A Complete History of The Great Flood at Sheffield
by Samuel Harrison
Web Page 26

(119 - contd.)

The return prepared by the Chief Constable of Sheffield shows the number of persons who lost their lives by the flood to have been 238, namely--138 males and 100 females, 105 being under 20 years of age, 132 of 20 years and upwards, and one whose age is not given The ages of those who lost their lives vary from two days to 87 years Some additional bodies have been found since the return was made. In addition to this loss of human life, 50 horses, 38 cows, 8 donkeys, 258 pigs, 267 fowls, and 72 tame rabbits, were drowned.

The Chief Constable of Sheffield also gives the following list of properties totally or partially destroyed, or flooded, by the bursting of Dale Dyke or Bradfield embankment:--

or Damaged
Manufactories, Tilts, &c .............
Rolling Mills and Corn Mills ............
Workshops, Warehouses, and Store Rooms.........
Drapers, grocers, and other Sale Shops ..............
Dwelling houses.............
Malt Houses, Breweries, Public and Beer Houses.. 
Churches, Chapels, Schools, &c..............
Tanneries, Skin Yards, &c. .................
Buildings not otherwise described...............
Bridges, &c............................
Cottage and Market Gardens..................
Lineal yards of Fence Walling ..................

The General Relief Committee in their report, dated 2nd May, 1864, give the number of houses flooded as 4357, and the number destroyed and abandoned 798.

The General Relief Committee in their report, dated 2nd May, 1864, give the number of widows caused by the flood as 11; number of orphans, 40; number of deaths, 250. The number of persons relieved (to 29th April, 1864) was 20,537.

Up to the 2nd May, 1864, the Chief Constable states that the sum of £2329 13s. had been expended in cleansing and carting away mud and refuse from the streets, roads, squares, yards, and alleys. Assuming that, on an average, the removal of mud and refuse cost one shilling per cubic yard, this would be equal to 46,593 cubic yards. A further sum of £260 had been expended in chloride of lime and for other disinfectants, to prevent any local spread of fever.

An investigation of the probable cause of failure at Dale Dyke reservoir embankment requires certain acquired knowledge, such as the geology and meteorology of the district. the general characteristics of the area of the gathering ground, the contour of the surface, and the gradient and peculiarities of the valley and river or brook; an examination of the ruined embankment, and works connected with it; as also a careful analysis of the evidence given before the jury at the inquest.

As previously stated, the district on which Dale Dyke reservoir embankment was formed, is millstone grit; and, at the site of the embankment, shale and beds of sandstone rock alternate; the dip of the strata being from north west to south east. The gradient of the brook is rapid, 90 feet, and 72 per mile, down to Owlerton, a suburb of Sheffield. The valley is narrow, and the sides are steep. The line of the valley to Sheffield is almost due east and west; the flow of the water being towards the east. As in the millstone grit generally, so in this Dale Dyke valley, there are strong springs of pure water. Rain and dew fall on the sides of the mountains, the water sinks into the porous and fissured sandstone surface, passes down to the impervious shale, and is thrown out in springs on the sides and bottom of the hills.

The stratified beds of rock are, for the most part, hidden beneath a superficial covering, consisting of the disintegrated material of the rock and shale above, washed down by rains and contoured by time. There are swamps of moss and bog in places where springs of water create and continue the moisture necessary to this form of vegetable growth, and in several places this superficial covering has slipped. Such a slip existed on the north side of the river, immediately below the line of the present embankment, which induced the engineers to recommend that the present line of embankment should be adopted. Such slips are, however, merely superficial, and cannot truly be designated "land slips," such as the Undercliffe, Isle of Wight, and "landslips" of this character. We did not notice any land slip at the site of the Dale Dyke embankment to lead to the conclusion that such contingency had caused the failure. The embankment would also, from its position and form, act as a buttress. The rupture and wash of the flood evidently disturbed strata on the north side of the valley, immediately below the foot of the bank; but this did not cause the rupture, but was caused by the violent scour.

The meteorology of the district shows a considerable rain fall, but not more than might have been anticipated. Mr. Gunson stated the fall of rain as under:--

Rain fall.   Inches.  
1859 ................
1861 ................
1862 ................
1863 ................

A range of five years is not sufficient (without knowledge from other sources) to lead to practical conclusions. A very dry year may be as low as 30 inches; a very wet year may exceed 60 inches, or in some such proportions. The driest year must be the test of the usefulness of the storage; the wettest year, the heaviest fall of rain, and greatest flood, must be the test of the works. Mr. Gunson stated that in August, 1856, the fall of rain was nine inches. If this fall had taken place on a full reservoir, and after previous wet weather, the volume of water to pass to unite would be more than the by wash and outlet pipes could have fairly discharged.

