American Society of Civil Engineers.

Transactions of the American Society of Civil Engineers (Volume 81) online

. (page 26 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 26 of 167)
Font size
QR-code for this ebook

ture has since been used exclusively during high water, eliminating
the danger of failure to a great extent. The weakest point in the
primary line of defense, in the writer's opinion, is still the imused
portion of the Hanlon gate. The weakness would be many times
greater if the structure was in the alluvial deposits along the river
bank in Mexico, where even an inferior grade of rock foundation is
not to be had.

Temporary Diversion Weir.

Mention has been made of the temporary rock weir built across
the Colorado River in the summer of 1910, and the statement has
been made that such a weir would not serve as a permanent structure.
H. T. Cory, M. Am. Soc. C. E., states that even the comparatively
small quantity of rock dumped from the trestle to form this weir was
not undermined, and did not settle, except to a slight extent and in
a few places, with the summer floods of 1911 and 1912.

In building the weir, 1 822 cu. yd. of rock were used, at a cost of
60 cents per cu. yd. The total cost of the weir, including this rock,
plus the cost of providing a trestle and maintaining it during its life,
amounted to $32 350. The greatest head developed by the weir was
3.4 ft. on October 19th, 1910, when the Colorado Eiver discharge was
5 000 sec-ft. The only way that this maximum head was developed
and maintained was by constant expenditure of money and labor in
obtaining rock and filling in the low places along the weir as they
developed. In addition to this maintenance cost on the weir itself,
there was the cost of maintaining the rock abutments on both ends
of the weir up and down stream. The Arizona abutment was main-
tained at great expense some time after the weir was removed, in
order to prevent cutting into the Arizona shore and injury to the
levee constructed by the United States Reclamation Service. All this
goes to show that such a weir is not practical, from a maintenance
point of view alone.


In March, 1911, the trestle was removed and all maintenance work
on the weir itself was stopped. For some time previous to March,
1911, no maintenance work was done because of instructions from the
War Department that no more material should be dumped into the
Colorado River. The average elevation of the crest of the weir was
about lOS in October, 1910, and 105 in February, 1911, showing that
as soon as the maintenance work was stopped, the rock began to slip
down stream. In January, 1913, further investigation of the weir
was made, and the elevation of the crest of the rock was found to be
below 94, there being only a few places where the rock could be
reached by sounding.

Very little good was effected by the weir in the way of securing
sufficient discharge through the concrete gate at the low stage of water,
after the summer of 1911. It is true that the piling butts, which re-
mained after the blowing of the trestle, caught and held a great quan-
tity of drift, and developed several sand bars in the location of the
old weir, but this obstruction caused only a slight increase in the
head at the intake.

Sole dependence for a sufficient water supply for Imperial Valley
was placed on the hydraulic dredge Imperial, operating below the head-
gates, and on the dredge El Centra, operating above the head-gate in
the intake canal and in the Colorado River channel, controlling sand
bar obstructions, etc.

The actual computed yardage of solid material moved by the
dredge Imperial from the canal below the head-gate in one season's
operations is equal to just one- tenth of 1% of the solid matter dis-
charged through Hanlon head-gate in the same interval, but it is
also a fact that it was just this one-tenth of 1% of solid matter which
had heretofore blocked the canal delivery below the head-gate before
the dredge was put in operation.

The major part of the one-tenth of 1% of solid matter removed
by the Imperial was shingle, ranging from 1-in. rock to ^-in. gravel,
and sand ranging from gravel down to materials which are retained
on a No. 10 mesh screen. It was the discovery of this fact that led
to the successful operation of the Hanlon Heading, from the time
of the entire removal of the rock weir in 1911 up to the present.
This grade of material has been removed systematically every year
since the dredge has been in operation, thus permitting the bed of


the canal below the head-gate to be maintained at an elevation which
gave a sufficiently low water surface in the canal at all times. All
other sands and silts not removed by the Imperial are carried down
the canal in suspension.

