Article - Issue 8, May 2001

The properties of steam: industrial engines before 1800

David Hulse

Download the article (1203 KB)

The cutting-edge technology of the 18th century was the atmospheric engine, driven by the pressure of the earth’s atmosphere helped along by steam. These engines first powered the pumps that removed water from mines but later became the major source of power for industries of all kinds. David Hulse takes a look at a 300-year-old breakthrough.

One of the most important industrial problems of the late 17th century was the drainage of mines, particularly coal mines. Coal had replaced wood as the major source of fuel for trades such as glass- and soap-making and metal working and demand for it was high. So coal mines were pushed to greater and greater depths and it became more and more difficult to remove the water that accumulated in them.

The water-extraction problem would be solved as soon as someone found a cheap source of mechanical power to work the water pumps. During the 16th and 17th centuries mines were drained mostly by horse power, sometimes requiring teams of horses to work the pumps at a single mine.

In 1663/4 when the water-extraction crisis was reaching its height, Thomas Newcomen was born at Dartmouth. After an apprenticeship as an ironmonger he began selling tools to the Cornish mining industry. On his visits to the mines he began to realise what a great financial reward there would be if he could develop a mechanical means for removing the flood water: ore could then be extracted much more quickly and safely from below the natural drainage level of the mines.

With his assistant John Calley, Newcomen experimented for many years, trying to harness the properties of steam. His work came to fruition in the year 1712 when he was able to demonstrate the first atmospheric engine, pumping water from a coal mine near Dudley Castle in South Staffordshire.

The eighteenth-century engines were powered by the weight of the earth’s atmosphere, acting against a vacuum that was created on the underside of the piston by the condensation of steam (see Figure 1). The engines were powered in this way because, in the early eighteenth century, technology had not advanced enough to make boilers which would restrain the forces of steam raised to a high pressure. Large, safe, steam-tight vessels simply could not be made at that time.

The first commercial steam engine

The Newcomen engine of 1712 had a boiler made from copper with a hemispherical dome made from beaten lead. The engine made 12 pumping movements in each minute, and on each movement brought 10 gallons of water from the mine workings 150 feet below to drain safely away at the surface. The powering cylinder, positioned vertically and made from brass, was 21 inches in diameter. The piston made a working stroke of almost 8 feet.

When the engine was set in motion at the beginning of the working period, the steam would have been raised within the boiler to a pressure of 1.5–2.0 pounds per square inch. With the overhead rocking beam of the engine stationary and the piston as near as possible to the boiler top, catches would have been released and the heavy pump rods at the other end of the rocking beam would then draw the piston to the other end of the vertical cylinder. The suction created by the ascending piston drew the steam into the cylinder from within the boiler underneath it. With the cylinder full, a spray of cold water would condense the steam and, on the formation of a vacuum, the piston would be drawn to the other end of the cylinder, completing a working stroke. The pump rods with the heavy weights attached then returned the piston to the top of the cylinder and the cycle continued.

The accumulated water from the condensing spray controlled the working speed of the engine: on each stroke of Newcomen’s engine about six gallons of water had to be removed from the powering cylinder before the piston could complete its full movement.

The Smethwick engine

Engines worked on the principle described above until James Watt patented the separate condenser in 1769. Watt created his vacuum in a water-cooled vessel away from the powering cylinder. The heating and then cooling of the powering cylinder on each stroke of the engine was eliminated: this single improvement doubled the power output of the engine for the same amount of coal consumed.

A Boulton and Watt pumping engine was built for the Birmingham Canal Navigation Company in 1779 and was in service on the canal until 1891. This engine raised water a distance of 38 feet from the bottom to the top lock on the canal and was situated at the top of Bridge Street in Smethwick. On each stroke of the engine 229 gallons of water were raised. The engine is to be displayed at the Think Tank in Birmingham: it will be the oldest working steam engine in the world.

The Smethwick engine was powered by low-pressure steam on the top of the piston and a vacuum on the underside. It represented a new generation of pumping engine; these engines were quickly accepted by the Cornish mining industry. Engines working with separate condensers and using the expansive force of steam developed more power. They were able to remove flood water from mines to a much greater depth than had been possible with the Newcomentype engine. The new design was so successful that the 75 Newcomen engines which were working in the Cornish mineral mines in 1779 had all been superceded within four years.

The Boulton and Watt organisation only sold their improved steam engines on an agreement with the customer that an annual premium was to be paid. This premium was based on calculating the amount of coal which would have been used had the customer installed a standard Newcomen-type engine. One third of the cost of the difference in fuel had to be paid annually to the Boulton and Watt organisation. For the Smethwick engine, the Canal Company paid £210 per year.

Rotary motion

The early steam engines produced linear motion, that is, the up and down movement of the pump rod used to move water. In order to power other machinery the engines would need to produce continuous rotary motion. This was not as easy to achieve as might be imagined.

The first engine in the world successfully to produce rotary motion, by the use of a flywheel and crank, was designed by Matthew Wasborough (a Bristol engineer) for James Pickard, a Birmingham manufacturer. It was originally installed in 1779 and at this time the rotary motion was achieved by a ratchet and paul arrangement. However, this proved to be unsuccessful and the engine was later fitted with the flywheel and crank mechanism. Pickard and Wasborough obtained a patent for their idea in 1780.

The granting of this patent started a long controversy, with the result that James Watt had to use a different arrangement (the ‘Sun and Planet’ method) to achieve rotary motion on his engines until the Pickard patent expired in 1792.

