Water Engines: Page 3


Updated: 2 July 2010

More on Keith compressor

Back to Home PageBack to The MuseumBack to Water Engine Index

W G Armstrong was one of the great pioneers in hydraulics, though as we have seen he was by no means the first to build a water engine in Great Britain. He was a very keen fisherman, and while he was angling on the River Dee at Dentdale in the Pennines, he observed in action a waterwheel that supplied power to a marble quarry. Armstrong felt that much of the available power was wasted, and on returning to Newcastle, he designed a rotary hydraulic engine. This was before he set up his famous Elswick works, (in 1847) and so a prototype was built in 1838 at the High Bridge works of his friend Henry Watson. As is usually the case with rotary engines, it was not successful.
Armstrong then adopted side-by-side cylinders with reciprocating pistons.

Left: Armstrong's rotary hydraulic engine: 1838.

The power water entered via the pipe on the left and exited at the right.

Judging by what looks like a description on a wooden tablet in front of the machine, this appears to be a photograph of museum exhibit with the background retouched out. If anyone can shed any light I would be most glad to hear from them.

From SciMusLib

Left: Armstrong's rotary hydraulic engine animated.

The engine had four circular paddles pivoted in a rotating annulus; tappets on the paddle shafts were operated by fixed protrusions on the inner ring visible in the photo above, and swivelled the paddles so they were flat in the annulus when they passed through a slot at the beginning of the part-toroid power section. They were then turned through 90 degrees so they sealed against the inner surface of the power section.
It is clear that Armstrong was aiming at a positive-displacement engine that could make better use of high water pressure than a waterwheel. This was many years before the invention of the Pelton turbine.

I am indebted to Bill Todd for this interpretation of how the rotary engine worked; I'm sure Bill is right.

This superb animation of the Armstrong rotary engine by Bill Todd.

Strangely, it was at first very difficult to find any details or images of Armstrong's reciprocating water engines, but this is now remedied:

Left: Armstrong three-cylinder hydraulic engine: 1850

There are three cylinders oscillating on trunnions, with cranks set at 120 degrees. In the upper drawing, the valves at left are operated by the lever L, which is driven by the oscillation of the cylinder. The supply and exhaust valves connected are connected by pipe to the trunnion of the cylinder.
The drawing shows what appears to be a substantial flywheel; it is a question as to why this would be required with three cylinders to smooth out the torque. Note that the engine is much more compact than the earlier beam engines shown on previous pages.

Drawing from Modern Steam Practice and Engineering by John G Winton p326: published Blackie & son 1883.

Above: Portrait of William George Armstrong

Left: A surviving Armstrong hydraulic engine at Newcastle-upon-Tyne Swing-Bridge.

There are two working Armstrong engines that move the Swing Bridge on the Tyne between Newcastle & Gateshead. The bridge operates on most days, still using the original hydraulic engines. The power fluid is driven by electric pumps; whether it is still water or has been converted to oil, (which has the merit of not freezing in winter) remains unclear at present.

Only two cylinders are visible in this photograph. The brass pipes leading to the cylinder trunnions carry the water in and out of the cylinders.

Left: A capstan powered by an Armstrong two-cylinder hydraulic engine.

This is Armstrong's original two-cylinder hydraulic engine. Note there is no flywheel, unlike the three-cylinder engine shown above. The cranks are set at 90 degrees, as in a steam engine, to prevent dead-centre problems.

The control valve is at A, and is operated by a foot-pedal B; push down to start, and release to stop, as the valve lever is weighted to return to the stop position. This provides a limited degree of fail-to-safe operation, which is highly desirable as powered capstans are dangerous things. It also minimised the number of obstructions above ground level, which is important when handling ropes.

This capstan was designed to turn upside down for maintenance, pivoting on trunnions, one of which can be seen in the upper drawing. At first sight this seems like an extraordinary way to construct a capstan. All the literature suggests that hydraulic engines were reliable, while this design seems to imply that they needed almost constant attention. The trunnions would have had to have been substantially built to withstand the forces on the capstan in use, and this must have added to the cost. The swivelling pipe joints would also have cost money, and introduced points of potential leakage.

