It is probably correct to state that most of the more obvious swallet cave entrances on Mendip have, at some time or other, either been investigated or entered by cavers. If therefore follows that future discoveries will, in general, tend to be the result of extended digging to a, greater extent than has occurred in the past.

Already, several digs and successful cave penetrations on Mendip have relied on shoring. In some cases (eg- St. Cuthberts) the shoring has proved itself effective. In others it has failed to keep the cave, or dig, open. One aspect of this report is therefore to review shoring methods and to describe those pract¬ices and principles which have so far proved practicable and -succ¬essful on Mendip.

Another trend which is becoming noticeable, is the recognition by some members of the caving fraternity that money, as well as effort, must sometimes be spent on a dig. The St. Cuthberts shaft already mentioned was constructed entirely from second hand mat¬erials at no cost. The recent Priddy Green shaft, on the other hand, is a concrete tube with cemented stone footings. This fine example of permanent, shoring may be followed by even more ambitious shafts in the future. Another aspect of this report is therefore to suggest methods of shoring which are not in use at present end which have not been used owing to expense.

It is hoped that this report may prove useful to any cavers faced with problems connected with the shoring of cave entrances involving either proved methods of constructing and installing such shoring or ideas for tackling situations which have not yet been attempted.


In order to open, and keep open, a swallet cave whose ent¬rance has become blocked by natural means, it is necessary to disturb the natural stability of the blocking material. It is necessary to restore this stability by means, if necessary, of shoring. A shaft is thus called on to withstand forces which may tend to destroy it by crushing, shearing or twisting. The most common of these is crushing and it is the properties of the subsoil through which the majority of such a shaft, will normally pass, which determine the magnitude of those forces.

2.1. Subsoil Friction.

This is the friction obtaining in a loose, dry, granular soil. It is this friction which prevents the soil from acting as a liquid and finding its own level, i.e. possessing a hori¬zontal surface when poured on to a flat plane. If such a soil acted in this way, the normal hydrostatic laws would apply and the pressure exerted on a vertical shaft would increase with depth according to the normal formula: –

Pressure = ρgh

Where ρ is the density, g the acceleration due to gravity and h the depth. ρg for water is 62.5 lbs wt/cubic foot and for clay about 115 lbs wt/cubic foot. Thus at 10 feet depth a pressure of 1150 lbs./ sq” ft would be pressing on a shaft in clay.

Luckily, soil friction reduces these pressures. A soil possessing friction will form a circular cone if poured on to a level surface whose sides will stand at an angle ϴ to the hori¬zontal. This angle ϴ is called the angle of repose or the angle of friction. Typical values of ϴ and ρg are shown in Table I below.

Material ϴ ρg
Water 62.5 lbs/ft3 
62.5 lbs/ft3Vegetable Topsoil 15-30   95 – 105
Clay 15-40  110 – 120 
Sand 25-35 110 – 120
Gravel 45 110
Shingle 30-35 115 – 120


Now. the lateral pressure exerted at depth “h” by a soil poss¬essing only soil friction is given by Rankin’s formula:-

                      Pressure = ρg x (1 – sin ϴ )
                                                (l sin ϴ )

which reduces to the previous formula when ϴ = 0.

Putting in the values, for ϴ into this formula for clay now produces a lateral crushing pressure at 10 feet depth of 383 lbs/sq ft instead of 1150 lbs/sq ft. Thus the effect of soil fric¬tion in clay is to reduce these pressures by two thirds.

However, the pressures are still high and still getting .larger at greater depths. If this was the case, a deep shaft would have to be built to withstand enormous pressures. Now surface friction only applies to a loose, dry soil. If the soil is neither loose nor dry the property of cohesion must be taken into account.

2.2 Subsoil Cohesion

The combined effects of compaction (the decrease of the av¬erage spacing between soil particles due to the pressure of over burden or top soil above the layer under consideration) and mois¬ture content give a soil the property of cohesion.

Cohesion is defined as the property which resists the tend¬ency for soil particles to separate under shear stress. The force of cohesion in a given plane is equal to the area of surf¬ace particles in contact times the coefficient of cohesion. A typical set of values is shown in Table II below.

