasing the boiler
pressure additional work could be gained from the engine as steam
pressure now assisted the vacuum. This primitive cycle endured, as
long as Cornish engines, until the 1950s in Great Britain.

In the early part of the 19th century, the southwestern part of
Great Britain was in the midst of a mining boom and bringing up
thousands of tons of tin and copper ore. Good pumping engines for
the mines were needed and as the engine’s major developments
took place there the name ‘Cornish engine’ was coined.
Engine houses by this time had progressed from the simple crude
wooden shacks of Newcomen’s day to massive stone buildings
three stories high.

The engine house was built on the lip of the mine shaft and one
wall now took the place of the cut stone beam supports. This wall
was invariably thicker than the other three and had an opening
built into it for the end of the beam to pass through. These by now
were of cast iron and usually consisted of two flat oval shaped
beams side by side. The piston rod was connected to one end of the
beam by Watt’s parallel motion. The other, ‘outdoor,’
end was usually simply connected to the uppr end of the pump rods
and these were long enough that the arc described by the beam end
as it moved up and down was absorbed by the spring of the wooden
pump rods.

These latter were massive affairs; in some cases 18′ or
24′ square at the top end, usually of baltic pine or from the
west coast of the United States. (Many old Cornish records refer to
shipments of ‘Oregon pine’, but the exact nature of this
type of wood seems to be in doubt. An English friend of mine has
asked about ‘Oregon pine’ while on a business visit to the
West Coast and been told no such designation exists today. Can some
reader offer information on what ‘Oregon pine’ was or what
it is called today?)

These rods ran to the bottom of the pit, regardless of how deep
it was, and pump rods several thousand feet long were not
unknown.

It was the great weight of these rods which actually pumped the
water, for except for the lowest pump in the sump which was of the
piston in cylinder type, the pumps were single acting outside
packed plumber units. The vertical plumbers were mounted on offset
arms bolted to the pump rods and of course had the same stroke as
the rods. They worked in ‘lifts’ of 150′ or 200′
thus minimizing the strains on the cast iron piping. The pump rods
were lifted by the steam cylinder pulling the ‘indoor’ end
of the beam which was the suction stroke. As the rods descended the
water was pumped up to the next pump’s sump and so on until it
was discharged from the mine. It was necessary to carefully balance
the weight of the pump rods so that they would not descend too fast
and ‘balance boxes’ were connected to the rods. These were
simply a pivoted horizontal beam with one end connected to the pump
rod and a box, loaded with cast iron or rocks at the other end. One
of these shows clearly in the photo of a Cornish engine model on
page 2 of the July/August ’81 issue of I.M.A. It is the
slightly tilted box to the left bottom of the photo.

The large frame over the shaft mouth was for the purpose of
pulling up the pump rods for maintenance work.

The high efficiency of the Cornish engine appealed to the
engineers who began to lay out the early water works system for
English cities and it also found a ready field for its use in that
capacity. There were slight differences of construction. The main
difference was that the entire engine was within the pump house
including the pumps. Without the weight of the mine shaft pump
rods, the engine was useless, so an artificial load was added. This
consisted of a cast iron case, usually with fluting or some other
ornamentation hung on the pump end of the beam. Tons of cast iron
would be packed inside this case and the pump plunger was fastened
underneath it. Now, the steam cylinder lifted this weight, and the
pump made its suction stroke. As the cylinder pressure on both
sides of the piston equalized, the massive cast iron weight
descended and the pump operated on its discharge stroke. The beam
center supports were no longer mounted on a masonry wall but
instead on a massive cast iron framing which was built into the
walls of the engine house. For water works engines, both steam
piston rod and pump rods were guided by parallel motion. The
condensors, usually of the jet type, were under the floor;
condensate and boiler feed pumps were also operated by links from
the beam.

‘Outdoor’ end of the engine. The two platforms would be
used at the engine ‘driver’ to lubricate or maintain the
upper pump rod bearing, or ‘nose bearing’ as it was known.
The iron plates which reinforced the massive wood pump rod can be
seen.

The date 1924, when this engine was installed in its present
location, is just visible in the keystone over the door.

As these engines had no crank shafts there were several unique
problems to be faced in their operation.

