The History of Groundwater Theory and Practice


Man has been aware of groundwater since prehistory, long before Biblical times. Over the centuries, the mysteries of groundwater have been solved, and man has developed an increasing capability to manipulate it to his will. This chapter describes some of the key stages in the development of the understanding and control of groundwater. The history of some of the technology now used for groundwater lowering is also discussed, especially in relation to early applications in the United Kingdom. Detailed knowledge of the history of groundwater control might not be considered essential for a practical engineer working today. Nevertheless, study of the past can be illuminating, not least by showing that even when theories are incomplete and technology untried, the application of scientific principles and engineering judgement can still allow groundwater to be controlled.


The digging of wells for the exploitation of water and primitive implementation of water management date back to Babylonian times and even earlier.

The source of the water flowing from springs and in streams was a puzzling problem and the subject of much controversy and speculation. It was generally held that the water discharged from springs could not be derived from rain. Ingenious hypotheses were formulated to account for the occurrence of springs. Some early writers suggested large inexhaustible reservoirs, while others recognized that there must be some form of replenishment of the supplying reservoirs. The Greek hypotheses, with many incredible embellishments, were generally accepted until near the end of the seventeenth century. The theory of rainfall infiltration was propounded by only a very few writers.

Though the Romans and other early cultures indulged in quite sophisticated water management projects, building imposing aqueducts to channel water from spring and other sources to centres of population, they had no understanding of the sources of replenishment of groundwater. The aqueducts of the Romans were remarkable and showed great appreciation of the value of water, but their methods of measuring or estimating water quantities were crude. Generally, they appear to have lacked any semblance of knowledge of either surface or groundwater hydrology.

According to Tolman (1937),

Centuries have been required to free scientists from superstitions and wild theories handed down from the dawn of history regarding the unseen sub-surface water, ... even in this century there is still much popular superstition concerning underground water.

An elemental principle, that gravity controls the motion of water underground as well as at the surface, is not appreciated by all. A popular belief exists that ‘rivers’ of underground water pass through solid rock devoid of interconnected interstices and flow under intervening mountain ranges.

Marcus Vitruvius, a Roman architect who lived about the time of Christ, produced a book describing the methods of finding water. He wrote of rain and snow falling on mountains and then percolating through the rock strata and reappearing as springs at the foot of the mountains. He gave a list of plants and of other conditions indicative of groundwater, such as colour and dampness of the soil and mists rising from the ground early in the morning. Vitruvius and two contemporaries of his, Cassiodorus and Pliny, were the first to make serious efforts to list practical methods for locating water, and this when geology was yet unknown!


At the beginning of the sixteenth century, Leonardo da Vinci directed his attention to the occurrence and behaviour of water. He correctly described the infiltration of rain and the occurrence of springs and concluded that water goes from rivers to the sea and from the sea to the rivers and thus is constantly circulating and returning. About the same time, Palissy, a French Huguenot, presented a clear and reasoned argument that all water from springs is derived from rain.

The latter part of the seventeenth century was a watershed in the beginnings of an understanding of the replenishment of groundwater. Gradually, there arose the concept of a ‘hydrological cycle’. This presumed that water was returned from the oceans by way of underground channels below the mountains. The removal of salt was thought to be either by distillation or by percolation, and there were some highly ingenious theories of how water was raised up to the springs.


It was in the seventeenth century that the quantitative science of hydrology was founded by Palissy, Perrault and Mariotte in France, Ramazzini in Italy and the astronomer Halley in Britain.

Palissy, a sixteenth-century potter and palaeontologist, stated that rain and melt snow were the only source of spring and river waters, and that rain water percolates into the earth, following ‘a downward course until they reach some solid and impervious rock surface over which they flow until they make some opening to the surface of the earth’.

Perrault made rainfall run-off measurements and demonstrated the fallacy of the long-held view that the rainfall was not sufficient to account for the discharge from springs. He also measured and investigated evaporation and capillarity. Mariotte verified Perrault’s results and showed that the flow from springs increased in rainy weather and decreased in times of drought. Halley made observations of the rate of evaporation from the Mediterranean ocean and showed that this was adequate to supply the quantity returned to that sea by its rivers. His measurements of evaporation were conducted with considerable care, but his estimates of stream flow were very crude.

Towards the end of the eighteenth century, La Metherie extended the researches of Mariotte and brought them to the attention of meteorologists. He also investigated permeability and explained that some rain flows off directly (surface water run-off), some infiltrates into the top soil layers only and evaporates or feeds plants, whilst some rain penetrates underground, whence it can issue as springs (i.e. infiltration or groundwater recharge). This is the first recorded mention of ‘permeability’ and so is the first link between hydrology and seepage to wells.

Seepage towards Wells

The robust Newcomen engine greatly influenced mining practice during the eighteenth century, but it was far too cumbersome for construction works. Generally speaking, until the early nineteenth century, civil engineers, by the use of timber caissons (Figure 2.3) and other devices, avoided pumping whenever possible. However, where there was no alternative, pumping was usually done by hand - a very onerous task - using a rag and chain pump (Figure 2.1), known also as ‘le chaplet’ (the rosary).

