Open Collections

BC Historical Books

BC Historical Books

BC Historical Books

Vancouver water works : excerpt minutes of the transactions of the [Canadian Society of Civil Engineers].… Smith, Henry Badeley 1889

Item Metadata


JSON: bcbooks-1.0222198.json
JSON-LD: bcbooks-1.0222198-ld.json
RDF/XML (Pretty): bcbooks-1.0222198-rdf.xml
RDF/JSON: bcbooks-1.0222198-rdf.json
Turtle: bcbooks-1.0222198-turtle.txt
N-Triples: bcbooks-1.0222198-rdf-ntriples.txt
Original Record: bcbooks-1.0222198-source.json
Full Text

Full Text

  The Society will not hold itself responsible for any
statements or opinions which may be advanced in the
following pages.
" The papers shall he the property of the Society,
and no publication of any papers or discussion shall
be made, except by the Society or under its express
permission."—By-Law No. 40. Paper No. 35.
By Henry Badeley Smith, M.Can.Soc.C.E.
Previous to the year 1886, the City of Vancouver, British Columbia,
had no existence. Where this city now stands, wa3 then a dense,
tangled forest of huge fir, cedar, spruce and hemlock; the only evidence of the presence of man being a clearing a few acres in extent, on
which low frame buildings, not more than a dozen in number, had been
erected, and which was vaguely known to the outside world as Coal
Harbour, Gas Town, and the Granville Town Plot.
At this date the Canadian Pacific Railway terminated at Port
Moody, a small town at the extreme head of Burrard Inlet, 18 miles
from the Gulf of Georgia. The Company, desiring a terminus nearer
the open sea, negotiated with the legislature of British Columbia for a
grant of land in the neighbourhood of the Granville Town Plot.
The Government, foreseeing that a large city would speedily be built
up at the terminus of this great trans-continental railway, were it
located on the best attainable site near the sea, voted the grant by a
large majority, stipulating only that the extension from Port Moody
westward to the lands granted should be constructed and in operatiou
by a stated time. When it became known that the terminus of the
railway would undoubtedly be at the Granville Town Plot, population
began to pour in so rapidly that, on April 6th, 1886, the Legislature
passed an act incorporating the locality as the City of Vancouver.
The population at that date did not exceed two thousand. So great,
however, has been the influx of all classes, that at the time of writing,
it is estimated on reliable data, that no less than ten thousand souls
are contained within the limits of the city.
The City of Vancouver is situated on the south shore of Burrard
Inlet, in Lat. 49°, 16', 31" N, Long. 123°, 05', 52" W, its western
boundary being 3|- miles east of the Gulf of Georgia. It is distant
from Liverpool on the east 6116 statute miles, and from Yokohama on the west 4991 statute miles. Prom Montreal to Vancouver
is 2905 miles, and from New York, via Canada, to the same point is
3162 miles.
JO JiJ Smith on Vancouver Water Works.
Burrard Inlet is the first harbour of magnitude on the Pacific
mainland north of the United States. It is easy of access to vessels of
the deepest draught, and safe anchorage can be found in any part.
English Bay, the entrance to the Inlet, is 4^ miles long and 4 miles
wide. At its head it divides into two branches,—False Creek on the
south, and the First Narrows on the north. False Creek is a narrow
arm 4J miles Ions:, extending due east from English Bay, midway
between the North Branch (Burrard Inlet proper) and the south
boundary of the City of Vancouver. Being almost uncovered at low
water, it is unsuitable for navigation.
The north branch, which leaves English Bay for the First Narrows,
extends due east a distance of 14 miles. The width of the Narrows
at extreme low water does not exceed 1086 feet, whereas a mile and a
half inland it reaches 12,210 feet. Soundings of 120 feet can be
obtained at the entrance, and 234 feet at the outlet opposite Vancouver.
The land between Burrard Inlet and False Creek, on which the
present Vancouver is built, is for the most part flat, the highest elevation above sea level not exceeding 145 feet. South of False Creek,
however, a rapid rise takes place, terminating in a table-land 200 feet
above sea level. A few small streams run down from this table-land
into False Creek; but these are insignificant, and cannot be utilized for
manufacturing or other purposes. The nearest river on the same side
of the Inlet on which Vancouver is built, passes 15 miles to the westward.
The subject of a good and sufficient water supply for the City of
Vancouver, or to write more accurately, for the place now known as the
City of Vancouver, was first taken into earnest consideration by Mr.
G. A. Keefer, M. Can. Soc. C. E., in June, 1885, nearly a year
previous to the incorporation of the city. Mr. Keefer, foreseeing at
that early date that the ultimate destiny of the Canadian Pacific Railway was to reach a point nearer the coast than Port Moody, and
knowing that the Granville town-site possessed all the requisites for the
foundation of a large city, interested himself in obtaining information
as^to the best source of a water gTippIy for that locality, should the
Railway Company decide upon it as the terminus of their system. He
speedily ascertained that no supply could be advantageously and economically obtained on the south side of the Inlet, where the city must
necessarily be located, no streams or lakes of any magnitude existing in
the vicinity. Smith on Vancouver Water Works. 5
He therefore directed his attention to the nojjb_jide of theJLnlet,
although confronted at the very outset by the fact that, never before in
the history of hydraulio engineering had a system of water mains been
laid across such a sheet of water as Burrard Inlet, and under such
conditions as pertained thereto.
Acting under instructions from Mr. Keefer, the writer placed a fully
equipped party in the field, in the winter of 1885-86, and thoroughly
examined all the streams flowing into the Inlet immediately opposite
the Granville town-site, from the lofty chain of mountains on the north
' The results obtained from this survey showed that of all the streams
available, the River Capilano. falling into the Inlet at the First Narrows
nearly opposite the western boundary of the present City of Vancouver,
was the most suitable, the discharge being much greater than that of
any of the others, and the average fall of the river so great that an
initial point for a gravity system of water supply could be obtained
within a reasonable distance upstream.
Having decided on utilizing the waters of the Capilano for the
supply of the future city, Mr. Keefer experienced no difficulty in obtaining the co-operatioiLofseveral prominent and enterprising capitalists of
Victoria, who were quite in accord with him in the belief that at a very
early day a large population would be located at the Granville town-site,
and that an immediate outlay for an efficient system of water works
- would be a remunerative investment.
— Accordingly, the extension of the railway to the Graiville town-site
being an assured fact, and the future name of that locality being definitely decided on as the City of Vancouver, these gentlemen applied to
the Provincial Legislature for an act of incorporation of a company, to
be known as the Vancouver Water Works Company, and proposing to
construct a gravity system of water works, tor the purpose of conveying
water from a point on the River Capilano, on the north side of Burrard
Inlet, to certain specified lots in the New Westminster district on the
south side of Burrard Inlet. About the same time, application was
made by the inhabitants of these lots for an act of incorporation under
the name of the City of Vancouver. Both requests were granted by
..the legislature on the same day, the 6th of April. 1886.
During the summer of 1886, the writer, acting under instructions
from Mr. Keefer, made detailed surveys, definitely locating the point of
supply on the River Capilano, and the crossing of Burrard Tnlafc Tn
June, 1887, the whole system was finally staked out, and contracts
entered into for clearing, close cutting and grubbing.    In December.
hmjWiws 6
Smith on Vancouver Water Works.
1887, a permanent Board of Directors was formed, comprising the
following gentlemen: President, Capt. John Irving; Directors, The
Hon. (now Sir) Joseph W. Trutch, Messrs. R. P. Rithet, G.
A. Keefer, Thomas Earle and D. M. Eberts; Mr. J. W. McFarland
being appointed Secretary; Mr. D. M. Eberts, solicitor; Mr. G. A.
Keefer, M. Can. Soc. C.E., chief engineer; and the writer, Mr. H. B.
Smith, M. Can. Soc. C.E., engineer in charge.
The River Capilano is a mountain stream of considerable magnitude.
Prospectors who have penetrated its canons, and claim to have reached
its source, estimate its length at no less than fifty miles, It rises in
the snow-covered mountains of the Howe Sound district, and flows
almost due south, emptying into Burrard Inlet at the First Narrows.
Although nothing definite is known as to its source, all accounts
agree that its origin is not a mountain lake, but the accumulated
waters derived from melted snow and ice falling from the mountain
summits. For a distance of seven miles from its mouth, the river has
been surveyed. Throughout this distance it flows at the average rate
of five feet per second over a bed of granite, basalt and conglomerate
boulders. Sand and gravel can be found only in a few sheltered bays.
It passes through several cafions of granite and whinstone rock, one of
which is only 15 feet wide at its base, 94 feet wide at its top, 500 feet
long, and 218 feet deep. Previous to the creation of this canon, the
whole valley to the north must have been one large lake. The wall of
rock through which the stream p nctrated ages ago, by some sudden
effort of the earth's hidden forces, stands like a huge gate at the south
end of the valley, the valley itself being but a strip of flat land from
1,000 to 1,500 feet wide, lying at the base of two parallel ranges of
mountains, which tower upwards to a height of 3,000 feet. The fall
that took place when the river flowed over the summit of this rocky
wall must have equalled the Niagara of to-day for depth, if not for
volume. Should the City of Vancouver increase to the magnitude pre-l
dieted, it may be that its people at some future day will cause a dam
to be constructed across the narrow gorge, and once again convert this
valley into a lake. Vancouver will then possess a reservoir from
whence to draw its water supply, which will not be surpassed by any
water works system on the continents These cafions are isolated, standing about a mile apart. Between them the river flows through low
lying flats, forming many islands. The immediate banks are but a few
feet above the level of the river, and from 100 to 200 feet in width,
aamiiiiiinninimmr* Smith on Vancouver Water Works. 7
the ground on each side rising in terraces until it is merged in the
uniform slope of the mountains. Both sides of the river are heavily
timbered with the huge trees peculiar to the British Columbia coast,
Douglas fir, cedar, hemlock, spruce, balsam and white fir being in
abundance. The Douglas fir and cedar grow to an enormous size.
One cedar in particular was measured by the writer, and found to be
€4 feet in circumference, 4 feet from the ground.
As a source of a city water supply, the River Capilano is an ideal
one. No purer water can be obtained from any source than that from
this mountain stream, flowing swiftly over a boulder bed, through deep
rocky cafions, and along shores as yet uncontaminated by the impurities which follow in the wake of settlement. The supply afforded, being
by gravitation, is superior to all other methods, whether by reservoir,
direct pressure, or stand pipe, and its permanence is beyond question,
careful gauging of the river at the initial point of the system having
demonstrated the fact, that at the lowest stage of water the river discharges 440 millions of gallons in 24 hours.
The first contract entered into by the Company was for clearing, close
cutting and grubbing. This work was done by a local firm at the
following prices : clearing, $59.00 per acre; close cutting, $95.00 per
acre; grubbing $200.00 per acre, under the conditions of the following
The pipe track is to be cleared a width of not less than 33 feet, and
all timber and brush, not required for the purposes of the work, piled
up and burned, as in clearing land for cultivation.
The dam site is to be cleared in the same manner, and to such limits
as may be directed by the engineer.
Whenever embankments, occurring on the line of pipe track or tram-
wa y, are less than two feet in height, all the trees, stumps and brush
immediately under the embankment, are to be cut close to the ground,
and whenever the embankments are from two to four feet high, they
shall be cut within six inches of the ground; but when the embankments exceed four feet in height, chopping as for ordinary clearing will
be allowed.
Grubbing shall be performed under the seats of the embankments
occurring on the line of pipe track, or tramway, that do not exceed one
foot six inches in height, and also all excavations for pipe track, tramway and dam embankment, less than three feet deep. The stumps and
ro ots from the grubbing shall be removed to such places as directed. 8
Smith on Vancouver Water Works.
No Chinese are to be employed, directly or indirectly, on the above
The point on the river selected as the source of supply is at a distance of 63 miles upstream from its mouth, where the river is confined
to one channel, and the banks on either side are sufficiently high to
admit of the construction of a dam.
The locality selected is the only point from the river's mouth upwards
where a dam could be safely and economically constructed, and give at
the same time a sufficient head to overcome the elevation of the high-
flats 'U- miles below it.
By reference to Plate XVI, which shews the dam site and its
vicinity, it will be seen that immediately south of the site the river is-
divided into two wide channels.
Still further south, all the way to the can on below, it is divided into
three and even four channels. Similarly, north of the dam site, the
river has two branches separated by a large, low, flat island. This
island is completely covered at high water, making the river at that
stage no less than 830 feet wide.
The cross section of the river at the dam site at low water gave a
current of 4J feet per second, a width of 100 feet, and an extreme
depth of 3 feet, the difference of level between low and high water being
6 feet. It has been subsequently ascertained, however, that during
occasional floods the water rose much higher, and covered the level flat
on the north side to a depth of 2 feet. This flat stands at an average
level of 12 feet above low water. The bed of the stream consisted of
large granite boulders, closely packed together, small stones and coarse
gravel filling up the interstices. The channel of the river in ordinary
floods was 210 feet wide.
On the north shore the immediate bank is 12 feet high, and extends
at the same level a distance of 140 feet inland. A sudden rise then
takes place, terminating in another flat 40 feet above low water, and
which stretches to the base of the mountains.
On the south shore, the bank rises abruptly to a height of 22 feet
above low water, and continues at that elevation for 200 feet. It then
rises rapidly in terraces till it reaches the mountain side hill. The high
land on the north shore trends to the northward immediately west of
the dam, and that on the south to the southward, immediately east of
the dam. Smith on Vancouver Water Works.
The dam site lies directly between these two high points. The contract for the construction of a stone-filled timber dam at the point
selected was let on the 24th of January, 1888, to Messrs. H. F.
Keefer and D. McGillivray of Vancouver, and was most satisfactorily
completed by them on the 18th of April following. The difficulties
encountered by the contractors in carrying out this work were of no
ordinary character. Inasmuch as it was the initial work of the system,
and located in a wilderness in which no roads existed, all supplies, tools
and machinery were of necessity packed to the works on the backs of
mules. The season was mid-winter, and unusually inclement. Chinook
winds and heavy rain-storms, melting the snow on the mountain summits, caused frequent freshets, in which the river would rise from 6 to
10 feet in a few hours time.
The formation of the banks in the vicinity did not admit of the river
being temporarily diverted, except at enormous cost. The foundations-'
of the structure had therefore to be excavated, and the first courses-
laid in from 3 to 4 feet of swift running ice cold water.
Plate XVII is a reduced copy of the working plan of the dam.
It will be seen that the structure is of continuous cribbing, stone filled,
planked and sheet piled. It consists of three principal parts, viz., the-
north abutment, the tumbling way, and the south abutment.
The north abutment is located well inland, owing to the tendency
of the river in high floods to over-run its channel, and spread over the
low lying land in the vicinity. For the purpose of description it may
be subdivided into the following heads: The abutment proper, the
well chambers, the settling pond, the pipe outlet, and the north wing..
The abutment proper is a right rectangular prism 41'. 2" x 20' x
18'. 9", constructed of round timbers, laid in alternate courses of cross
ties and longitudinals, dove-tailed at the angles, and forming 28 cribs,
which are filled up with heavy stone filling and coarse gravel, the latter
being rammed into all interstices between the stones and under the timbers. A space equivalent to four cribs, in the exact centre of the abutment, is floored and walled, from the foundation upwards, with double
2" planking over-lapping. A perfectly watertight chamber 10'. 6" x
7'. 10" is formed. This chamber is subdivided into two smaller and
equal ones by parallel walls, 4" apart, of double 2" planking overlapping,,
and placed at right angles to the length of the main chamber. These
constitute the well chambers, by means of which the water from the
reservoir formed by the dam is conveyed into the mains. An influent
conduit of double 2" planking overlapping, 15' 5J" long, and of area
sufficient to admit a larger volume of water than can be discharged by
ess 10
Smith on Vancouver Water Works.
the mains, connects the first of these chambers with the settling pond,
and consequently with the reservoir in front of the dam. In the 4"
space between the double central walls, close to the floor of the chambers, are placed double fish screens of the same area as the influent conduit, and so arranged that they can be easily removed, one at a time,
for the purpose of cleaning. The first or outer screen is coarse, being
of No. 12 copper wire, woven into meshes one inch square. The second
or inner screen is finer, being of No. 15 copper wire, 6 meshes to the
inch. The rear of the second chamber is pierced exactly opposite the
fish screens to admit of two bevelled 22 inch rivetted steel pipes, the
mouths of which are opened or closed at will by means of timber gates
eliding in vertical uprights attached to the walls of the chamber.
Two trap doors cover the top of the chambers, and over all, resting
on the top courses of the abutment, is built a compact water-proof shed
12' x 13' x 13'. This shed serves for a tool house, as well as effectually preventing " the access of strangers to the gates which control the
In front of the influent conduit is a triangular shaped settling pond,
measuring 15J feet at the base, 16 feet from base to apex, and 14' 2"
•deep. It is constructed of longitudinal timbers and cross ties, laid one
above the other, the whole being firmly bolted to the face of the abutment. At the apex the ends of the longitudinals are dressed, so as to
fit closely, and bolted together. The triangular space between the apex
and the apex cross ties is filled with large boulders, for the purpose of
giving weight to- the structure, and retaining it in position.
At the base of the pond, the entrance, of water into the influent conduit is controlled by means of a timber gate, sliding in vertical runners
bolted to the sheet piling on the face of the abutment. Immediately
behind this gate covering the mouth of the conduit is placed a cast iron
grating with 4-inch openings. The water from the river has free access
to the settling pond through the spaces between the longitudinal timbers
of the walls. The main object of its construction is to prevent logs and
.floating debris from accumulating in front of the influent conduit. It
will thus be seen, that, in order to reach the mains, the water must first
enter the settling pond, then pass through the iron grating at the mouth
of the influent conduit, then, by means of that conduit, enter the first.
Wfell chamber, then through the double fish screens in the central walls
into the second chamber, and finally into the mains in the pipe outlet.
The pipe outlet at the rear of the north abutment is a crib continuation of that abutment, serving as a protection for the mains against the
action of the water flowing over the tumbling way, until a safe point Smith on Vancouver Water Works.
is reached on the flat below. It is 138 feet long, 15 feet 3 inches wide,
10 feet high on the side facing the river, and 6 feet on the land side.
It has three parallel rows of longitudinals supported on cross ties, the
two outside rows, or the rows nearest the river forming cribs 4'. 9" x
3'. 5" x 10', which are heavily loaded with boulders. Between the cribs
and the third row of longitudinals on the land side, is a space 8 feet'
wide, in which the mains leading from the well chambers are laid.
Provision is made for two mains, but only one is in use at present,
the other being capped at its lower end, and closed at its mouth by
means of its gate in the second well chamber. The space containing
the two mains is filled with coarse gravel, well packed. Above the
filling is a covering of 15 inch logs close laid.
In the immediate rear of the abutment the timbers of the pipe outlet are continued upwards in steps to the top of the abutment, forming
a " lean to," which prevents the water, flowing over the tumbling way,
} from flooding the top of the pipe outlet.    The  " lean to, "  as well as
fthe entire face of the pipe outlet, is planked with 3 inch planking, sunk
3 feet below foundation level.
The low lying porous nature of the ground on the north side of the
river rendered necessary the construction of an extensive land wing,
with deep foundations. This wing is 155 feet long, and 10 feet wide.
'J he first 20 feet out from the abutment is 16 feet 11 inches high, and
ps in reality part of the abutment proper, its longitudinals being a continuation of the longitudinals of that structure. The remaining 135
feet, being built on higher ground, has a uniform height of 7' 9"
Both portions are built in rows of parallel longitudinals, 3 in number,
and in lengths of 31 feet, supported on cross ties 10 feet long, and 5
feet apart. These form 62 cribs, which are filled with stone and gravel
as previously described.