From Dock & Harbour Engineering by H F Cornick, pub Charles Griffin, 1960

Left: A cutting from Scientific American on Armstrong's introduction of hydraulics.

Confusingly, the hydraulic engine referred to is a double-acting one "applying the water power alternately on both sides of the piston" whereas the Armstrong engines shown here- and all the other drawings I have seen- are very much single-acting. Possibly the reporter simply got it wrong; if not there are some engines that remain to be discovered.

I'm not sure that the bit about "a woman's hand" is strictly Politically Correct.

From Scientific American New Series Vol 10, Iss 4, Jan 1864

Left: A summary of one of Armstrong's patents.

For some reason he is referred to here as George Armstrong rather William George Armstrong.

From Mechanics Magazine July-Dec 1848

When Armstrong established the famous Elswick Works on Tyneside, one of the first major orders was for several hydraulic engines, ordered by his friend Thomas Sopwith for the lead mine at Allenheads; they were used for pumping and line winding in the workings. This was at some point after 1845. Allenheads Heritage Centre now has one of the very few surviving Armstrong hydraulic engines; it was made at the Elswick works in 1852, and originally powered the saw mill at the Allenheads lead mine.

Left: Hydraulic engine at Allenheads Lead mine on the Cumberland/Durham border.

Information on this machine is currently in short supply. It is a horizontal engine allegedly built at Elswick in 1864, which presumably means it is an Armstrong design; however it appears to have a conventional arrangement of connecting rod and fixed cylinders, rather than the oscillating-cylinder engines shown above. It is not possible to see the cylinder arrangement, but there appear to be vertical structures that contain the inlet and exhaust valves.

I am aware the date contradicts that given above. I'm still trying to sort this out.

From SciMusLib

In 1894, Armstrong's Elswick Works constructed and installed the steam-driven pumps, hydraulic accumulators, and hydraulic engines to operate London's Tower Bridge.


Hydraulic engines were widely used on large ships, especially on naval vessels where it was necessary to rotate gun turrets weighing hundreds of tons. In naval language this is called "training" the turrets.

Sir William Armstrong, Mitchell & Co produced a single-crank double-acting engine that powered capstans with a pull of up to five tons. These had a fixed bedplate, avoiding the complications of the turn-over capstan system described above.

Left: An elevation of a turret-training engine installed in HMS Inflexible.

The engines shown here appear to be of the Armstrong type, and are apparently vertically mounted. The three-cylinder engine drives a vertical shaft through bevel gearing, and this shaft carries a pinion that engages with a circular rack around the bottom of the turret. The turret could be rotated in about a minute.
HMS Inflexible was launched in 1876, and eventually scrapped in 1903. She had two gun turrets each weighing 750 tons in total; each turret carried two massive muzzle-loading guns weighing 81 tons each. Loading and ramming were also hydraulically powered.

Apologies for poor picture quality- it's an old drawing.

From Birth of The Battleship by John Beeler, 2001. Original drawing is in the National Maritime Museum at Greenwich, London.

Left: Plan of the turret-training engines in HMS Inflexible.

This makes it clear that there were two hydraulic engines mounted side by side, presumably so gun operation could continue if one broke down.

Part of the hydraulic piping can also be seen. The pressure water was provided by a steam-powered pump.

Once again, apologies for poor picture quality.

From Birth of The Battleship by John Beeler, 2001. Original drawing is in the National Maritime Museum at Greenwich, London.

Some parts in the drawing above are numbered; this is how they are described on the original drawing:

  • 61 From turret turning engines to return pipe
  • 65 Exhaust from turret pivot and main exhaust
  • 66 Pressure to rear(?) of turret presses and exhaust when gun is run in


    Left: The Sinclair Engine.

    This description of a double-acting water engine, with what appears to be oscillating cylinders, was found in a year-book for 1856. So far no other reference to it has been located anywhere.

    From The Year-Book of Facts in Science and Art by John Timbs FSA. Pub David Bogue, London 1856


    Left: A three-cylinder Ramsbottom engine

    The Ramsbottom was one of the most popular water engines in English use. This version has three oscillating cylinders swivelling on a hollow trunnion at the bottom of the frame, and driving a three-throw crankshaft at the top. The trunnion was divided longitudinally by a central partition into inlet and exhaust sides; exhaust and inlet ports to the cylinders were covered and uncovered as the cylinders oscillated.

