Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 56, 1926
This text is also available in PDF
(553 KB) Opens in new window
– 363 –

Results given by Wells at Christchurch Pumping station, Beckenham.

[Read before the Philosophical Institute of Canterbury, 3rd December, 1924; received by Editor, 31st December, 1924; issued separately, 31st March, 1926.]

A Little while ago I was afforded facilities for examining the records of certain wells at the Christchurch pumping-station at Beckenham, and, as they seemed to afford data of purely scientific interest, I asked permission from the Mayor (Mr. J. A. Flesher) to embody some of the material in a short statement of the facts and of the conclusions to be drawn therefrom. This permission was kindly granted, and hence these notes.

The location of the pumping-station is on the east side of Colombo Street South, about half a mile from the foot of the Port Hills. On an area of approximately an acre, one 9 in., five 8 in., and two 6 in. wells were sunk, being spaced over the area at an average distance apart of about a chain. The water from these was allowed to discharge freely into a tank sunk 12 ft. below the ground-surface, which corresponds with the level of the engine-room floor of the station. This is taken as a datum for reckoning levels at the station, specially those in connection with the flow of wells. The floor of this tank is 3 ft. 9 in. above high

– 364 –

water, spring tide, at Heathcote. From the tank the water is pumped directly into a reservoir on the hills, or into the mains.

These wells were all sunk to the first stratum, and water was obtained in quantity from a depth of approximately 80 ft., the range being between 75 ft. and 85 ft. In order to augment the supply a 14 in. well was sunk at a point 5 chains east-south-east of the tank—that is, 2 ½ chains from the nearest 8 in. well—and yet another 6 in. well about 2 chains farther on in the same direction. In order to obtain a supply from a greater depth a bore was sunk to a depth of 610 ft. at a spot 5 chains north-east of the tank. The hope of striking a reasonable supply from a deeper stratum was not realized, and the pipes were drawn up to the 75 ft. level, from which a flow of over 300 gallons a minute was obtained even while the other wells were being drawn on.

The first point concerning which comment appears appropriate is the record of the beds passed through in the case of the deep bore. The log is as follows:—

Surface clay 3
Blue gravel 15
Blue sandy clay 41
Sandy gravel 64
Water gravels 86 Gave flow of water 250 gallons per minute at 78 ft.
Sandy gravels 113
Sand 162
Sand and gravel 169
Sand 174
Gravel and clay 189
Sand 193
Clay and yellow sand 229
Volcanic rock 234 Containing pebbles of basalt.
Fine blue sand 236
Rough gravel and sand 251 Water rises 9 ft. above surface, flows 30 gallons per minute.
Grey sand and gravel 278
Soft blue sand 305
Stiff blue clay 347
Soft blue sand and pebbles 355
Fragmentary volcanic rock 370
Soft blue sand 385
Sandy blue clay 443
Volcanic rock 445
Blue clay and sand 450
Sandy blue clay 536
Soft rock 538
Blue sand 550 No water. End of pipe.
Blue sand to 610 No, water. Last 60 ft. merely drilled.

This log shows a considerable difference in the beds as compared with those disclosed in the records given by ordinary wells of the Christchurch area. First of all, there is a marked deficiency in the thickness of the gravel and a great increase in the amount of sand and sandy clay. (For example, in record of well at Sydenham Water-tower, given in a paper by

– 365 –

the present author entitled “Preliminary Account of the Geological Features of the Christchurch Artesian Area”—Trans. N.Z. Inst., vol. 43, 1911, pl. 12, col. 4, facing p. 429.) An occasional shell-fragment was reported from the sandy beds, confirming the records of adjacent wells in which shells have also been encountered (loc. cit., pl. 12), and this almost certainly shows that the beds are marine in origin. It is possible that such shells may be fresh-water, but well-sinkers have told me that they are the same as those now found on the beach. I have seen none myself. This occurrence almost certainly shows that the sea extended over the area where the beds were being laid down, while gravels were being deposited either on a land surface or near the mouths of fast-flowing rivers in adjacent parts of the area. In the vicinity of the station they were probably laid down on a land surface, since totara-trees with their roots in position were encountered below high-water mark while the tank was being sunk. The marine beds just referred to follow right round the base of the hills towards Heathcote, and show that before the upper-stratum gravels were laid down a lagoon or arm of the sea, or perhaps an estuary, extended all round this sector of Banks Peninsula.