Material  Moisture  Coefficient of cohesion
Sand Wet
400 lbs/sq ft
Gravel  Wet
Clay  Optimum  900


The water, in the case, of sand, with its small particle size, acts as a binder by capillary action and accounts for the great cohesion of wet sand with respect to dry. The sand in an egg timer is completely dry and possesses almost zero cohesion. A long as the angle of the waist of the timer exceeds ϴ for dry sand, the. sand will act as a liquid. On the other hand, sand castles made from wet sand possess a surprising, degree of stab¬ility. In the case of gravel (with its large particle size) the capillary effects are small. The water here acts as a lub¬ricant and lowers the cohesion properties of the dry material.



In clay, a much more complex series of events occur. Under some circumstances the effective particle size itself varies. It is sufficient to note that there is an optimum moisture content giving a high cohesion which is close to that naturally found in Mendip clay subsoils. The graph in Figure III shows the effect of cohesion. It will be seen that the curve tends to turn back¬on itself. Thus, in theory, no shoring might be necessary at great depths in clay. However, a further property of this type of subsoil is involved.

2. 3. Subsoil Slumping.

This property (which is peculiar to clays and materials of a similar nature) is less easy than either of the other properties to describe in a quantitative manner. However, the process is roughly as follows:-

If a mass of clay is supported solidly over some of its bulk and relatively or completely unsupported over the remainder, stresses will be present in the mass of clay. These will in general be present oh the free surface as a series of tensile, and compressive stresses. The tensile stresses will produce strains tending to pull the surface material apart (i.e. to produce a crack). Where the tensile stress is great enough and applied for long enough, this will occur. The crack so produced will not run normal to the face but will be bent towards regions of maximum tensile stress. This will lead to slumping when, as in most cases, the process is unstable (i.e. the formation of a crack leads to an increase in the forces producing the crack). Thus the crack builds up very slowly at first and then progressively speeds up until a section of the clay suddenly slumps. The process is completely silent and the slumped mass of clay exerts a heavy and uniform pressure on anything beneath it. The author has had the experience of being pinned by the legs under a large clay slump during the excavation of Browne’s Hole. .A painful throbbing occurs quite quickly and one supposes that it would not take much clay over the chest to make breathing impossible.

Clay which is unsupported from below over a circular area will slump to form a slump chamber having a nearly hemispherical roof, such a slump chamber was dug into at Vole a Hole – its roof apex being some three feet below the ground and its diameter about four feet. Observations of slumping in St Cuthberts and Alfie Hole suggests that a slump crack forms as shown in Figure I and it is suggested that the ideal sequence of events in the collapse of an unsupported vertical hole would be as shown in Figure II.



In the first figure (below in FIGURE II) the hole is shown as dug. In the second Slumping has occurred and a slump chamber formed. In the third figure the slump chamber has reached the soil level and the roof is mainly held by grass roots, etc. In the fourth figure the roof has collapsed, thus allowing drying out and subsequent crumbling to occur. In the last figure a wide, shallow depression results since all the clay has recompacted. In practice this would be terraced (as in some shake holes) due to the binding effect of surface grass.

Since slumping will occur, not only must any permanent hole in clay be supported by shoring, but any horizontal tunnel must be built to withstand the entire weight of the overburden. A powerful argument against the use of other than vertical shoring.




There are two main types of initial excavations in subsoil during the search for a cave entrance. The sinking of a small shaft (some three to four feet in side length) has the advantage of retaining nearly all the wall moisture and hence all the co¬hesion. On the other hand, a wider excavation (defined as a pit in this report) has the advantage of exposing more rock face at the bottom and reducing the danger of slumping. Since no two situations are alike, the choice of excavation will depend, on factors present at any particular dig. In some cases the stability of either a pit or shaft may be judged sufficient to stand unsupported until an entrance is found. This was the case in St Cuthberts. In others shoring must be commence at once (as in Alfie’s Hole). In the latter cases the best plan is to install temporary shoring until the nature of the dig warrants something more permanent.