The fact that there was deposited below the head-gate, at times
when the velocities in the intake exceeded 6 ft. per sec, material
of a kind which could not be carried down the canal by the velocities
below the head-gate, which are never greater than 4 ft. per sec,
led to the discovery of a means by which a hydraulic-fill dam could
be placed across the Colorado River at a point directly below the
intake, to serve as a means of diverting sufficient water into the
Imperial Canal, should the necessity ever again arise. This time
did arrive during the summer of 1915, because of the combination of
an exceptionally low Colorado River during August and September,
and an exceptionally great demand for water in Imperial Valley,
due to the large increase in irrigated acreage over the previous years.

A study of the cross-section of the channel of the Colorado at
Hanlon Heading, and at the gauging station of the United States
Reclamation Service at Yuma (Fig. 2), shows that the Colorado —
like other rivers forming deltas — scours and refills its channel bed
with the increase and decrease of quantities of water, and that a
direct relation exists between the character of the materials forming
the bed and the velocities of the water. For example: The river bed
is scoured to an elevation of 85 at Yuma, Ariz., by a discharge of
140 800 sec-ft., whereas it is refilled to an elevation of about 110 by
a discharge of 3 400 sec-ft. The only reason that the bed does not
scour deeper than an elevation of 85, with a discharge of 140 800
sec-ft., is that at that depth the materials constituting the bottom
are so heavy that the velocities, registered as 6.8 ft. per sec. for this
discharge, will not move them. It follows that, in refilling the bed,
the materials are graded as to weight, and deposited in direct relation
to the velocities. It also follows that the refill in all cases is in strata
of materials decreasing in weight as they approach the elevation corre-
sponding with the lesser velocities.

Though this is by no means a new discovery, the experience with
the dredge Imperial below the head-gate served to bring out more
strikingly the real relations between the weights of materials and



the velocities, and showed a way to make a practical use of this

The scour line of the bed of the Colorado at Hanlon Heading is
at an elevation of about S4. The ladder of the 10-in. suction dredge
El Centro, operating in the river and above the concrete heading,
was designed so as to make it possible to pump from an elevation of
80 when the surface of the river was 110. The materials in this area
below the scour line had withstood the scouring effects of velocities
as high as 7 and 8 ft. per sec. Those materials, when brought to the
surface and deposited in the water where the velocity had decreased
to 2 and 3 ft. per sec, due to falling discharge, were certain to stand
in the position placed by the discharge of the dredge, making it



0— :^«— 60 ' > e ^ C O- ^'t C O ' >< 60' >t< 60-

130 r






S 90




"•June 20





August 1_



.Jaa.13 ,



\ t




- —









- -"































\ ^-



, V


1 — ■"^







May 30


13 040

G8 80O





June aO


20 740

140 SOO



Aug. 1



39 100





110 bp




Fig. 2.

possible to construct a hydraulic-fill dam under the water of a running
stream varying in depth from 2 to 15 ft.

Hydraulic-Fill Diversion Dam.

On August 12th, the El Centro commenced the construction of such
a dam below the intake, and in 14 days had raised the fill to an eleva-
tion of 12 in. above the water surface three-fourths the distance across


the river. This fill reached from the Arizona bank of the river to
within 250 ft. of the California shore, or a total length of fill of 650 ft.
Due to the constriction of the channel during this period, a head of
somewhat more than 1 ft. was developed, which resulted in a corre-
sponding increase in velocity through the remaining opening. The
velocities resulting from this head became as great as those developed
during flood discharge of the river, and consequently material pumped
from below the scour line could no longer be deposited in the dam,
with any degree of economy, without adopting some method of de-
creasing the velocities. In closing the last 250 ft., therefore, a great
quantity of brush and sacks filled with gravel was used.

