Water Engines: Page 5

Gallery opened: Nov 2006

Updated: 15 Mar 2018

London Hydraulic Company pump
Back to Home PageBack to The MuseumBack to Water Engine Index




Left: A domestic water-powered water pump.

Pumps like these were used when the mains pressure was not enough to raise water to the top of the house. The pump was placed in the cellar, and the power section (the larger diameter cylinder on the right) was connected between the main and the kitchen, where much of the water would be used. When water was drawn, a percentage was pumped up to the top of the house by the smaller cylinder on the left.

The two cylinders were connected by the horizontal slotted lever, which has a movable pivot on the vertical lever with the handle on top. This was presumably used to control the amount of water pumped up.
In fact, the control of this device presents an interesting problem; if the water tank in the roof was prevented from over-filling by the usual ball float and valve, when the valve closed the pump would stop and no water could be drawn in the kitchen. The control lever probably circumvented this impasse by allowing the engine cylinder to work without operating the pump, but if so it is not clear from this illustration.

From Knight's American Mechanical Dictionary, 1881.


The Roberts pump was produced in America, probably between 1890 and 1912; US patent No 715,871 was issued to George J Roberts of Dayton, Ohio in December 1902. The pump was intended to use high-pressure mains water, of unacceptable quality, to pump purer water from another source, such as a well or rainwater storage up into a tank as an alternative supply. Rain-water would be much softer and better for domestic use. It had two cylinders and very much resembled a duplex boiler feed-pump. It was patented and introduced by George J Roberts of Dayton, Ohio. Other things were happening in Dayton, Ohio at the time.

Left: Drawing of the pump

A is the engine end and B the pump end. C is the water supply to the pump, and C' is the delivery from the pump. D is the water supply to the engine. The engine/pump was of a familiar type, and the patentable feature was a new kind of control valve, contained in the cylinder E. This cut off the power water when the pump delivery pressure passed a threshold, for example due to a ballcock closing.

From US Patent 715,871, 16 Dec 1902

You can see an example working (powered by compressed air rather than water) on Youtube. The version seen has duplex power and pumping cylinders and dates from around 1910.


A major- probably the major- application of water engines was the draining of mines by water-powered pumps. This was particularly useful for mines that produced minerals other than coal; coal at a coal mine was so cheap that it was used to ballast railway tracks, so steam pumping engines, even crude and inefficient ones, made economic sense. At other mines, the cost of transport, especially before the coming of the railways, made coal very expensive, and water engines were used whenever a suitable head of water was available. Lead mines made great use of water engines.

Another example is the Stennack copper mine in Cornwall (later incorporated into the North Levant) which used a water-pressure engine for pumping. In 1815 it employed about 100 men.

Info about the Stennack mine from A History of Tin Mining & Smelting In Cornwall by D B Barton. Published by D Bradford Barton in 1967.

Left: A horizontal water-powered pump

The engine section A is on the right, and the pump B on the left. The engine has a larger piston diameter than the pump so the water can be delivered with a larger head than it enters with. The engine has a slide valve operated by the slanting lever I connected to the piston rod; this is apparently arranged so that it can turn the machine on and off.

Judging by its horizontal format, this illustration was probably intended to be representative of pumps for mining applications.

From Knight's American Mechanical Dictionary, 1881.


The Sir Francis level was a lead mine at Gunnerside in Swaledale, England. It was begun in 1864, being driven horizontally into the hillside. From this passage a shaft was excavated to reach lower deposits of ore. Just above this adit, 43 fathoms from the surface, was installed a hydraulic engine for winding (moving men and material up and down in a lift cage) and a dual hydraulic pump for clearing water from the shaft. The hydraulic system was designed by Henry Davey of Hathorn Davey & Co of Leeds, and they installed it between 1880 and 1881. (See next section for more on Henry Davey) By December 1881 the hydraulic system was operational but work at the mine had almost stopped due to financial difficulties. The engine and pumps worked until summer 1882 but then remained idle for 8 years, because of a considerable drop in the price of lead. Work recommenced under a new company in 1889.

The Sir Francis also had a compressed air supply generated by a waterwheel and compressor on the surface; the air receiver can still be seen nearby.