A reservoir and its works must be sufficient to store safely, or to pass, harmlessly, the greatest possible flood during the wettest season. Mr. Gunson stated "that 12 or 14 inches of water flowing from the 4300 acres would fill the reservoir;" as also, that "the only effect which the greatest known flood could produce would be to put 50 or 60 feet in depth of water in the reservoir." Mr. Leather stated "that the two outlet pipes of 18 inches diameter each would discharge about 10,000 cube feet of water per minute, and that it would take 190 hours to empty the reservoir." These statements are both alike--incorrect. A depth of water or rain equal to 7 3 inches, from 4300 acres, would have filled the Dale Dyke Reservoir to its top water level, and not more than 5000 cube feet of water per minute would pass out of the pipes under their greatest pressure, whilst 30,000 cube feet per minute might be flowing into the reservoir during an extreme flood. The by wash, at 12 inches in depth, would discharge about 14,000 cube feet per minute, so that 11,000 cube feet per minute might be accumulating more than the outlet pipes and by wash could discharge. Such contingency ought not to have existed, as a flood on a full reservoir at any time must have seriously injured the bank, if it did not entirely destroy it.

An examination of the ruined embankment and empty reservoir showed that the stratification had been stripped of the superficial, and, in some respects, water tight covering, and that several acres of open-jointed sandstone rocks are bared. Into some of these joints and seams surface water flowed readily during an examination made a few days after the failure. Mr. Gunson stated--"the outer part of the embankment, to prevent slips, was formed of rubble stone, to a height of 50 feet at the apex; 3d. per cube yard extra was paid to the contractor for selecting and placing such stone. This was done to prevent slips."

An examination of the ruined embankment showed that this selecting of rubble for the outer slope or toe of the embankment had, with the first stripping of the strata, insured more water tight material for the first 50 feet in depth of the inner part of the embankment: hence, the comparative tightness of the reservoir under 50 feet of water during the summer, autumn, and winter of 1863, and spring of 1864. The upper part of the embankment, on both sides of the puddle wall, is evidently composed of rubble material, in no respect water tight.

Mr. Gunson stated that the embankment was formed by wheelbarrows, three wheeled, or "Dobbin carts;" by ordinary two wheeled carts, and by contractors' railway tipping wagons. "Tips," three feet to five feet in depth were "allowed," as also that in every respect the work and material of the Dale Dyke embankment corresponded to the mode of working and material used at the Agden embankment. This work was then in course of formation with contractors, wagons, tipping, loose and coarse rubbly material, in layers of from six feet to nine feet deep, which had the effect of rolling all the largest stones to the bottom, and forming open stratification from the inside, or water surface of the embankment, to the puddle wall, and from the outside of the puddle wall through the mass of the embankment. The material is not of a good character for the construction of a water tight embankment, and the mode of depositing and working is objectionable.

The puddle trench could not, of course, be examined, nor the outlet pipes and valves, as these were buried beneath the ruins of the embankment. The inner bay of masonry, where the pipes commence could be seen, and here planks and rough timber props had been left into form a temporary guard of timber in front of the pipes. This work was in an unfinished state, although the reservoir was in the course of being filled for use; and it could not have been completed without entirely emptying the reservoir. No means had been provided to close the pipes on the inside of the reservoir, or subsequently to examine them.

Sinking the puddle trench appears to have been a tedious and costly affair. The strata made water sufficient to keep two steam engines at work for upwards of two years, part of the time night and day. During wet weather the engines and pumps were "drowned out." It was only intended to excavate the puddle trench some 9 or 10 feet deep; an impervious foundation was not met with, and it was necessary to sink in some parts to a depth of 60 feet. No special provision was made to block out this water on the inside, or to remove any springs from the bottom of the puddle trench, or from the seat of the embankment outside by drains. When the puddle trench was considered complete, Mr. Leather examined the bottom, "found it tight," and gave orders for the trench to be filled with puddle. A trench which had occupied upwards of two years in sinking, under continued steam pumping, required puddle of a superior character to be used, and extraordinary care to prevent any wash of water from beneath the puddle. The evidence does not record any such arrangements having been made, other than placing rubble stone at the toe of the embankment. Continued pumping seriously injures the sub strata for reservoir purposes, as all beds, joints, and fissures are opened, and the subsoil is rendered more porous. The Hindoos, in forming embankments, far larger than this at Dale Dyke, had no means of sinking deep puddle trenches, as they had not steam pumping power to use, neither did they form any puddle wall, as in English reservoirs.