The first and perhaps most dangerous was that there was no crank
to limit piston movement at the end of the stroke. This was
controlled solely by the valve gear; should the engine overstroke,
the piston would hit the cylinder head. In an effort to keep the
engine from smashing itself to pieces, (and more than one did),
massive timbers were bolted across the ends of the beam, at right
angles. These were supposed to strike ‘stops’ or catch
beams built into the engine framing and so limit piston travel. In
some cases the wood was replaced by forked iron ‘ears.’
When the engine was at rest the pump end of the beam rested on its
stops. It may seem to us to be an incredible way to build a
machine, yet the Victorian engineers had a direct approach to
solving problems!

Some of the engines preserved today still show traces of
accidents, either in repairs to cylinders or other metal parts or
extra timbers and strapping to repair cracked catch beams.

The biggest dangers the engines faced was a bursting of the
discharge water mains which allowed the pump end to descend too
fast. It was said that when a big Cornish engine hit its stops,
anyone within a couple of blocks of the engine house knew it!

Another peculiarity of the Cornish engine was its speed control.
It was not regulated by throttle opening but by stalling it between
strokes! With everything else about the engine being oddball, why
not the valve gear as well?

As there was no eccentric the valve gear was operated by a
‘plug rod’ which hung from the walking beam between the
center bearing and the cylinder. The cylinder itself had four
valves, three on top, one at the bottom.

The upper valve chest housed throttle, admission, and
equilibrium valves; the bottom valve chest had the exhaust valve.
The cylinder had a framing on the side facing the beam center which
mounted several horizontal shafts about 4 to 6 feet above floor
level. These were connected to the valves by offset arms and links
so that as the horizontal shafts or ‘arbors’ as they were
called rotated slightly in their supports, the valves opened or
closed. Each arbor also had a large handle fixed to it which served
a dual purpose. The handles were struck by pins connected to the
plug rod as it rose and fell thus opening and closing the valves,
but they were also used by the engine driver when starting an
engine, to open and close the valves until the engine was self
acting.

Cylinder or ‘indoor’ end of the beam. Notice the massive
construction of the beam and the links which connect it to the
piston rod cap, the latter being visible between the links.

The ‘ears’ which prevent the engine from overstroking
are fitted across the upper end of the beam. The timbers on which
they strike are located just below the wooden fence. An interesting
point is that the upper link pin, in the beam eye, works only in a
half bearing. There is no top brass as the strain is always in one
directiondown.

The wire cable holds up this end of the beam as the pump rods in
the shaft have been removed, removing the weight which would
normally keep the cylinder end up.

The valve action wasn’t quite as simple as this, however,
for gravity also played a key part. To better understand the
system, let’s follow the whole cycle starting with the piston
at the top of the cylinder.

The steam and exhaust valves open which puts boiler pressure
above, and vacuum below, the piston. The piston descends, raising
the pump weight or shaft rods. The plug rod also descends, and at a
predetermined point an adjustable bar on it strikes the admission
valve handle closing the admission valve and holding it shut as the
bar continues to slide down along the handle. The bar’s
position on the plug rod is adjustable by means of a screw so that
the point at which it hits the handle and cuts off the steam can be
set to suit operating conditions.

When the piston reaches the bottom of its stroke, another plug
rod pin strikes and closes the exhaust valve. This valve is
automatically ‘latched’ shut by a catch device on the arbor
and as the exhaust valve arbor rotates, it in turn releases a catch
which has locked the equilibrium valve shut allowing it to pop
open. The opening of the equilibrium valve allows the steam from
above the piston to flow under the piston thus equalizing pressure
in both ends of the cylinder.

The equilibrium arbor’s rotation also locked the admission
valve in its closed position against the force of a counterweight
which would otherwise open it. The plug rod on reaching the bottom
of its stroke lifted the cataract, still another part of the valve
gear which we will come to presently.

With cylinder pressure equalized, the pump end of the beam
descended, the discharge stroke took place and the piston neared
the top of its stroke. The plug rod was rising as well, and just
before the end of the stroke it hit the handle on the equilibrium
valve arbor and closed it, forming a steam cushion to bring

This machine was built by Boulton and Watt in 1S20. In 1840 it
was installed in its present location and in 1848 the valve sear
was rebuilt into its present style. This engine, with a 64′
bore ran in regular service until 1943. Today it is in operation
every weekend at the Kew Bridge Steam Museum, London.