Left: Hydraulic winding-engine and dual hydraulic pump at Sir Francis mine

A and A are the two linked cylinders of the hydraulic pump. Clearly the chain connection means that the ram/piston of one cylinder can pull the other one up, but obviously not push it down. The exact functioning of this is unclear at present, but it seems likely that it allows the weight of one pump-rod to balance the other. The operation of the two cylinders is synchronised, presumably so that one is taking in power water while the other is expelling it. This would make the flow of power water in the supply pipe more or less constant, eliminating the need for large air accumulators to cushion water-hammer.

NOTE: This drawing is identical with that in The Principles, Construction and Application of Pumping Machinery by Henry Davey, where it was stated that it was installed in the A D lead mine near Richmond, Yorkshire. However the drawing closely echoes the Sir Francis photographs below, and it would appear that the two installations were virtually identical.

One of the lift cages can be seen lower right. The vertical features behind it appear to be the pump rods, presumably driving force pumps at the bottom of the shaft.

Left: General layout at Sir Francis mine

By kind permission of Nick Catford and Subterranea Britannica. Photo 1996.

Left: The dual hydraulic pump at the Sir Francis mine

The suspension chains for the lift cage can be seen lower right. There is another cage at the bottom of the flooded shaft.

A chain can just be discerned at the top of the further cylinder; this passed over a pulley above and connected to the other cylinder. The small-bore pipes carry water from the pilot valves near the top of the cylinders to operate the main valves. The tappet-rod that operated the pilot valve can be seen at the top of the nearer cylinder. The main valves were fitted in short casings behind the pump cylinders; one can be seen here on the left.

Date of photograph unknown but before 1975.

Left: Section of one of the pump cylinders at the Sir Francis mine

As in many hydraulic pumping engines, a small pilot valve was used to control power water which operated a much large main valve.

The single pilot valve is at C, towards the top of the pump; it was operated by the tappet rod shown dotted behind the main ram A. It appears that this valve is what synchronises the push-pull operation of the two cylinders. Water from the pilot valve acted on main valve B, which allowed the flow of power water from the inlet pipe, pushing the ram A upwards and pulling up the pump rod. When the ram was at the top of its stroke the pilot valve was switched over and the main valve then allowed the water under the ram to escape via the outlet pipe, and into the shaft being pumped out. As explained elsewhere, this apparently daft procedure actually makes perfect sense as the high-pressure power water carries much more energy per volume than the water being pumped out.

Many water engines were double-acting, but this is single-acting, pushing up the pump-rod only. Presumably it fell back under its own weight.

Illustration from the book: The Principles, Construction and Application of Pumping Machinery by Henry Davey, pub Charles Griffin & Co, 1900. As noted in the illustration, this pump was installed in the A D lead mine near Richmond, Yorkshire. That is a long way from Gunnerside (although still in Swaledale) and so must be a separate installation. However this drawing very closely resembles the photographs from the Sir Francis mine.

Left: Hydraulic pump at Sir Francis mine: 1996

The purpose of the valve at the bottom of the two cylinders is currently unknown.

By kind permission of Nick Catford and Subterranea Britannica. Photo 1996.

Left: Hydraulic pump at Sir Francis mine: 1996

The control valve for starting and stopping the pump can be seen to the left.

By kind permission of Nick Catford and Subterranea Britannica. Photo 1996.

Left: Hydraulic winding-engine at Sir Francis mine: 1996

Two power cylinders drove a layshaft that was geared to the winding drum. One of the power cylinders and its associated valve-chest can be seen lower right. The control lever is in the mid position, presumably meaning 'stop'.

By kind permission of Nick Catford and Subterranea Britannica. Photo 1996.

Left: Hydraulic winding-engine at Sir Francis mine: 1996

One of the power cylinders can be seen at the far left. The control lever and the water supply pipes can be seen.

Note that the winding ropes go onto the drum on different sides so that one cage goes up as the other goes down.

By kind permission of Nick Catford and Subterranea Britannica. Photo 1996.

The engine, winding drum, and pumps are still in position today.


Left: Water-powered water pump for mine drainage: 1877

This engine-pump combination was made for the Seaton Delaval colliery. It was designed to raise 100 gallons per minute to a height of 60 feet, working from a water pressure of 165 psi. Since water can only be lifted 28 feet or so by suction, this sort of pump was mounted at the bottom of the mineshaft and pushed the water upwards to the surface. The two figures at the bottom show the valve chest of the engine section.