The outlet pipes, Mr. Gunson stated, "are laid in a straight line, but diagonally to the line of the embankment." A trench was excavated in the solid water tight shale some eight or nine feet in depth, and so wide as to allow the pipes to be about 18 inches from each side, and 2 ft. 6. in apart. Where the pipes cross the puddle trench, in order to guard against fracture of the pipes, the trench was sloped down from the ordinary depth, 100 feet on each side, so as to intersect the bottom of the puddle trench, which, at this point, was some 30 feet in depth. The space so excavated was filled in with puddle, and 18 inches in depth of puddle was placed in the remainder of the pipe-trench on either side. The pipes are ordinary socket pipes, in 9 feet lengths, but of extra thickness (l¼ inch metal); the joints are made with lead in the usual manner. The sockets are six inches deep, are towards the inside of the reservoir, and break-joint in the trench. Mr. Gunson thought, as the pipes were so strong, and the joints so carefully made, that if a subsidence of two feet took place, the pipes and joints would remain uninjured. A provision was undoubtedly made to allow of subsidence, but Mr. Gunson does not seem to have thought about the consequences to the puddle wall at the point of intersection. If the least subsidence took place, a fracture in the puddle wall above the pipes was inevitable, as the puddle, from its character and form, could not follow the pipes down. Pipes so arranged, so jointed and so laid, could not possibly retain the position in which they were first placed, but must inevitably go down; and in such case would fracture the puddle above the pipes, leaving a cavity in proportion to the depth of the subsidence. Water, under the pressure of a reservoir nearly full, would, at first, carry away small portions of the puddle and of the bank; and, as a cavity was formed, subsidence would commence, cracks would show, and the whole would rapidly wash and sink to ruin. The rubble toe of the bank would distribute and conceal the overflow of water. The mischief, once commenced, would rapidly increase, the subsidence of the outside material of the bank would bend the upper part of the puddle wall outwards, until the top line of the bank came below the water level. The work of destruction setting in at the top and bottom of the bank, would destroy it as rapidly as it is recorded to have been destroyed. The greater portion of the mischief was effected in from 15 to 30 minutes, in which time 92,000 cube yards of the material were swept away.

An examination of the pipes will not now be evidence against them, as it may be said the rupture disturbed and injured the pipes, rather than that the pipes originally were the cause of the failure. Destruction so rapid and so entire, requires, however, adequate power for its accomplishment. The facts stand as under: The puddle-trench was unfavourable; the outlet pipes were laid in a most objectionable manner, so as, in fact, to insure a fracture somewhere; the puddle-wall is much too thin, and the material placed on either side of it is of too porous a character, and was placed by railway tip wagons, which is the worst manner of constructing a water tight embankment. No puddle wall should ever be placed betwixt masses of porous earth, as puddle, under such conditions, will crack, and is also liable to be fractured by pressure of water. A wall of puddle 16 feet thick at the ground line, and 95 feet in height, could not remain entire in the midst of such material; any injury to any part of it would be permanent, and, if exposed to pressure of water, might prove dangerous. A puddle wall requires to be backed up, on both sides, to at least double its own thickness, with fine selected material, so as to prevent direct pressure of water from the inside, or any drying, and consequent cracking on the outside. Six feet in depth of water will puncture 12 inches of sound puddle, if laid hollow over rubble-stone and loose earth like this bank; so that 16 feet of puddle would be liable to be punctured by a head of water of 90 feet, if the least flaw existed in any part, and the full pressure could act on it. The objectionable mode of laying the outlet pipes most probably fractured the puddle wall at the point of crossing. The loose state of the material at the top of the bank let in the water. As it rises in the reservoir, this water has most probably found its way down the face of the puddle to the fracture in the puddle wall above the outlet pipes, and hence the destruction so swift and terrible in its effects. Cast iron pipes ought never to be laid under such conditions as these were. A culvert of masonry, with an inner valve well, as in the Bradford reservoirs, should have been provided. This culvert should have been on one side of the valley, and in solid ground, free from the loose earth of the embankment. The lower 20 feet of any reservoir, formed on the plan of this at Dale Dyke, may, if required, be drawn down by a syphon arrangement, and all the valves may be within reach for examination or for repairs. The by wash arrangement at Dale Dyke was inadequate for the drainage area. The length provided, 64 feet, ought to have been not less than three feet for each l00 acres of drainage area, or 129 feet; extra power for lowering flood water during a storm should, even with such a length of by wash, be provided. The embankment was not properly designed. The material and mode of construction were alike defective. The following recommendation of the jury, "That in our opinion the Legislature ought to take such action as will result in a Governmental inspection of all works of this character, and that such inspection should be frequent, sufficient, and regular," has received our serious attention. We cannot, however, recommend it for adoption. Any approval of plans or casual inspection of waterworks embankments cannot insure ultimate safety in such works. The responsibility must remain, as at present, with the engineer and persons immediately connected with the works. Magistrates have jurisdiction under clauses inserted in recent Waterworks' Acts. In our opinion, a longer period than is usually inserted in such Acts for the construction of works of this character should be allowed, and arrangements should be made gradually to test the strength and soundness of the work. For this purpose, ample means to draw the water down should be provided, considerably below the full water level. We have, &c.

(Signed)                    ROBERT RAWLINSON.


   Web Page 25 Web Page 26 Web Page 27   
   Sheffield Flood - 1864   (Main page)    Web Page / Book Contents List
26  -  June  -  99   
Mick Armitage (e-mail)