Another point concerns the presence of pebbles of volcanic rock at several horizons. These prove to be basalt or related basic rock similar to that occurring on the neighbouring Port Hills. Its presence shows that, with the exception of the upper gravels, a very important part of the coarse detrital constituent of the beds was derived from a source entirely different from that which supplies the gravel for the other part of the area. It also shows that these beds are later than the Lyttelton volcanics, a point which is apparently not very clear to some people. The beds of volcanic pebbles probably formed beaches along an old shore while finer-grained sands and sandy clays were being laid down in slightly deeper water. There does not appear to be any definite association of these beds with artesian water.

The levels at which such volcanic pebbles occur are to some extent accordant with those of other wells along the base of the hills. For example, such pebbles occur in the Beckenham bore at 234 ft. and 355–370 ft.; in the well at Heathcote Valley pumping-station at 355 ft.; and in well at Rat Island, in the Estuary, at 238–247 ft. and at 385–386 ft., the last-named according to the log of the bore kindly furnished me by Job Osborne and Co., well-sinkers. The presence of pebbles at these two levels perhaps indicates that the depth of water along the base of the hills was about the same for a considerable period in order to allow a general drift of coarse material for some distance off-shore. Volcanic pebbles occur at other levels in the sandy beds besides those just mentioned, specially just above the solid rock, where they are associated with weak flows of water, which probably finds its way down from the hills along its upper surface. The persistence of the volcanic pebbles at the levels just mentioned may, after all, be a mere coincidence and have no special significance.

The depths at which water is encountered in the fringe round the hills is also fairly close, allowance being made for slight variations in the level of the surface of the ground. In the Beckenham bore they are at 78 ft. and 251 ft., and similar occurrences can be seen in the records of wells given in my former paper (loc. cit., pi. 12, cols. 6, 8, 11, 12, 15). This may also be only a coincidence, but it is what might reasonably be expected, since marine conditions must have been much the same at the same levels off an old shore, in contradistinction to conditions when gravels and associated beds are laid down on a land surface. All the same, the shallow

– 366 –

stratum at Beckenham does not belong to the system fringing the hills, but is due to the invasion of the marine area by land beds. That this is really the case is shown by the roots of totara-trees in position, which were discovered when the excavations for the tank were being made.

It is thus fairly clear that the arm of the sea or an estuary was gradually encroached on by gravels brought down by rivers, which covered up the marine deposits and caused a gradual extension of the “first stratum” series of beds up to the base of the hills; so that first-stratum water is generally obtainable, except in such places as Heathcote, where the gravels have not extended far enough and where the topmost beds still consist of sands and sandy clays similar to those now being laid down in the adjacent estuary. The top layers in this place are merely the extension landwards of these later estuarine deposits.

This fringe round the hills, not intimately connected with the general artesian system of Christchurch, cannot extend out very far beyond the pumping-station, since a well 5 chains farther north-east beyond the deep 6 in. well gives a plentiful supply of water from a deep stratum not existing in the 6 in. well. In this case a 2 ½ in. pipe gave, when the well was first sunk, a flow of 100 gallons per minute from a depth of 365 ft., and water rose 25 ft. above ground-level. The first-stratum water of this well reached to only 2 ft. below the surface of the ground; but this level must be fully 7 ft. above the ground-level at the 6 in. well, since it is on a high terrace of the river, and the apparent failure may thus be accounted for. This prolific 365 ft. stratum has no counterpart whatsoever in the well at the pumping-station, for at that depth there occurs a bed of fragmentary volcanic rock, but no water. Similar records can be got from other wells just outside the station limits. which confirm the results from the well just cited.