This is best carried out in timber with any shuttering mat¬erials available. Long life is not important, but access to the work, at the bottom is and so crossbeams, etc should be arranged so that they do not interfere with the work. Compared to per¬manent shoring, temporary shoring has usually only to prevent crumbling at the top of the hole or slumping lower down and hence need not in general, be very strong. Sheets of corrugated iron will act well as temporary shuttering. (Note: “shuttering” on Mendip is normally used to describe the sheeting, planking, poling, etc needed to retain the earth and form the outer wall of a shoring). In addition there is little or no need for temporary shafting to be footed (i.e. to have a solid foundation to rest on) or for any of the beams retaining the shuttering in place to be firmly enough, wedged in to permit their use for climb¬ing in and out (since only the diggers will be using the shaft at this stage).

Apart from the reluctance to spend much time and/or money on permanent shoring before the effectiveness of a dig is known, it is not a sound plan to install permanent shoring until suitable rock has been uncovered so that footings for the final shaft may be constructed. It is difficult to over emphasise the importance of footings. Very few shoring jobs on Mendip (one exception be¬ing the shaft in Fairman’s Folly) have failed through any other reason than that of inadequate footings. Therefore, as soon as solid rock has been reached this should be carefully examined. If there is any doubt the job should be delayed as long as possible, consistent with safety, until more is known of the underly¬ing rock. It is very difficult to re-foot a shaft once it has been installed and so this stage is perhaps the most critical of any from the standpoint of lasting success of the job.


Footings- have to provide three functions and provide them all adequately.

  1. To rest the base of the shoring on to a solid and stable foundation so that the shaft is able to resist any shear¬ing forces tending for drag it, or parts of it, down¬wards. Since most shafts are only proof against crushing forces providing they do not move vertically, this is doubly important.
  2. To increase the stability of the “transition zone” where artificial shafting meets natural cave.
  3. To prevent “running in” behind the completed shaft. By reducing the pressures locally behind a portion of the shaft, running in will create forces tending to move the shaft sideways. In addition, many shafts are designed to rely for their strength on an even, external pressure. Thus, running in can seriously weaken a shaft.

There is, unfortunately, no golden rule for footing or trans¬ition zone shoring as each case is so different. In the compara¬tively solid rock of Priddy Green Sink, cemented stone walls have been used to build up the irregular rock to a level base for a shaft. In St Cuthberts the transition zone extends from the shaft base down to the top of the entrance pitch about eight feet or so. The stone revetting in the small chamber was built here to prevent running in from the floor of the entrance shaft.

In Alfie’s Hole, small rocks and timbers have been wedged in between the large rocks to prevent running in and to prevent movement of the rock.


The system of keyed timbering evolved by the author during the construction of the St Cuthberts shaft has since been used with equal success elsewhere. In fact, no case of collapse of a keyed shaft .has yet been recorded. The system is, perhaps, best explained by describing an imaginary shoring job and choosing a situation where most of the variations on this theme may be ill¬ustrated. Let us assume then that a subsidence in clay some ten feet in diameter and six feet deep has had a shaft dug at its’ base, again in clay, which is three to four feet square and a further ten feet deep, at which point rock has been struck, and a hole leading to a cave system cleared. Let us further assume that no temporary sharing has been found necessary and that a keyed timb¬ered shaft is to be installed.


Four long timbers (the main uprights) are lowered into the hole in each corner of the shaft. These should if possible be long enough to reach the surface. In the example above the SW and SE posts can be footed directly on the bed rock at B, the SE post being higher than the SW owing to the dip. The suspected flake at K can be used for the NW post after jamming the gap below it with suitable rocks and cement. Probing in the NE corner has failed to find rock and so the NE post must be rested on the clay floor and constructed as a hanging corner.

The uprights should be held apart by temporary spacers which can be removed as the shoring progresses.

The work now starts at the bottom by the insertion of a ring or set of crossbeams, which should be fitted in the lowest poss¬ible position (in this case at fourteen feet down) consistent with their being horizontal. Each crossbeam should be cut slightly too long and hammered into place by means of a sledgehammer. A little practice will determine the amount of extra wood to allow when cutting. Each beam should require considerable force to force it into place. Rings of crossbeams should follow at about three foot intervals up the shaft (say one at eleven feet down and one at eight feet down). At this point work should cease as the top portion of the shaft is not pressing against the clay. A different technique should be used on this, section at a later stage.