The California shore had in years past been thoroughly revetted
with a blanket of rock 25 ft. thick, making it safe to drive the channel
directly against it, where the final .closure was made. It is true that
before the final closure took place the terrific velocities along the
California shore scoured even the rock revetments to a certain extent,
due to the large percentage of fine material composing the revetment,
but a careful study of records, and the materials themselves, indicated
the presence of a sufficient quantity of large rock in the revetment to
make it a dependable protection against the cutting of the river
on this side. The dam was commenced on the line of the old
rock weir on the Arizona side, and the rock abutment served as pro-
tection there.

The closure of the final 75 ft. of the channel represented the most
difficult operation in the construction of the dam, and the means de-
vised for making this closure again demonstrated the advantage of
close observation of the action of a river when left to itself, and the
application of the results of this study in a practical way.

From August 12th to September 12th was the period of principal
construction. The time subsequent to September 12th was spent prin-
cipally in making the final closure and in maintaining the structure.
On August 12th the river was discharging 8 800 sec-ft. Of this dis-
charge, 2 529 sec-ft. were flowing into the Imperial Canal, all the gates
in the heading being fully opened. The requirement for irrigation in
Imperial Valley was 5 000 sec-ft. By September 12th, the river dis-
charge had dropped to 4 000 sec-ft. During this period of 30 days
there was used in the construction of the dam:


40 000 cu. yd. of sand and gravel fill, at 2.7 cents

per yd $1 080.00

21 300 burlap and canvas sacks, at 3 to 6J cents each 1 140.00
450 cords of willow brush and poles, at 73 cents per

cord 328.50

800 lb. of wire 32.00

Cables and clamps 100.00

Labor, filling and placing sacks, carrying and plac-
ing brush, poles, etc 2 000.00

$4 680.50
Plus 10% for superintendence and tools, etc 468.05

Total $5 148.55

Since September 12th there has been an additional expense of
$2 000 for raising and strengthening the dam by dredging, and for
labor in closing and opening the dam, and general maintenance of the
structure, making the total final cost $7 148.55.

The following methods were used in constructing the major part
of the dam: Pumping was commenced on the Arizona side, the dredge
pontoon being some distance above the site of the dam, and the dis-
charge being carried astern through a 10-in. pipe floated on pontoons
to the fill. Owing to the admixture of lighter grades of material en-
countered in burying the suction pipe to the scour line depth, and to
the subsequent caving in on the suction of the surface materials when
the depth was once reached, the base of the dam, in a great many
places, became as wide as 150 ft. The heaviest materials pumped —
in some instances stones and cemented adobe chunks 6 in. in diameter —
remained exactly where placed, but the lighter sands and gravel were
carried by the current some distance down stream before they lodged
in permanent position. When a point near the water surface was
reached with the fill, the conservation of the material was made pos-
sible by constructing brush fences to help retain the material at the
slope required. (Fig. 3.) By this process the structure was carried
across the stream to the point of final closure. Three re-runs across
were made later with the dredge, bringing the crown to an elevation
of 114, which is about 5 ft. above the water surface of the river when
5 000 sec-ft. are passing into the canal. (Figs. 4 and 5.)

On September 10th, 2 300 sec-ft. were passing through the opening
in the dam, and 3 300 sec-ft. had been diverted into the Imperial



Canal. The head on the dam was 2 ft., and the resulting velocity
through the remaining opening was S ft. per sec. The bottom of the
opening had been paved with about 10 000 sacks, filled with gravel
pumped by the dredge, and stoutly tied, decreasing the depth to about
5 ft., and thereby preventing scour. Great quantities of sacks were
used in protecting the Arizona end of the opening, the California
shore, as before stated, being protected by its rock facing. Additional
sacks were used in forming the overpour riffles as fast as the head de-
veloped. A great number of the sacks were of no avail, either because
of their lodgment in the wrong place, or because of breaking and wash-
ing out; and, throughout the whole process, constant replacements
were required because of the boring effects at the toe of the sack riffles,
causing the embedding of the sacks.


Pig. 3.