The original Newcomen-type engine was assembled at Snow Hill in 1779 and after its conversion to use a crank and flywheel the engine continued to work until about 1879, driving a mill for grinding metals. The model in Figure 5 represents what this engine most probably looked like. All that is known is that the powering cylinder had a diameter of 30 inches and the throw of the crank was 3 feet and 7 inches. The powering cylinder was positioned vertically above the haystack boiler, which is unusual for the late 1770s: by this date boilers were usually positioned in a separate building adjoining the main engine house.

Boulton and Watt’s Lap engine

In 1788 James Watt designed and built an engine that would directly produce rotary motion. By this time many other beam engines had been adapted to provide rotary motion, but they had not been designed and built with this aim in mind and were usually adaptations of single-acting pumping engines. Watt’s new engine was powered by vacuum and also by the expansive force of steam. The vacuum was first directed to the top and then to the underside of the piston; low-pressure steam was also applied to the piston on both up and down strokes. This combination of vacuum and low-pressure steam produced a continuous output of energy which powered the engine.

The new rotative beam engine was erected at Boulton and Watt’s Soho Manufactory in Birmingham where it became known as the Lap engine because it was used to lap and polish small manufactured components, such as the large buckles worn on court shoes. The large flywheel of this engine, almost 16 feet in diameter with 304 wooden teeth, drove more than forty lapping and polishing machines and the rotary power was transmitted to each individual machine by a system of belts and pulleys.

The engine was the first in the world to have its rotational speed controlled by a centrifugal device which later became known as the ‘Watt Governor’. The Lap engine was one of the first engines to have its power output rated in horsepower – calculated by James Watt to be 10. It provided the rotary drive to the factory machinery for 70 years until the Soho Manufactory ceased production in 1858. The engine is now on display at the Science Museum in London.

The Arnold Mill engine

This engine provided the rotary power to drive Robert Davison and John Hawksley’s worsted mill at Arnold, Nottingham, in 1797. The engine, designed by an engineer from Ashover in Derbyshire called Francis Thompson (1747–1809), has some unusual features intended to achieve good performance without infringing a patent held by James Watt on steam engine design. The most outstanding of these features is the double-acting cylinder – the powering force of the engine, achieved by the condensation of steam.

Francis Thompson used the principles first pioneered by Thomas Newcomen in 1712. It is believed that only eight double-action rotary engines were ever made that operated by atmospheric pressure in this way. Two cylinders were used, positioned vertically one above the other. The upper cylinder contained a piston that was pushed vertically up by the atmospheric pressure. The piston in the lower cylinder was pushed vertically down, again by the atmospheric pressure. The powering force of this engine was a vacuum created by a water spray condensing the steam within each cylinder.

The combined force from the two connected pistons was transmitted on to the main oscillating beam of the engine by three chains, all adjusted to be in tension. Francis Thompson achieved rotary motion on this engine by fitting a large crank pin on the side of the countershaft gear.

When the mill at Arnold was first built in the early 1790s, a waterwheel was the only source of power for the factory. With the outbreak of the Napoleonic Wars the demand for worsted for military uniforms increased. In order to meet this increased demand the mill started working throughout the day and night. However after two complete days of non-stop working, the mill pond had emptied and production had to stop. To maintain production for the remainder of the working week the atmospheric engine was used, thus allowing the mill pond to refill.

When the mill pond had refilled, the spinning and weaving machines were again driven by the waterwheel. A line shaft 105 feet long connected the two independent drives and large flanged couplings were engaged or disengaged to select the drive which was to be used. The waterwheel and the engine were never used together.

The engine worked at 18 strokes a minute, driving the factory countershaft at 50 revolutions per minute, and is thought to have generated about 45 horsepower. The original engine stood 47 feet tall, tall enough to pass through all five floors of the textile mill. The mill was demolished in 1811 and the engine was dismantled and sold for scrap; all that now remains is the mill pond.

A high-pressure steam engine

The pre-1800 engines were powered by a vacuum; this vacuum was created by the condensation of low-pressure steam. Our story ends with the development around 1800 of engines operated by high-pressure steam. One of the first of these was designed by Richard Trevithick, a Cornish engineer, and installed in 1804 to drive the machinery of a dye house at Lambeth in London.

The cylinder which powered Trevithick’s 1804 engine was a completely new innovation as it was in a horizontal position. It is the first of a new generation of engines which became known as ‘high-speed horizontal engines’.

The boiler supplying the steam to the engine was 6 feet in diameter and was originally made from cast iron, cast at Abraham Darby’s foundry in Coalbrookdale. The engine was powered by a double-acting cylinder 8 inches in diameter with a working stroke of 48 inches. It was rated at 6 horsepower. The high-speed engine made 24 revolutions per minute when working at a boiler pressure of approximately 45 pounds per square inch.

So in just under 100 years the state-of- the-art in industrial power sources had moved on from horse or water power, through the power of low-pressure steam and atmospheric pressure, to steam produced at 25 times the pressure of the early engines. One generation’s technical impossibility – the large, steam-tight boilers that could not be made in 1700 – became another generation’s staple source of power.

David Hulse

David Hulse retired in 1998 from his position as Chief Development Engineer of the Royal Doulton group of potteries. He holds 17 patents for his designs for automated machinery used in pottery making. During his spare time he has researched and constructed in miniature the most important of the early steam engines. He is available to give illustrated talks on the early engines; for more details visit the web site Copies of David Hulse’s book on the development of the steam engine during the 18th century can be obtained from the author at 133 Oulton Road, Stone, Staffordshire, ST15 8DS for £12.50 (including postage and packing). His second book, describing how beam engines provided rotary motion to industries, is published in October 2001. Email:

[Top of the page]