    From The Hydraulic Age by B Pugh, pub Mechanical Engineering Publications Ltd, 1980

    Mr Knight says "The Ramsbottom engine is largely used in England for operating printing-presses, circular saws, lathes, cranes etc."

    Left: A two-cylinder Ramsbottom-type engine

    This version has a somewhat different construction to the three-cylinder model above, but once again there is a divided trunnion and the exhaust and inlet ports are covered and uncovered as the cylinders oscillate.

    From Knight's American Mechanical Dictionary, 1881.

    The differences are such that it is a bit surprising that the model above is actually a Ramsbottom. For example, the trunnion and its valve-ports are in the middle of the cylinder rather than at the end, and there are conventional pistons rather than plungers. Nonetheless Knight clearly states it to be a Ramsbottom.

    Left: A contemporary article about the Ramsbottom engine: 1865

    Note the high pressure of the water mains in Bradford at the time; 60 to 70 psi.

    From Scientific American Vol 13, Iss13, Sep 1865

    Left: A three-cylinder Ramsbottom engine still in use at Twyford Waterworks

    There are two working Ramsbottom water engines at Twyford Waterworks near Winchester in Hampshire, UK. One is used for mixing lime for water-softening; the other powers a winch used for hauling wagons of chalk from the quarry up an incline to a set of lime kilns. This is believed to be the last site with working water engines in southern England and one of only three sites in the UK. (Allenheads Lead Mines in Northumberland and St Munns church in Scotland being the others)
    The head of water from Twyford Reservoir provides a pressure about 20 psi to drive the engines.

    The red livery is non-authentic, and there are plans to repaint the engine in the correct dark green.


    These instructions on how to run a Ramsbottom engine like that shown above were kindly provided by 'An anonymous operator'.

    "One of the simplest and most efficient engines ever devised, and yet one which has had the least recognition, is the humble water engine. From its inception and development by such notables as Armstrong and Brotherhood in the nineteenth century, the water engine has continued to work with a minimum of fuss and attention. The water engine lacks all the glamour of the steam engine for it makes but little sound, and, due to its simplicity there it very little to go wrong or that which needs tinkering with.

    When running it can be left alone for hours or days at a time with just the occasional visit to check on its oil and grease levels. Even an electric motor, which has the ability to shock, make noise and get warm, has more 'soul' than this engine - however, it would be quite difficult to be killed by a water engine.

    At the start of the day, two minutes before you need the engine to start work, the operator needs to fill the bearing oilers and to refill the crankshaft grease cups. The main valve is opened and the engine will start to move. If the engine has been standing for a few days or longer, the leather cup washers, which seal the ends of the piston against the flow of water, will have dried out and will be leaking. With old, worn washers, this can be quite dramatic and a jet of water can be sent fifteen feet into the air, much to the discomfort of anyone who happens to be standing in the way. After a few revolutions, provided that the washers are not past their useful life, it then settles down: the only sound being the gentle and barely audible rush of water and the clunk of a worn bearing that requires a new shim.

    Cylinder, valve and trunnion lubrication is by water and it is imperative that a little is allowed to leak from around the washers and at the point where the cylinders oscillate. When the engine is in mechanically perfect condition this should equate to a thin trickle of water from the front drain found at the bottom of the engine. This is the only water to be seen when the engine is running as both the inlet and the outflow are piped."

    A highly successful range of radial water engines made in England.

    Left: The Brotherhood radial engine: section

    A three-cylinder radial engine with inlet and exhaust controlled by a rotary valve.

    In 1871 Peter Brotherhood (1838-1902) patented a radial three-cylinder steam engine. The cylinders were at angles of 120 degrees and the connecting rods were linked to a common crank pin. The design was also used as a water engine, as here; and also as a compressed-air motor for driving torpedoes, and as an air compressor.