The main steam line conies in at the middle right of the
photo.

This engine is unusual in that one arbor (horizontal valve sear
rock shaft) is mounted just below the upper valve chest. Usual
practice had all three arbors at about chest level.

The two fluted columns are side pipes, one of which carries the
steam from the upper valve chest to the bottom of the cylinder.

The vertical rod with a crank at the bottom is the throttle. The
larger rod running top to bottom just to the left of center is the
plus rod.

The beam center bearing and pumps are to the extreme left, out
of the photo.

Now springs into play the cataract! This is simply a weighted
hydraulic plunger which is lifted about 6-10′ by the plug rod
and whose rate of descent is controlled by a needle valve. When the
cataract nears the bottom of its stroke, a linnale trips the
catches holding exhaust and steam valves shut. These are both
counter balanced with weights and as the catches are released the
arbors rotate, the handles fly up and the valves open beginning the
cycle.

It was this pause, controlled by an adjustable cataract, that
determined actual total gallonage output of a Cornish engine, and
they could be adjusted by it to run from 2 or 3 strokes per minute
to 12 or 14. The actual ‘stroke’, that is one down and up
of the piston, took the same length of time regardless of how many
strokes per minute the engine was set for.

The arbors had several interlocking features built into them to
prevent the possibility of the wrong valve dropping open at the
wrong time. All in all it is an ingenious valve gear and especially
unusual to see in operation, especially when the engine stops, and
then suddenly almost as if at their own volition the valves
suddenly pop open and the engine starts again.

Yes, I did say ‘see in operation’ for there are a number
of preserved Cornish engines in Great Britain, and two locations
still operate them.

The oldest of these is the Crofton Pump house on the Kennet and
Avon Canal in Wiltshire. (British Rail to Great Bedwyn, pleasant
hour walk past thatched roof farms to pump house) and this location
has two engines, the no. 1, 42′ bore built 1812, and no. 2
built in 1846 also 42′ bore. These engines are steamed several
times a year and the no. 1 engine is the oldest known steam engine
in the world which is still on its original site and handling its
original load. (Water for the summit level of the canal.)

The other site to visit is in the city of London itself. It is
the ‘Kew Bridge Steam Museum’ and has engines which have to
be seen to be believed. There is an excellent collection of engines
which have been moved to the site from other areas but the
‘originals’ consist of 5 Cornish engines. Two of these, a
Boulton and Watt built in 1820 with a 64′ dia cylinder, and a
90′ bore engine built by Sandys, Carne and Vivian, have been
restored to operation and are run, together with the other
aforementioned engines every weekend.

The Boulton and Watt engine, with a 8′ stroke pumped 2
million gallons a day from 1820 to 1943, an incredible service life
of 123 years.

The ’90’ as the Sandys engine is known as an 11-foot
stroke with a capacity of 6 million gallons a day. It ran only from
1846 to 1943 so missing the century mark by 3 years! The station
also has a 100′ x 11′ Cornish engine which is not yet
restored to service. The cast iron beam on this monster weighs
almost 50 tons!

These engines, Crofton, and Kew, are all of the ‘Water
works’ pattern, where the entire engine is within the house.
There are several preserved Cornish engines on old mine sites which
show the type illustrated by the photo in I.M.A. July/August
’81.

One site is the Preston Grange Mining Museum near Edinburgh,
Scotland. Several others are in Cornwall but not all are accessible
to the general public as some are located on still active sites and
require permission to visit.

The Cornish working cycle described in this article is a typical
one and as with all other forms of engines there were minor
variations. Of the 26 preserved engines listed above there are
probably close to a dozen variations in the engine ‘gear
work’ as the old Cornish men called it.

Cornish engines were used in this country as well but never to
the extent that the British used them.

Of all forms of the steam engines made by man, the Cornish
engine must certainly rate as one of the more unusual types and
without doubt impressive in concept, finish and operation.

For those wishing additional information on location of other
Cornish pumping engines, send 50 to: Copies, Stemgas, P.O.
Box 328, Lancaster, PA 17603, for a 3-page Xerox copy of a
list published in February 1981 by the Trevethick Society
Newsletter. Listed are all known Cornish pumping engines still in
existence throughout the world.