The use of pumps for mining in the sixteenth century is described in two seminal references by Agricola and Ramelli. De Re Metallica (Latin for On the Nature of Metals) by Georgius Agricola (translated by Hoover and Hoover, 1950) was a review of the state of art in mining and remained a key text on mining for nearly two centuries. Agricola describes several designs of piston pumps, which are either man or animal powered or powered by waterwheels. The suction lift limits of these pumps meant that multiple stages of pumps were required for the deepest mines. Le diverse et artificiose machine del Capitano Agostino Ramelli (1588) translates as The Various and Ingenious Machines of Captain Agostino Ramelli. Ramelli, an Italian military engineer, spent the latter half of his life and career in France as Ingenieur du Roi (Engineer to the King) in the service of Henri III and Henri IV. His book describes several machines and methods to pump water from excavations (Figure 2.2) and even includes the use of timber cofferdams to exclude groundwater (Figure 2.3).

Some idea of the magnitude of pumping problems is given by de Cessart in his book Oeuvres Hydrauliques. Speaking of the foundations for the abutment of a bridge at Saumur in 1757, he says that 45 chain pumps were in use, operated by 350 soldiers and 145 peasants. Work on this type of pump was, of course, most exhausting, and the men could only work in short spells. Pryce, in his Mineralogea Cornubiensis in 1778, said that work on pumps of this sort led to a great many premature deaths among Cornish miners.

For permanent installations such as graving docks, large horse-driven chaplet pumps were used. Perronnet, the famous French bridge builder, made use of elaborate pumping installations in the cofferdams for the piers of his larger bridges; for example, under-shot water wheels were used to operate both chaplet pumps and Archimedean screws.

According to Cresy, writing in 1847, the first engineer to use steam pumps on bridge foundations was Rennie, who employed them on Waterloo Bridge in 1811. In the same year, Telford, on the construction of a lock on the Caledonian Canal at Clachnacarry, at first used a chain pump worked by six horses but replaced it by a 9 horsepower steam-engined pump. From then on, steam pumps were used during the construction of all the principal locks on that canal. In 1825, Marc Brunei used a 14 horsepower steam engine when sinking the

Rag and chain pump, manually operated, 1556

Figure 2.1 Rag and chain pump, manually operated, 1556. (From Bromehead, C N, Mining and quarrying in the seventeenth century. In A History of Technology (Singer, C, Holmyard, E J, Hall, A R and Williams, T L, eds), 1956, by permission of Oxford University Press.) The balls, which are stuffed with horsehair, are spaced at intervals along the chain and act as one-way pistons when the wheel revolves.

shafts for the Thames Tunnel. By this date, steam pumping seems to have been the common practice for dealing with groundwater, so below-ground excavations for construction were less problematical.

Land Drainage in the Eighteenth and Nineteenth Centuries

Notable contributions to British practice from the later eighteenth century through to the mid-nineteenth century were the work of British land drainers such as Joseph Elkington, John Wedge and James Anderson (Stephens and Stephens, 2006). Their approach involved using an understanding of shallow groundwater conditions to drain waterlogged ground. The methods they developed are comparable to sumps and relief wells (to bleed ground- water pressure from confined layers), which would be recognized by practitioners today. Interestingly, because of the successful and widespread application of his knowledge to horizontal and vertical drainage in the latter part of the eighteenth century, Stephens and Ankeny (2004) make the case that Joseph Elkington was perhaps the first professional consulting hydrogeologist.

Device for two men to drain water from a foundation

Figure 2.2 Device for two men to drain water from a foundation: mechanics: hand crank; force pump. (Plate 103 from Ramelli, A., Le Diverse et Artificiose Machine del Capitano Agostino Ramelli, Paris, France, 1588. With permission from Science History Institute.)

Kilsby Tunnel, London to Birmingham Railway

For civil engineering excavations, there appear to have been no important advances in pumping from excavations until the construction in the 1830s by the renowned civil engineer Robert Stephenson of the Kilsby Tunnel south of Rugby on the London to Birmingham Railway (Preene, 2004). He pumped from two lines of wells sited parallel to and on either side of the line of the tunnel drive (Figure 2.4).

It is clear from Stephenson’s own Second Report to the Directors of the London, Westminster and Metropolitan Water Company, 1841, that he was the first to observe and explain the seepage or flow of water through sand to pumped wells. The wells were sited just outside the periphery of the construction so as to lower the groundwater level in the area of the work by pumping from these water abstraction points. This is most certainly the first temporary works installation of a deep well groundwater lowering system in Britain, if not

Cofferdam used to exclude water from an excavation

Figure 2.3 Cofferdam used to exclude water from an excavation. Workmen use buckets to drain water from the section walled off by the cofferdam in order to secure foundations for construction (Plate 111 from Ramelli, A., Le Diverse et Artificiose Machine del Capitano Agostino Ramelli, Paris, France, 1588. With permission from Science History Institute.)

in the world. The following extract (from Boyd-Dawkins, 1898, courtesy of the Institution of Civil Engineers Library) quotes from the report and shows that Stephenson had understood the mechanism of groundwater flow towards a pumping installation:

The Kilsby Tunnel, near Rugby, completed in the year 1838, presented extreme difficulties because it had to be carried through the water-logged sands of the Inferior Oolites, so highly charged with water as to be a veritable quicksand. The difficulty was overcome in the following manner. Shafts were sunk and steam-driven pumps erected in the line of the tunnel. As the pumping progressed the most careful measurements were taken of the level at which the water stood in the various shafts and boreholes; and I was soon much surprised to find how slightly the depression of the water-level in the

Pumps for draining the Kilsby Tunnel

Figure 2.4 Pumps for draining the Kilsby Tunnel. (From Bourne, J C, Drawings of the London and Birmingham Railway, 1839, with permission from the Institution of Civil Engineers.) A pumping well is shown in the foreground, with the steam pumphouse in the distance.

one shaft, influenced that of the other, notwithstanding a free communication existed between them through the medium of the sand, which was very coarse and open. It then occurred to me that the resistance which the water encountered in its passage through the sands to the pumps would be accurately measured by the angle or inclination which the surface of the water assumed towards the pumps, and that it would be unnecessary to draw the whole of the water off from the quicksands, but to persevere in pumping only in the precise level of the tunnel, allowing the surface of the water flowing through the sand to assume that inclination which was due to its resistance.

The simple result of all the pumping was to establish and maintain a channel of comparatively dry sand in the immediate line of the intended tunnel, leaving the water heaped up on each side by the resistance which the sand offered to its descent to that line on which the pumps and shafts were situated.

As Boyd-Dawkins then comments,

The result of observations, carried on for two years, led to the conclusion that no extent of pumping would completely drain the sands. Borings, put down within 200 yards [185 m] of the line of the tunnel on either side, showed further, that the water-level had scarcely been reduced after 12 months continuous pumping and, for the latter six months, pumping was at the rate of 1,800 gallons per minute [490 mVh], In other words, the cone of depression did not extend much beyond 200 yards [185 m] away from the line of pumps.

In this account, ... it is difficult to decide which is the more admirable, the scientific method by which Stephenson arrived at the conclusion that the cone of depression was small in range, or the practical application of the results in making a dry [the authors would have used the word ‘workable’ rather than ‘dry’] pathway for the railway between the waters heaped up [in the soil] ... on either side.

If is astonishing that neither Robert Stephenson nor any of his contemporaries realized the significance of this newly discovered principle: that by sinking water abstraction points, and more importantly placing them clear of the excavation so that the flow of water in the ground would be away from the excavation rather than towards it, stable ground conditions were created. For many decades, this most important principle was ignored.

Early Theory - Darcy and Dupuit

In the 1850s and early 1860s, Henri Darcy made an extensive study of the problems of obtaining an adequate supply of potable water for the town of Dijon (Freeze, 1994). He is famous for his Darcy’s law (Darcy, 1856) postulating how to determine the permeability of a column of sand of selected grading when the rate of water flow through it is known (Figure 2.5). In fact, Darcy’s experiments formed only a small part of his treatise. He compiled a very comprehensive report (two thick volumes) in which he analysed the available sources of water from both rivers and wells - some of them artesian - and how economically to harness all these for optimum usage.

Darcy investigated the then current volume of supply of water per day per inhabitant for about 10 municipalities in Britain - Glasgow, Nottingham and Chelsea, among others - as well as Marseille, Paris and many other French towns. He concluded that the average water provision in Britain was 80-85 litres/inhabitant/day and more than 60 litres/inhabitant/ day for Paris. Darcy designed the water supply system for Dijon on the basis of 150 litres/ inhabitant/day - no doubt his Victorian contemporaries this side of ‘la Manche’ would have applauded this philosophy.

In the mid-1860s, Dupuit (1863), using Darcy’s law to express soil permeability, propounded his equations for determining flow to a single well positioned in the middle of an island. Dupuit made certain simplifying assumptions, and having stated them (i.e. truly horizontal flow), he then discounted their implications! For this, Dupuit has been much castigated by some later purists, but most accept that the Dupuit concept, later slightly modified by Forchheimer, is acceptable and adequate in many practical situations.

Exchange of information was not as simple in Darcy’s and Dupuit’s time as it is now. Much of the fundamental work of these two French engineers was duplicated by independent developments shortly afterwards in Germany and Austria and a little later in the United States.

In about 1883, Reynolds demonstrated that for linear flow - i.e. flow in orderly layers, commonly known as laminar flow - there is a proportionality existing between the hydraulic gradient and the velocity of flow. This is in keeping with Darcy’s law, but as velocity increased, the pattern of flow became irregular (i.e. turbulent), and the hydraulic gradient approached the square of the velocity. Reynolds endorsed the conclusion that Darcy’s law gives an acceptable representation of the flow within porous media - i.e. the flow through the pore spaces of soils will remain laminar save for very rare and exceptional circumstances. However, this may not always be true of flow through fractured rocks (e.g. karstic limestone formations).

< Prev   CONTENTS   Source   Next >