This invention relates to barriers of the kind that are pile-driven into the ground.

It is common practice to drive sheet-metal pilings into the ground, to provide a barrier. It is common practice to provide the sheet-metal piling elements with complementary edge forms, whereby the elements, when in place, interlock with each other to provide a strong mechanical tie between adjacent elements.

In many applications, there is a requirement that the barrier should not leak. Edge forms have been developed whereby the interlock between adjacent sheets can be made to resist or prevent the throughflow of water.

In some applications - for shoring when excavating foundations for a new building for example - it is not necessary that the barrier be completely leakproof, but merely that the inflow of water be reduced to a value which the pumps can cope with.

In other cases - where the purpose of the barrier is to contain a spill of a toxic chemical, for example - the requirement that the barrier be completely leakproof is much more stringent.

It is a very demanding task for the designer to provide an edge form between the pile-driven elements which not only is economical, but is also capable of providing a leakproof seal over the whole height of the barrier, especially considering that the seal facility must survive being pile-driven into the ground.

Patent publication no GB-2228760 provides an example of a barrier formed of pile-driven elements, which is leak-resistant to a very high standard.

GENERAL FEATURES OF THE INVENTION

In those cases where a leakproof seal of a very high standard is demanded, there is generally also a requirement for some means of monitoring the barrier, and for confirming whether or not the barrier is leaking and allowing contaminated water through, even at a very small rate. That is to say, that even an absolutely perfectly watertight barrier might not be satisfactory if the barrier did not provide the authorities with some means of determining that it was indeed watertight.

In some cases, an amount of monetary damages might depend on whether it can be shown that no toxic chemical can have leaked through a barrier. Here especially, it is an advantage, in terms of evidence value, if the barrier not only is leakproof but can be seen to be leakproof.

The invention is aimed at providing a barrier which is leakproof to a very high standard, and also which is amenable to being readily monitored for leaks.

In the invention, the piling elements are so arranged as to form a two-wall barrier, with a space defined between the two walls. Preferably, each wall is made leakproof to a very high standard.

The space between the walls preferably is divided into separate compartments, and preferably the compartments are also made leakproof relative to each other to a very high standard.

When the barrier is in place, any contained water may be pumped out or partially pumped out of the compartments. An on-going monitoring exercise may be undertaken, in which the level of the water table inside the compartment is periodically checked, along with periodic analysis of samples to determine whether contaminants are present.

With a conventional barrier, it is usually very difficult to determine whether the barrier is leaking. But apart from that, one of the key difficulties in respect of leakproof barriers of the conventional type lies in the fact that it is almost impossible (economically) to re-seal a barrier that has developed a leak. Sometimes, the only economical solution to the problem of a leaking barrier is to provide a complete new barrier.

The double-wall structure of the invention, with the compartments, provide a means by which an authority can determine whether a barrier is leaking. Periodic measurements can be carried out of the height of the water table within a compartment, since a change in the water table within the compartment would indicate a leak. Also, samples can be taken of the water within the compartment to see whether any contaminant is present. If desired, a tracer can be added to the water in the aquifer immediately outside the barrier, to make it easier to confirm the presence or absence of a leak.

The invention also provides a means whereby a leak can be rendered harmless.

If the water table is seen to change in one of the sealed compartments, this is an indication that a leak has developed in that compartment. In that case, the engineer can arrange to lower the water table in that compartment, by pumping water out of the compartment. Because the compartment is leaking, this pumping out will have to be repeated periodically.

In the event of a leak, the water table within the compartment should be lowered to a level below the water table in the surrounding aquifer; the effect of reducing the water table inside the compartment is that, even though the compartment is leaking, the flow of water through the leak is constrained to be inwards into the compartment, and not outwards.

The water table within the compartment should be lower than both the water table in front of, and the water table behind, the barrier. The fact of the presence of the waterproof barrier of course will tend to make the "front" water table A differ substantially from the "back" water table B. If the water table in the compartment were allowed to become intermediate between A and B, it could not be ruled out that water might be passing right through the barrier.

If the compartment water table C is kept below both A and B, it can be affirmed that all contaminated water that leaks into the compartment is removed from the compartment, and that therefore the contaminated water has not passed through the barrier.

If, because of a leak, contaminated water does start to leak into the compartment, that water, having been pumped out, will of course have to be disposed of. This can generally be done economically, however, as compared with the cost of replacing the barrier.