The engine is on the left and the pump on the right.

From the British journal Engineering for June 1877, p488

The company had made earlier engines in which the engine had slide valves made of lignum vitae; these wore out quickly as they handled muddy water at high pressure, and the arrangement shown here was designed by Mr Davey, one of the partners. In Fig 1, the exhaust valve is the hollow cylinder, and the admission valve seats on the right end of it, being attached to a piston at the extreme right of the valve cylinder. These parts are moved by water pressure controlled by a pilot valve which can be seen protruding to the right in Fig 2. The pilot valve is operated by the tappet rod visible above the main piston-rod. The pump section was double-acting. The scale at bottom right reads up to 9 inches; this is a relatively small pump, designed for mounting on a wooden baulk and dragging to whichever mine dip needed pumping out.

So what happens to the exhaust water from the engine side? There is not much point in pumping water out of a mine if you're simultaneously filling it up with the exhaust water. The answer is that the exhaust was fed into the pump delivery, and returned to the surface. This obviously means that the exhaust water was still at considerable pressure. A 60 foot lift represents about 26 psi, so there was still plenty of pressure differential to work the engine side. The efficiency was claimed to be between 70 and 80%.

Left: Cross-section of a Davey pump for mine drainage: 1922.

This appears to be a development of the engine-pump combination shown above. The valve chest V on top of the engine section at left is quite different. The pump is on the right.

From "Modern Pumping & Hydraulic Machinery" by Edward Butler. 2nd edition 1922, p237

Key to drawing above:

H Power cylinder
P Pump cylinder
S Suction intake
D Delivery outlet
R Pump valve cover
T Tappet rod
V Engine valve chest

Left: Cross-section of the engine valve-chest of the Davey pump just above: 1922.

The valve system is another example of water engine valves being water-driven and controlled by a small pilot valve.

E is the inlet for power water and X the outlet. V and V1 are two auxiliary cylinders holding the valve pistons G and G1; their movement is controlled by the pilot slide valve S. Sliding on the spindles of these pistons are two annular valves N and N1 which control the exhaust. The inlet is controlled by the mushroom-headed valves M,M1.
As piston G descends, annular valve N falls with it, its outside seat closing the path from cylinder to exhaust port X. G descends further and M opens the path from inlet port E to the cylinder. This arrangement prevents inlet and exhaust valves ever being open at the same time.

From "Modern Pumping & Hydraulic Machinery" by Edward Butler. 2nd edition 1922, p237

Left: Detail of the Davey pump valves: 1922.

The valve seats were faced with rubber or leather.

From "Modern Pumping & Hydraulic Machinery" by Edward Butler. 2nd edition 1922, p237

Left: A Hathorn-Davey duplex double-acting pump : 1920ish.

All these pumps have a common feature; they are in a low-height horizontal format to fit into a mine heading.

From "Practical Coal Mining" ed Prof Boulton. 1920-ish, Volume 4, p403

Water-powered mine pumps persisted well into the 20th century, being definitely in use in the early 1930's.


Left: Water-powered water pump at Wapping: 1893

For many years power was distributed around London by a high-pressure water supply, provided by The London Hydraulic Power Company. (Wikipedia page) Similar hydraulic networks were built in Hull, Liverpool, Birmingham, Manchester, and Glasgow.

Several pumping stations were built to provide high-pressure water for the London network, and the largest and most powerful was at Wapping in east London. It was equipped with steam-driven water pumps that delivered at 700psi. The water was obtained from a well sunk for the purpose, the water level being about 40ft below ground level. The well was sunk to the level of the London Clay; on top this is a layer of gravel about 8ft thick, from which the water was pumped. The well yielded about 16,000 gallons per hour, when pumping day and night. That was only about half the water required, and the rest was syphoned from the adjacent London Dock into the well, under valve control.

Two hydraulically-driven pumps made by the Hydraulic Engineering Co of Chester (under Ellington's patent) were placed in the well and pumped the water into storage tanks built on top of the boiler-house. After filtering this water went to the main high-pressure pumps. While there is an excellent picture of one of the pumps in the Engineer article, (see left) practical details are rather lacking. The main pump plungers were 20 inches in diameter, working with a 4-foot stroke at ten strokes per minute.