    From The Hydraulic Age by B Pugh, pub Mechanical Engineering Publications Ltd, 1980

    Left: The Brotherhood radial engine: another drawing

    The rotary valve can be seen at the right of Fig 1. The lower port in the valve, labelled V1, opens the inlet water coming in from the vertical pipe at the bottom to the rather convoluted passage to the cylinder. The upper port V connects the cylinder to the exhaust, the water leaving through the axial pipe going off to the right.

    Note the three-foot long scale at the bottom.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Henry Robinson worked as an assistant to Armstrong, and built the first hydraulic power network, in Hull. We may be confident that he knew what hewas talking about in his book.

    Left: A typical application of The Brotherhood radial engine: driving a capstan. This is the underside of the mounting plate.

    The control linkage operates a balanced mitre valve just above the three-cylinder engine which controls the incoming power water. There are two control pedals, both of which move the linkage in the same direction and start the engine. A powered capstan is a dangerous thing and you need to be able to stop it quickly; here the engine stopped as soon as the pedal was released. There appears to be no way to reverse the engine; presumably this not necessary for this application. You just have to make sure you wind the rope round the right way.

    The water outlet is in the middle of the circular plate mounted on the engine.

    Image from Knight's American Mechanical Dictionary, Supplementary Volume, 1884

    Left: The capstan right way up

    The two control pedals can be seen protruding through the mounting plate.

    Something seems to have gone a bit wrong with the perspective of this drawing.

    Image from Knight's American Mechanical Dictionary, Supplementary Volume, 1884

    To underline that powered capstans are dangerous, in 1903 or thereabouts, an inquest was held on a 19-year old lad who got caught up in a rope being wound onto a hydraulic capstan at the Bishopsgate goods station (external link) in London, a facility that made extensive use of hydraulic capstans for moving wagons. The control pedals apparently jammed and the capstan continued rotating for something like a minute, with the unfortunates victim's head striking an iron stanchion once a second on each revolution; an experience he did not survive. I have to confess that I have mislaid the original source of the this story, and some of the details may be a little off, but the main story is correct. I will find it again one day.

    Left: Another capstan driven by a Brotherhood engine

    The diameter of the capstan head was 26 inches and it was capable of a three-ton pull on a hawser. Note that there are still holes in the top of the capstan to allow the insertion of hand-spikes for manual operation.

    The control lever is weighted, (in fact very heavily weighted) presumably to return it to the 'stop' position when released.

    Image from Knight's American Mechanical Dictionary, Supplementary Volume, 1884. p162

    Left: Traversing saw powered by Brotherhood hydraulic engine (?)

    Designed for cutting metal.

    While Knight makes it clear that the table is traversed by an hydraulic cylinder, he is rather ambiguous as to whether the engine itself is worked by steam or water; the implication is that either can be used.

    Image from Knight's American Mechanical Dictionary, Supplementary Volume, 1884

    The Brotherhood company is still very much alive:


    While a water engine may look very much like a steam engine, its operation is actually rather different. Steam expands in a cylinder, while water, being virtually incompressible, does not. There is no equivalent to an early inlet valve cutoff- if you want the piston to keep moving then the inlet valve must be open. Care is required in setting the valve timings at the end of the stroke- any attempt at compression, as is common in steam engines, would simply result in knocking off the end of the cylinder. Relief valves were often fitted to prevent excessive pressures.
    Because of this, it is difficult to make a water engine that copes well with a variable load. It is straightforward to control the speed- you just use a valve to control the flow rate- but the same amount of water is expended whether the engine is lightly or heavily loaded, so at light loads it would be very wasteful. One solution would be the use of an infinitely-variable transmission between the engine and load, so that while the output shaft turned at the same speed, the engine could run more slowly on a light load; however such transmissions are not simple or cheap. The solution was the variable-stroke water engine, such as the Brotherhood-Hastie engine, or the engine patented by Arthur Rigg. (see below)

    Left: A Brotherhood engine with variable output

    This version incorporates the automatic stroke adjuster introduced by John Hastie, which reacted to the torque the engine was producing. When the load increases, a strong volute spring H inside the output pulley U is wound up, and the rotation is transmitted to a cam F which adjusts the position of the crank pin from the shaft centre. The power output was said by Engineering to be variable over a 3:1 range. This seems like a more useful method of control than that of the Rigg engine below, where the stroke was controlled by output speed rather than torque.