The connection between the wing and the high land at its extremity
is protected by a gravel enbankment, extending 57 feet along the face
of the wing. This embankment is made of picked material, and effec
tually prevents all seepage round the end of the wing. The face of
both abutment and wing is protected from leakage by a doullo row of
sheet piling, the lower ends of which are embedded in a conci ete trench
sunk 3 feet below foundation level. The inner sheet piling is 2 inches
thick, while the outer and overlapping piling is 1 inch.
The main body of the dam, technically named the Tumbling Way,
is 165 feet in clear length, 41' 2" broad, and 13' 9" high in the deepest part of the original channel of the river. Great difficulty was
experienced in excavating foundations for this portion of the dam.    At
I 12
Smith on Vancouver Water Works.
first an effort was made to partially divert the river by excavating a
new channel, between high and low water mark on the south shore,
the intention beins;, if this succeeded, to excavate the foundations and
build the sub-structure up to the toe of the front slope ; then to return
the river back to its original channel, allowing it to flow through the
row of horizontal openings provided in the design of the structure for
that purpose. It was found however that the bed of the proposed
diversion, being entirely composed of loose boulders, was too porous to
admit of the water being confined within the excavation ; and as, at
that time, no clay, fit for puddling, was known to exist in the near
neighbourhood, this project had to be abandoned. The method then
adopted and which proved successful, though carried out under great
difficulties, was as follows :—
Both abutments having been p irtially constructed, the foundations
for the end divisions of the tumbling way were excavated as far as
possible from the abutments towards mid-channel. As much of the
structure as the excavations could contain was rapidly built up, and
loaded with stone filling. An embankment of gravel and sand was
then run out from each extremity, meeting about 20 feet up stream and
forming a V, the apex of which divided the current of the river, and
forced it through the horizontal openings in the sections already built.
This had the effect of leaving still water three feet deep behind the
embankment, and as this could not be removed, nor lessened in depth, the
foundations were excavated an 1 the mid lie section built under these
exceptionally difficult circumstances.
The sills of the north and south sections are on the same level,
while those of the middle section in the deepest part of the river bed
are 2' 2" lower. The cross sections of the three portions are similar.
Plate XVII shews that of the middle section.
The ground sills, 10 in number, in lengths of 32 feet, are placed at
right angles to the stream, at distances varying from 5' 5" to 6' apart,
the distances varying in order t) secure a row of longitudinals under
each vertical angle of the surface of the tumbling way. Above the
sills and at right angles to them are placed a row of cross ties parallel
with the stream, each 53 feet long, and from 5' 8" to 6' apart. These
project 11' 10" to the rear of the main body of the dam, resting on two
of the sills of the ground course. The spaces between these projections
are filled in with round timbers laid close. A solid close laid platform,
to the rear of th? main boly of the tumbling way, is thus formed, which
serves to dissipate the force of the water flowing over the tumbling way
before it reaches the bed  of the  river.    The next or third course Smith on Vancouver Water Works.
consists of eight longitudinals, above which, on the fourth course, are
the horizontal openings previously mentioned. These arc 28 in number,
five feet wide, 12 ins. deep, and extend entirely through the structure
from its upstream face to the open river in the rear.
They are formed by flooring the spaces between the cross ties of the
4th course with double one inch planking, and close laying the longitudinals of the 5th course to serve as a covering. Above the 5th cour-e the
longitudinals and cross ties are so arranged that the front face slopes
upwards to the ridge at the rate of 2 '3 \" to 1'. The longitudinal which
constitutes the ridge is placed at a horizontal distance of 17' 2J" from
the front face, and is at an elevation of 415 feet (surface planking not
included) above high water mark of Burrard Inlet. The rear slope
extends downwards from the ridge at the same rate as the front slope,
and terminates in a level bench 12 feet wide.
In the tumbling way there are 196 cribs, formed by the intersections
of cross ties and longitudinals.    Especial care was exercised in filling
or ©
these cribs. As each course was completed, the largest boulders
obtainable were placed in the cribs by hoists. The spaces between
were filled up with smaller stones and coarse gravel, the latter being rammed into every crevice. In excavating the foundations, certain huge
boulders, which were found to be firmly anchored in the river bed, were
blasted into a columnar shape, so that the bed sills and cross ties when
laid would enclose them. These not only served as stone filling, but al~o
securely locked the whole structure to the bed of the river in a much
more substantial manner than any artificial means..
The whole surface of the tumbling way is covered with 3 inch
planking, jointed and laid close. The upper half of the front slope,
being exposed to floating logs, is laid double. The verticil part of the
front face is protected by 1" and 2" sheet piling, embedded in a concrete
trench three feet deep, extending over the whole length of the structure.
Inasmuch as it was necessary to keep the horizontal openings open
until the whole dam was completed, the placing of this sheet piling was
done in two operations.
The lower portion of the piling below the level of the floor of the
openings was placed in position in the usual manner, the tops being
dressed to a uniform level. A longitudinal 12" by 3"plank, extending
over the whole length of the tumbling way, was spiked to the tops of
this sheet piling, projecting one inch above, and forming a groove into
which the upper sheet piling would fit when placed in position. When
the proper time arrived to close the openings, a sufficient number of
men were ranged along the toe of the front slope,  provided with the mBts
Smith on Vancouver Water Works.
proper lengths of sheet piling, spikes and hammers. On a given signal
each plank was pushed home into the groove below the openings, and
the necessary spikes driven into the top ends. It required only five
minutes to complete the whole operation, and by that time, the water in
front had not risen above the toe of the front slope.
Immediately in front of the tumbling way is an apron of brush,
gravel and boulders. This apron extends from the settling pond in
front of the north abutment clear across the face of the tumbling way
to the gate of the sluiceway. In cross section, it begins at a point halfway up the front slope, and extends horizontally a distance of 9 feet.
It then slopes down to the bed of the river at the rate of 3 to 1.
The south abutment, being partially let into the high land, required
no wing extension. Properly speaking, it consists of three distinct parts,
viz., the abutment proper, connecting with the tumbling way; the
land abutment, connecting with the shore; and the sluiceway, which
lies immediately between the two. The foundations of all three are on
the same level as those of the north abutment, and being above low
water mark were excavated without trouble.
The   abutment   proper is a  rectangular  prism 41'  2" x 15'   x
18' 9" constructed of longitudinals and cross ties in alternate tiers, bolted
together and dove-tailed at all four corners.    As in   the  north  abutment, the longitudinals of the tumbling way at regular intervals project
into  the abutment, and  are securely  bolted to  it,  thus forming an
absolute and immovable connection between the three structures.    In
this abutment, there are in all 21 cribs, each 5' 8" x 4' 7" x 18' 9,"
filled and rammed as previously described.    In the  rear of the abutment is a | lean to, " 31 feet long, and tapering from 15 feet at the abutment to 11 ft. 7 ins. at its extremity.    This also is a stone filled crib
structure, the object of which is to prevent any scouring that might take
place, by guiding the water flowing over the tumbling way beyond the
rear of the abutment, and into the original channel of the  river.    It
may be here mentioned that one year after the completion of the dam,
a large scour did take place in  the angle formed by the foundation
courses of the | lean to "  and the rear platform.    During a sudden
freshet the bed of the river at this point scoured out to a depth of 4
feet below foundation level.    The end  cribs of the   " lean to "  were
completely undermined, the stone filling carried away, and the timbers
left unsupported.    A somewhat similar occurrence had taken place a.
few months previously at the anjjle formed between  the  rear platform
and the pipe outlet on the north side.    The latter was readily repaired
by filling in and constructing a triangular extension of the rear platform. Smith on Vancouver Water Works.
as shewn in drawing.    In this case the extension could be easily bolted
to the existing platform and the pipe outlet.    But in the case of the
first   mentioned  scour it was quite different.    The "lean to" being
an  addition to the rear of the abutment and not a part of it, timbers
extending from its extreme end to the rear platform, so as to cover the
large scour made, and prevent further injury, would have been insecure.
Instead,   therefore, the damage done was repaired by refilling the
scour with a mixture of large boulders and concrete, the latter being in»
the proportion of 1 part of pure cement to 7 of coarse gravel and sand.
Over this filling, and extending three feet beyond the rear of the " lean
to," was placed a covering of almost pure cement, 1 foot thick.    Twenty-
one barrels of Portland cement, each weighing 400 lbs., were used in
making these repairs.    The total length of the abutment and I lean to"
combined is 71 ft. 11 ins.    It therefore projects beyond the rear of the
tumbling way, a distance of 31 feet.  Both sides and rear, as well as the top
of the "lean to," are planked with 3" planking laid close.
The sluiceway is 73 feet long and 14 feet in clear width.    From wall
to wall it is 15 feet wide, and at the upstream end is the full height of
the abutments.    Both walls and face are planked with 3" planking, laid
close.    It is opened and shut by means of a stop log gate, consisting of
17  stop logs 17' 4" x 12" x 12", placed horizontally one above the-
other, each capable of being moved vertically in a groove formed by
vertical 12" | 12" uprights, let into the walls of the abutments on eaeh
side.    On the upstream face the uprights are single, connected at the
base by a 12" x 12" sill.    Behind the stop logs the uprights are double,
while  midway  between is a triangular truss of framed 12" x 12"
timbers, planked with 3" planks, the sill of which extends back from the
rear of the stop logs, a distance of 17| feet, and is securely bolted to the
ground  flooring.    The floor sills beneath the truss are close laid on a
concrete bed, forming a solid apron, on which the force of the water
falling over the gate when partially open is spent previous to discharge
into the channel of the river.    From the end of the truss to the outlet
of the sluiceway, sills are laid four feet apart, extending underneath and
bolted to the sills of the walls, or in other words to the sills of the abutments on each side.    The two sills immediately behind the rear uprights-
of the gates, and the three sills at the end of the close laid flooring are
squared 12" x 12" timbers, 43\ feet long, and pass under the whole-
width of both abutments.    Similarly two caps 43J feet long are laid
across the top of the sluiceway, behind the rear uprights of the gate.
These sills and caps are securely bolted to every intersecting timber of P§s
Smith on Vancouver Water Works.
the abutments on each side of the sluiceway, thus making a solid union
between the three parts.
Above the stop logs is a powerful windlass, with supports on each
abutment, the roller being directly above the stop logs. The upper
surface of each stop log is provided with a wrought iron ring at each
end, the stop log immediately above it being grooved on its under face,
so as to admit the rings, when the stop logs are in position, and the
gate is closed. The extremities of the chains connected with the windlass are provided with clutches which can be readily guided so as to
hook on to the rings, when it is required to open or close the gate.
The sluiceway abutment, or that portion of the south abutment
which connects directly with the land, having to withstand much less
pressure than other portions of the dam, is not of uniform height, but is
built in steps. At the upstream end it is of equal height, 18 ft. 9 ins., with
the main portion of the abutment on the other side of the sluiceway,
and 13 feet wide, while at the extreme rear, the height is only 5 feet,
and the width 8 feet. It consists of 16 separate cribs, loaded with
stone and gravel, as previously described.
The whole abutment, including the sluiceway, is protected in front
by 1" x 2" sheet piling overlapping and imbedded in concrete, as in
the case of the tumbling way and north abutment. This concrete is
in the proportion of 1 part of cement to 5 of gravel and sand. The
manner of its preparation was as follows : moist gravel of suitable nature
obtained from the river bank was deposited on a plank platform 10
feet square. This was thoroughly worked with shovels, and all stones
larger than 1^- inch diameter eliminated, leaving the mass spread over
the platform about 9 inches deep. The proper proportion of cement
was then spread over the gravel, in a dry state. Very little water was
used, the moisture in the gravel being sufficient for the purpose. Six
men with shovels then energetically worked the whole mass, shovelling
from the outside edges towards the centre. When satisfied that the
mass had been completely turned over once, it was flattened out on the
platform, and again turned over in the same manner. This operation
was repeated three times, the mixture being then considered fit for use.
The concrete trench mentioned above, extends along the whole face
of the dam below the level of the sills, forming a perfectly watertight
connection between the foundations and the bed of the river, through
which no seepage can take place. Seepage round the extremities of the
abutments, where they penetrate the banks, is prevented on the north
side, as previously stated, by a gravel embankment. On the south
side the same purpose is served by a hand-laid stone wall, built in the
'^S Smith on   Vancouver Water Works.
angle formed by the extremity of the abutment and the natural bank
of the river, fine gravel and earth being filled in behind and well
The reservoir created by this dam is, in the high water season, 380
feet wide by 700 feet long, and contains approximately fourteen millions
of gallons.
At low water the elevation of the water flowing over the crest of the
tumbling way is 483 feet above the lowest depression in the pipe line,
417 feet above the lowest level in Vancouver, 317 feet above the average, and 201 feet above the highest. These elevations correspond to a
maximum pressure of 210 lbs., an average pressure of 138 lbs., and a
minimum pressure of 87 lbs. per square inch.
The wrought iron drift bolts used were of ■£" and %" round iron, and,
of lengths varying from 12" to 32y.    Spikes for 3'' planking were 6"'
long, weighing 11 per pound, and nails for \" planking were 4§-" long
weighing 19 per pound.
From the above description it will be seen that the extreme length
of the dam, from land connection to land connection, is 384 feet, the
clear tumbling way 165 feet, supplemented by an additional 14 feet o
sluiceway, when required, and the breadth of base, not including rear*
platform 41 ft. 2 ins.
The total cost amounted to $15,039.26.
The country traversed by the mains, from the dam to the central
point of the city was, from a hydraulic point of view, of a very rough
nature, and presented many engineering difficulties.
From the dam, for a distance of 12,716 feet in a downstream direction,
the ground passed over is a gradually descending flat, the total fall in
this distance being 164 feet. The flat is a narrow strip of land, composed of hardpan and granite boulders, lying between the base of the
mountains on the one side and the river on the other. At two points,
the river, in former heavy floods, has invaded the flat and the adjoining
side hill, scouring off portions 500 feet in length, and leaving a bare
boulder bottom only a few feet above the low water level of the river.
Several streams running down from the adjoining mountains, intersect
the flat at right angles. Two of these are of considerable size, one
being 47 feet, and the other 212 feet from bank to bank. Both flow
over rough boulder bottoms.
At the termination of the flat is the rock wall through which the
river has cut the deep cafion previously described.    Owing to the rug Smith on Vancouver Water Works.
ged nature of the walls of the cafion, it was not deemed advisable to
carry the mains along its face, and its great height prevented their
being laid over the summit. A tunnel therefore was rendered necessary.
This tunnel is 280 feet long, 4 feet wide, and 6 feet from floor to centre
of roof. In cross section, the walls rise vertically 4 feet from the floor,
and are surmounted by a semicircular roof of 2 feet radius. The
floor elevation is 27^ feet below the crest of the dam.
Inasmuch as the hydraulic grade line of the whole system passes considerably below the floor of the tunnel, it was necessary that the main,,
from the dam to the tunnel, should be of larger diameter than that
from the tunnel to the city. It having been decided that the discharge of a 16 inch main was necessary for the city's supply, a 22
inch main is laid between the dam and tunnel, connecting in the centre
of the tunnel with the 16 inch main. The total length of the 22 inch
main is 13,530 feet, the total available head 29 feet, and the discharge at the tunnel 5,853,600 U. S. gallons in 24 hours.
The 16 inch main, connecting with the 22 inch main at the centre of the
tunnel, for the first 8000 feet of its length, passes over a rough, irregular
side hill, composed of earth, gravel and boulders. The sinuosities of
the side hill are closely followed, all great vertical depressions or elevations being avoided. In one instance, 1400 feet below the rock tunnel,
where the side hill juts out in the form of a steep " Hog's back," it was.
found expedient to pierce it with a timber lined tunnel, 108 feet long,
4 feet wide, and 6 feet high.
At the termination of the side hill, a series of flats, composed of
hardpan, gravel and boulders, descending in broad terraces is reached.
These are followed by the 16 inch main to ordinary high water mark of
Burrard Inlet, the total distance from the centre of the tunnel being
19,320 feet, and the total fall from the floor of the tunnel 388 feet.
At Burrard Inlet the 16 inch main is divided by a cast iron Y breech
into two branches of 12" diameter. One 12 inch branch has already
been laid across the Inlet, and preparations are in progress for the laying
of the second, which will take place at an early date. Plates XVIII.
and XIX shew plan and profile of the First Narrows of Burrard
Inlet, at the point selected for crossing. It will be seen that this is at
*/he narrowest part of the Inlet, where the tidal current runs with the
greatest velocity. It would naturally be supposed that the greatest
depth of water would be obtained here, but this is not the case. The
bed of the Inlet at this point, being soft sandstone rock, partially
covered with mud, gravel and cobblestones, forms a broad flat ridge,
extending from shore to shore.    The greatest depth of water on the Smith on Vancouver Water Works.
summit of this ridge at extreme low tide is 56 feet, gradually increasing
on each side till soundings of 120 feet and over can be obtained.
In extreme low tides the width of the crossing is 1086 feet. These
thles^_ho_wever, are very rare, occurring in May and June. In ordinary
tides the width at low water is 1237 feet, and at high water 2140-
feet. At extreme high water, which occurs in December and January,
the width is 2680 feet.
The north shore is extremely low and flat. From low water mark
for a distance of 6750 feet inland, the total rise does not exceed 63-
feet. Between high and low water mark, the surface covering consists
of cobblestones, small boulders, and coarse gravel, underneath which
is a stratum of hard pan overlying sandstone rock. The south shore
rises abruptly at high water mark to a height of 12 feet, terminating
in a level flat, which extends some distance inland. Immediately
west of the crossing on this side of the Inlet, is a steep rocky headland,,
which rises to an elevation of 216 feet above sea level.
This is the highest elevation within the limits of the city of
Vancouver, and may at some future day be utilized, as the site of a level
reservoir, of sufficient capacity to supply the city for 20 or 30 days.
Between high and low water marks on the south shore, and for nearly
three-quarters of the distance across the Inlet, the surface formation is
soft yellow sandstone rook, which, when blasted and exposed to the air,,
rapidly disintegrates. The contour of the bottom is an almost perfect
curve, the value of which railway engineers would express as 1\
Skilled divers made three different examinations of the bottom, and
reported fully thereon,   agreeing with each other in every  particular.
The substance of their reports was to the effect that no crevices existed in the rock ledge on the pipe lino, or in its neighbourhood, and that
the bottom from shore to shore was perfectly smooth and free from
boulders of any magnitude.
These reports were verified to a certain extent, by soundings taken
by the writer, at intervals of five feet apart, the lead, which weighed
15 lbs., never being allowed to leave the bottom all the way across.
The greatest depth recorded is, as before stated, 56 feet at low water,
increasing to 70\ feet at high water. The "Bore" or tidal current
varies from 4J to 9 miles per hour, the greatest velocity occurring in
the out-going tide, 2£ hours after low water. In a volume of water
like that flowing from the broad basin of Burrard Inlet through the
restricted channel of the First Narrows into English Bay, this velocity
of 9 miles per hour is terrific in its effects on any body opposing it. 20
Smith on Vancouver Water Works.
Some idea of its force may be gathered from the fact that a new 9
inch manilla hawser of 20 tons ultimate tensile strain, which, in the
preliminary operations of laying the submerged mains, was stretched
across the inlet, was snapped like pack thread by being suddenly
lifted to the surface, and allowed to float on it.