It has been described that the water table C inside the compartment is lower than the two water tables A,B outside the barrier. In point of fact, however, the essential feature is that the water table C must not be intermediate between A and B: it would be equally satisfactory, from the point of view of ensuring that contaminated water cannot pass through the barrier, if the water table within the barrier were maintained above, rather than below, the two outside water tables A,B.

However, keeping the compartment water table low is generally preferred, because (a) no source of water is then needed; and (b) the water that is drawn off in lowering the water table provides a ready source of water for sample analysis purposes. On the other hand, if the water table is kept high, the contaminant is held at bay in the ground, and there is no contaminated draw-off water to be disposed of.

The manipulation of the water table within the compartment is .described hereinafter as it relates to pumping water out of the compartment.

It may be noted that the manipulation of the water table within the compartment is only necessary in the event of a leak. The compartment water table need not be controlled if measurements show it to remain unchanged.

In this specification, the expression "leakproof to a very high standard" is used in relation to barriers. This expression should be taken to refer to the degree of watertightness found in barriers which have substantial seals running the height of the joints or junctions between the elements of the barrier. No form of seal can be expected to be 100% reliable: the expression refers to the kind of seal that only leaks perhaps once in several barriers. A barrier which has leaks at several of its junctions is not "leakproof to a very high standard" as construed herein.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENT

By way of further explanation of the invention, an exemplary embodiment of the invention will now be described with reference to the accompanying drawings, in which:

Fig 1 is a cross-section through an aquifer in which a barrier which embodies the invention has been placed;

Fig 2 is a pictorial view of some elements of the barrier of Fig 1 ;

Fig 3 is a close-up, in plan, of a portion of the barrier of Fig 1 ;

Fig 4 is a plan of another barrier;

Fig 5 is plan of yet another barrier;

Fig 6A is a cross-section of an aquifer in which a barrier has been inserted;

Fig 6B is a cross-section corresponding to Fig 6A, but showing another barrier.

The structures and procedures shown in the accompanying drawings and described below are examples which embody the invention. It should be noted that the scope of the invention is defined by the accompanying claims, and not necessarily by specific features of exemplary embodiments.

As shown in Fig 1, a barrier 2 has been pile-driven into an aquifer 3. The barrier 2 extends from above the ground surface 4 down into a clay or other layer 5 below the aquifer 3.

A spill of contaminant forms a plume 6, and the plume is travelling through the aquifer 3 with a velocity V. The purpose of providing the barrier 2 is to contain the spill. Depending on the circumstances, the barrier may be arranged to encircle the plume, or the barrier may be arranged simply to lie in the path of the plume.

The barrier 2 comprises two walls 7, which are separated by spacers 8. The spacers 8 and the walls 7 divide the barrier 2 into a series of separate compartments 9. The walls 7, the spacers 8, and the joints and junctions between them, are all arranged to be leakproof to a very high standard.

It is not necessary that the aquifer material contained within the compartments be removed, but this can be done if desired.

A sample of the water is taken for analysis periodically from each of the compartments 9. The levels of the water table in the surrounding aquifer 3, in

front of and behind the barrier, as indicated at 1OA.10B and the level of the water table 10C within each compartment are measured from time to time.

If contaminant is discovered within a compartment 9, or if the water table in the compartment should change unusually, water can be pumped out of the compartment. In fact, the level 10C of the water table in the compartments can be kept lowered all the time, as a precaution in case of leaks, if desired.

If one compartment does develop a leak, only that one compartment is affected and only that one compartment needs to be treated, as by pumping; nothing needs to be done to the rest of the compartments. Thus a leak in the barrier as described only gives rise to a small isolated contaminated place, and this can be handled comparatively inexpensively. In any case, it is usually possible to ensure that the contaminant, even though it may leak into the compartment, cannot leak through the wall.

Fig 2 shows details of the structure of the two-wall barrier of Fig 1.

A front wall component 24 is formed from sheet steel, which is roll-formed to the shape as shown. The roll-forming process is arranged to provide edge forms along the respective side edges of the component 24. The left edge form 25 is arranged to be the complement of the right edge form 26.

When the left edge form 25 of one component 24 is assembled into the complementary right edge form 26 of the adjacent component, they define together an enclosure, which constitutes a tube 27 which extends down over the whole height of the barrier, and which is open at the surface.

When the barrier has been driven fully into place, this tube 27 can be flushed out with water, using a hose inserted down to the bottom of the tube. The tube can then be filled with sealant, again using a hose inserted from the surface down into the tube 27. This procedure is explained in GB-2228760,

which was referred to previously. As is also explained, the junior one of the edge forms, being the one that is driven in second, preferably should be provided at its foot with a scraper 30 (Fig 3), the function of which is to scrape out dirt etc which has accumulated inside the edge form of the senior, and which would lie in the tube if not scraped out.