    From A Textbook of Mechanical Engineering, by Wilfred J Lineham, 1912.

    D is the rotary valve controlling water inlet and exhaust.


    Left: A Hastie water engine at Wanlockhead

    The Hastie engine, patented by John Hastie & Co of Greenock, had three cylinders arranged radially about a central crank; this ensued the engine could not get stuck on a dead point. Each cylinder and its plunger oscillated about a bearing at its end, the movement of the cylinder causing ports in its end to open and close appropriately, with water entering and leaving through ducts cast in the frame.

    This rusty Hastie is in the abandoned Lochnell lead mine workings at Wanlockhead in Scotland. Judging by the drum or pulley behind it, it was used for underground haulage. It does not have the variable-output feature described in the section below. The control lever can be seen to the left of the frame.

    The Lochnell mine produced galena (lead sulphide) until it finally closed in 1861.

    Picture courtesy Chris Cowdery

    Left: A Hastie water engine at Wanlockhead

    The other side of the engine, showing the output shaft.

    The Lochnell mine was also drained of water by a hydraulic pump. By 1861 the mine workings had reached a depth of 500 feet below the entrance (called Thomson’s Drift) and closure was caused by the failure of the hydraulic pumping engine to cope with the amount of water entering the workings. Whether this pump was the original Dean engine installed in 1830/31 is not currently known.

    Picture courtesy Chris Cowdery

    Left: A Hastie water engine at Wanlockhead

    None genuine without maker's label.

    Picture courtesy Chris Cowdery

    The Hastie variable output engine was a development of the Hastie engine shown just above, and resembles
    the Brotherhood-Hastie engine above in that a variable crank throw is automatically adjusted to alter the effective gear ratio. The output torque is sensed in the same way by means of heavy springs. The major difference is that the Hastie engine has oscillating cylinders, like the Ramsbottom engine, rather than the trunk pistons and connecting rods of the Brotherhood engines.
    The advantage of the Hastie engine over the Rigg engine shown below is that it has fixed instead of rotating cylinders, eliminating movable pipe joints that are likely to be a source of leakage and extra maintenance.

    Left: The Hastie variable output engine

    The Hastie engine had three cylinders D, disposed radially about the central crank H. Each cylinder and its plunger G oscillated about a bearing at its end. The movement of the cylinder caused ports in its end to open and close appropriately, with water entering and leaving through ducts E cast in the framework.

    A is the water inlet pipe, Q is the exhaust water pipe.

    B is the rotary control valve, with handle C. With the lever moved to either side, the engine would run forwards or backwards. When the lever was vertical, the cylinders were connected only to the exhaust pipe Q, and water was alternately forced into or drawn out of Q, the frictional losses serving as a brake.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Left: The Hastie variable output engine

    Here the engine is used to drive a hoisting drum, and is proportioned for relatively low-pressure water; probably 80 psi. (The London Hydraulic supply was considered high-pressure, being delivered at 800 psi) The engine crank drives the sleeve M, which is mounted coaxially on the output shaft P. M turns the spring-box S, and the springs inside it turn shaft P, which is rigidly connected to the drum.

    The crank pin I was moved in slides by the two cams K, attached to the shaft P. This can be seen in detail below. A drawback of this sort of load-sensing system is that the whole output torque is transmitted by the springs, and they are correspondingly heavy. The spring casing S in the diagram above is almost as big as the engine itself.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Left: The Hastie engine torque sensor

    The central shaft P is the output; the engine turns the outer casing, and when a load is thrown on the output, the chains R are wound up and compress the springs T,T. When their resistance balances the pressure put on them by the load, the spring-box S begins to revolve, carrying shaft P round with it.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Left: The Hastie variable output engine

    This shows the details of the variable-throw crank. The two cams K were fixed on the output shaft P, and acted on the two rollers J and L to move a sliding frame H that constituted the crank.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    When the Hastie engine was run from high-pressure mains with hydraulic accumulators, the springs were replaced by hydraulic rams, connected to the supply by a hole in the central shaft.

    Here are some test results for a Hastie engine working a hoisr with a 22-foot lift, and running from a supply at 80 psi.