South of Burrard Inlet, at high water mark, the single 12 inch main
connects with a Y breech similar to that on the north side. A 16 in.
main leads out from this breech, passing over a uniform boulder and
gravel flat, known as Stanley Park, the greatest elevation of which
above sea level is 73 feet. South of Stanley Park, at a distance of
5041 feet from Burrard Inlet, is a long, narrow, shallow bay of Burrard Inlet, known as Coal Harbour. This bay lies directly south of
and parallel to, the First Narrows. The extreme length from east to
west is 6720 feet. The entrance to the bayis 3,730 feet wide. This
width gradually decreases till the head is reached at a distance of only
1,500 feet from English Bay, and separated from it by a low lying
strip of land, the highest elevation of which above sea level is not more
than 17 feet. The bottom is of soft mud, thickly studded with boulders. Half a mile from the head of the bay, the shore on each side
cuts out in long narrow promontories, leaving a waterway 870 feet
wide at high water, and 250 feet at extreme low water. This
is the point selected for the crossing of the 16 inch main. The bottom
is of uniform contour, and consists of tenacious mud and small boulders. The greatest depth at low water, which recurs in mid-channel,
is 5 feet.
Immediately south of Coal Harbour the City of Vancouver is
reached. The 16 inch main is continued along the graded streets to
the centre of the City, a distance of 39,211 feet from the centre of the
tunnel, or almost exactly 10 miles from the well chambers of the
The total fall from the level of water in the reservoir at the dam to
the termination of the 16 inch main is 384 feet, and from the floor
of the tunnel to the same point 355 feet. The total available discharge is 5,103,000 U. S.gals. in 24 hours.
South of Burrard Inlet, all works of excavation, refilling, culvert
building, etc., were done by the company by day labor. North of
Burrard Inlet, between the First Narrows and the dam, such works
were done by Messrs. H. F. Keefer and D. McGillivray, of Vancouver,
nder a lump sum   contract, based on a table of quantities furnished by Smith on Vancouver Water Works.
the Company. The trenches were excavated to regular grades, the
average depth for 12" pipes being 3' 6", for 16" pipes, 3' 10", and
for 22" pipes 4' 4", this gave a covering over all pipes of not les.? than
2' 6", an amply sufficient depth in the climate of Vancouver, frost
never being known to penetrate the soil deeper than 14 inches.
When the nature of the ground was uneven, and the grade line
laid down gave excavations less in places than these depths, the
difference was made up by embankments, 3 feet wide on top, with
slopes of lj to 1. In certain small gullies, embankments 6 feet wide
on top, were built under the mains, instead of timber trestling, there
being danger of bush fires during the summer months. The mains on
top of these embankments, and also under all streams, are protected
from injury by being enclosed in timber culverts. (See Appendix,
p. 358.)
Previous to describing the rivetted mild steel mains used by the Vancouver Water Works Co., it may be of interest to trace the origin of
steel pipes, and exemplify the many advantages possessed by them over
cast iron pipes.
TJp to the year 1845, cast iron was in universal use for the manufacture of water pipes; but in that year, Mr. Jonathan Ball invented and
laid in Saratoga, N. Y., a wrought iron pipe, coated inside and out with
hydraulic cement. This is the first instance on record in which wrought
iron water pipes were laid on this continent. Owing to the great saving
effected by this invention, it rapidly rose in favour, and was adopted by
many cities in the Union. It was soon, however, discovered that these
pipes required to be laid on a perfectly solid and unyielding foundation. If
laid on made ground, the slightest settlement caused the cement linings to
crack and leakage took place. The method of lining and laying in the
trench was cumbersome, and could only be employed to advantage near
thecentres of civilization, where transport was cheap and labour abundant.
When it was required to carry long lines of water pipes over mountainous country, in wildernesses entirely unsettled, and without roads or means
of conveyance, engineers were confronted with the task of devising another
and still more economical pipe. In California and the Pacific States of the
Union, this problem wa3 successfully solved by the invention of asphal-
tum coated rivetted wrought iron pipes. The cheapness of construction
of these pipes, and the facility with which they could be handled, and more
especially in the mining districts, brought them at once into general use.
In design and construction they are exactly similar to the rivetted mild 22
Smith on Vancouver Water Works.
steel mains described further on in this paper. Between 1870 and
1885, the Risdon Iron Works Company, of San Francisco, furnished
various water and mining companies with over 150 miles of these pipes
varying in diameter from 12 to 52 inches. Among the more notable
examples may be mentioned the following :
Spring Valley Water Works Co.—36 miles of pipe from^l8 to
inches diameter, and from
to $■ in. thick.
The Virginia and Gold Hill Water Works Co.—3 miles of
pipe 11-^r inches diameter, and from £ to & in. thick. This main
crosses a deep valley lying between its point of supply at Lake Mar-
lette and Virginia city. The bottom of the valley is 1750 feet below
the level of the lake. Therefore this main is subject to a constant static
pressure of 750 lbs. per square inch at its lowest point
The White Pine Water Works Co.—2 miles of pipe, 12 inches
diameter, y-^ to. ^>6 in. thick.
The Portland Water Works Co.—4£ miles of pipe, 30J inches
diameter, and yifu in. thick.
The Cherokee Flat Mining Co.—3 miles of pipe, 30 inches diameter, and from ,-|-, to -3- in. thick.
The great success of asphaltum-coated rivetted wrought iron pipes
led to still further researches. Manufacturers of water pipes directed
their attention to the adaptability of mild steel for hydraulic purposes,
and arrived at most gratifying results.
The writer, in seeking information on this subject, received from
Messrs. Duncan Bros., of London, England, a pamphlet on mild steel
mains, of which only a few copies were published by that firm for private circulation. The following extracts, giving a comparison between
mild steel, wrought iron, and cast iron for water mains, may be of
" Scientific investigation proved that in addition to being more ductile,
it (wrought iron) had greater tensile strength than cast iron, the relative
tensile strengths of cast iron and wrought iron being approximately 1
and 2.7. Mild steel is refined wrought iron, being nearly pure meta.-
lic iron, and when rolled into plates its strength compared to cast iron
is as 4 to 1. In consequence of its strength and ductility, it is eminently
adapted for all purposes to which cast iron has been formerly applied.
" With regard to strength, the ultimate tensile strength usually mentioned in specifications for cast iron pipes is 18,000 lbs. per square inch
mild steel, however, is now made with an ultimate tensile strength of
72,000 lbs. per square inch. It follows, therefore, that if pipes are
made of steel plates of the same thickness as would be employed in cast Smith on Vancouver Water Works.
iron, they are approximately lour times as strong. The actual strength
is not exactly four times, because it is not customary to calculate resistance to internal pressures with the same co-efficient or factor of safety
for both materials.
" The factor of safety usually employed for cast iron is 10, that is to
say, the working strength of the material is taken as only one-tenth of
the actual strength, which, in the case of pipes, means that if the
internal working pressure is to be 100 lbs- per square inch, the strength
of the pipes is calculated to resist 1000 lbs. per square inch. For
wrought iron, the factor is 6, and for mild steel 5. The reason for the
differences in the factor of safety, is because iron and mild steel are more
homogeneous, and thus more reliable than cast iron.
" The impurities which are present in cast iron are of less specific gravity than metallic iron, and consequently the specific gravity of the
mixture called cast iron is less than that of pure metallic iron. Mild
steel is the nearest approach to pure metallic iron, which commerce and
science combined have yet produced on an extensive working scale.
The average weights of the metals are:
Cast Iron.
Wrought Iron.
Mild Steel
lbs. per cubic foot; the average weight of water is 62^ lbs. per cubic
foot; therefore the specific gravities average
Water. Cast Iron. Wrought Iron. Mild Steel.
1 7.20 7.68 7.83
table op relative thickness por equal strength.
Cast Iron.  Wrought Iron. Mild Steel.
Weight of plate in lbs., per sq. ft.
1 inch thick         37.5 40 40.8
Tenacity per square inch       18,000      48,600 72,000
Relative strength for equal thickness  1 2.7 4
Factor of safety  10 6 5
.Relative strength due to factor of
safety  1 4.5 8
^Reduction in strength due to rivetted joints  — 30 p.c.        30 p.c.
Relative strength after reduction
for rivetted joints  1 3.15 5.6
Relative   thickness  for plates of
equal strength ,   1      0.3174 0.1786 24
Smith on Vancouver Water Works.
Cast Iron.  Wrought Iron. Mild Steel.
Thickness of plate in inches, 401bs.
weight per sq. ft     1.066 1.00 0.9804
Relative strength for equal weight    1 2.533      3.678
" due to factor of
safety     1 4.22       7.356
Relative strength after reduction
for rivetted joints     1 2.955      5.149
Cast Iron.  Wrought Iron. Mild Steele
Weight of plain cylinders of equal
strength      1 0.3384    0.1942
Increase in weight of pipes due to
joints     5.8 p.c. 15 p.c.    15 p.c.
Relative weight of pipes of equal
strength     1 0.3678    0.2111
" The relative thickness for plates of equal strength for materials of the
ultimate tenacity under consideration are given on the last line of the
first table. In the next table, the results obtained shew the relative
weights of pipes of equal strength, having socket and spigot joints, made
from materials of the ultimate tensile strength specified.
"Applying these results to an ideal case, we find that, if it is specified
that cast iron pipes, to stand 300 feet working head of pressure, and 24
inches internal diameter, are to be -| inch (= .875) thick, then wrought
iron pipes of the same diameter would be .875 x .3174 = .2778 inches
thick, and mild steel pipes would be .875 x .1786 = .1563 inches thick
or say ^ in., ~ in. and 3\ in. thick respectively, for equal internal
working pressures.
"Then again, if one mile of 34 inch cast iron pipes, ^ inch thick, made
up of pipes in 12 feet lengths/weighing 24.8 cwt. each length, weighs
545.6 tons, the corresponding weight of one mile of wrought iron pipes
will be 545.6 x 0.3678 = 200.6 tons,
and one mile of mild steel 545.6 x 0,2111 5= 115.2 tons.
" These results shew that for equal diameters, 24 inches, equal working
pressures of 300 feet and equal lengths of one mile, the weights are
Cast Iron. Wrought Iron. Mild Steel.
545.6 200.6 115.2 tons.
The price per ton of mild steel pipes averages about 4|- times the current price of cast iron pipes ; as the relative weights for equal strength
are as 1 : .2111, it is therefore apparent that the relative costs for a given
length are as 1: 0.90, or in other words, length for length, at a cost of Smith on Vancouver  Water Works. 25
10 per cent, less than cast iron pipes. With regard to carriage, the
rate per ton by rail is the same for either cast iron or mild steel pipes,
and as the saving is in the direct ratio of dead weightfbr a given length,
the cost of railway carriage is 78 per cent, less than on cast iron pipes?
and a like saving can be effected in handling the pipes at the site of the
track in which they are to be laid.
"The next point to which attention is directed is the jointing. As
mild steel pipes are so much lighter than cast iron pipes, it is clear that
they may be conveniently handled in longer lengths. The system of
construction also favours this, and in fact the pipes may be made in
one continuous length, built upon the site if it is desired. The customary methods are to make them in lengths of 24 feet, this being twice
the usual length of cast iron pipe, and there are, consequently, only half
the number of joints.    Taking the 24 inch pipes before mentioned, the
lengths and weights would be
Cast Iron.    Mild Steel.
Diameter         24 inches      24 inches
Length of each pipe ,         12 feet 24 feet
Weight do          24.8 cwt.    10.47 cwt.
Relative weights per pipe  1 0.42
"     lengths      "   1 2
" Again, taking the case of one mile in length, 440 pipes would be
required in cast iron, and only 220 in mild steel, consequently, there is
a saving of 50 per cent, in the labour and cost of jointing a given
length. Then with regard to each joint, the mean circumference of the
space for lead in an ordinary cast iron socket joint is greater than in a
mild steel pipe, in consequence of the greater thickness of cast iron-
The reduction in the circumference of a mild steel socket is equal to a
saving of 9J per cent, upon the weight of lead required for a 24 inch
cast iron pipe socket; assuming that the depth of lead is the same in
each case, the total saving in lead is therefore 59£ per cent.
"To shew the final economical result in the case of one mile of 24
inch pipes previously mentioned, the several relative costs are :
Cast Iron.    Mild Steel.    Saving
Internal diameter, inches         24 24
Length, mile  1 1
Number of pipes       440 220
Weight of each pipe, cwts       24.8 10.47
"        one mile, tons     545.6 115.2
Relative cost per ton  1 4.25
" of carriage, per ton  1 1 '26 Smith on Vancouver Water Works.
Cast Iron.    Mild Steel.    Saving.
Relative cost of Carriage on total.... 1 0.2111 78 p.c.
"           of laying per yard  1 0.7 30 p.c.
Relative number of joints  1 0.5 50 p.c.
"    weight of lead, each joint.... 1 0.905 9£"p.c.
"        "        "        each mile.... 1 0.405 59 j- p.c.
"    cost of making each joint ... 1 0.8 20 p.c.
"        "     jointing one mile  1 0.40 60 p.c.
"    cost of total for one mile.... 1 0.9 10 p.c.
"      "    of pipes and carriage... 1 0.84 |16 p.c.
"      "    of carriage and laying. 1 0.834        16.6 p.c.
"      "    of pipes, carriage, laying and jointing one
mile  1 0.788 21.2 p.c
" The saving actually effected in the total outlay for one mile of 24
inch pipes, is therefore :
Cost of pipes.     Cost of carriage.     Cost of laying.     Cost of jointing.
10 p.c. 6 p.c. 0.6 p.c. 4.6 p.c.
•or a grand total of 21,2 p.c."
It will be seen that the above extracts treat of a comparison between
cast iron mains, and mild steel mains fitted with faucets and spigots.
This is a cumbersome arrangement, and has been entirely discarded on
the Pacific coast, the Moore and Smith joint, a description of which
will be given further on, taking its place. This joint is specially adapted to all pipes between the diameter of 12" and 24". When of larger .
sizes the pipes are made in plain lengths of 24 feet 6 inches, and rivetted together in the trench.
The rivetted mild steel mains in use by the Vancouver Water Works*
Company are of three diameters, 22 inches, 16 inches, and 12 inches.
The 22 inch is laid from the dam to the tunnel, a distance of 13,530
feet, the 16 inch from the tunnel to ordinary high water mark of Burrard Inlet on the north shore, and from ordinary high water mark on
the south shore to the centre of the city, a total distance of 39,211 feet.
The 12 inch are laid on both shores of Burrard Inlet, between ordinary
high water marks, and the submerged 12 inch flexible main across the
Inlet, a total distance of 747 feet.
; The 22 inch and 16 inch pipes are j^-0 in. in thickness, and the 12
inch A in. The latter, being laid below high water mark, require
greater   thickness   of   metal   to withstand  the   corrosive    influence Smith on Vancouver Water Works.
•of salt water. These pipes were manufactured from plates imported
from England by the Company, and rolled, rivetted, coated with
asphaltum, and laid in trench by the Albion Iron Works Company of
Victoria, B.C. Plate XX shews a longitudinal section of the 16
inch pipe. The 22 inch and 12 inch pipes are constructed in an
exactly similar manner. It will be seen that the pipe is made in 7
courses, 4 large or outside courses, and 3 smaller or inside courses, rivetted together, and having a projecting nipple at one end. At the
foundry, the plates were trimmed to the exact sizes required, and the
rivet holes punched with multiple punches at one and the same
time. Absolute uniformity in size and spacing of rivet holes was thus
secured. Each plate was then rolled in the usual manner, by means of
three parallel revolving cylinders, which gave it the circular form of
the required diameter. It was then made to encircle the vertical cylinder of a hydraulic rivetting machine, which cold rivetted the straight
or longitudinal seams. When 7 plates had been treated in this manner
and converted into cylinders 3 ft. 6 in. long, and of diameters differing
sufficiently to allow the ends of the smaller cylinders to be passed into
the ends of the larger, they were rivetted together, so as to form one
length. On the lap, between two thicknesses of steel at the end of ea
course, the plate was scraped down to a fine edge, and a rivet driven
through. Where three thicknesses of metal came together, as when the
longitudinal seams of the large course overlap the smaller course, extra
heavy lap rivets were used. The edges of each sheet for 3 inches from
the laps were chipped and caulked. Straight and round seams were
split caulked. The whole length was then heated in an oven, and
immersed in a bath of hot asphaltum. This bath was an iron trough,
26 feet long and 3 feet wide, supported on brickwork, and so arranged
that a fire could be kept constantly burning underneath. In preparing
the mixture, the trough was filled to within a few inches of the top
with asphaltum broken up into small cubes of about an inch to the side.
Coal tar, devoid of all oily matter, was then poured in till the
asphaltum cubes were completely covered. The mixture was then
allowed to boil for three hours, being constantly stirred during the process. As many pipes as the mixture would cover were then dipped and
allowed to dry. The coating obtained was smooth, tough, free from
brittleness, and of uniform thickness.
The form of joint used in connecting these pipes is, as before stated,
that invented by Joseph Moore and Francis Smith, employees of the
Risdon Iron Works Co., San Francisco. Plate XX shews a longitudinal section of this joint.    In making the joint in the trenches, the 28
Smith on Vancouver Water Works.
nipple end of one length of pipe was forced into the larger end of the
adjoining length, by means of hammering on wooden blocks placed
against the end opposite the nipple. The abutting ends of the two
lengths were not driven up tight, a space of from J to \ an inch being
left, for the purpose of allowing for any expansion or contracti&n that
might take place. The outside surface of the pipes was then scarped
clean for about 1\ inches back from the junction of the two ends. A
band or ring of diameter sufficiently great to allow of tV inch play between its inside surface and the outside surface of the pipe, was then
made to encircle the junction. The space between was filled up with
lead in the usual manner, and carefully caulked. Joints made after this
pattern, have been in use for 15 years, and have given entire satisfaction. Care must be taken in making this joint, that no angle
greater than one degree is made at the junction of the two lengths of
pipe, otherwise the lead packing will be of unequal thickness, and will, in
all probability, result in a leaky joint. Caulkers, accustomed to jointing
cast iron pipes, must be cautioned, when making for the first time, a
Moore and Smith joint, that the steel pipe will only admit of the lead
being packed to a certain firmness, the degree of which can only be
ascertained by actual trial. If the lead is beaten in between the ring
and the pipe too tightly, the shell of the latter will bend inward, and
render good work impossible.
As before stated, steel mains of more than 24 inches diameter, when
subject to heavy pressure, are usually made in specified lengths at the
foundry, and rivetted together in the trench. To accomplish this, it is
necessary that each length shall have a large course at one end, and a
small one at the other. The large course has its extreme end punched
for rivets at the foundry, while the small course at the other end of the
length is unpunched.
The pipes being placed in the trench, the small course of one length
is forced by hammering, or other power, into the punched large course
of the adjoining length. The position of the rivet holes on the small
course, to correspond with those on the large course, are then marked
and screw punched after separation. This being done, the two lengths
are again united, their surfaces pressed firmly against each other by
means of a set stool, and cold rivetted from the outside. The seam is
split caulked in the usual manner. This makes the most desirable connection for pipes of large diameter.
However, it may be mentioned, that a pipe of 41 inches diameter, and
subject to a pressure of 300 feet, was laid, ten years ago, in the Sandwich
Islands.    The lengths were connected by Moore & Smith joints, and
are in active service to this day. Smith on Vancouver  Water Works.
The Vancouver pipes were laid in the trench with the straight seams
upwards, so that any leakage might readily be detected, and repaired
by further split caulking. In most systems, however, the straight seams
are laid downwards, the advantage of which is that in course of time,
sediment gathers on the bottom of the pipe along the edges of the seams,
and tends to prevent leakage.     (See Appendix, p. 361.)
Inasmuch as the steel mains described in the foregoing pages were
constructed with a view to securing absolutely tight joints, the outside
surfaces of the nipples fitted tightly against the inside surfaces of the
adjoining lengths. Consequently, no deviation from a straight line
greater than one degree, could be made between any two lengths with
out special bends. By means of specially adapted machinery, steel
elbows and bends are made by certain manufacturers, but these lack
stability when the angle of curvature is large. All bends in the Vancouver-system are of cast iron, one inch thick. They are segments of a
circle, the axis of the bend being the circumference, and the radius
five feet. Previous to leaving the foundry, they were individually
subjected to a pressure of 300 lbs. per square inch.