In paving the bottom, the point above the opening at which the
sacks should be dropped in order that they might lodge on the bottom
in the proper place, was determined by constantly experimenting with
sacks thrown overboard attached to wires, thus making it possible to
locate the sack when it had finally settled by following the line of the
wire. Only those filled with the coarsest shingle and gravel and of
greatest weight would remain in position and withstand the velocity
for any length of time.

In the final closure, when the bottom had been protected carefully,
cables were stretched across, securely anchored to deadmen on the
California side and in the body of the dam on the Arizona side. The
first cables used were of f-in. plow steel, but proved to be too light
when maximum strains were put on them. After one failure in making
the closure, due to breaking the cable, a double strand of 1^-in. cable
was used for the lower run.

■ M:


Willow poles were cut from the Arizona and California shores
near by, where the supply is abundant. The poles were 30 ft. long, the
diameter of the large end being 10 in., and the small end tapering to
5 or 6 in. By the skillful use of barges and lines working above the
openings, these poles were maneuvered with the current in such a
manner as to cause their large ends to lodge against the cable stretched
across the opening above the water surface, while the small ends rested
on the bottom at an angle of 30°, much as a drifting snag will lodge
in a swiftly running stream when the heavy end is dragging the
bottom and the upper end encoimters a surface obstruction of trees
or other large drift. (Figs. 6 and 7.) These poles were placed against
the cable, in this manner, with enough space (sometimes 4 or 6 in.
between them) to give stifficient strength to the structure and yet not
decrease the area in the opening so greatly as to produce much addi-
tional head. When all the poles were in place, large bundles of willow
brush were made up, weighted heavily with sacks, and rolled into the
water from the barges, anchored some distance above the opening, in
such manner as to cause them to strike the bottom of the channel
and roll against the lower ends of the poles. As additional bundles
were thrown in, the pressure against the poles increased, forcing them
to bend, until in some instances, before final relief to the strain was
obtained, the upper parts of the poles were standing vertically, and
the lower sections were bowed to the extreme. Green poles were always
chosen because of their flexibility. (Fig. 8.)

The placing of the bundles of brush was carried on in sections
only, and only in sufficient quantity to decrease the velocity through
the section to a point where material deposited by the dredge could
be retained in place. In this manner the strain against the poles in
any section of the opening was of short duration, as the dredge would
give relief in a few hours by extending the dam above. Of course,
a great many sections failed during the work, because of poles break-
ing, but means were at hand for speedy replacement when such acci-
dents occurred. Because the impounded area above the dam was
so great as to require several hours to increase its elevation when flow
through the opening was reduced, the extra heads created by stop-
page with the brush were not felt on the poles to any great extent
before the dredge had given relief.


After numerous local failures, due to small freshet rises in tlie
river, and the breaking of cables and poles, the opening was entirely
closed on September 20th, the head then being 6.4 ft. (Figs. 9 and 10),
and the discharge of the river, measured at Hanlon Heading, 2 958 sec-ft.
Between September 20th and 27th the river discharge fluctuated be-
tween 3 000 and 4 000 sec-ft., all of it being required for irrigation
in Imperial Valley. On September 27th, because of advance notice
from Needles of a rise in the river, the dam was cut at a place pro-
vided, and all water, with the exception of 4 000 sec-ft. was passed
through into the channel of the Colorado Eiver below. The waters
thus let through the dam were forced into the California bank below
by local obstructions, so that on October 2d the opening in the dam
was again closed, and all the water was diverted into the Imperial
Canal. The discharge of the Colorado Eiver on October 2d, as meas-
ured at Hanlon Heading, was 4 406 sec-ft. The second closing head
was 8.2 ft. (The closing head of the Hind rock-filled dam, on November
4th, 1906, was 15.8 ft., with, a river discharge of 9 275 sec-ft.*)