Now shuttering should be hammered into place at the bottom of the shaft. In this case, horizontal planking will be best for the lower half of the shaft as this can be easily hammered behind the uprights. Other types of shuttering, such as horizontal poling, will have to be “dug through”.

With the lower part shuttered, the first set of ring spacers should be cut to size and fitted below the bottom ring, which is then hammered down onto them. At this stage, keys are fitted to the bottom ring. Figures 1, 2, and 3 of the Appendix show the principle of keying and ring spacers. Note that cappings or spacers may be necessary to support the cross keys and prevent the from slipping downwards. When the first ring has been spaced and keyed, work proceeds to the second ring and thus to the third. The lower half of the shaft is now complete.

Additional rings of crossbeams are now cut to length and nailed temporarily in position say at four feet and ground level and shuttering is nailed in position outside the shaft. All the space between the Shaft and the subsidence is now filled with clay and stamped down tightly. Spacers and cross keys are then fitted as before.

Finally, bracing must be added to prevent the NE corner from sinking into the clay on which it rests. This must be added to the north and east faces of the shaft in between the rings of tim¬ber diagonally so that the upper ends of each brace is at the NS Corner. The south and west faces need no bracing as there is no tendency to slip.

Now consider the completed shaft.

  1. The shuttering withstands the forces exerted laterally by the surrounding clay and is prevented from collapse by the main uprights.
  2. The main uprights are prevented from moving towards each other by the -crossbeams, which must be kept horizontal and between the uprights.
  3. The ring spacers keep the crossbeams horizontal and also (by capping or spacing) prevent the cross keys from downward motion.
  4. The cross keys stop the crossbeams from moving horizontally and are in turn locked by the next set of ring spacers,
  5. Finally, the bracing ensures that the only way in which the NE corner can move downwards is by also moving out¬wards. It is prevented from so doing by the pressure of the clay.

It is possible to build a keyed timbered shaft (using mitre keys) so that it is impossible to move any timber in the entire shaft except the topmost ring, WITHOUT THE USE OF A SINGLE NAIL OR SIMILAR DEVICE: Such a shaft would fall apart if the surrounding clay were removed. It is for this reason, apart from the others given, that it is so important to prevent running in by good footings.

In many cases, shoring materials must be improvised and timbering of the sort illustrated in Figures 1, 2 and 3 of the appendix cannot be obtained. In such cases construction must be modified to suit whatever is to hand.

A wide variety of shuttering materials have been used on Mendip. Whole doors, airfield landing strip and even a set of polished mahogany table legs (these may be seen in the SE shaft at Brownes Hole, although the top of the shaft is sealed it may be entered from Coronation chamber below). Similarly, it is not always possible to obtain long enough pieces of timber, or a rock may prevent standard methods from being employed. In all cases, however, the general principles of letting the subsoil hold the shaft together and ensuring adequate footings, apply, although these often call for ingenuity of the highest order. A point to remember is that one of the main sets of forces acting on the shaft are those caused by a caver using it to climb in. or out of the cave and rapidly and clumsily applying his whole weight to the timber in question.


The use of concrete piping as vertical shoring has been pioneered in the Priddy Green dig. Adequate cement footings are, of course, necessary as there is no inherent continuity of the shaft. A point here is that care should he taken to ensure that the shaft is not waterproof; leaks through the joins or through holes deliberately made should be encouraged. Otherwise, full hydrostatic pressures will build up. Reference to the graph on Page 3 and comparison with water pressures of 625 lbs/ sq ft at ten feet down will illustrate this point.

Manholes and underground concrete pipes as constructed by the GPO and other public bodies are designed on an even more pessi¬mistic basis and the relationship of ρ=90h or 900 lbs/sq ft at ten feet depth is often used.

The effect of heavy loads adjacent to the shaft is also all¬owed for. Boussinese’s formula is normally used and an example is that a load of 10 tons on the surface at a horizontal distance of four feet produces a lateral pressure on a vertical shaft wall at five feet depth of 100 lbs/sq ft. This sort of thing could well apply when a spoil heap is situated close to the shaft.

Normally, however, it is assumed that sections such as those used at Priddy Green will be employed. All such “bought out” components have stressing specifications which can be relied on and the subject then devolves into one of picking the right type of pipe for the job in question.