The two interlocking edge forms 25,26 together form a complete enclosure: that is to say, the tube 27 is a complete encirclement. Each edge form on its own, however, does not provide a complete encirclement: the completeness of the encirclement only arises when the two edge forms are interlocked. The purpose of this arrangement is explained in the said GB-2228760.

With the elements in place, a potential leakpath exists between the edge forms, whereby water can potentially pass through the wall. The entry mouth of this potential leakpath is shown at 36 and its exit mouth is shown at 38.

The potential leakpath exists of course because the steel of the piling elements does not fit together perfectly. Any joint between two surfaces of steel, even when heavily pressed together, can never be expected to form a watertight seal: a sealing compound must be introduced into the joint.

In the junction that is formed by the interlocking of the two edge forms 25,26 the sealing or caulking compound is introduced through the tube 27, and in fact acts to fill up the tube 27. The mouths 36,38 of the potential leakpath are therefore sealed from the inside of the tube 27.

As described in GB-2228760, the tube 27 should be of such dimensions as will accommodate the flush-out hose and sealant-injection hose that are to be passed into it. Consequently, the shape of the edge forms 25, 26 should be so arranged as to provide an inscribed circle inside the tube 27 of about 2 cm diameter. The joints between the elements when constructed in the manner

described can be expected to be watertight and leakproof to a very high standard.

Because of the great degree of interlock between the edge-forms, the tube 27 can be expected to remain as a tube all the way down the height of the barrier, and for the tube to have the same dimensions at the bottom of the barrier as at the top. This is true even though the barrier will tend inevitably be distorted and diverted by the fact of being pile-driven into the ground. It is the fact that the dimensions and integrity of the tube can survive being pile- driven that gives the joint as described the ability to be leakproof to a very high standard, over its whole height.

As described, the sealant may be inserted into the tube by injecting the sealant in liquid or foam form by means of the sealant-injection hose; alternatively, the sealant can be provided in the form of an elongate "snake" which is passed down into the tube from the surface. In this case, the sealant is of the kind which swells upon coming into contact with water.

It is understood that no manner of sealing is absolutely perfect, especially in such extensive seals as are present in a pile-driven wall. The seals extend over the whole height of the wall, and there are seals between every pair of edges of the elements. The provision of the compartments 9, as described, assists the engineer in detecting whether a leak has occurred, and also makes it possible for him to take remedial action.

The rear wall component 28 is similar to the front wall component 24. The spacer 8 is of sheet metal and is welded to the two wall components, as shown. The spacer 8 and the two wall components 24,28 are of rigid material (usually steel is selected) and are robust enough to survive being pile-driven into an aquifer. Together, the wall components and the spacer form an H-shaped sub-assembly, which is able to be driven into the ground as a unitary whole.

If necessary, a suitable prepared cap or hood may be placed over the H-shaped sub-assembly during driving, whereby the pile hammer strides the cap or hood, not the spacer or the wall components. The spacer 8 is welded to the wall components to form the H-shaped sub-assembly. The welds of course must be robust enough that the H-shaped sub-assembly survives the pile-driving process.

As described, the wall consists entirely of the H-shaped sub-assemblies. However, as shown in Fig 5, if desired the compartments may be longer than those shown, in that the walls may include singleton roll-formed pile-driven elements between the H-shaped sub-assemblies.

The designer of the barrier should see to it that the compartments are not too large. If the distance (ie the distance measured along the length of the barrier) between the dividers is large, the walls between the dividers become less well-supported; it could then happen that the walls might become distorted or diverted, upon insertion, whereby the walls are no longer spaced a good distance apart. Even though the walls may be held firmly say 30 cm apart near the dividers, if the walls are allowed to almost touch together between the dividers then the barrier cannot be trusted to be leakproof. It may be noted that it is difficult to tell, from the surface, whether, at the foot of the barrier, the walls have been distorted in this way. Placing the dividers close together ensures that the compartments can be expected to have the same shape and dimensional integrity over the whole height of the barrier, even after driving into an aquifer material that contains stones or other non- homogeneities that might tend to divert the walls.

The spacer 8 comprises the divider between adjacent compartments, and ideally the welds should be continuous, so that the compartments are isolated from each other in a leakproof manner. However, watertightness between adjacent compartments is not essential. In many cases, it will not matter too

much if slight gaps exist between the spacer and the wall components. This would be the case, where, for example, the spacer is welded to the wall component only at intervals, and not in a continuous line.