In certain parts of the pipe line, north of Burrard Inlet, the ground
traversed, being contiguous to the river, is irregular horizontally and
vertically, and required bends ranging from 5 to 70 degrees angle of
deflection. That portion of the pipe line immediately south of the
tunnel, and following the irregularities of the side hill for a distance
of 8000 feet, required no less than 80 bends of all angles of deflection,
being an average of one bend to every 100 feet of length. The total
number required by the system from the point of supply to the centre
of the city were 179.    (See Appendix, p. 363.)
The other castings connected with the mains, not including the connections with the city distribution system, are as follows : two miles and
a half below the dam, at the lowest depression between the dam and the
tunnel is placed a blow off, 8" off 22". This is controlled by an
eight nch valve, leading into a 12" x 12" box drain, which in turn
leads to the river. To the middle pipe length in the tunnel is affixed
a self-acting Chabot air valve, the air passage of which is 2J inches
diameter, and is controlled by a brass valve, so that the upper part containing the rubber ball may be taken off for examination at any time
without the necessity of shutting off the main at the dam. 30
Smith on Vancouver  Water Works.
At Burrard Inlet, on the north side is placed a blow off, 8" off 16""
and on the south side, 12" off 16", reducing to 8", both controlled by
valves, and emptying into Burrard Inlet. The ends of the 16 inch
main, on both sides of the inlet, are provided with "Y" breeches, two 12
inch branches off 16 inch. These branches connect with the double
line of 12 inch mains, that will ultimately cross Burrard Inlet, and are
individually controlled by 12 inch valves, so that each main can be
shut off independently if required. Between the Inlet and Coal
Harbour, on the highest elevation between the two waters, is placed
another Chabot air valve, arranged in a manner similar to the one
already described.
On both sides of Coal Harbour are placed blow offs, 8" off 16" discharging into Coal Harbour, and finally a 16 inch valve is located at
the point where the mains enter the inhabited part of the city. It
will thus be seen that in case of necessity the supply to the city can
be shut off at five different places, viz., at the entrance and outlet of
well chambers at the dam, on both sides of Burrard Inlet, and at the
entrance to the city.
Inasmuch as the pipe line between the centre of the City and Coal
Harbour follows well graded streets, the distribution of steel mains,,
lead and castings was attended with little or no difficulty. Ordinary
four-wheeled drays, drawn by two horses and accompanied by two-
teamsters, accomplished this work in a most satisfactory manner at a
cost of $3.00 per ton.
The flexible mains for the crossings of Coal Harbour and Burrard
Inlet, were transported on scows and discharged on the beach between
high and low water mark at a cost of $5.00 per ton.
Between Coal Harbour and Burrard Inlet the first difficulties were
encountered. The land between these two waters being heavily timbered, and only accessible by waggon road at both ends, rendered
necessary the construction of a temporary road parallel to the
pipe trench, and within easy reach of it. This was of the simplest
character, being a roughly graded track, 8 feet wide, along which and at
right angles to it were placed, at regular intervals of ten feet, rough
undressed skids. Sleighs similar to those used by loggers in winter and
drawn by two powerful horses, carried two lengths of pipe per load,
along this road and deposited them where required. The cost of this,
service, including the construction of road, averaged $5.00 per ton.. Smith on Vancouver Water Works,
Between Burrard Inlet and the Dam the work of distribution was
accomplished under very great difficulties. As before stated, the
country traversed by the pipe line is very irregular and heavily timbered-
No roads exist in the vicinity, and the construction of an economical
mode of conveyance for the 480 tons of steel mains, lead and castings,
which were to be laid continuously along the pipe trench, was a
problem, the solution of which involved considerable ingenuity. The
mode adopted was a combination of waggon road and tramway. A
tramway 15,400 feet long was built from the Inlet to a point on the
side hill ground, about 4,000 feet south of the rock tunnel. This
tramway is of three feet gauge with 4 " and 5 " track timbers, supported
on ties placed four feet apart. At four points in its course there occur
sudden breaks in ground level. The first takes place 5,800 feet north of
the Inlet, the ground rising 37 feet in a distance of 135 feet; the second!
at 9,400 feet north of the Inlet, the ground rising 54 feet in a distance
of 260 feet; the third at 11,100 feet north of the Inlet, the ground
rising 27 feet in a distance of 82 ; and the fourth at the termination of
the tramway, where it leaves the pipe trench and climbs the face of the
side hill to the flat above. The total rise at this point is 80 feet in a
distance of 150 feet. In distributing the pipes along the tramway, each
car, pulled by one horse, carried three lengths of 16-inch pipe. On.
arrival at one of the above-mentioned changes of level, the horse was
removed from the car to the top of the rise, where it was made to haul
on a cable connected with the car, until the car .also had reached the
top. At the termination of the tramway a waggon road was built following the edge of the flat, a distance of 4,000 feet, to the rock tunnel.
At suitable intervals clearings were made from the waggon road to the
pipe trench on the side hill below, down which the pipe lengths were
lowered, one at a time, by cables.
At the rock tunnel the face of the Cafion slopes downwards 164 feet
in a distance of 400 feet. A short tramway was built down this steep
descent, and loaded cars lowered down by cables. From the foot of
this Cafion a waggon road was built and operated in a manner similar
to that between Coal Harbour and Burrard Inlet.
The prices paid for the distribution of mains, lead and castings north
of Burrard Inlet were, for the first 1268 feet $7.64 per ton, for the next.
18,859 feet $12.87 per ton, and for the next 13,516 feet $14.85 per
ton.    These prices  included  the building of the tramway and the
waggon roads as well as the cost of the distributing mains. 32
Smith on Vancouver Water Works.
Having in view the difficulty of effecting repairs in pipes laid under
water, and the disastrous consequences that might result from a temporary stoppage of the city's water supply should a break take place,
through unavoidable causes, the design for crossing the first narrows,
instead of being one 16 inch main, comprised its equivalent, two
separate lines of 12 inch mains, 50 feet apart, and capable of independent
action by means of stop valves placed at high water mark on each side
of the Inlet. Up to the present only one of these lines has been laid in
position on the bed of the Inlet, made up of 746 feet of plain rivetted
steel pipes; 261 feet of rivetted steel pipe, fitted with cast iron flexible
joints,   and 1236 feet of cast iron flexible joint pipe.
The plain rivetted steel pipe is placed at each end of the line, 584
feet on the north shore and 162 feet on the south shore. The
rivetted steel pipe with flexible joints is placed on the north shore between the plain pipes and the cast iron flexible pipes, and the latter are
placed on the bed of the Inlet, reaching from low water to low water
The construction and details of the plain pipe have been already
described. The flexible steel pipe is in lengths of 22' 2" over all, and is
exactly similar to the plain pipe, but provided with cast iron spigots and
faucets, bored and turned in the same manner as the cast iron flexible
pipes. The latter are of the pattern known as the Ward patent flexible
joint pipe. They were manufactured in Scotland, and are of hard close
grained white cast iron, thoroughly coated with Dr. Smith's coal
pitch varnish. Each length is 12' 4" over all, -J^in. thick, weighs
1280 lbs., and is warranted by the manufacturers to stand with
safety the pressure due to a column of water 600 feet high. Each
joint required 70 lbs. of the best Spanish pig lead. Plate XX shews
a longitudinal section of this joint. The larger portion of the inside
surface of the bell or faucet forms a spherical zone, the centre of which
is a point on the axis of the faucets at such a distance from its mouth,
that the inside diameter of the latter is greater by half an inch than
the inside diameter of the shoulder. The extreme end of the spigot is
turned truly, and exactly fits the inside surface of the faucet. The
outer end, or the end encircled by the mouth of the faucet, is of smaller
diameter, so as to allow half an inch of space between the two surfaces Smith on Vancouver Water Works.
for lead packing. At the middle of the spigot is a circular groove, a
quarter of an inch deep and an inch and a half wide, which serves the
purpose of retaining the lead packing, and prevents the joint from pulling asunder, when exposed to tensile strain. This joint is capable of
motion through an angle of 12-degrees, and a complete circle can be
made with 30 lengths.
The contract for furnishing and laying the single line of cast iron
flexible joint pipe was let on the 1st of November, 1887, to the inventor
and patentee of the joint, Mr. John F. Ward, late chief engineer of the
Jersey City Water Works. The price agreed on, which covered all risks
and contingencies, was nine dollars per lineal foot.
Mr. Ward has devoted many years of his life to laying submerged
pipes of all diameters, and has, hitherto, met with unfailing success..
Among some of the more prominent works standing to his credit, may
be mentioned the six inch pipe crossing the Delaware River at Easton,
Pa., the 12 inch pipe, 963 feet long above the dam, at Lawrence, Mass.,
and the two lines of 8 inch pipe crossing Shirley Gut, Boston Harbour,
a channel 400 feet wide, and 37 feet deep, through which a tidal current flows at the rate of 7£ miles per hour.
Mr. Ward, on his arrival, made a thorough inspection of the crossing,,
and expressed himself as confident of being able to complete his contract
with ease and rapidity. Accordingly on the 21st of April, 1888, he
began operations, his plan being to joint the pipes on a suitable platform stationed at low water mark on the north shore, and by means
of a stationary engine on the south shore, to haul them across, length by
length. Inasmuch as Mr. Ward Jailed to carry out this plan to completion, the writer, without expressing any opinion as to its practicability
will merely describe his mode of procedure.
The structure erected on the north shore of the Inlet, on which the
pipes were jointed, was a frame work staging of sufficient height to
reach above extreme high water, and strong enough to resist the force of
the incoming and outgoing tides. In the middle of this stage was constructed a sloping platform, extending from the front face, 4 feet below
the top, down to the ground at the rear face, or the face fronting the
Inlet. The object of the platform was to admit of the pipes being
jointed in an inclined position, and therefore sliding easily to the ground,
when the hauling power was applied. The 104 lengths of pipe required
to reach from shore to shore were piled within easy reach of the platform. The engine on the south side of the river, opposite the platform
and at a distance of 1400 feet from it, was of 30 H. P., and revolved
at ordinary drum, to which was attached a  hundred feet of wrought,
c 34
Smith on Vancouver Water Works.
iron chain, connecting with a continuous wrought iron rod of 1J inches
diameter. This rod reached clear across the Inlet, and was attached
to the rear end of the first length of pipe lying on the sloping platform
of the staging. The rod was made from round iron in lengths of 51
feet, jointed together by common sorewunions, its whole tensile strength
being that due to the resistance offered to stripping by the threads of
the unions.
When Mr. Ward had completed these arrangements, he began without delay to joint the lengths together. To the length lying on the
platform, the spigot end of which faced the Inlet, a second length was
jointed in the usual manner.
The engine on the south side was then put in motion, and the first
length hauled forward a distance equal to its own length, leaving the
second length -to fill the place previously occupied by the first. A
third length was then jointed to the second, the engine again pulled
forward, until the third length occupied the place vacated by the second.
It was intended to repeat this operation until the whole 104 lengths
had been dragged across the bottom of the Inlet. However, after 18
lengths, covering La distance of 216 feet, had been submerged, Mr.
Ward concluded to substitute a steel wire cable for the wrought iron
rod. In stretching this cable across the Inlet, it unfortunately fouled
on a small boulder, about 200 feet above the pipe line, and such efforts
as were made to dislodge it proved unavailing. Mr. Ward then notified the company that urgent private business compelled him to leave
the works for St. Paul, Minn. He did not return, but shortly afterwards officially abandoned the contract.
On July 9th, more than a month after Mr. Ward's failure, the
company contracted with Messrs. H. F. Keefer and D. McGillivray,
the gentlemen who already held the contract for trenching and refilling
to complete the work according to certain specifications, from which
the following clauses are extracted:
" The total length of the crossing to be made is 1248 feet, extending
from low water mark on the south shore to low water mark on the
north shore. These points will be defined by stakes placed by the
company's engineer, and the whole main when finally laid shall be in a
perfectly straight line between them.
" Each pipe length, previous to being placed in position, shall be well
and carefully tested for flaws in manufacture, cracks, air-holes, and
other defects, by the usual process of suspending in slings and tapping
with hammer. Should any be found defective, they shall be discarded,
and the engineer notified of the same. Smith on Vancouver Water Works.
" The lead to be used in jointing shall be that known as 'Best Spanish
" The whole number of pipe lengths, previous to being placed in final
position on the bed of the first narrows, shall be jointed, leaded, and
made perfectly water-tight on dryland, and on such a structure as will
admit of the whole length of 1248 feet being of easy access for the purpose of inspection.
" A test pressure of not less than 300 lbs. per square inch shall then
be applied by the contractors, in the presence of the Company's Engineer,
the leakage under which, throughout the whole length of 1248
feet, shall not exceed one cubic foot per minute. Such joints as may
prove defective under this pressure shall be made good by the contractors at their own expense, and such pipe lengths as may leak or give
evidence of flaws shall be removed by the contractors, and replaced by
sound lengths, the cost of which shall be defrayed by the company.
" The Engineer's approval of the main, after the application of the
above test being given, the contractors shall be at liberty to place it in
position on the bed of the first narrows, which being done, a similar
test pressure of 300 lbs. per square inch, subject to the same conditions,
shall be applied.
" A diver will be appointed by the company to inspect the main when
finally laid in position, and on his report such alterations in its position
as may be rendered necessary by reason of its resting on boulders or
sharp irregularities of the bed of the Inlet, shall be made by the contractors, and at their expense, provided the total cost does not exceed
five hundred dollars. All costs over this amount shall be defrayed by
the company."
Messrs. Keefer and McGillivray entered into the fulfilment of their
contract with energy. A 30 H. P. engine was stationed on the north
shore of the Inlet, between high and low water marks. With this the
18 lengths submerged by Mr. Ward were hauled back to dry land. A
trench, 4 feet wide, 4 feet deep, and 1300 feet long, was excavated on
the line of the crossing on the north shore. Parallel continuous runners
of barked fir, three in number, were placed in the bottom of the trench,
in such a manner that the bell end of each pipe when jointed would
rest on the central runner, and be supported on each side by the.
other two runners. A frame work staging, similar to that employed
by Mr. Ward, was built over the trench and supported on rollers, on
which it could readily be moved over the whole length of the trench.
On this staging with its sloping platform, the whole number of
pipe  lengths were jointed, the operation being very similar to that of 36
Smith on Vancouver Water Works.
paying off a cable from a moving ship. As soon as the first joint was
made, the staging was moved forward till the first pipe length rested on
the runners in the trench, leaving the second in the place vacated by
the first. A third pipe was then hoisted up by winches, its spigot end.
inserted into the bell of the second, and carefully adjusted in exact line..
Molten lead was then poured in and caulked in the usual manner..
This done, the staging was again moved forward and another pipe
adjusted, the operation being repeated day by day, till one hundred
lengths had been connected. As before stated 104 lengths were provided, but during the process of jointing, four, shewing evident signs
of fracture, were discarded.
Immediately on the completion of the work of jointing, both ends-
of the chain of pipes were securely capped, and the stipulated test
pressure of 300 lbs. per square inch applied.
A first attempt was made to apply the pressure by means of a hand
pump, worked by six men, forcing a stream of water into a circular
opening, one inch in diameter, provided for that purpose in the cap on
the north end. It was speedily found, however, that owing to the
leakage at the joints, slight as it was, this method was not powerful,
enough to keep the chain of pipes full and attain the required pressure.
The stationary engine, situated midway between high and low water
mark, was then brought into requisition. The middle length of the-
chain of pipes was tapped, and by means of the engine, water was
pumped in until the first defective pipe manifested itself, which occurred
when the gauge registered 30 lbs. per square inch. This length was
immediately broken up by sledge hammers, the bell cut by a cold
chisel, split open, and the lead removed.
The two portions of the chain of pipes were then hauled together by
means of the engine, and re-jointed. Pressure was again applied until
the second injured pipe gave way.
This operation was repeated until no less than eight defective pipes
had been removed. The remaining 92 sustained the required pressure-
of 300 lbs. per square inch for a period of five minutes, during which,
each length was subjected to heavy blows from a 12 lb. hammer. As-
the joints sustained this severe pressure without exceeding the specified!
amount of leakage, and as every length seemed to be absolutely free
from defects, the test was considered eminently satisfactory. The
following table shews the pressures at which the different pipe lengths
gave evidence of the fractures they had sustained during their repeated
handlings, and which were not detected by the process of " ring:-
ft Smith on Vancouver Water Works.
Number of Leng
reckoning fr
north   end
pipe chain.
Pressure per square
inch under which
pipe gave way.
Nature of fracture.
Longitudinal crack 12
Notwithstanding the additional loss of these 8 pipes, it was deemed
advisable to proceed with the submersion of the remaining 92, the
shelving nature of the north shore being such that the north end of the
chain of pipes, when laid in position, would not be covered by more
than two feet of water at low tide, and, therefore, it would be no difficult
matter to raise that end at any future convenient time, and add the
whole 12 lengths necessary to complete the crossing as planned.
The plan adopted for placing this long line of heavy flexible pipes
in position on the bed of the Inlet was direct hauling from shore to shore,
during the half tides which occur in the Inlet during the months of July
and August. For the purpose of lessening the weight as much as possible, each length was encircled by a wrought iron ring, to each of which
floats of 500 lbs. buoyancy were attached. To prevent as much as
possible the forward end of the chain of pipes from ploughing a deep
furrow in the bed of the Inlet during the process of hauling, it was
buoyed up by a number of cedar logs laid lengthways. The hauling
gear was as follows—(See Plate XIX). To the rear end, that is
the end farthest from the water, was attached a 9 inch manilla cable
of 44,800 lbs. ultimate tensile strength, and 600 feet long, which was
connected with the 30 H. P. Engine, stationed on the same shore,
midway between high and low water marks. To the middle length
was attached a 4 inch steel cable of 52,000 lbs. ultimate strength, and
1880 feet long, which connected with a 30 H. P. engine stationed on
the south or opposite shore. Midway between the middle length and
the forward end of the chain of pipes, a similar steel cable 1,600 feet
long was attached, which also connected with a 30 H. P. engine on the
opposite shore. A third steel cable of the same strength, and 1,325
feet long, was attached to the forward end of the chain of pipes. This
latter connected with two 30 H. P. engines on the opposite shore. It
will thus be seen that there were no less than three 4 inch steel wire
as 38
Smith on Vancouver Water Works.
cables, and one 9 inch manilla cable attached to the chain of pipes, the
total ultimate strength of which was very nearly 90 tons. The total
effective strength of the engines pulling the tackle connected with these
cables aggregated 150 horse power.
The four engines on the south side were stationed on the beach a
high water mark. The blocks and tackle were arranged in three parallel rows 10 feet apart on the flat immediately to the rear of the engines.
This flat being densely timbered with the huge trees peculiar to the
Pacific coast, the space cleared in which to operate the tackle was
necessarily limited. The blocks were securely anchored to huge
stumps in the vicinity by heavy wrought iron chains. The pulleys, one
of which was four sheaved and two three sheaved, had a clear distance
of 56 feet in which to operate. The manilla cables passing through
the sheaves were connected to the wire cables by wrought iron grips
invented for the occasion by the contractors.
All arrangements having been satisfactorily completed, the engines
were set in motion on the 28th of August, 1888, at 10 a. m. The steel
cables straightened out and remained taut and stationary, but only for
a minute. A sudden slackening took place, and the whole chain of
pipes took a forward motion of several feet, and from that instant the
success of the undertaking was an assured fact. There had been a
question as to whether the joints would withstand the enormous tensile
train brought to bear on them, but it now beci me certain that the
lead packing would remain intact as long as the East iron bell held
Owing to the extreme distance between the blocks and pulleys being
no more than 56 feet, the tackle connecting them had to be overh auled
every advance of 56 feet made by the chain of pipes. The process of
hauling was therefore necessarily slow; but being kept up without intermission, at 7 p. m. the forward end of the chain of pipes arrived at its
destination on the south shore.