The use of iron or steel shuttering has been employed in Browne’s Hole and several other cave digs. Oxidation is the drawback here and in this connection ordinary corrugated iron should be avoided. It has a short life below ground. A very strong shaft could be made from girder and would be useful against rock. Providing that the structure is simple in form, strains could easily be measured and stresses calculated from them. Local reinforcement would then be applied until an adequate safety factor had been obtained. Although the cost of such a shaft would be high, it would possess the advantages of ease of assembly and plurality providing it was given a suitable protective finish.


The life of a timbered shaft is quite long. In St Cuthberts the shaft has stood for nearly eight years but will eventually have to be replaced. A method of dismantling evolved during the digs at Vole and Alfie’s Holes is described below.

i. Seal the bottom of the shaft firmly with boards or steel plates to prevent any collapse from entering the cave.

ii. Starting from the bottom of the shaft, remove all keys and rings spacers, bracing, etc; keeping if possible above the unsupported lower section.

iii. Drive nails into the crossbeams with a pole and attach ropes to them.

iv. From above lurch out crossbeams with a pole and haul out on ropes.

v. Push in main uprights from the top and haul out if shaft does not collapse.

vi. Re-excavate hole and install new shaft.

vii. Remove seal from the bottom of the shaft.


It seems worthwhile to include three suggestions for tackling shoring problems which, so far, have not been successfully carried out on Mendip. All these suggestions may be seen as ambitious, but it is worthwhile remembering that only two years ago many people did not consider the use of concrete tubes to be within the practical limits of cave digging expense and effort.

10.1 Shoring Against Torsional Forces.

A keyed timbered shaft could further be strengthened by the addition of torsion resistant beams. The method suggested would also be useful if it was ever necessary to construct a wide timbered shaft of adequate strength without of having to use timbers of very large sections. The scheme consists of placing central spacers and keys between each ring of timber and wedging torsion beams between them. Looking down a ring you would have the arrangement as shown in FIGURE IV. A picture of the corner of such a shaft, with a list of timbers, will be found in Figure 4 of the Appendix.

10.2 Boulder Ruckle Shoring

Sooner or later the problem of digging through an unstable boulder ruckle will be tackled. It would seem that a steel girdered shaft bolted together on the Meccano principle would be a solution, the shaft being extended from below as work progresses. This could occur by prising rocks out, hammering corners off, using screw jacks or by blasting. Strains in the shaft could be measured by gauging rods. Shuttering would not be necessary and indeed would be a disadvantage since the position and movements of rock could be noted through the shaft girders.



10.3 The Hanging Shaft.

Such a shaft as described in l0.2 would have footings and this leads us to the consideration of shafting through a ruckle which forms the roof of a chamber. It would be most unwise to construct a shaft which was liable to fall when roof keystones were removed,

A solution is to use the surrounding surface ground as foot¬ings and to hang the entire shaft from top supports which are spread over on to safe ground. An illustration of such a shaft will be found in the Appendix.


Cave shoring is an occupation in which much depends on the individual concerned; the materials at his disposal and the part¬icular patterns of the dig and/or Cave. Of the shoring which has already been done on Mendip, some examples have fared better than others and in some cases conclusions may fairly safely be drawn (e.g. insufficient cross timbering, laying and bracing in Fairman’s Folly). In other cases, the skill of the builder would be difficult to put on paper – an example here being the shaft in Browne’s Hole. It is hoped, however, that some of the principles and suggestions offered in this paper may be of use to cavers whose work necessitates the installation of shoring in swallet cave entrances.



PLAIN KEYING (Keys nailed to beams)


MITRED KEYING (No nailing necessary)



Use of cappings to suit timber section



Use of spacers to suit timber section






A – Main Upright
B – Crossbeams
C – Ring Spacers
D – Crosskeys
E – Centre Spacers
F – Centre Spacer keys
G – Cross Bracing
H – Torsion Beams
J – Horizontal plank shuttering





(A) Horizontal Planking
(B) Horizontal Poling
(C) Cross Planking
(D) Sheeting

Also possible, amongst others are: vertical poling, diagonal and cross diagonal planking and poling.


© 2024 Bristol Exploration Club Ltd

registered in England and Wales as a co-operative society under the Co-operative and Community Benefit Societies Act 2014, registered no. 4934.