In the picture the power water for the pump can be seen coming in at top right. The pump cylinder (which also contained the power cylinder, coaxially) is in the centre of the drawing; directly above it can be seen the valvegear. At upper left there is a big tank of unknown function in the delivery pipe; it does not appear to be an air-vessel. There is a float that automatically stops the pump, presumably if the water level falls too low, in which case I imagine the pumps would draw air rather than water and overspeed.

Note the break in the drawing towards the bottom, and the given depth to the wate of 38 ft 7.5in.

Image and info from The Engineer, for 20 Jan 1893, p46.

Many thanks to Tom Bates for bringing this application of hydraulic pumping to my attention.


"Less talk, Moore pump"

Moore's pump used two columns of water running in pipes down the pit shaft; the columns were made to oscillate in anti-phase by a double-acting piston pump at ground level, and then powered a double-acting piston engine down at the pumping level. An interesting problem with this technique was that water leakage from one column would displace the engine piston sideways until it knocked off the cylinder cover. This was overcome by an ingenious system of automatic valves that controlled the amount of water in each column.

Left: Elevation of Moore's hydraulic pump: 1900.

The left half is sectioned. The mine shaft is at the right.

From "Lectures On Mining" by W Galloway, 1900. Lectures VIII and IX, p32

Left: Plan of Moore's hydraulic pump: 1900.

The left half is sectioned. The small plungers (four in total) form the engine part, while the two large plungers form the pump. All of them move together.

From "Lectures On Mining" by W Galloway, 1900. Lectures VIII and IX, p31

The Moore company still exists, or at least it appears to; the name may be coincidence. See www.moorepump.com


In Westphalia, Germany, the Kaselowsky system was used, which employed water pressures as high as 3000 - 3500 psi. At such pressures there were potential problems with efficiency losses due to the elasticity of the steel pipework, but the Kaselowsky system was still claimed to be 70% efficient.

Info in section above from "Modern Practice in Mining" by Sir R A S Redmayne. Longmans, Green & Co, 1932

The water was conveyed from the actuating pump on the surface to the underground pump in pipes 2.25 to 2.75 inches in diameter, depending on the capacity of the pump. The underground installation consisted of two pumps mounted on a common bedplate; each pump actuated the valves of the other.

Left: A Kaselowsky duplex double-acting pump : 1900.

Note the massive construction needed to cope with the high pressures used.

From "Lectures On Mining" by W Galloway, 1900. Lectures VIII and IX, p35

This table gives some details of Kaselowsky installations in Westphalia, Germany, in 1900:

Name of collieryPump depthWater pumped
König Wilhelm1541880

Left: Plan view of a Kaselowsky pump installed underground: 1900.

The mine shaft is the semi-circular structure on the left.

a Distributing chamber
d Pump plungers
g Suction-valve chambers
h Discharge-valve chambers

From "Lectures On Mining" by W Galloway, 1900. Lectures VIII and IX, p36

Kaselowsky pumps were also installed outside Germany; see this link: whc.unesco.org/en/tentativelists/5139

...which describes a coal mine near Seville, Spain. I quote:
"Special mention should be made of the remains at shaft No 5 which include the metallic extraction crane powered by the Bollinclx (1922) steam engine, the engine house, sieves and washing tanks, a Babcok Wilcox boiler, Kaselowsky drainage pump house, Schlamms tanks (coal sludge decantation reservoirs) and the electical power plant (1926). Shaft No 7 features the water tower (1928), the crane and engine house (1926-28), energy distribution tower (1929), electrical power plant and workshop buildings."

Left: End elevation of a Kaselowsky pump installed underground: 1900.

g Suction-valve chambers
h Discharge-valve chambers
k Suction pipe
w Pump well
x Conduit to rest of mine
y Sluice valve for shutting off conduit

From "Lectures On Mining" by W Galloway, 1900. Lectures VIII and IX, p36

Left: Side elevation of a Kaselowsky pump installed underground: 1900.

The mine shaft is on the left.

a Distributing chamber
g Suction-valve chambers
h Discharge-valve chambers
k Suction pipe

From "Lectures On Mining" by W Galloway, 1900. Lectures VIII and IX, p36


This is an extract from an article in the Scientific American Supplement, #508, September 26, 1885.