On the day following, at slack tide, a skillful marine diver walked
across the bed of the Inlet, following the chain of pipes, entering on the
south shore and emerging on the north. His report was to the effect
that the whole line of pipes was lying on the bed of the Inlet in a perfectly straight line, without sag or bend, that the heavy projecting bells
of the pipes, as they were being drawn over, had scooped a deep groove
in the soft sandstone rock, and that the whole chain of pipes was resting
in a rock trench of its own excavating; that above this trench silt was
rapidly gathering, and that in his unqualified opinion the pipes would Smith on Vancouver Water Works.
in a few weeks be entirely covered over, rendering their permanency and
safety beyond question.
The day following this examination, the contractors applied the
final test pressure of 300 lbs. per square inch as called for by the
specifications. An opening was made in the cap on the end length, the
pipes filled with water by steam pumps, and the required pressure
steadily maintained for five minutes of time, without perceptible leakage.
The enormous strain on the joints apparently had no other than a beneficial effect, having compacted the lead, and rendered the whole line
perfectly water-tight. Eleven of the 12 pipes which had been discarded
were subsequently replaced by pipes cast by the Albion Iron Works Co.
of Victoria, tested to a pressure of 300 lbs. per square inch before
leaving the foundry. No difficulty was experienced in attaching these
to the main already submerged. The end of that main having been
lifted up was buoyed on the deck of a small scow. The additional
lengths were added one by one, the scow being moved forward as each
length was jointed, until the whole eleven rested in position on the bed
of the Inlet. It was found, however, at a later date that owing to the
shelving nature of the north shore, and the fluctuations of the tides, a
satisfactory connection between the end of the cast iron flexible pipe and
the plain rivetted steel pipes could not be made. Twelve of the latter
were accordingly fittedwith flexible cast iron spigots and faucets, similar
to those shewn on Plate XX, and connected with the cast iron pipes,
making a total length of 1496s feet of flexible pipe, covering a horizontal distance of 1483^ feet.
When the project for supplying the city of Vancouver with water
from the River Capilano, by means of a submerged main across
Burrard Inlet, was first made public, considerable interest was evinced
by both engineers and civilians. Printers' ink was called into requisition
and many articles published demonstrating the utter impracticability
of the project.
The complete success of the undertaking is an irrefutable answer to all
the adverse theories advanced. However, it may be of interest, even
at this late day, to mention some of the objections urged and believed
in up to the successful completion of the work, and the published answers
Objection 1.
That the  known force of the  current  in the first narrows would
cause the chain of pipes to sway up and down the bed of the Inlet with
each change of tide, and eventually result in separation of the joints.
Answer—That it could be mathematically demonstrated (calculation 40
Smith on Vancouver Water Works.
shewn), that the force of the current was altogether insufficient to produce the results stated, and that the proposed method of laying the
pipes by "direct hauling" from shore to shore would result in the
sharp-edged bells of the pipes cutting a groove, sufficiently deep to
embed the whole chain, and thus effectually destroy the possibility of
Objection 2.
That the current would create a friction that would scour off any
coating that might be put on to protect the pipes from corrosion.
Answer.—That the pipes being embedded in the bottom of the inlet,
and covered by silt, would be absolutely free from frictional action.
Objection 3.
That vessels might accidentally drop anchor on the pipes, or that vessels, finding themselves in danger of drifting ashore, through stress of
weather or other causes, might be obliged to drop their anchors on the
bottom, and as a result hook on to the chain of pipes and break it asunder. /?a?^e.
Answer.—That the thickness of the pipe shells if exposed to the
shock of a falling anchor would be sufficient to keep them intact, and
that if the anchor fluke of a drifting vessel were to bury itself under
the chain of pipes, the vessel would be securely anchored, and would be
obliged to wait for the turn of the tide to free herself, such cases
occurring daily in Boston Harbour and elsewhere.
Objection 4.
That salt water would cause galvanic action of a destructive nature
to take place at the joints where lead and cast iron were in close contact.
Answer.—That there is no instance on record of destructive galvanic
action having occurred in the case of lead and cast iron in contact under
salt water.
Objection 5.
That the chain of pipes, being of cast iron, would, owing to the
action of salt water, speedily become soft like plumbago, and in a few
months become utterly worthless.
Answer.—That softening of cast iron exposed to the action of salt
water takes place only in castings of inferior metal, and that it is on
record that castings of close grained, hard, white metal had resisted the
corroding action of salt water for 40 years and upwards.
Objection 6.
That in the case of a Narrows, connecting a large inland basin with
the sea, where the tide has a rise and fall of 12 feet, the counter cur- Smith on Vancouver Water Works.
rents in such a restricted passage defied calculation, and were more
likely to be greater at the bottom than at the surface.
Answer.—That the laws of nature are unchangeable, and that the
future experiments of the company's engineers would amply demonstrate
that it was impossible for a current exposed to the influence of a vast
friction bed, like the bottom of Burrard Inlet, to be greater than the free
-and unrestricted current of the surface.
Objection 7.
That the great force of the current rendered it imperative that the
whole chain of pipes should be laid in the short interval of slack water
between two tides, which did not exceed twenty minutes duration, and
that no means could be devised to perform such an arduous undertaking in suoh a short period of time.
Answer.—That the method proposed by the company, of jointing the
pipes and hauling them in a continuous chain across the inlet, would,
as before stated, entrench the pipes, and cause a resistance to motion
which would render it immaterial whether the pipes were laid in twenty
minutes or twenty hours.
Objection 8.
That the method of laying the pipes proposed by the company, viz.—
Jointing and hauling in one continuous chain, was impossible, as no
pipe joint could be made strong enough to withstand the enormous tensile strain this method would entail.
Answer—That the construction of the Ward flexible joint was of suoh
a nature that the lead packing could not be pulled out, and before a
joint could break asunder, it would be necessary for the cast iron bell
to give way, and that in consequence the strength of the joint was
limited only by the sectional area of cast iron exposed to the tensile
Objection 9.
That there were no instances on record of pipes laid in salt water
subject to a tidal current of 9 miles per hour, where the depth of the
channel was 60 feet, and the width 1240 feet.
Answer—That this was most certainly true, and that when the
Vancouver Company's submerged main was laid, it would serve as a
precedent for similar works on a more gigantic scale.
The above objections and answers, and many more of a like nature,
were publicly discussed and argued upon by professional men. Elaborate and specious mathematical calculations were produced in support
of each theory. However, as the work is now an accomplished fact,
all opposing theories are thereby proved worthless. 42
Smith on Vancouver Water Works.
In regard to the ninth objection, the writer is well aware that no
similar work of a like magnitude has ever been attempted. Greater
lengths of flexible pipes have been laid in lakes, rivers, and ocean
bays; but previous to the laying of the submerged main across Burrard
Inlet, no pipe of 12 inches diameter and 1100 feet in length had been
laid in salt water 60 feet deep, on a smooth rock bottom, and exposed
to a tidal current of 9 miles per hour. The nearest approach to it is
the Shirley Gut pipe, 8 inches diameter, laid by Mr. Ward many
years ago, which, as before stated, crosses an arm of the sea, 400 feet
wide, 37 feet deep, and subject to a tidal current of 7£ miles per hour.
The double line of 16 inch flexible pipe laid across San Francisco Bay
for the San Francisco Water Works Co. is the longest chain of submerged pipes yet laid. The pipes are seamless wrought iron tubes, 5-16"
thick, fitted with*cast iron faucets and spigots after the Ward pattern.
The bay, where the pipes cross, is 6300 feet wide, and entirely free
from currents. A thousand feet out from the Alameda shore it is 60
feet deep, but at two thousand feet it is only 15 feet, and this latter
depth gradually decreases till the San Francisco shore is reached. The
pipes were jointed on a large scow, fitted with a derrick and sloping
platform, and paid out from the rear as each successive length was
added. The whole time occupied in jointing and paying out the double
ine was 40 days.
The following table shews the more prominent instances of submerged
pipes, known to the writer as being laid previous to the laying of the
Burrard Inlet pipes.
Le     h.
Water Wor
Where laid.
Single 36 in
Toronto Water
Lake Ontario.
"      36
Lake Michigan.
"       36
Jersey City
Hudson River.
"       36
Delaware River
"      12
Double 16
San Francisco
San Francisco Bay.
"        8
Deer Island
Shirley Gut.
Single    8
San Diego
San Diego Bay.
"       6
Delaware River.
Coal Harbour, being shallow and its bed easy of access at all stages
of the tide, is crossed by a 16 inch rivetted steel main, 3-16" thick,
fitted with cast iron flexible joints, and costing $3.50 per lineal foot at
the   foundry.      Plate  XX shews the form of joint used.      Three Smith on Vancouver Water Works.
hundred lineal feet of flexible pipe were provided, but at the time it was
necessary to effect the crossing, it was found that unusually high tides
prevailed, and that this amount was insufficient. This difficulty was
overcome by rivetting two plain lengths to two flexible lengths, the
compound lengths, each 48 feet long, being placed at the ends of
crossing, the whole covering, when jointed, a distance of 348 feet.
The submerging of the pipes was effected without difficulty in the
following simple manner:
The total number of lengths were jointed in one continuous straight
line on the south shore, between high and low water marks, the forward
end resting on and firmly secured to a small scow.
The whole line was buoyed on each side by cedar floats, capable of
sustaining the entire weight. On the rising of the tide, the scow and
the chain of pipes rose with it, and when well afloat, a dozen men
stationed on the opposite shore hauled on a small rope attached to the
scow, pulling it forward, till the line of pipes was directly above its
destined position on the bed of the Bay. The floats were then cut off,
and the pipes allowed to sink to the bottom. At low water the ends
of the chain were exposed, and connection with the 16 inch mains on
each shore was effected without difficulty. The whole operation
occupied three days from start to finish. .
The general plan of the distribution system was designed by Mr. T.
C. Keefer, C. E., C.M.G., Past President of the Canadian Society of
Civil Engineers. Its excellence is therefore beyond question. Subjoined are a few of the more important details.
The city of Vancouver is laid out on the rectangular system, the
streets being 99 and 66 feet wide, forming blocks 260 feet wide by 500
feet long. The 16 inch steel main is carried under the principal streets
into the centre of the city. Branching from it, at suitable intervals,
by means of special castings, the larger sub-mains, 8" and 6" diameter,
form rectangles, from the sides of which the smaller sub-mains, 4", 2-J"
and 2" diameter, branch out in any required direction. The system is
liberally supplied with stop valves. Each pipe feeding direct from the
main, and each small sub-main feeding from the larger sub-mains, can
be closed independently, when required. In the case of breaks and
necessary repairs, a single street or part of a street can be shut off without interfering with the supply to other parts of the city. Should it
ever become necessary to shut off the whole system, a 16 inch valve is
provided on the main for this purpose, outside the limits of the distri- -44
Smith on Vancouver Water Works.
bution system. In all cases the valves have been placed at a distance
•of four feet from the initial point of the sub-main, or from the intersecting centre of the two sub-mains. The sub-mains are laid at a distance of 20 feet from and parallel to the street lines, so that the exact
locality of the valves can be found without difficulty, even in winter
when the ground may be covered with snow and ice. In most cities
the practice followed has been to locate the valves uniformly on the
iines of the street boundaries, the disadvantage of which is that a break
in a sub-main may occur between the valve and the feeding pipe, in
which case the valve is rendered useless.
To resist the severe water hammer, due to the great pressure in the
system, the valves are made unusually heavy.
The bodies, caps, and nuts are of cast iron; the spindles, stuffing
boxes, glands and followers are of composition metal.
The plugs are of cast iron with composition faces, and spindle bushings. The following table gives their dimensions, weight and cost in
Diameter in inches.
Shoulder to shoul-
Diameter of Beil in
Aver, weight in lbs.
Cost at Victoria....
The body of each valve is enclosed in a square brick chamber, built
to such a height that the top of the valve chamber (a small, square
cast iron box, weighing llllbs., and protecting the nut of the spindle),
when placed upon it, is flush with the street.
The system is provided with 75 double valves, two hose Matthew's
fire hydrants, with 4 inch valve openings. This hydrant is in
general use throughout the United States. The manufacturers claim,
and the claim is conceded by all cities using them, the following
advantages over all others.
There being two main valves, possible leakage is reduced to a
minimum. The lower valve, working independently of the upper
valve, the hydrant can be disconnected for repairs, without the
necessity of excavation, and without shutting off the feeding sub-main. Smith on Vancouver Water Works.
The rod and automatic waste valve, attached to the upper induction
valve, work in such a manner that the opening of the lower induction
valve involves the closing of the waste valve, and vice-versa. Waste of
water cannot therefore take place, and no water can remain in the
stock of the hydrant, when the upper valve is closed.
The lower valve being capable of independent action, the temporary
removal of the upper valve for repairs does not interfere with the
utility of the hydrant.
As previously stated, the works of excavation and pipe laying main
included south of Burrard Inlet were carried out by the company
by day labour. The average depth of trench for the mains was 3' 10,',
and for the sub-mains 3 feet. The cost, including tools, laying pipes,
placing specials, erecting hydrants, refilling and tamping trenches,
taking up and replacing crossings, and works of a like nature, did not
exceed 17 cents per lineal foot.
On Wednesday, the 20th of March, 1889, the gate in the well chambers-
of the dam was partially raised, and water allowed to flow for the first
time into the 22" main. The 8" blow off near the rock tunnel was
kept open, and the water was not allowed tor several days to fill up to
he level of the tunnel, and flow into the 16" main. On March 25th
at 4 p. m., the gate in the well chambers was opened wide, and a full
head of water turned on. At 6 p. m. the 22" main was filled, and
began flowing through the tunnel into the 16" main. At 9.45 p. m.
the water reached the closed 12" valves, on the uorth shore of Burrard
Inlet. At 10p.m. the valve controlling the 12" submerged main
was opened three-quarters full. At 10 minutes past 10 the water
reached the south shore. At 3 a. m. it had reached the termination
of the 16" main in the centre of the city, and at 4 a. m. it was
discharging fully into False Creek, by means of an 8 inch sub-main
opened wide.
It is worthy of note that in the whole length of the mains, not a
single joint was found to leak. Such leaks as were discovered occurred
at the seams, where the rivetting and split caulking had been imperfectly
done. These were speedily repaired by encircling the mains by steel
rings, 4 inches wide, made in two halves, and provided with " Lugs. "
The lugs were bolted together, above and below the main, the space
between the ring and the pipe being filled up with lead, and carefully
caulked in the usual manner.
From the drawings accompanying this paper, Plates XVI to XX
have been prepared.   48
Vancouver   Water   Works-—Appendix.
Description of Work.
27+00 |Hauling and distributing
as above 19 lengths of
16// steel rivetted main
■^s// thick, in lengths of
. 23/ 9§// each, weighing
sleeve   included
Hauling and distributing
lead for same	
215+59|Hauling and distributing
as above 767 lengths of
16" rivetted steel main
T10-10-// thick, in lengths of
23/, 9f//,each weighing
Isleeve included,  550-j1,.
Hauling and distributing
as above, castings, lead
and extra sleeves for 16"
main as above	
215+59JPartially distributing 6
lengths of 16" steel main
(rivetted) as above
350+89 Hauling and distributing
as above 568 lengths of
22// steel rivetted main
T1Ij^// thick, in lengths of
23/ 9f// each weighing,
Hauling and  distributing
as above 2 rivetted steel
thimbles for connection
at dam, each weighin
242 lbs	
Hauling and distributing,
as above, castings, lead
and extra sleeves for 22"
Partially distributing as
above 7 lengths of 22"
rivetted steel main	
Partially distributing as
above 6/ 22// castings-
4  67
2-3 4-
O  3.8
7 64
7 64
12 87«
6 44
14 85
14 85
14 85
7 43
•2425 43
594 38-
9 53-
2812 14-
3 27'
458 57"
$38,674 25-   Vancouver Water Works—Appendix.
Angle of
of each.
Length of chord
from bell to bell
Distance covered
when in place
in trench.
Cost of each
at Foundry.
5 degrees
310 lbs
K 7 'I
10        "
409 "
15        "
487 "
20        "
580 I
25        |
674 I
30        «<
772 §
35         "
870 1
40        "
939  "
1    41£"
50        "
1188 «
55        "
1267 |
60        «
1347  "
4'. 11"
6 21|"
70        «
1490 |
90        "
1860 "
Angle of
5 degrees
10       |
15       "
of each.
470 lbs
630 "
785 "
910 "
1050 "
1270 "
1435 "
1895 "
Length of chord
from bell to bell
2/ 2"
3'. 2 J"
4'. 2i"
Distance covered
when in place
in trench.
Cost of each
at Foundry
-w     I
Mr. E. Mohan
The author, in his very interesting and comprehensive paper, gives
the average velocity of the Capilano River for the first seven miles from
its mouth at five feet a second. A difficulty has been frequently
encountered,—notably in the precipitous mountain ranges of the Pacific
Coast,—in maintaining impounding reservoirs in somewhat similar
situations. In these torrential streams it has been found that the
boulders and gravel washed from the banks and bed above the dam are
liable to be swept down and gradually fill the reservoir. A velocity of
five or six feet a second will move good sized boulders, and less than
half that velocity will move gravel. From the author's description of
the bed of the stream, it would appear that the velocity has been great
enough to remove most of the gravel from the channel, except at points
where from the formation of the banks slack water was encountered.
On the other hand, as even the most sanguine Vancouverites do not
anticipate a population of four or five millions requiring a daily supply
of 440 millions of gallons, a storage reservoir, as such, is not an essential,
provided means are adopted to keep the entry conduit clear.
The steel mains which were built and laid by the Albion Iron
Works Company of Victoria did not when first laid give satisfaction.
Many were leaky, and considerable damage was done to the streets.
These defects have, it is believed, been since remedied, and the writer
is informed that the same Company subsequently laid a similar main
with complete success for the. Victoria Water-Works. The writer
understands that the Iron Works Company was solely responsible for
making and laying the steel mains in Vancouver.
On the 15th November, 1889, a serious accident happened to the
submerged main, by which the city was deprived of its water supply for
eight days. One of the 12 in. cast iron pipes, lying in 40 feet of water,
was badly fractured, and it can hardly be doubted, from the position of
the pipe, that the break was caused by a water ram. The writer understands that Mr. G. A. Keefer had recommended automatic blow offs
at each end of the submerged main, to guard against just such an accident ; but unfortunately his advice had been neglected. The blow offs
referred to by the author are not automatic. The break was repaired
by a diver, who covered the fracture with a wrought iron sleeve,
in two parts bolted together and lined with vulcanite. In justice to
the Water-Works Company, it should be added that it did all in its
BHJaw Discussion on Vancouver Water Works.
power to reduce the inconvenience to a minimum, bringing water across
the Inlet, and delivering it free to its customers by cart.
Having been resident in Vancouver at the time of the discussion of
the rival water-works schemes, the writer is aware that the main
objections raised to the Capilano project were based upon the supposed
difficulty of crossing the Narrows. One party said that it could not be
done, because no similar work on the same scale had yet been attempted.
It is to be hoped that there are but few engineers with whom such a
reason would have weight. The more reasonable opponents of the
work did not dispute the practicability of laying the pipe, but thought
that, in only having a single main, in the event of an accident to it,
the safety of the city would be imperilled, putting on one side the
terrible inconvenience of a short water supply. Prom Mr. Smith's paper
it appears that the duplicate submerged main was a part of the general
plan which hitherto the company has failed to carry out.