As mentioned, when a leak occurs between the wall and the compartment, the compartment is pumped out, in order to keep the water table in the compartment low, and in turn in order to ensure that the contaminant cannot pass right through the compartment to the other side of the barrier. The size of the leak in the wall would determine the rate at which water needed to be pumped out of the compartment. If there is leakage between compartments, due to non-continuous welding, or for any other reason, then water will need to be pumped out of the contaminated compartment at a greater rate if water is also leaking into that compartment from an adjacent compartment. However, given that pumping equipment will have to be provided for the leaking compartment, the extra rate of pumping required because of the inter-compartment leaking will generally be insignificant.

On the other hand, of course it is preferred that there be no inter-compartment leaking.

The distance apart of the two walls is important. The space between the walls should be large enough that water can be pumped out of the compartment: if the compartments were very small (in plan area) then constrictions and confinements might exist which would interfere with the ability to draw water out the compartment.

The aquifer material may be left in the compartments, ie in the space between the walls. If there were to be no material between the walls, that would mean that water could be pumped out without difficulty but the disadvantage might then arise that the walls would have to made stronger in order to resist a tendency to cave in. The space between the walls may be filled with a

porous, permeable material such as sand/gravel from some source other than the surrounding aquifer material.

(The term "aquifer" as used herein refers to any body of ground which contains, or which can contain, groundwater: the term is not used in a narrow sense to refer only to an in-ground supply of drinking water. An aquifer usually includes sand and gravel as a major component, but need not do so.)

One of the advantages of the presence of the spacers and of a good spacing between the walls is that the barrier is thereby rendered more rigid.

One of the requirements for a barrier is to retain the material into which the barrier is inserted against lateral movement, and the more rigid the barrier the better it can perform this function. The barrier of the invention is intended mainly for use in situations where the main purpose of the barrier is to contain a plume of contaminant, in which case there is little requirement for the barrier to act to shore up the material; nevertheless, the barrier of the invention may be used to advantage in shoring applications, in view of the enhanced rigidity imparted by the spaced double-wall arrangement.

The distance apart or spacing of the walls preferably is in the region of 30 cm.

The spacing between the walls need not be identical all along the barrier, nor need the walls be parallel. However, wherever the walls are close together, the possibility arises that a leak may pass right through the barrier. The main safeguard that prevents contaminant passing right through both walls is that the large compartments exist between the two walls of the barrier.

If the walls are to be spaced only a small distance apart over some portion of the walls, then at least the walls should be substantially spaced between the junctions in the front wall and the junctions in the back wall, ie between the

potential leakpaths in the front wall and the potential leakpaths in the back wall.

Fig 4 shows an arrangement of the elements in a barrier that may be acceptable in some cases, even though the spacing between the walls is locally quite small. Here, there is no spacer, as such. The division between the compartments 9 is effected by bringing the front wall and back wall components 45,46 together. At the junction 47 between the two, a tube 48 is formed, which is similar to the tube 27, and which similarly can be sealed to a very high standard. The front wall junction 49, and the back wall junction 50, also can be sealed to a very high standard, as before, and as described in GB-2228760.

The spacing distance D1 of the walls as measured between the front wall junctions and the back wall junctions is substantial, as shown in Fig 4. However, the distance D2 between the walls at the point where the compartments are divided is very much smaller. On the other hand, given that leaks can only occur at the junctions between the elements this reduced spacing over the portions of the wall remote from the junction is not important, and would be acceptable.

In some cases, however, leakage underneath the barrier cannot be ruled out. Where the aquifer lies over hard rock, for example, leakage underneath the pile-driven elements is all too possible. In that case, if the spacing between the walls were small, contaminant leaks might pass right through the barrier and not be detected. Therefore, where the substrate is rock, for example, good spacing should be maintained over the whole barrier. Where the aquifer lies over soft clay, on the other hand, the bottom of the barrier can be expected to be sealed very well to the clay, in which case there is less need for the wall spacing to be large, at least over the portions of the wall remote from the junctions, and the arrangement of Fig 4 would be acceptable.

The front and back walls of the barrier should not be sealed to each other at the bottom. If contaminated water leaks underneath the barrier, the contaminated water must be able to pass into the space between the walls in order to be detected.

Also, the walls should not be sealed to each other because the leakage rate into or out of a compartment can be used as a measure of the degree of the seal between the wall and the underlying stratum.