But first, some background information:

The Sutro Tunnel is a drainage tunnel located under the Comstock Lode in Northern Nevada. It begins at Virginia City and empties approximately 6 miles to the southeast, near the town of Dayton, Nevada. Despite the upbeat tone of the article, at this date the output of the mines was already in decline.

The Comstock Lode was the first major USA deposit of silver ore, discovered under what is now Virginia City, Nevada on the eastern slope of Mount Davidson, a peak in the Virginia range. The discovery was made public in 1859.

The tunnel was first planned by Adolph Sutro in 1860, in order to ease the daunting problems of draining the Comstock to allow access to deeper mineral exploration. The main tunnel was completed in 1878. Unfortunately this was a little late as Comstock production peaked in 1877, the mines producing over $14,000,000 of gold and $21,000,000 of silver that year. After that production decreased rapidly, and by 1880 the Comstock was considered to be worked out. The deepest excavation was reached in 1884, in the Mexican Winze, 3300 feet below the surface. (A winze is a shaft that is started underground and goes straight down- Ed) Underground mining continued sporadically until 1922, when the last of the pumps were shut off and the mines allowed to flood.

Now read on...

"The Sutro drain tunnel (nearly four miles in length) connects with the shaft at a depth of 1,600 ft, up to which point all the water encountered below is pumped. The shaft was sunk to the depth of 2,200 ft. before more water was encountered than could be hoisted out in the "skips" with the dirt. At the 2,200 level two Cornish pumps, each with columns fifteen inches in diameter, were put in. At the 2,400 level the same pumps were used. On this level a drift was run that connected with the old Hale & Norcross and Savage shafts, producing a good circulation of air both in the shaft and in the mines mentioned. At this point, on account of the inflow from the mines consequent upon connecting with them by means of the drift, they had more water than the Cornish pumps could handle, and introduced the hydraulic pumps, which pumps are run by the pressure of water from the surface through a pipe running down from the top of the shaft, whereas the Cornish pumps are run by huge steam engines.

"By means of the hydraulic pumps they were enabled to sink the shaft to the 2,600 level, and extended the Cornish pumps to that point, where another set of hydraulic pumps was put in. They then sunk the shaft to the 2,800 level, when they ran another drift westward, and tapped the vein. The prospects at this depth in the Hale & Norcross and Chollar mines were so encouraging that the management decided to sink the shaft to the depth of 3,000 ft. On reaching the 3,000 level, they ran a third drift through to the vein. The distance from the shaft to the east wall of the vein was found to be only 250 ft. At the depth of 3,000 ft. they put in one of the pair of hydraulic pumps that is to be set up there. The second pump is now arriving from San Francisco, and as soon as the several parts are on the ground, it will be at once put in place alongside its fellow on the 3,000 level. This additional pump will increase the capacity from 600,000 to 700,000 gallons in twenty-four hours, or about forty-five miners' inches.

"Owing to the excellent showing of ore obtained on the 3,000 level by the Hale & Norcross Company, and to the continuation of the ore below that level (as shown by a winze sunk in the vein), the management determined to sink the shaft to the vertical depth of 3,200 ft. It is now 3,120 ft. deep, and it is safe to say that it will reach the depth of 3,200 ft. early in September, when it will lack but eighty feet of being as deep as the shaft at Przibram was at the time of the great festival. Although the shaft is of great size--about thirty feet by ten feet before the timbers are put in--the workmen lower it at the rate of about three feet a day, in rock as hard as flint.

"The hydraulic pump now working at the 3,000 foot level of the shaft is the deepest in the world. In Europe the deepest is in a mine in the Hartz Mountains, Germany, which is working at the depth of 2,700 feet. It is, however, a small pump not half the size of the one in the Combination shaft. Although these pumps were first used in Europe, those in operation here are far superior in size, and in every other respect, to those of the Old World, several valuable improvements having been made in them by the machinists of the Pacific coast.

"The capacity of the two Cornish pumps, which lift the water from the 2,900 foot level to the Sutro drain tunnel (at the 1,600 level), is about 1,000,000 gallons in twenty-four hours, and the capacity of the present hydraulic pumps is 3,500,000 gallons in the same time. They are now daily pumping, with both hydraulic and Cornish pumps, about 4,000,000 gallons, but could pump at least 500,000 gallons more in twenty-four hours than they are now doing. The daily capacity with the hydraulic pump now coming, and which will be set up as mate to that now in operation at the 3,000 foot level, will be 5,200,000 gallons.