Consideration was also given to the source of the water at the 251 ft. level. It appears to me that this cannot be credited to water which has come down from the Port Hills as distinct from that which supplies wells generally. The 6 in. well rises 9 ft. above the surface, and a similar rise is recorded in other wells—13 ft. in the case of the well at Mrs. Cocks's, and 15 ft. in the case of the Heathcote Racecourse well. Now, it does not seem to me at all possible that this rise can be obtained by the pressure due to a head of water from the hills. This would demand that the water-bearing bed should rise against their flanks for a distance of quite 20 ft., or even more, above the level of the plain, and there is no evidence of this. Rather it appears to me the water comes from the ordinary supply, but that, owing to the high coefficient of friction in such waterbearing beds as extend into the marine beds in this area, the supplies are comparatively poor. No doubt tongues of gravel, gradually diminishing in thickness and perhaps increasing in compactness as the percentage of finer material increases, were pushed forward from the land-surface which existed to the north and west, and along these the supplies of water found their way, but in diminishing quantity. Although these beds do not obtain their supply from the hills, there is every reason for thinking that the supplies obtained from just above the solid rock substratum do come from there; but solid rock was not reached in the deep bore at Beckenham, although it was in all probability not far away, so the question of a supply from just above it does not arise in connection with this discussion.

The next point to be considered is the amount of flow from the wells at Beckenham. This is extraordinarily great considering that it is from the first stratum. The records of the Venturi meters installed in the station show that when the surface of the water is lowered till it stands at 5 ft. 6 in. above the floor of the tank the yield from seven wells (not

– 367 –

including the 14 in. well) was 1,623 gallons per minute: with the depth of water in the tank standing at 4 ft. 6 in. the flow was 1,669 gallons per minute; and with the level of the water standing in the tank at 3 ft. and the 14 in. well flowing the total reached 2,686 gallons per minute, and this was while a 6 in. well 5 chains away was running to waste over 300 gallons per minute. As far as can be seen, this amount can be maintained, because, strange to say, the sinking of the 14 in. well resulted in increased flows being given by the others.

This amount is very great, seeing that it is practically the yield to be obtained from a little over an acre of ground; still, it is of the same order as is at times obtained from first-stratum wells in the Christchurch area: e.g., a 3 in. well in the Botanic Gardens gives a flow of 80 gallons per minute, and this flow has been maintained for several years. But there appears to be an extraordinary concentration at Beckenham, which may be explained in two ways:—

(1.) The major portion of the supply is obtained from a leak from a deeper stratum. Now, if this is the case it must be from a deep stratum not under the present wells but some distance farther from the hills. It might appear quite feasible that water should leak up the contact between the beds of the fringing area and those of the Christchurch area, were the contact definite, and thus fill the overlying bed. But if this is the case, then the pressure to be obtained from the first stratum should be much higher, unless friction is very great, or unless there is a ready outflow from somewhere near to lower the pressure. This may be furnished by the big springs in Beckenham Park, which lie approximately 37 chains southeast of the pumping-station and yield over 1,000,000 gallons per day. This water may, after all, not be a true spring, but a leakage from the river at a higher level coming directly across a loop. Also, I do not think there is a definite line of demarcation between the two sets of beds, but one dovetails into the other along their line of junction.

(2.) The increased supply is due to the obstruction afforded by Banks Peninsula to the movement of the underground streams. This may be explained in some such way as the following: In that sector of the plains lying behind the mass of Banks Peninsula the water is continually finding its way seaward, following along the different underground permeable beds. North and south of the peninsula there is no obstruction to this movement except the friction of the beds themselves. Owing to this friction they give a flow which diminishes progressively as the beds are followed north parallel to the coast-line past Kaiapoi to the line of the Ashley, and on following then south towards the mouth of the Rakaia. The prolific yields are in the vicinity of the volcanic mass, and especially in close proximity to it. This may be explained by the water being concentrated in more or less well-defined streams as it forces its way past the obstructions, and specially is this the case on the northern flank of the volcano; and where such concentrations are tapped great yields are obtained. This would also account for the springs at Beckenham Park and at Hoonhay. The latter he just where the obstruction is placed right across the normal path of underground water-movement, and the yield is very great—viz., 3,000,000 gallons per day. This appears to me the reason for the abnormal concentration at Beckenham. All the same, both causes may be operative at the same time, and the second cause may promote abnormal concentration of the flows in deeper beds as well as in the shallow ones; and the springs at Hoonhay may be due to an upward leak from a deep stratum, perhaps following initially along the contact of the gravel-beds with the underlying volcanics.