    Weight lifted (lbs)Chain only42763374585798910811193
    Water used (gall)7.510141617202122

    This shows effective the Hastie principle was; without the variable output system, lifting just the hoist chain would have also consumed 22 gallons of water.

    The engines manufactured by Arthur Rigg inverted the Brotherhood principle by making the cylinders revolve around a stationary crankshaft, as in a rotary aircraft engine. This apparently perverse construction in fact allowed the engines to be constructed with a variable power output.

    Left: The Rigg variable-speed Engine. (Fig 3)

    The Rigg engine was an ingenious solution to the problem of power delivery into varying loads. This is a 30-HP engine capable of using water at 700psi. The two straight arms inside the output pulley are part of a centrifugal governor.

    Left: The principle of the Rigg variable-speed Engine. (Fig 4)

    Versions with three or four cylinders were produced. This is a three-cylinder engine.

    Three cylinders were pinned to a centre C, with their piston rods connecting to the periphery of a large flywheel which rotated around a fixed axis at D. The centre C could be moved to vary the eccentric distance C-D, and hence the piston stroke, which was twice C-D. The engine could be stopped by moving C in line with D, when the stroke was zero, or reversed by moving C to the other side of D.

    Water was supplied to the cylinders through a hollow shaft at C, with water distribution controlled by ports in the central boss of the cylinder block.

    In small engines the movable centre C was moved by a screw and handwheel, but larger versions used a hydraulic servo in which a small control valve determined the position of a powerful stroke-control hydraulic ram fixed to C. This could be controlled by a governor to maintain a constant speed. This approach differs from the Brotherhood-Hastie engine which was governed by output torque and not output speed.

    Left: Side elevation of a four-cylinder Rigg engine.

    The drawings here from the Robinson book show a Rigg engine that was used for winding waggons up a steep incline. It was worked with water at 200psi.

    Each cylinder was a separate gun-metal casting. The pistons were 6.5 inches in diameter, with a maximum stroke of 9 inches.

    The valve ports can be seen grouped around the centre bearing.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Left: Side elevation of the Rigg variable-speed Engine.

    Showing the outer casing of the stroke-control ram with the control valve on top of it.

    From Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Left: Side elevation of the Rigg variable-speed Engine.

    Showing the stroke-control ram in section. This makes it clear how water got in and out through the movable centre bearing. Water enters at the left, and flows through the hollow left plunger of the ram to the valve ports around centre C. The exhaust water exits through the lower valve port and drops through the open bottom of the ram, and runs to waste through the hole in the bottom of the bedplate.

    Note that the right hand plunger has twice the area of the left-hand one. Thus when water was applied by the control valve to the left plunger only, the ram pushes to the right. When water was applied to both plungers, the right-hand one overpowered the left, and the effective force was to the left.

    from Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Left: Plan view of the Rigg variable-speed Engine.

    The stroke-control ram is in the lower part of the drawing.

    from Hydraulic Power & Machinery by Henry Robinson. Pub Charles Griffin & Co, London 1893

    Below is an edited extract from an article on hydraulic power in the 11th edition of Encyclopaedia Britannica (1911); the Rigg engine was clearly well-known at the time.

    Direct-acting Water Motors.
    "Owing to the difficulty of securing a durable motor with a simple and trustworthy means of automatically regulating the quantity of water used to the power needed at various times from the motor, not much advance has been recently made in the use of water motors with reciprocating rams or pistons. Probably the most successful one has been a rotary engine invented by Mr Arthur Rigg. In this engine the stroke, and therefore the amount of water used, can be varied either by hand or by a governor while it is running; the speed can also be varied, very high rates, as much as 600 revolutions a minute, being attainable without the question of shock or vibration becoming troublesome.
    The cylinders are cast in one piece with a circular valve, and rotate about a main stud C (Fig 4), while their plungers are connected to a disk crank which rotates above the point D, which is the centre of the main crank; C-D being the crank length or half stroke of the engine, any variation in its length will vary the power of the engine and at the same time the quantity of water used. The movement of C is obtained by means of a relay engine, in which there are two rams of different diameters; a constant pressure is always acting on the smaller of these when the motor is at work, while the governor (or handpower if desired) admits or exhausts pressure water from the face of the other, and the movements to and fro thus given to the two rams alter the position of the stud C, and thus change the stroke of the plungers of the main engine. Fig 3 gives an outside view of a 30-HP engine capable of using water at a pressure of 700psi; the governor is carried within the driving pulley shown at the right-hand end, while the working revolving cylinders are carried inside the boxed-in flywheel at the left-hand end, the relay cylinder and its attachments being fixed to the bed-plate in front of the flywheel. On a test one of these engines gave an efficiency or duty of 80%."