The writer has always believed in the feasibility of laying the submerged main, and has never hesitated to express this opinion to the
opponents of the scheme. Looking, however, to the vast interests at stake,
he has always insisted upon the necessity of the submerged main being
in duplicate, and is glad to learn that since the accident the company
has taken active steps towards laying an additional main across the
Narrows. Prom the reports of the divers it would seem that the
pipe has not been moved by the current from the position in which
it was first laid.
With regard to the dangers to be apprehended from vessels' anchors,
the writer is inclined to think they have been exaggerated; nevertheless,
were he responsible for the maintenance of a single line of pipe across
the Narrows, upon the integrity of which the very existence of the
city might at any moment depend, he would be very uneasy until all
human precautions had been taken to secure its safety. A simple
mode of obviating any risk from ships' anchors would be to cover the
pipe with concrete in sacks laid by a diver. This if done properly
would also add greatly to the strength of the pipe; and if the diver's
report, that the pipe was lying in a groove in the sandstone rock, is
correct, the pipe, by the means suggested, would be completely incased
in a shell of rock.
The unfortunate accident referred to has, for a time, put a stop to
submitting to the ratepayers a scheme for the purchase of the Water
works by the city; and though the vast majority of the citizens are
agreed that the city should own its water supply, no by-law approving
of the purchase would pass, unless it was felt that such  another 54
Discussion on Vancouver Water Works.
Mr. P. Sum-
failure was practically impossible. The burnt child dreads the
fire. Vancouver has once been wiped off the face of the earth, and
it cannot be wondered that her citizens should refuse to purchase a
system, which, under the same circumstances as occurred a month or
two ago, would leave them at the mercy of a conflagration which
might occur at any moment.
The conception of the Capilano Water-Works for the supply of the
city of Vancouver was of its very nature one of the best examples of
hydraulic engineering to be met with on the Pacific Coast of the
Dominion. The great difficulty to be encountered was the crossing of
an arm of the sea under conditions entirely unparalleled in water-works
engineering, and in some respects the pipe line across the Narrows of
Burrard Inlet is almost without an equal.
The design of Mr. G. A. Keefer, M. Can. Soc. C. B., for the Capilano Water-Works was not the only design for Water-Works for the
city of Vancouver.
The Coquitlam Water-Works, designed by Mr. B. A. Wilmot,
M. Can. Soc. C. E., was also under consideration at the same time that
the Capilano Water-Works was being matured, and as both projects received a great deal of consideration at the time, their merits and demerits being frequently discussed both in the press and elsewhere, it may
not be out of place for an outsider to present the features of both schemes
to the Society of Civil Engineers. The data before the writer are the
reports by Col. W. R. Bckhart, of San Francisco, on the Capilano River
Scheme, and which is now constructed, and the Coquitlam Lake
Scheme, which is now being matured for the supply of the city of New
Westminster, reported upon by Mr. H. Schussler, chief engineer of
the Spring Valley Water-Works of San Francisco.
The elevation of the pipe inlet as given by Mr. Eckhart for the Capilano was 422 ft.
The elevation of the pipe inlet, as given by Mr. Schussler for the Coquitlam, was 435 ft., so no important difference obtained between the
elevations of the two proposed systems.
The entire length of the Capilano scheme, as given by Mr. H. B.
Smith in his valuable report, is 52,741 ft. from the dam to the centre
of the city, or very nearly 10 miles.
The entire length of the Coquitlam scheme as given by Mr. Schussler
is 105,600 ft., or 20 miles.
It will thus be seen that the laying of the submerged pipe across the
Narrows of Burrard Inlet was the direct means of saving 10 miles of
piping so far as Vancouver water supply was concerned. t)iscussion on Vancouver Water WorJcs.
Mr. Thomas C. Keefer, C.M.C., Past President of the Society, reported upon the Capilano scheme; and looking now upon the accomplished work, the only feeling is one of admiration for the design of so
bold an engineering feature as the submerged pipe line across Burrard
Inlet Narrows. For a country or carrier main, mild steel enables the
engineer to undertake works of magnitude, that, if he were compelled to
use cast iron mains, would almost render the cost of such works prohibitive.
At present ruling prices, a steel main of equal dimensions can be laid
complete in the trench for about the same figure that a cast iron main
could be landed on the wharf in Vancouver or any other port on the Pacific Coast; thus shewing clearly that it is possible to effect a saving of
nearly 40 per cent, in the cost of the pipe line. But valuable as mild
steel is for mains of large dimensions, there appears to be a tendency
among engineers to exact too much from it.
Thus in the Vancouver Water-Works system, the ultimate strength of
the 16 inch main, allowing a tensile strength of 60,000 lbs. per inch for
plates, and .7 for strength rivetting, would be 580 lbs. per sq. inch, and
the same data for the 22 inch would permit of 420 lbs. per sq. inch. Now,
according to Mr. Smith's report, the 22 inch main has to sustain a head
of water equal to 164 feet in height or a pressure of 71.17 lbs. per sq.
inch, as a maximum.    The factor of safety in this case is therefore 6.
In the case of the 16 inch main, the head of water sustained is
equal to a column of 415 feet, and equal to a pressure of 180.11 lbs.
per sq. inch as a maximum. The factor of safety in this case equals
a little over 3 or nearly 3J. It is apparent therefore that no very
great amount of corrosion can take place without very materially
reducing the factor of safety, and that moreover the workmanship and
material must be the best obtainable.
From Mr. Smith's report there is a flush valve on the 22 inch main,
and likewise one air valve in the tunnel. On the 16 inch main there
is a flush valve on the north side and another on the south side of the
Inlet, one air valve located between the Inlet and Coal Harbor, and
two flush valves on opposite sides of Coal Harbor.
The total number of vertical and horizontal bends are given by Mr.
Smith at 179. It would be interesting to know how many of these
are horizontal and how many vertical bends; for no matter how
free from sediment the waters of the Capilano may be, there are
certainly a great number of elevations and depressions without either
air or flush valves, and doubtless there will be a great difference of
opinion as to the capacity of the 16 inch main after accumulations of
air and silt have taken place. 56
Discussion on Vancouver Water Works.
From the description of the turning on of the water, it is very evident that water-ram and air compression were treated as a myth, and
it would afford a great deal of valuable information if'Mr. Smith would
give a detailed account of the behaviour of steel mains, having a factor of
safety ranging from 3 to 5, a large number of angles of elevation and
depression along the pipe line, few air valves and very few flash
valves, with a full head of water flowing with a velocity of 1.4
feet per second. The water seems to have taken 3f hours to traverse
19,320 feet from the tunnel to the north shore of Burrard Inlet.
That something occurred may be inferred from the last two paragraphs of
Mr. Smith's paper.
It appears that a tank is now placed on the north side of the Inlet at
an elevation of about 230 feet above tide water mark, thus reducing
the pressure upon the lower portions of the system by 200 feet of
head, and likewise reducing the pressure upon the upper portion of the
system by the difference between the velocity and static pressures, as
the tank is allowed to flow over and to discharge the surplus water
into the river. Some kind of"provision is made for inserting a plug so
as to utilize the full head of 430 feet in case it should be required.
Taking Mr. Smith's description of the Vancouver Water-Works, one
cannot help feeling that there can be no difference of opinion about the
works as projected by Mr. Keefer, for they are without a doubt unrivalled in this section of country: but there is a wide field for discussion
relative to the manner in which Mr. Keefer's plans have been carried
out. Doubtless other engineers who are conversant with the question
will give us the benefit of their extended and valuable experience on this
very important branch of engineering,—water-works construction.
Mr. D J. Rub- -^^e wrfter has read the advance proof of Mr. H. B. Smith's paper
eeii uunoan. on ^g Vancouver Water-Works with very great interest, and as copious extracts have been made from a small pamphlet, which the
London Steel Pipe Co. published a few years ago, and the joint referred
to in that pamphlet is described on page page 338 (Vol. Ill) of the
author's paper as a cumbersome arrangement, the writer desires to bring
under the notice of the Society some particulars with reference to this
The advantages of a single socket, as compared with the Moore &
Smith joint, are the great reduction in the quantity of lead consumed,
the adaptability of the joint to overcome angular deviations in the
pipe track, and the increased strength of the spigot, which must be
of sufficient firmness to resist the strains due to caulking by inexperienced workmen. Discussion on Vancouver Water Work
The Moore & Smith joint, although it has no doubt been very
largely employed, is inferior to the joints referred to, because the nipple
reduces the internal diameter of the pipe, and tends to create any
increased velocity of the flow of water through the pipe and the joints ;
whereas the joint now referred to is so constructed, that the pipe is
actually larger in internal diameter wherever the joints occur, and this
system of construction is of decided advantage by reducing the friction
of flowing water.
It cannot be fairly said by the author that the joints to which the
writer refers have been entirely discarded on the Pacific Coast, for the
simple reason that they have never yet been employed there, with the
exception of the pipe line constructed by the Steel Pipe Company,
Limited, London, for the water supply of the town of Mazatlan, Mexico.
This is the first pipe line on the Pacific Coast fitted with the improved
joints, and is a line 20 miles in length and 14 and 12 inches internal
With regard to the 22 inch and 16 inch pipes, it is a matter of
some surprise that these pipes were specified so thin. The pressure
upon the pipe line is about 400 feet near to the Burrard inlet, and
the pipes at this point are 16 inches diameter, and .11 of an inch
in thickness. So far as the writer's investigations go, the ordinary
working pressure on a Californian wrought iron pipe of this thickness
and diameter would not exceed about 265 feet head, and according to the
practice of this company, pipes of this thickness would not be made of
mild steel for a higher pressure than 285 feet. It is clear therefore
that the pipes laid by the Vancouver Water-Works Company are
subject to much higher strains than is customary, and there is every
reason to believe that the life of the pipe will thereby be injuriously
affected. The paper gives no information regarding the test pressures
imposed upon the steel pipes, indicating that the cast iron connections
only were tested to a pressure of 300 lbs. per square inch.
It is unfortunate that the author has not given details of the rivet-
ting. On the 2nd of May, 1889, this Company received a long letter
from him, indicating that the pipes leaked to such an extent that when
the water was turned on it was found impossible to stop them. In his
letter dated April 13th, 1889, he says: " In two days time it was
discovered that the leaks were enlarging, and the plates being actually
cut by the jets of water forced out between the laps. The holes made
in tins manner are just beyond the edge of the outside lap, and are of
all shapes, some being circular and nearly half an inch in diameter.
After the pressure the appearance presented by the pipes was as if they 58
Discussion on Vancouver Water Works.
had been acted on at several points by a sand blast." He then goes on
to state that attempts were made to repair the leaks by cast iron rings;
but on again subjecting the pipes to pressure, new leaks were developed,
and at the date of his letter 5,700 feet of the pipes were leaking so
badly that their removal was ordered and new pipes decided upon.
It is somewhat misleading to read in the latter paragraphs of the paper
that the leaks discovered at the seams were speedily repaired by steel
rings four inches wide, made in halves, which were made to go round
the pipes, in the face of the information contained in the letter of the
13th April, to which reference has been made, and the writer would
suggest that the author should correct the latter part of his paper.
In replying to the author's letter about the leaks, the following
questions were asked, and it is desirable that the information
should be communicated to the Society :
Were the rivets put in hot or cold ?
Were they closed by power, and if so by what power ?
How were the longitudinal seams distributed round the circle ?
What was the length of the rivets ?
Were any sections of pipe rivetted up and tested by hydraulic
power in excess of the working pressure, to prove the quality of
the rivetting, and was any test whatever made of the rivetting ?
With what were the pipes coated, and how was the coating
applied ?
How were the pipes jointed together ?
The length of pipes being 23 feet, 9f in., seven plates   was   a   very
large number to employ. Four would have been preferable, thus saving
a number of rivetted seams.
The rivetting was apparently faulty and the plates not properly laid
together at the seams, and possibly burrs were left at the rivet holes
which prevented the perfect contact of the surfaces.
Although a rivetted joint may be made strong enough to resist the
strain brought to bear upon it by a tension testing machine, and may be
proved to be very nearly equal in tensile strength to the original plate,
yet the rivets may not be close enough to secure water-tightness under
pressure. Experience has proved that there is considerable difference
between pipes being strong enough to resist a certain tensile strain
across the joint and being water-tight under fluid pressure, which
may exert no greater tensile strain across the joint than that for which
it was designed. The evidence submitted at the time led this firm to
the opinion, that the defects in the pipe were due to faulty workmanship
and inexperience on the part of the contractors. Discussion on Vancouver Water Works.
^No information is given in the paper as to the means adopted for
i letting off air in the pipe, and possibly the leakages which were found
were aggravated by the accumulation of air in the mainsbefore the pipes
were properly filled with water. Neither is any evidence given as to
the hydraulic tests upon specimens of the pipes to prove the tightness
of the rivetting.
The writer considers it bad practice to make up thejpipes of so many
small plates. It is found much more satisfactory to make the plate
cylinders in lengths of from 6 to 8 feet; and if this system had been
adopted for the steel pipes described by the author, there would have
been a great reduction in the number of circumferential rivetted
Caulking, in the manner described by the author, tends rather to
aggravate leaks than to prevent them whenever mild steel plates are
employed. A better system for preventing leakages and to make perfectly water-tight joints is to employ suitable plate closing apparatus
upon the rivetting machines, the effect of which is to close the plates
perfectly together, so that the laps are thoroughly overlaid and the
surfaces brought into close contact before the rivets are inserted ; then
when the rivets are closed by hydraulic pressure there is no liability to-
leakage, and caulking is entirely dispensed with.
The description of the method in which the lead joints were made is
such as to indicate that the work of making a Moore & Smith joint is
at least twice as much as that of making the patent socket joint
recommended by the writer. The great care which has to be taken in
making the Moore & Smith joint in a perfectly straight line is a very
serious disadvantage. A particular feature in an efficient joint should
be its facility for overcoming angular deviation in the pipe line, and
thereby reducing the number of bends.
It may be mentioned that a 9 inch steel pipe, which was laid by this
firm in the Tay Viaduct some years ago, was laid round a very sharp
curve on the bridge, without any special connections whatever, the
flexibility of the joint permitting the curvature of the pipe line. The
joints were made above ground, and 360 feet of jointed pipe in a single
curved line lowered into the trench.
Nor is there any necessity in a joint such as is recommended for the
caulkers to be cautioned with regard to the packing of the lead. The
spigot ends of the pipes are welded into cylinders, and made about 50
p.c. thicker than the pipe, so that any pressure the caulker imposes upon
the lead will not indent the pipe, or in any way cause it damage. The
sockets for pipes up to 24 inches diameter are stamped by hydraulic 60
Discussion on Vancouver Water Works.
pressure on the system invented by Mr. James Riley, General Manager
of the Steel Company of Scotland, Limited. This is known as the 1
Riley Patent Socket. This socket has an external lip or flange which
adds very greatly to its strength, and with this joint it is not found that j
the shell of the pipe is bent inward nor the mouth of the socket bent
outward, as is the case with the Moore & Smith joint, described by the
author on page 339 (Vol. Ill) of his paper.
With regard to the system of constructing pipes with inner and outer
courses, this method is entirely discarded by this Company on pipes \
of all diameters made of plates under .2 of an inch in thickness. A
much more correct and reliable method of making perfect circumferential seams is obtained by the use of suitable machinery for expanding
the circumferential lap at one end of the plate cylinder, after the plates
have been punched and bent into cylinders; all the circumferential
seams are carefully punched by multiple punching machines precisely
to the same pitch, and the increase in pitch due to the enlarged diameter of the overlap is obtained by stretching the plate in the manner
For pipes made of plates exceeding .2 of an inch thick, the circumferential seams of the outer course are punched, when the plate is flat, by
dividing machines which divide the pitch into the most minute fractions
necessary to overcome the irregularities between the diameter of the
outer and the diameter of the inner course of plates. The system
described by the author, of leaving one end of the plate cylinder to be
punched while the pipe is rivetted up in the pipe track, is not only one
which causes enormous delay in the execution of a pipe contract, but
ene which is liable to very great inaccuracies of workmanship. The
method described of fitting the pipes together in a trench, drawing
them apart, punching the holes, putting them together again, and rivetting them up is most unsatisfactory.
With regard to the distribution of the longitudinal or straight seams,
a pipe made of very thin plates, such as those employed for the Vancouver Water Works, is made stronger and better by a uniform distribution of the longitudinal seams of the several courses around the
circumference, instead of having them all in one straight line. Sometimes in constructing pipes with the longitudinal seam all in one line
along the pipe, much greater care has to be taken in straightening the
pipe, as there is a tendency to curvature.
The paper is one of very great interest to all engineers interested in
water-works, and the author has displayed conspicuous ability in the
execution of the works under his charge. Discussion on Vancouver Water Works.
Mr. Henry Badeley Smith's able paper on the construction of the dam Mr. G. H. Hen-
and pipe line of the Vancouver Water Works leaves little for the critic
to say, beyond expressing admiration for the successful manner in which
difficulties both novel and otherwise have been met under exceptional
circumstances. The moderate height of the dam no doubt permitted
the foundation (in the absence of puddling clay) to be laid with safety
in the manner described; nevertheless, the accident to the "lean to,"
and one of a similar nature that occurred previously to another part
of the structure, would seem to be a warning in ease it should be proposed
at any future time to increase the height of the dam.
With regard to the objections and answers to the plan of laying the
pipe across the Narrows of Burrard Inlet, a few remarks may be in order.
The first objection begs the question by claiming as known a force
that is not known, while the answer claims a mathematical demonstration not given, but doubtful in its fundamental basis, as all calculations
of the forces of the sea must be without actual previous experiment.
The true reply to the objection is philosophical not mathematical.
The bed of the inlet at the point of crossing is stated to be composed of
soft sandstone, partially covered with mud, gravel and cobble stones. Now
a stream with no friction on the bottom, other things being equal, would
flow with equal swiftness at every depth, hence the difference in speed
between the bottom and top would be a measure of the friction at the
bottom. But the present case is one of a tidal current, whose efflux and
reflux are governed by so many and complicated conditions, extending
even to the existence, occasionally at least, of an outward current at
bottom and an inward current at top, that even the most careful observation would fail in producing reliable formulae. It is plain then that where
the erosive force is actually unable to remove mud, gravel and cobble
stones, a 12 in. metal pipe .would lie undisturbed, quite independent of its
self-made trench. This argument also answers the second and sixth objections. The third has a little more in it, and is more weakly replied to. Itis
decidedly doubtful whether such a pipe could resist the direct impact of a
falling anchor, or that any engineer would invite a trial; nor is it likely
that he would be free from anxiety should a vessel, whose anchor fluke
embraced the pipe, be caught in a violent gale. It would be interesting to
sec what mathematicians would make of the problem. The wisest course
would seem to be to protect those parts which are not sufficiently trenched
already in the bottom. The fourth and fifth are speculative, and need
facts capable of clear verification in their support before they have a
right to claim notice. The seventh and eighth are mathematical, to which
a very decided practical answer has been given.    The ninth objection is 62 Discussion on Vancouver Water Works.
of unascertained antiquity, and is still voiced wherever old settled habits
of thought are disturbed by new ideas. " Who ever heard of such
a thing ? "    " Don't take much stock in these new fads," etc.
It is true that it is sometimes difficult, without giving more attention
than is convenient, to decide whether a certain scheme is the work of a
" crank" or of a scientific mind, but if it is worth considering at all, it
is surely the duty of critics to meet arguments with something more
than mere assertion, and to support their own statements on facts and
premises which are undisputed or capable of ready proof. Mr. Smith
i* to be congratulated on having contributed such an interesting
paper to the Society.