The 36 hours following the second closure were spent in clearing
obstructions in the channel immediately below the dam, and in pro-
viding a new waterway to train the flow away from the California
shore. On the evening of October 3d, the dam was again cut at the
place provided, because of a rise in the river in excess of the water
requirements in Imperial Valley. During the period from October 3d
to the present writing, the opening in the dam, 100 ft. in width, has
been left, and through it the discharge has been as great as 8 000 sec-ft.
In constructing the dam, provision was made for accommodating the
freshet discharges of the river, which commonly occur during October
and November. These sections will be sacrificed only as required,
because of the fact that during December and January the discharge
of the river commonly reaches the minimum, and this being a year
of exceptionally low water, it is quite likely that the irrigation re-
quirements of Imperial Valley will again make it necessary to close
the dam fully. Entire removal probably will be advisable in March
or April, 1916, and by that time it will have fulfilled its purpose as
a means of diverting sufficient water to meet the irrigation demands
of the Valley. In any event, the annual flow period of May and Jime,

• "Irrigation and River Control in the Colorado River Delta," by H. T. Cory,
M. Am. Soc. C. E., Transactions, Am. Soc. C. E., Vol. LXXVI, p. 1204.

^nf^^k^^^ 'j^.




Fig. 7. — Hydraulic-Fill Dam, Hanlon Heading, Neaking Completion,
September 4th, 1915.

Fig. 8. — Remaining Opening in Hydraulic-Fill Dam, Hanlon Heading,
September 17th, 1915.

Fig. 9. — Final Closure, from Below Hydraulic-Fill Dam, Hanlon Heading,
September 20th, 1915.

Fig. 10. — Dredge, El Centra, Pumping in Final Closure, Hydraulic-Fill Dam,
September 20th, 1915.


1916, will remove the structure entirely. If the conditions which de-
velop before this time should require it, the whole structure could be
speedily removed by mining, as it is composed entirely of brush, sacks,
and gravel, with the exception of a few willow poles driven with the
aid of a jet from the dredge. The structure contains absolutely noth-
ing of a permanent nature.

The experience in building this dam has developed naturally a
great many improvements which would be adopted in a second con-
struction of similar nature; so that such a dam could be built in a
much shorter period and with much less expense. For instance, by
constructing an aerial tramway, anchored to towers on each bank of
the river, dredge lines can be carried across the stream during its
flood, thus enabling the dredge to venture into the river during much
higher water, resulting in earlier completion of the work. The ladder
of the dredge El Centro is to be lengthened 15 ft., enabling it to pump
material at a greater depth below the scour line. These two improve-
ments alone will permit the commencement of the dam while the
river is discharging 56 000 sec-ft., and will permit the complete
diversion of a river carrying 15 000 sec-ft. if necessary. More econom-
ical use of sacks and the labor attending can be achieved by a more
general use of cable and pole manipulation in closing the final opening.
Fewer accidents in the final closure will be assured by anticipating
the material requirements in time to accumulate sufficient quantities
of poles and brush. By lengthening the pontoon and pipe line of
the dredge, thus providing an increased radius of operation, a large
supply of material suitable for construction can be obtained.

Closure of Bee River.

The application of a hydraulic-fill diversion dam as the fundamental
factor in the closure of the Bee River break in the first line of defense
is easily recognized. It is not possible to use the hydraulic fill to close
the break itself and to force the river waters back into the old channel
of the Colorado, because, in the vicinity of the Bee River break, the
Colorado has never maintained a permanent channel bed for any
length of time. Each flood, instead of scouring out a well-defined bed
and refilling this with material suitable for dam construction, has
widened and changed its channel to accommodate the larger volumes
of water, deriving its increased sectional area from width, rather than


from depth. Very seldom is the river channel at the end of any
flood season in exactly the same location as in the beginning, and
for this reason there is no certainty that proper materials for con-
structing a hydraulic-fill dara at Bee River could be found. On the
other hand, in the stretch of river channel from Yuma to a point 4
miles below Hanlon Heading, the river has always, within the memory
of man, remained practically in the location it is in to-day, and the
successful construction at any place within this radius is assured.

Engineers acquainted with the Colorado Delta agree that the final

Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 26 of 167)