The water table 10A in front of the barrier may be at a quite different height from the water table 10B behind the barrier. The water table 10C inside the barrier should be lower than the lower of the two water tables front and back of the barrier, if it is to be ensured that contaminated water cannot pass right through the barrier. As mentioned previously, it is also possible to arrange that water be pumped into the compartment, in order to maintain the compartment water table 10C higher than the two outside water tables.

As mentioned, it is important that the walls of the barrier be spaced a good distance apart, and 30 cm has been mentioned as a typical minimum distance. The purpose of keeping this dimension large may be understood from Figs 6A and 6B.

Due to miscalculations, faulty measurements, and other reasons, it is possible that the walls of the barrier might not quite reach the sub-layer of clay or other impervious material underneath the aquifer. Therefore, a leakpath might be present beneath the barrier. If the walls were to be close together, in relation to the gap between the bottom of the walls and the sub-layer, then it could happen that water could pass from front to back of the barrier, through the gap 60 below the two walls 62,63, as shown in Fig 6A. When the walls are too close together, this leakage can occur irrespective of how strongly water is being sucked out of the space 65 between the walls.

When the walls are wide apart in relation to the gap beneath the barrier, as shown in Fig 6B, the possibility that water could flow right through underneath the barrier is eliminated.

In the kind of pile-driven sheet metal barrier described, which is inserted into a normally-homogeneous aquifer of sand and gravel, down to a sub-layer of reasonably soft clay, at a reasonably constant depth in the five or ten-metre range, the distance apart of the walls can be dropped to a minimum of about 20 cm. If the aquifer material contains rocks etc, or for greater depths, or if the sub-layer is known to be hard and/or uneven, the minimum distance apart of the walls should be increased to a minimum of about 30 cm. If the gap is too small, the condition shown in Fig 6A can occur.

In addition to the walls being widely spaced apart, the cross-sectional area of the compartments also should be large. Of course, the wider the walls are apart, the greater the area of the compartments will be - but the purpose of making the compartment of large area is different from the purpose of avoiding the Fig 6A leakage.

If the compartment were of very small cross-sectional area, then the engineer could not be sure, upon drawing water off from the compartment, that a corresponding suction was being felt lower down the compartment, ie at the foot of the barrier. In order for the engineer to be able to give worthwhile evidence regarding leaks to a tribunal, it is important that the engineer be sure that the spaced-apart shape and dimensional integrity of the compartment is maintained over the whole height of the barrier.

It is important for the engineer to be able to state, where he has reduced the water table in the compartment, that he must in consequence have reduced the pressure head of the groundwater at the foot of the barrier, in proportion. If the compartments were small in cross-sectional area, he would not be able to state that. It would not be important whether any particular compartment

happened to have a blockage: the important point is that it is only when the compartment is of large area that the engineer can declare that there is no possibility of a blockage. The engineer wishes to be able to state that, if the water table in the compartment is stable, that therefore there are no leaks through the barrier; the engineer does not want to have to admit that the reason the water table is stable could be that there is a blockage in the compartment.

Also, the engineer wishes to be able to draw water out of the compartment at a sufficient volumetric rate of flow to ensure that the Fig 6A condition is not occurring. Again, making the compartment of a large cross-sectional area helps to achieve this required flow rate. If the compartment were of small area, such that only a trickle were extracted at the top, the engineer could not be sure that the Fig 6A problem was not occurring

The minimum size of the cross-sectional area of the compartment again depends on the nature of the aquifer, the depth of the barrier, etc: the kind of barrier which required a minimum width apart of the walls of 20 cm would require a minimum compartment area of about 1000 sq cm. For the kind of barrier that requires a minium wall spacing of 30 cm, the compartment should have an area of at least 3000 sq cm.

If the compartment area is smaller than the above values, the engineer cannot be sure that an effect at the top of the compartment is being reproduced at the foot of the compartment. Above these values, on the other hand, the engineer can be sure, bearing in mind the height of the compartment and the permeability of the material contained therein, that substantially no pressure gradients can exist over any horizontal cross-section of the compartment, and substantially no pressure gradients, other than the regularly depth-proportional pressure gradients due to gravity, can exist over the depth of the compartment.

Although the invention has been described as it relates to pile-driven sheet metal walls, the invention can be applied to barriers made of other materials. For example, the wall elements may be of plastic. Plastic elements are not inserted into the ground by pile-driving, but by excavating a trench, filling the trench with a suitable liquid or slurry, and then lowering the elements down into the slurry. The wall elements may alternatively be made of concrete, although a concrete barrier is often very expensive. When the barrier is not of sheet piling, it is important to note that the compartments of the barrier still must not be closed at the bottom, whereby water could pass underneath the barrier without that fact being detected.