Mr. Peterson. As the paper is a long one, it will be better to discuss that portion
relating to the dam this evening. If the paper is all read through at
once, the points relating to the dam will probably be forgotten by next
Mr. Biackweii. Was there not some mistake about the size of the tree ? It would
be hard to get the stumps out.
Mr.c.L.Smith. The author was perhaps a little misleading. The tree must have
been one badly formed at the root. He did not think there was a tree
of that size solid right through.
Mr.c.E.Goad. It was a rare thing to see a tree solid right through there. The
paper is well written and the subject elaborately gone into. There
seemed to have been great difficulty in regard to the carriage of material, etc., but this matter could not be followed out properly without
being able to see how the plans were adapted to the peculiar formation
of the ground.
Mr. Peterson. It would have been much better had a large sketch been made
showing the most important points of the dam, etc., and the methods
taken to prevent the water following through the longitudinal timber,
which is a great source of trouble in all works of this kind. As he
understood it, in this case the water was kept back by sheet piling.
Then there was the question of getting round the end, also a difficulty
that had been overcome by a brush and gravel bank, but he did not
understand how the concrete was put in. It would be a good plan to
make a rule, such as was in force in the Institution, that all papers
should be accompanied with large drawings which could be hung on the
walls.    This would very materially tend to aid the discussion. Discussion on Vancouver  Water Works.
In reading the paper over, it would seem that some of the concrete prof. Bovey.
had been made by mixing a certain amount of cement with gravel, out
of which stones over a certain size had been eliminated. This was
contrary to the practice recognized by many eminent authorities, who
said that the best concrete was made of stones of irregular sizes—not
limited to 1^ inches.
The material was different from what we had here. Take the London Mr. Gower.
gravel, which was similar to that they had in Vancouver, and the
practice was the same. All big gravel was taken out. It was
more of a sea gravel than anything else, and generally it ran 1J" to
f," and with a fine kind of grit was mixed six to one with Portland
cement. This plan was invariably adopted by engineers in the south of
The best concrete was made of gravel of different sizes—the more prof. Bovey.
irregular the sizes the better.    As to limiting the size of the gravel,
some engineers say that large blocks of stone can be put in without
injuring the concrete in the least.
Col. Hayward, City Engineer of London,  in his specification for Mr. Gower.
gravel, specifies for  a small gravel nothing larger than 1^," mixed in
proportion of six to one.
The idea of having stones larger than 1^-' taken out was probably Mr. Peterson,
because the concrete had to go into a small space and had to be
rammed. In filling large spaces there was no question but that large
stones, two or three feet square, might be put in. In the Lachine
bridge they put large masses of concrete—20' x 40' x 15'—in the
middle of which they might have put large stones and filled in all round
them, and in many cases stones larger than 1-|" were put in. In the
case under discussion, however, they were quite right in taking out
everything over 1^," for the reason that they had to pack the cement,
and it would be much easier to ram and make a much better job with
small stones than with large ones.
The timber in Burrard Inlet would in all probability be affected by Prof. Bovey.
worms.    He did not know to what extent the Teredo  Navalis had
appeared on the Pacific coast.    Some member present might be able to
say whether any information had been obtained on this subject.   • 64
Discussion on Vancouver Water Works.
Mr. Biackweii. Had seen specimens in the Library of the House of Commons at
Ottawa, showing the effect of the " Teredo Navalis " on the timber of
the Pacific coast, which is very disastrous. It had occurred to him what
a fortunate city Vancouver was as compared with other towns having the
average system of water-supply and cost of same. In Vancouver they
had an abundant supply of cold, clear, running water for almost nothing. He imagined that if these works had been economically carried
out, there would be a supply of water for all time at a cost of from
one to two dollars per head per annum. There was a pressure of
801bs. at the highest part of the town, which was remarkable.
Mr. Peterson. In the harbour of Vancouver, on Burrard Inlet, piles of 18" diameter
are eaten through sometimes in a year, and in a sawn section one can
scarcely find half a square inch in any one place of solid wood. The
holes are \" in diameter, and some probably larger. The wood is just
as if it had been perforated with a series of auger holes. He felt certain the teredo would never reach the dam, there is a great elevation
with a very heavy current. They did not find them even where the
water was brackish.
Prof. Bovey. It had been stated that good cast-iron would resist the action of salt
water, but he did not know of a case on record in which it had been
found to do so. In a case that had come under his own notice columns
taken up from foundations in salt-water were just like sponge.
Mr. Biackweii. He had not had much experience with cast-iron in salt water, but had
seen cast-iron taken out of columns or piers just like cheese. That was
the ordinary gray cast-iron of foundry casting. He would like to know
something more about these pipes; which he believed were made in Glas--
gow. The sheet steel pipes had ends rivetted on which were described,
as " white cast-iron." There must be something more than that, because if simply white cast-iron they would be of no use at all; they
would belike glass, having no tensile strength. In Glasgow there were
several firms which made a specialty of producing a metal called
" McCaffie's metal"—a sort of white metal. In the first place when,
cast it was like white pig, and was then decarbonized in a furnace, which
converted it into a kind of semi-steel. If simply white metal they
would not be fit for the purpose. It might be assumed that these joints
were made of a material superior to ordinary "white pig " or " white
cast-iron." A little more information regarding these joint castings
would be interesting. Discussion on Vancouver Water Works.
Plenty of white metal water pipes could be obtained from Scotland Mr. Petersoa..
and the north of England.    He had got some out of pure white metal,
which were nearly useless.
White metal was useless, as it had so very little tensile strength.       Mr. Biackweii
When in Vancouver last year he saw this cast-iron pipe, a portion of Mr. Peterson,
the wrought-iron pipe, and also the valve that was placed between the cast
and wrought iron pipe. There was a very heavy stream of water pouring-
out, and he was told that the wrought-iron pipe was not sufficiently
strong to stand the pressure. Its thickness was less than J"—about
i", and the rivets, he imagined, were about 4" or 5" apart. It was a
very thin plate, and the water was flowing in little streams out of nearly
all the joints of the sides. He believed that they had since put straps
round th pipe to hold it together, but they had reduced the pressure
by means of a safety valve. He was astonished to see the small size of
pipes which had been put in for mains, viz. :—4" and 2^-", and he
thought this was a mistake. He'considered streets should have nothing
less than 6" pipe, that is in places where there was to be fire protection.
A 4" pipe was not large enough, and a hydrant on a 2" pipe, in the case
of a large fire with three streams going, would be utterly useless—there
would be no water. It was, however, a case where the water supply
was furnished by a company, which he thought should never be done.
There can be no possible reason why every city or town should not build
and own its water works. The corporation can certainly borrow
money at a lower rate of interest than any company, and can afford to
put in just such distribution of pipes as are wanted, and the corporation
will then have complete control of its own streets. He knows of no
instance where corporations have been supplied by water companies
in which the supply has not been unsatisfactory, or in which the cost to
the citizens has not been greater than if they had had a much more
liberal supply under their own control.
In reply to the remarks of Mr. Mohun, the bed and banks of the Mr. H.B. Smith
Capilano having been, for ages past, swept by flood currents of great
velocity, all loose boulders have long since been carried away, leaving
the larger boulders so interlocked that it is rarely, and only in extreme
floods, that one is loosened and carried down stream. In the course of
years, it is probable that detritus from the banks may lodge itself in
front of the dam; but, should this accumulate to any inconvenient
extent, it can be readily removed, and at small expense. In the ordin- 66
Discussion on Vancouver Water Works.
ary quiescent state of the river, when the current does not exceed 5
miles per hour, it is not probable that stones or good -sized boulders, as
suggested by Mr. Mohun, are in motion along the bed of the stream.
Smeaton's observations go to prove that a velocity of 11£ feet per
second will not derange quarry rubble stones not exceeding half a cubic
The considerations which decided the construction of a dam and
reservoir at the point of supply were as follows:—A pressure as nearly
uniform and invariable as possible would be obtained. There would
be a direct gain of 11 feet of head, which would allow the system to be
carried through the rock bluff at the end of the 22 inch main, by means
of a tunnel of moderate length. To obtain the same head without a
dam, the 22 inch main would have to be extended at least 1,500 feet
further up stream, under conditions which presented features especially
unfavourable for pipe laying, and for the permanency of the pipes
when laid. By reference to plate XVI., it will be seen that, at a
point 1,500 feet above the present dam, the channel of the river is not
confined to one permanent position. Moreover, as the eastern branch,
at the present period, is dry at low water, and the island between the
channels is invariably flooded at high water, an extension of the 22 inch
main above the dam must necessarily cross the western branch to its
western side, and following that side, where the banks are high and
rock in situ is exposed, reach a point where works for an entry conduit
of a safe and reliable character, during all stages of the river, would be
of a costly nature.
Finally, as the Capilano is a stream which has only become known
since the construction of the water works, reliable data as to the lowest
possible stage of water are still wanting. It is impossible to say with
certainty what changes in its volume may occur in the future; more
especially when it is taken into account that its banks maybe denuded
of timber by forest fires and the improvements of settlers, and that at
any season the snowfall in the mountains, on which its very existence
depends, might be very much less than it has ever been known to be.
Giving due weight to these considerations, no doubt Mr. Mohun will
acquiesce in the opinion that the dam is a very valuable addition to
the system.
Several lengths of steel main in the low lying parts of the system
subject to full head, when first laid, developed leaks more or less annoying. These leaks were, however, overcome in the usual manner, by
means of cast-iron lugs, leaded and caulked, and by the substitution
of new lengths where the originals were leaky enough to warrant this Discussion on Vancouver Water Works.
step. The Iron Works Company, who furnished the pipes, under contract for manufacturing, laying, and maintaining the line of mains for
two months after completion, were solely responsible for the leaks; and
it is worthy of note that the substituted lengths, made by them from
the same design as the original leaky ones, were found to be entirely
satisfactory, showing that the leaks in the original lengths arose, not
from the character of the pipes used, but from some fault in the construction of these lengths. The Victoria Water-Works, for which the
same company subsequently made and laid a main exactly similar in
all its details, is no parallel case, the total head of that system being
under 200 feet.
When the break in the submerged main occurred, the writer was in
England, and can give no account of it from personal knowledge. All
the evidence in his possession goes to prove that it was a break pure
and simple, due to a severe water ram created by a too sudden closing
of the hydrants, and acting upon a defective pipe casting. The outer
and inner skins of the metal at the point of rupture were found to be as
perfect as on the day of immersion, 15 months previously, proving conclusively that the action of the salt water had had no deteriorating
effects. There appeared, however, in the centre of the fracture, a
globular mass, showing a check by cooling or some other cause, which
had prevented a proper fusion of the metal during the process of casting. The fracture beyond the flaw showed the hard, close-grained,
white surface of a casting of good quality.
Inasmuch as there has been a great diversity of opinion, and so
much has been written and spoken as to the possibility of repairing a
break in the submerged main, it is fortunate that this accident occurred
at the time it did, as it has demonstrated beyond question, even to the
most sceptical mind, that this part of the system can be got at and
repaired almost as readily as any other part. Although, on this occasion, the city was deprived of its water supply for a period of eight
days, this was not due to the difficulty of effecting actual repairs, but
to the unusually stormy weather which unfortunately prevailed at the
time of the occurrence, preventing the company's divers from locating
the leak. As soon as this was done, the necessary repairs were accomplished within 24 hours. Under the ordinary summer and winter conditions appertaining to Burrard Inlet, any similar leak can be repaired
in the same or possibly a less period of time.
There can be no doubt that had Mr. Keefer's recommendation (that
automatic blow-offs be placed at each end of the submerged main) been
adopted, the possibility of a break from water ram, even under the 68
Discussion on Vancouver Water Works.
conditions described, would have been greatly lessened; and had. the
general plan of the system been carried out in its entirety before the
works were put in operation, such a break would have caused no inconvenience whatever to the city. It is even doubtful whether the citizens
would have been aware of its occurrence. By referring to plate XVIII,
it will be seen that the general plan provides for duplicate mains at the
Narrows—one of cast-iron (laid) and one of steel (to be laid). Without presuming to criticise the action of the company in not having
completed the crossing of the Narrows, as designed, up to the time of
the break, there can be no doubt that it would have been well had they
done so. At the same time, it must be conceded by all disinterested
parties, that, as Vancouver is a young city, with as yet a small population, tiie~&itiai'p3 deserve every credit for the energy already displayed,
and for the confidence they have expressed in the future progress of the
city, by risking .so much private capital in completing the system all
but this one link, the more so as the chance of a break was infinitesimal.
Since the occurrence of the accident, pipes have been ordered and
tenders invited for the laying of the duplicate main.
The method proposed by Mr. Mohun, for protecting the single submerged main from ships' anchors, is not a new idea. From the inception of the scheme, it was intended, should circumstances render it
advisable, to insure the safety of the pipes by a covering of concrete in
sacks. In the present instance, however, the pipe shell being of great
strength, and to a large extent imbedded in the soft sandstone bottom
of the Narrows, it is a matter of opinion whether such a costly addition
to the works would have been advisable; the more so, as the author is
not aware that this or any other protection has been adopted in the case
of the submerged mains of other systems.
Mr. Summerfield is labouring under a misapprehension when he
quotes the Vancouver Water Works system as an example of the
tendency amongst engineers to exact too much from mild steel. To make
good this assertion, he has endeavoured to show that the factors of
safety of the 22 inch and 16 ineh mains are respectively 6 and 3£,
assuming the ultimate tensile strength of mild steel at 60,000 lbs. per
square inch, with a loss of 30 per cent, for double rivetting, and that, with
these factors of safety, no great amount of corrosion can take place.
The tensile strength of the mild steel employed in the construction
of the mains in question is, as per contract, 72,000 lbs. per square
inch; 60,000 lbs. being little more than is ordinarily allowed for rolled
wrought-iron plates of the best quality. The double rivetting of the
horizontal seams has been so designed that 75 per cent, of the full Discussion on ■ Vancouver Water Works.
-11   of an inch thick, is
The maximum pressure it is called upon to withstand
strength of the plates has been retained.    The ultimate strength of the
22 inch main -H- of an inch thick is, therefore, 540 pounds.    This
loo . ' .
main, being subject to a maximum pressure of 71 lbs. per sq. inch, has
consequently a factor of safety against rupture at the horizontal seams
of 7 j~ instead of 6.
The ultimate strength of the 16 inch main,
743 pounds
at its lowest levels (which it may be here mentioned form but a small
part of its whole length, and are so situated as to be easy of access
both for inspection and repairs) is 180 pounds per sq. inch. Its
factor of safety against rupture at the horizontal seams is, therefore,
4^j, instead of 3J.
These calculations refer to the ultimate strengths of the mains at
their weakest part, namely, the double rivetted horizontal seams. As
far as corrosion is concerned, and this is what Mr. Summerfield more
particularly refers to, the seams, being of two thicknesses of metal,
are the strongest parts of the pipe. The shell will be the first to give
way under this cause; and, in calculating factors of safety with reference to corrosion, the full strength of the plate must be allowed.
These factors may readily be ascertained to be 10 and 5^ for the 22
inch and 16 inch mains, respectively. The author is of opinion that
these results give satisfactory evidence of the permanent safety of the
mains, and in support of this opinion would mention, that in the
Pacific States of the Union long systems of rivetted wrought-iroh
pipes have been in constant and satisfactory use for many years, under
pressures sometimes as great as one-half their ultimate strength.
In connection with this subject, the following table, showing the
strengths and pressures in the various parts of the Virginia City,
Nevada, supply main, 11 £ inches diameter and 37,000 feet in length,
may be of some interest.
Ultimate strength 55,000 lbs. per sq. inch, less 30 p.c. for double rivetting.
Head in feet.
Extreme pressure in lbs.
per inch	
Thickness in inches...
Ultimate strength in lb
per sq. inch	
Factor of safety	
200 330 430 570
871 143
.065 .072
435 482
5 3.4
330 430
to  to
187 247 304
.083 .109 .120
556 724
2.96 2.93
to and
1400 over
608 750
.259 .340
1734 2276
2.85 3 70
Discussion on Vancouver Water Works.
It may be here incidentally mentioned that, in a recent publication
of the London Steel Pipe Company, issued under the auspices of
Messrs. Russell Duncan and H. C. Mylne, Associate Members Inst.
C.E., it is stated : " Wrought-iron pipes have been extensively used in
the States, and it has been proved that no risks are run from fear of
corrosion. Hundreds of miles of wrought iron pipes have been used
for periods extending to over 30 years, and no indications are shown
of injurious oxidation. In some cases the pipes have been used without coating of any kind, in others they have been coated with natural
asphaltum, which is the common method of coating pipes in the
United States. Under ordinary conditions, the coating of natural
asphaltum is not affected by water, earth, or atmosphere."
In May of this year, 1890, 14 months after the opening of the
Vancouver system, an official examination of the steel mains was
made for the purpose of ascertaining to what extent corrosion had
taken place. The earth covering was removed at numerous points
along the line, and the pipes and castings closely examined without an
indication of oxidation being found.
The number of horizontal bends in the 22 inch main is 33, and in
the 16 inch main 103. Of vertical bends the 22 inch main has 20,
none of which exceed 10 degrees deflection; and the 16 inch main 23,
none exceeding 15 degrees deflection. As is the case in all properly
constructed systems, every elevation of any magnitude is tapped by an
air-cock; a fact, which, though not mentioned in the paper on the
Vancouver Water Works, was so obviously necessary, that it might
have been taken for granted. The profile of the 22 inch main from
the dam to the blow-off at the tunnel, a distance of 2^ miles, is very
uniform. Naturally the greater part of any sediment brought down
from the well chambers will be deposited in this main. The 8 inch
blow-off is amply sufficient to flush it when required.
The greater number of the vertical bends in the 16 inch main are
located in the first 8,000 feet beyond the tunnel, where the pipe line
passes over unavoidable sidehill. Whatever small proportion of
sediment may be carried past the deep depression, where the blow off
at the tunnel is situated, into the 16 inch main, will be of minute
particles, and will have little opportunity to settle in the bends, inasmuch as the velocity of flow in that main is 5 -i feet per second, a
velocity amply sufficient to keep fine particles in a constant state of
progression towards the 8 inch blow-off at Burrard Inlet. However,
should an accumulation of silt from some unexpected cause take
place at any of these bends, the deposit may readily and with reason- Discussion on Vancouver Water Works.
able expense be removed by the ordinary means adopted in such cases,.
such as self-acting scrapers, spring gouges, etc.
Mr. Summerfield unhesitatingly states that it is very evident, from
the description given of the turning on of the water, that | water ram.
and air compression were treated as a myth." This is, to say the
least, a hasty and ill-considered assertion, in no wise warranted by
the account given. It is evident that, in this instance, he has pictured
to himself a system of water works, being opened for the first time
with a full velocity of water flowing, and every air-valve and blow-off
closed. This portion of the paper was necessarily brief, and did not
contain details minute enough to give a full and complete description of how the work was accomplished. Moreover, Mr. Summerfield, although he makes no mention of the fact, may have
been led somewhat astray by a clerical error which appears to have
crept into the manuscript or print. In the latter it is stated—" at
9.45 p.m. the water reached the closed 12 inch valves on the north
shore of Burrard Inlet." For "valves" should be read valve. If
this faulty paragraph had anything to do with the misunderstanding
which seems to have occurred, it is to be regretted. But, in any case^.
in order that a clear conception be obtained of the manner in which the
mains were first filled, the following more minute description is given..
When it was decided to make a test trial of the line of mains, the
gate at the well chambers of the dam was lifted high enough to expose
a segment of the 22 inch main 3 inches deep; the 8 inch blow-off at
the tunnel and all the air-valves being wide open. By means of
these, the air was expelled, the incoming water occupying its place,,
and very gradually filling up the main until it was discharging freely
through the blow-off. The 22 inch main was allowed to continue in
this state for a period of 48 hours. No breakage or sign of leakage
having taken place, the gate at the well chambers was lifted from time
to time, until, on the 5th day from the first partial opening, it was
opened wide, the water in the main discharging freely through the 8
inch blow-off.