"The water which feeds the pressure pipe of the three sets of hydraulic pumps is brought from near Lake Tahoe, in the Sierra Nevada Mountains. The distance is about thirty miles, and the greater part of the way the water flows through iron pipes, which at one point cross a depression 1,720 feet in depth. The pressure pipe takes this water from a tank situated on the eastern slope of Mount Davidson, 3,500 feet west of the shaft. At the tank this pipe is twelve inches in diameter, but is only eight inches where it enters the top of the shaft. The tank whence the water is taken is 426 feet higher than the top of the shaft, therefore the vertical pressure upon the hydraulic pump at the 3,000 foot level is 3,426 feet. The pressure pipe is of ordinary galvanized iron where it receives the water at the tank, but gradually grows thicker and stronger, and at the 3,000 level it is constructed of cast iron, and is 2˝ inches in thickness. The pressure at this point is 1,500 pounds to the square inch.

"In the early days of hydraulic mining in California the miners thought that with a vertical pressure of 300 feet they could almost tear the world to pieces, and not a man among them could have been made to believe that any pipe could be constructed that would withstand a vertical pressure of 1,000 feet; but we now see that a thickness of two and a half inches of cast iron will sustain a vertical pressure of over 3,400 feet.

"There is only one pressure pipe for all the hydraulic pumps. This extends from the tank on the side of the mountain to the 3,000 foot level. It is tapped at the points where are situated the several sets of hydraulic pumps. The water from the pressure pipe enters one part of the pump, where it moves a piston-back and forth, just as the piston of a steam engine is moved by steam. This water engine moves a pump which not only raises to the surface the water which has been used as driving power, but also a vast quantity of water from the shaft, all of which is forced up to the Sutro drain tunnel through what is called a return pipe. Each set of hydraulic pumps has its return pipe; therefore there are three return pipes--one from the 2,400, one from the 2,600, and another from the 3,000 level.

"Some idea may be formed of the great size of these hydraulic engines when it is known that the stations excavated for them at the several levels where they are placed are 85 feet long, 28 feet wide, and 12 feet high. All this space is so filled with machinery that only sufficient room is left to allow of the workmen moving about it. One of these stations would, on the surface, form a hall large enough for a ball room, and to those who are unacquainted with the skill of our miners it must seem wonderful that such great openings can be made and securely supported far down in the bowels of the earth; yet it is very effectually done. These great subterranean halls are supported by timbers 14×16 inches square set along the walls three feet apart, from center to center, and the caps or joists passing overhead are timbers of the same size. The timber used is mountain spruce. Not one of these huge stations has thus far cost one dollar for repairs. The station at the 2,400 level has been in use five years, that at the 2,600 three years, and the one at the 3,000 level eight months. Room for ventilation is left behind the timbers, and all are still sound. Timbers of the same kind are used in the shaft, and all are sound. The shaft has cost nothing for repairs. Being in hard andesite rock from top to bottom, the ground does not swell and crowd upon the timbers.

"If it shall be thought advisable to go to a greater depth than 3,200 feet, a station of large size will be made on the east side of the present shaft, and in this station will be sunk a shaft of smaller size. The reason why the work will be continued in this way is that in a single hoist of 3,200 feet the weight of a steel wire cable of that length is very great--so great that the loaded cage it brings up is a mere trifle in comparison. In this secondary shaft the hoisting apparatus and pumps will be run by means of compressed air. As it is very expensive to make compressed air by steam power, the pressure pipe will be tapped at the level of the Sutro tunnel, and a stream of water taken out that will be used in running a turbine wheel of sufficient capacity to drive three air compressors. As there will be a vertical pressure upon the turbine at this depth of over 2,000 feet, a large stream of water will not be required. The water used in driving the wheel will flow out through the Sutro tunnel, and give no trouble in the shaft.

"By means of this great shaft and its powerful hydraulic and Cornish pumps the crust of the earth will probably yet be penetrated to far greater depth than in any other place in the world."

Back to Home PageBack to The MuseumBack to Water Engine IndexTop of this page