In order to test the interference of one well on the flow of another I suggested to the Engineer that tests could be made by shutting down sets

– 368 –

of wells progressively and noting the effect. The following results have been obtained:—

First of all a comparison was made between the original four wells put down and a subsequent three wells. The former constitute an inner ring consisting of two 6 in. and two 8 in. nearest the station; included in this should be a 9 in. well, but this is unsatisfactory, and yields only 20 gallons per minute, and for that reason is excluded from the test. The wells are uneven in yield, and the best results are obtained from one of the 6 in. wells. The latter set of wells consists of three 8 in. wells placed on a line approximately 3 chains from the tank. The tests were made with the water at an average depth of 2 ft. 6 in. in the tank, the measurement made by noting the time the water-level took to rise from 2 ft. to 3 ft.

The four original wells (two 6 in. and two 8 in.—all other wells shut off) gave 1,538 gallons per minute; the three 8 in. wells (all other wells shut off) gave 1,205 gallons per minute: a total of 2,763 gallons per minute. The whole seven (all other wells shut off) gave 2,205 gallons per minute. This means a loss of about 20 per cent. due to interference of one set with the other. The 14 in. well and the 6 in. coupled with it (all other wells shut off) gave 846 gallons per minute. All wells open except the 9 in. gave 2,686 gallons per minute. This amount combined with the previous total from the former set of seven wells should be 3,051 gallons per minute, which shows an interference of only 8 per cent. The small amount is probably due to the last two wells being an average distance of 3 chains away from the set of 8 in. wells. This interference is comparatively small.

Both these results are apparently lower than the values given by Schlichter (Theoretical Investigation of the Motion of Ground-waters, 19th Ann. Rep. U.S. Geol. Sur., 1897–98), and are indicative of the largeness of the supply.

Considerable value is attached in some quarters to the amount of water yielded per square inch of well cross-section. Of course, there is a falling-off in the quantity reckoned this way in the case of larger wells, and therefore it may appear at first sight that the advantage lies with a number of smaller bores rather than with a smaller number of larger bores of equal total sectional area. No doubt the smaller bores exhaust an area more thoroughly if a sufficient number are sunk, but then the question of cost arises, and the larger wells may be more economical in the long-run as giving a greater return per money expended. There is another advantage in the case of wells in the Christchurch area—viz., that large-bore wells, if not so prolific per unit of area, yield their water quietly, and so a smaller quantity of sand, and at times of stones, is carried into the settling-tank. If no settling-tank is provided, then the wear on pumps, valves, &c., must be serious. Oscar Edward Meinzer, in his paper on the Occurrence of Ground-water in the United States (U.S. Geol. Surv., Water-supply Paper 489, 1923 p. 119) says: “By coming into wells after they are finished, it [the sand] damages the pump, sometimes erodes the casing, and frequently clogs the wells.” The first two reasons appear to be of special importance in the case of wells in the Christchurch area, so that any means of diminishing the disability arising therefrom, unless too great an expenditure is incurred, should be followed out. By sinking wider bores the rush of water is reduced, and the objection to the presence of the sand in pumps, &c., partially or completely removed.

In concluding this brief statement I have to express my indebtedness to the Mayor for kindly allowing me to use the records; also to Mr. C. Dawe, the City Surveyor, and to Mr. F. Mellish, Engineer at the pumping-station, for substantial and ready assistance whenever they were called on.