This trial being so satisfactory, it was then decided to test the 16
inch main. The 8 inch blow-off, before mentioned, was partially
closed, and the water allowed to rise up the steep incline to the tunnel,
until that part of the 22 inch main passing through the tunnel, and
connecting with the 16 inch main, was filled one quarter full. How
very slowly this was accomplished will be understood from the fact
that, although the 22 inch main was nearly full up to the 8 inch blow-
off, it required two hours from the time of partially closing that. 72
Discussion on Vancouver Water Wor^s.
blow-off to reach the floor of the tunnel. The 22 inch main was not
allowed to exceed about one-quarter full, but was maintained at this
level, slowly and uniformly filling up the 16 inch main, until it was
discharging freely by means of one of its 12 inch valves and the 8
inch blow-off at Burrard Inlet, the other 12 inch valve being kept
closed in order to prevent the water entering the 12 inch main under
the narrows. As in the case of the 22 inch main, all air-valves were
kept wide open, allowing free vent for the escape of air, and affording
absolutely no opportunity for air compression or hydraulic ram.
It is much to be regretted that Mr. Summerfield cannot be gratified
in his desire for information as to the behaviour of the 16 inch main,
under such (very contradictory) conditions as " a full head of water
flowing with a velocity of 1.4 feet per second." As Mr. Summerfield
calculated the velocity with which the water in this main travelled
during the test trial, using the time and distance given in the paper,
it is somewhat surprising that he did not also calculate the velocity
due to the head and length, and thus avoid the error into which he
appears to have fallen. From the data at his disposal, he might
readily have deduced, that, under a full head of water, the velocity of
discharge at the termination of the main in the city is 5 -£- feet per
second. But, as the portion referred to in the present instance is
only that between the tunnel and Burrard Inlet, if a full head of
water had been allowed to enter the 16 inch main at the tunnel, the
velocity of out-flow would have been found to greatly exceed 5 ■$- feet
per second, using as data the total fall 388 feet, and total length
19,320 feet, between these two points.
Had Mr. Summerfield ascertained these facts, he could not have
fallen into the rather extraordinary belief that water ram and air
compression had been treated as a myth; and it may be here mentioned, that, had both 12 inch valves been closed, and only the air-
valves along the line and the 8 inch blow-off at Burrard Inlet been
open, these latter would have been amply sufficient for the discharge
of air and the prevention of hydraulic ram.
The day following the turning on of the water, a thorough examination of the whole line of mains from Burrard Inlet to the dam was
made. Throughout the whole distance, no sign of leakage was
apparent. Mr. Summerfield's inferences that "something occurred''
are without foundation.
The reducing tank referred to is entirely temporary, having been
built solely to relieve the pressure on certain defective pipe lengths in
the lower levels of the system, until new lengths could be manufactured and substituted. Discussion on Vancouver Water Works.
With reference to the remarks of Mr. Russell Duncan, in a comparison between steel pipes jointed by the Moore and Smith joint, and
those jointed by means of faucets and spigots attached to the shell,
the word " cumbersome " was used in reference to the latter. This
was a general expression, and not intended to particularize any
individual joint, such as the Riley patent referred to by Mr. Duncan,
and it may have been used without due consideration.
It is merely a matter of opinion, which is the simpler joint of the
two; but on comparing them, it certainly seems that a joint consisting
of a plain nipple and ring is simpler and more easily handled than a
joint made in imitation of ordinary cast-iron joints.
Without details of the particular joint referred to by Mr. Duncan,
it cannot be determined whether it requires a less quantity.of lead, as
he avers, than the Moore and Smith joint; but it is open to question,
whether any safe spigot and faucet joint can be designed for 16 inch or
22 inch pipes, which will require less lead than 27^, lbs. and 37 _i-
lbs., respectively, the amounts used in the Vancouver water works for
these diameters. In the pamphlet issued by the London Steel Pipe
Company, it is claimed that the steel spigot and faucet joint of a 24
inch pipe requires 9^- per cent, less lead than the joint of a cast-iron
pipe of the same diameter.    (See author's paper, page 337.)
According to American practice, a cast-iron joint for a 24 inch pipe
would require not less than 50 lbs. of lead; consequently, Mr. Duncan's
joint for the same diameter would require 50 lbs. less 9^ per cent., or
45J lbs. But it can be easily calculated that a Moore and Smith joint
for a 24 inch steel main -1-1- of an inch thick with i- inch thickness of
loo . 16
lead, the usual allowance, would require only 39^- lbs. It appears,
therefore, that in the case of a 24 inch pipe, according to the London
Steel Pipe Company's pamphlet, as applied to American practice, the
Moore and Smith joint requires 5§ lbs. less lead than the spigot and
faucet joint recommended by them.
As for great flexibility at the joints, it is not advisable in any system of water works to make with large mains much angular deviation,
without special castings. It is not necessary, however, with the Moore
and Smith joint to lay the pipes in a perfectly straight line, as claimed
by Mr. Duncan, contrary to the description given. It will, on occasion, admit of deflection to the extent of one degree per length. In
the case of 24 feet lengths, this would form an 8 degree curve of 716
feet radius. Without actual knowledge of such curves having been put
in practice, it would be unwise to express an opinion as to their suitability ; but there can be no question, that by means of this joint, such 74
Discussion on Vancouver Water Works.
a curve as that on the Tay Viaduct, alluded to by Mr. Duncan, round
which his 9 inch spigot and faucet pipe was laid, can be easily overcome, it being only a 4° 10' curve with a radius of 1386 feet.
The hippie of the Moore and Smith joint in the Vancouver Water
Works pipes does not decrease the internal diameter of the pipes at
the joints; but in thickness of pipe shell over 11- of an inch actually
increases it. Had careful consideration been given to the drawing
(Plate XX), there could have been no misconception on this point. It
is there plainly shown, that the end courses of each pipe length are
large or outer courses. The nipple being inserted in the end of a large
course must, therefore, when of the same thickness as the shell, have the
same inside diameter as the inner or small course, which, it is unnecessary to state, is the governing diameter of the pipe. The nipple adds
greatly to the stiffness of the shell at the joint, but having no water
pressure to withstand is, for economy sake, usually made of thin material, so that, in the case of a thick pipe shell, the inside diameter at
the joint is greater than the governing diameter.
The reply given to Mr. Summerfield fully answers Mr. Duncan's
doubts as to the strength of the 16 and 22 inch mains in comparison
with ordinary Californian practice. When he reviews this subject, it
would be well to bear in mind that there is no equality of strength
between wrought-iron and steel.
The omission of mention of the test pressures imposed on the
mains was an oversight. The engineer's specification, and the company's agreement with the contractors for supplying and laying the steel
mains, provided that the pipes should withstand safely and without
leakage a column of water 600 feet high or 260 lbs. pressure to the
square inch; and in order to insure the application of proof tests by the
contractors, a special clause -bound them to maintain and keep the
whole system perfectly tight for two months after completion. So far as
the engineers are concerned, these are the terms of the contract to this
day. It will be admitted that no more stringent agreement for the
laying of properly tested pipes could have been entered into.
Full details of the rivetting of the 12 inch, 16 inch, and 22 inch mains
- in the form of a table headed, | Details of Rivetted Steel Mains,"
accompanied the paper under discussion, when forwarded to the Society.
That this table did not appear in the advance proof is to be regretted.
However, it has now been embodied in the appendix to the paper, Vol.
III., page 361.
Mr. Duncan's inferences and quotations from the author's letter to
him are liable to mislead, Discussion on Vancouver Water Works.
Immediately after the successful turning on of the water, the concluding portion of the paper, in which the account appears, was written,
and the whole forwarded to the Society. There is nothing in this
account at variance with the facts contained in the letter of April 13th.
The primary leaks which were discovered up to the time of forwarding
the paper to the Society were being readily repaired in the manner
described. But, on the second day after the main had been put under
full pressure, leaks developed themselves in the low-lying parts of the
system, of an altogether different nature, and too numerous to be economically repaired in the same manner. The extracts quoted by Mr.
Duncan were not intended by the author to reflect in any way on the
character of the mains,- as would be plainly seen were the latter given
in full. The information afforded him was to serve as data on which
to answer the questions, whether in his experience he had ever known
a jet of water forced out between the laps of a rivetted pipe to actually cut the shell of the pipe beyond the laps, and, if so, what remedies
were adopted by him ? This was what had occurred in the Vancouver
system, after the paper had left the author's possession.
Mr. Duncan in his reply could state no similar case; but very
kindly advanced many theories, among which were the following :—
I That an hydraulic ram had occurred; that the rivetting of the seams
was faulty; that the plates were not laid flat together at the seams;
that there may have been burrs left at the rivet holes; that in caulking,
the edge of the plate may have been upset so as to spring one plate
more than another ; and that the rivets may have been pitched too far
apart." Any one of these causes would have been sufficient reason for
leakage between the laps, but would not account for the cutting of the
steel plates beyond the laps, which was the point upon which information was desired. The most probable explanation of this phenomenon,
of a water jet apparently cutting steel, is that, in rivetting the plates
together, too much power was used, the effect being to compress the
plates at the rivets, and to buckle them in the intervals between, thus
allowing the water when under full pressure to be forced out with great
velocity. As the soil in which the injured pipe lengths are laid is
nearly pure sand, the sand particles mixed with the escaping jets, and
being kept in constant motion, acted as sand blasts at the points where
the jets struck the shell just beyond the laps. This theory is supported
by the fact, that the pipe shells were not cut where the excavations were
in earth or clay, and that part of the 12 inch steel main which was uncovered, though leaking at several laps, showed no sign of abrasion. It
has since been learned that similar occurrences have taken place in 76
Discussion on Vancouver Water Works.
eastern cities, where the practice is, either to box the mains, or remove
the sand and substitute clay.
Answers to the queries proposed by Mr. Duncan as to rivetting,
coating, etc., etc., are to be found in the paper.
With reference to the statement that the pipe lengths being 23/ 9|-",
four courses would have been preferable to seven in the manufacture of
each length, it may be briefly replied, that four courses would not have
answered in the kind of pipe used. In order to admit of a nipple which
would not lessen the governing diameter of the main, it was necessary
that a large or outer course should be at each end of each length. This
could only be (see Plate XX.) when each length consisted of;8f:5, or 7
courses. The " preferable pipe " of 4 courses would have had a small
course at one end, which would not have received a nipple, except of
lese diameter than itself. If a nipple were dispensed with, and the
smaller course forced into the larger, this latter, overlapping the other,
would make the thickness of lead, at each side of the junction, unequal.
As to its being bad practice to make up each pipe length of 7 courses
of so-called small plates, this is a matter of opinion, and not warranted
by custom on the Pacific Coast. The pipes of the Vancouver system
have been endorsed by the best authority in California, and as a notable example of similar design, the 36 inch wrought iron main of the
Spring Valley Water Works, Cal., may be cited. The specification of
this pipe calls for plates 116" x 44".
It is a questionable statement, and unsupported by proof, to say that
caulking the seams as described tends to aggravate leaks. Experience
on this coast has shown that, to obtain perfectly tight seams, chipping,
caulking and split caulking are necessary. A reference to the specification of the Spring Valley Water Works, California, and the Portland
Water Works, Oregon, will give further information on this point.
The spigot and faucet joint, having only one surface to be caulked,
requires less labour in laying than the Moore and Smith joint, which,
owing to its simple construction, has two surfaces. This advantage,
however, is more than counterbalanced by the original labour of making the spigots and faucets, and attaching them to the shell. Had
the cost of these details been given, a satisfactory comparison could
have been made, which might not have been unfavorable to the Moore
and Smith joint.
The statement made, that in the Riley patent socket joint it is not
found that the shell is bent inward by caulking, as is the case with the
Moore and Smith joint described in the paper, is misleading. The
paper merely says: " That if the lead is beaten between the ring and Discussion on Vancouver Wdter Works.
'the pipe too tightly, the shell of the latter will bend inward; and that
it is necessary to caution inexperienced caulkers with regard to this."
But in the hands of caulkers who have learned their business, the
Moore and Smith joint is as free from liability to defective workmanship as any other joint; an assertion which is amply proved by the
fact that in the Vancouver Water Works system, after the pipes had
been subjected to a full pressure for several weeks, no joint in the
whole length of 10 miles was found to leak.
Mr. Duncan claims that owing to the spigot end of the socket joint
pipe being made 50 per cent, thicker than the shell, no pressure
imposed in caulking will indent it. This is somewhat strange, seeing
that he admits the possibility of the shell of the Moore and Smith
joint pipe being bent, where the extra thickness exposed to the pressure
of caulking is often, as in the case of the Vancouver mains, 100 per
cent. In the latter joint, a nipple or ring 5 inches wide, and of the
same thickness as the pipe shell, is closely fitted to the interior surface
of the pipe, directly under the 4 inch ring of lead to be caulked, thus
doubling the thickness of metal exposed to the pressure of caulking.
Mr. Duncan's concluding observations afford very valuable information as to the manufacture of pipes by the London Steel Pipe Company. When in London some months ago the author had the pleasure
of seeing various specimens of these pipes, and was much struck by
their excellence. For finish and workmanship, they are not likely to be
surpassed; and it is very desirable that full information regarding
them should be well known.
In reply to Mr. Henshaw, as to the question of an anchor falling on
the submerged main, it is conceded that it would be undesirable to
invite a trial; but in the event of such a test being made, it is by no
means certain that the disastrous effects hinted at would take place.
In any case, such an occurrence is a very remote possibility, even with
one submerged main; while, that an accident should happen to two at
the same time is almost an impossibility.
With reference to the remarks of Messrs. Biackweii, Smith, Goad,
Peterson, Gower and Bovey, there is no mistake about the dimensions
of the Cedar in question. Should Messrs. Biackweii, Smith and Goad
care to visit British Columbia, they can have the pleasure of verifying
the measurements. The Cedar is dead, but is very uniform from the
ground upwards, and has no abnormal enlargement at the base. Such
trees are rarely solid, and this is no exception. In trenching for pipe
lines, it is not usual to extract such stumps, it being much more economical to tunnel under them. 78
Discussion on Vancouver Water Works'.
A working plan of the dam, drawn to the scale of 8 feet to the
inch, accompanied the paper on the Vancouver Water Works. It was
considered that this would, if hung on the wall, have afforded all
necessary information.
As the whole front of the dam is protected by double sheet piling,
overlapping, with the ends imbedded in a trench of concrete sunk in
the bottom of the river, and the shore connections protected on the
south side by a hand-laid stone wall with a well rammed gravel and
earth backing, and on the north side by a gravel embankment of
picked material, the chance of water percolation through the body of
the dam, and seepage round the ends, has been reduced to a minimum.
Certainly, none has taken place during the two years of the dam's
existence. The trench in which the sheet piling is imbedded, owing
to the shallowness of the river, was filled with concrete in a simple
manner. The concrete was wheeled over gangways extending along
the face of the dam, within a few inches of the water, was then
deposited over the sides, and allowed to settle in the trench. Twenty-
four hours after the trench had been filled, the concrete was firm
enough to resist moderate probing.
The definite object in view in the manufacture of the concrete was
imperviousness and a thoroughly secure connection between the foundations of the dam and the bed of the river. The space in which it
was to be placed was very confined, being a narrow trench, not
exceeding at any point a depth of 3 feet; and being for the most part
entirely under water, it would not permit of the filling being packed
and rammed. These objects could only be obtained by making the
concrete of such small materials that the whole would form as nearly
as possible a homogeneous mass.
The author's experience ~in the manufacture of concrete is not
extensive enough to permit of his expressing an opinion as to the
advisability of employing large stones. But it is worthy of note, that
the specifications for the water works systems of Gosport, Lambeth,
Bideford, Halifax, and Aberdeen distinctly require that either sharp
gravel be employed, or that all stones shall be broken so as to pass
through a ring 1^ or 2 inches diameter. Also, Trautwine (page
680, edition 1886) states " that broken stone for concrete is generally
specified not to exceed about 2 inches on any side; but if well freed
from dust, all sizes from \ inch to 4 inches on any side may be used."
It is therefore apparent that if eminent authorities advocate the use
of large stones, many quite as eminent do not.
The Teredo Navalis, not being gifted by nature with the acrobatic Discussion on Vancouver Water Works.
powers of the salmon, is not likely to ever find a home in the timbers
of the dam.
Corrosion of metal exposed to the action of salt water is a subject
\ of great importance; and it is a matter for surprise that really reliable
j information with regard to it is difficult to obtain.    But experience,
i so far, goes to prove that steel and wrought or cast iron of proper
quality resist this action for long periods.    Trautwine says "that
certain cast-iron sea piles of hard white metal showed no deterioration
• after 40 years immersion." (page 218, edition 1886.)    The double 8
inch cast-iron pipes laid across Shirley gut nearly 20 years ago are in
active operation to this day.    In the discussion on the Avon Bridge,
published in the last volume of  the Society's Transactions,  Mr.
Uniacke in reply to Mr. Biackweii (page 286) says : " That bolts and
plates used in the construction of the old bridge, 50 years ago, and
exposed to salt water at different stages of the tide, showed not the
slightest corrosion."    Also, in constructing the new steel dock gates of
Limerick Floating Dock, the pintles of the old gates were found to be
so little affected by their 38 years' exposure to salt water, that they
were used again.
It is to be noted that the cast-iron in the Vancouver submerged
main is of the quality approved of in Trautwine; and, also, that there
are many instances of similar mains submerged in salt water, designed
by eminent engineers; as, for example, the aforementioned Shirley gut
pipes, the Bournemouth storm outfall, the San Francisco Bay mains
and the San Diego Bay main. Full information as to the manufacture, specification, quality, and proof tests of the submerged main
are given in the paper on pages 344, 348, 351.
It is difficult to believe that Mr. Peterson's remarks relating to his
observations when in Vancouver are uttered in any other than a
jocular spirit. No one can suppose that because an irresponsible
individual volunteered the information that the steel mains (wrought
iron, he calls them) were not sufficiently strong to withstand the
pressure, he credited that statement without investigation. It is hard
to understand his imagining the rivets to be 4 or 5 inches apart,
seeing that this is nearly 4 times the extreme distance. The round
seam rivets of the 12 inch and 16 inch mains, referred to by him, are
1.2 inches and 1.08 inches apart, from centre to centre, respectively;
and the straight seam rivets, which are in double rows, 1.3 inches and
1.4 inches. It is impossible he can believe that any steel or wrought
iron pipe, which had so yielded to pressure that water was flowing in
little streams put of nearly all the joints of the sides, could be held 80
Discussion on Vancouver Water Worlcs.
together and rendered serviceable by straps placed round it. All new
rivetted pipes weep more or less at the laps, when first put under
pressure. The great majority of the leaks he observed, being on the
south side of the Inlet, where there was little or no sand to cut the
plates, " took up " of their own accord, without special repairs.
The submains are 8 inches, 6 inches, and 4 inches in diameter; not
4 inches and %\ inches. 2^- inch welded wrought iron tubes have been
laid for short distances, as a temporary means of supply for certain
buildings.    No hydrants are connected with 2^ inch pipes.
The distribution system was designed by the most eminent authority
on water works in Canada, Mr. T. C. Keefer, C.E., C.M.G.
In the case of Vancouver, there is a very substantial reason why
the water works should be in the hands of a private company. They
were projected before the city had an existence, and constructed at a
time when the city's resources were being fully employed in making
absolutely necessary improvements, and it is doubtful whether even at
this day the city would care to incur the expense of water works,
construction.   VANCOUVER WATER WORKS,


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items