    The Rigg engine was also referred to in a paper by the eminent engineer H S Hele-Shaw, "Theory of a New Form of the Chamber Crank Chain" (Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character, Vol. 87, No. 592 (Jul. 26, 1912), pp. 62-69.
    Note that chain here refers to a kinematic "chain" not the physical thing you find on a bicycle)
    You can read it here:
    Proceedings of the Royal Society 1912.

    According to a history of Chester: "Economic conditions remained difficult during the 1880s, and in 1889 Arthur Rigg's Victoria Engine Works failed." Since his engine was still being discussed in 1912, he presumably bounced back.

    Left: Mr Arthur Rigg surfaces at a meeting of The Society of Engineers: 1901.

    This was a discussion about the drainage of Ilford, East of London. Note his reference to revolving hydraulic engines for driving sewage pumps.

    from Transactions of the Society of Engineers 1901


    Left: Bad news for Mr S.B. of New York: 1856.

    from Scientific American Vol 11, Iss 3, Apr 1856

    Left: A Water Engine in Stirling, Scotland: 1857.

    I can vouch from personal experience that Stirling Castle is on top of a substantial hill.

    from Scientific American Vol 12, Iss 42, Jun 1857

    Left: A Water Engine down a British mine: 1857.

    This is the first reference I have come across to the use of water engines for wagon haulage underground.

    from Scientific American Vol 12, Iss 49, Aug 1857

    At the turn of the century some houses were lighted with a mixture of petrol vapour and air. This had to be compressed to give a workable density for burning, apparently to "always greater than 0.4 of an atmosphere" though this seems like a very high pressure compare with that of town gas.

    Left: The Keith water-powered compressor, small size: 1900.

    Conventional reciprocating compressors created pulsations that made the lights flicker, and required elaborate pressure regulators. Domestic installations did not want these complications, and the Keith company avoided them by introducing a combined compressor/gas-holder driven by mains water pressure. A small Keith compressor could supply 100 - 150 ft3 of compressed gas per hour, and the large size 390 - 495 ft3 per hour, depending to the water pressure available. A great advantage of the Keith design was that it had no external stuffing boxes, eliminating a major cause of dangerous gas leaks.

    Operation was fully automatic. The inner cylinder is connected to a water control valve in the supply pipe at the right.

    Scientific American Supplement, 9 June 1900, p20433

    By 1912 Keith had apparently joined forces with a Mr Blackman, and James Keith and Blackman Co Ltd were still making water-powered petrol-air gas compressors used for domestic lighting purposes. A contemporary description reads: "...the pressure of the water actuates a ram attached through a stuffing-box to the compressing holder. The two ends of the ram are of different diameters so that the pressure on the larger ram overcomes that on the smaller ram, and the working is continuous so long as water is supplied."

    Info from "Petrol-Air Gas" by Henry O'Connor. Second edition 1912.

    The Keith and Blackman company still exists but is now a part of Woods Air Movement Ltd of Colchester; see here (external link)

    The lighting of houses with a petrol-vapour/air mixture appears to have been a relatively short-lived business. If you lived in the English countryside around the 1900's, out of reach of mains gas and electricity, there were several options for lighting your house, none of them very alluring. Steam-engine-driven dynamos required a boiler which demanded constant attention from a competent and reliable (and therefore reasonably well-paid) employee; even then there was always the risk of an explosion. Oil lamps also required wick-trimming and refilling and were dangerous if knocked over.

    The essence of the petrol-air approach was to have a gas producing plant that required minimal attention. To this end petrol-air gas plants were powered by:

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