American Society of Civil Engineers.

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

. (page 25 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 25 of 167)
Font size
QR-code for this ebook


tity in comparison with that applied artificially. The average settle-
ment of this portion of the embankment, in the 2 years following its
completion, was less than f in.

The writer believes that any one who observed the action of the
moistened sand under the rollers would readily admit that they per-
formed a very important part in the consolidation of the material.



DISCUSSION ON THE YALE BOWL 295

The height of the ridges formed by the grooved roller was noticeably Mr.
less with each successive passage of that machine; and the use of ^"^'
the two kinds of roller, passing alternately, was much more effective
than that of either a smooth or grooved one alone would have been.

The Sherardized bolts, referred to by Mr. Lynch, were lag-screws
between the threads of which was spun a flat wire spiral to serve as
reinforcement in transmitting the thrust from the screw to the concrete.

These bolts were mechanically defective in two respects. The wire
spiral was likely to become slightly bent or sprung, in which case,
only one thread of the screw would have a bearing, thus reducing
the effective strength of the fastening. The defect which gave the
most trouble, however, was caused by the machine which fashioned
the bolt. This machine, in gripping the rod, changed the shape of
the shank of the bolt, making the cross-section slightly oval instead
of round, and, in addition, it formed a slight ridge along the sides
of the shank along the major axis. The result was that, when the bolt
was being unscrewed, it acted as a wedge which split a spall, several
inches long, from the face of the step, thereby releasing the bolt. The
only way in which this could be avoided was to give the bolt one or
two turns just after the concrete had taken its initial set, thus forming
a hole large enough to permit the bolt to be turned. If it was not
turned until after the concrete had become hard, the head twisted off.

To repair the damage, dove-tailed recesses were chiseled into the
steps, and the bolts were re-set in mortar. Several hundred bolts had
to be re-set in this way before the writer succeeded in getting an
order to discontinue their use.

In order to insure the head of the bolt having a firm bearing on
the standards, the bolts were set about i in. lower than the position
they were expected to occupy finally. In spite of this precaution,
however, the concrete broken off in unscrewing the bolts filled up
the holes to such an extent that it was impossible, in many cases, to
screw the bolts down again to a bearing, and it was necessary to drill
out the holes and blow out the dust with a blast from an air pump,
the threads of the screw interfering with the use of a spoon.

In the writer's opinion, whenever it is necessary to use screws for
attaching fixtures to concrete it should be done by building the bolt
into the concrete, rather than the reverse. If, however, it is desirable
that the bolt shall be the movable piece, then the nut should be in
an enclosing shell, thereby preventing the thread of the screw from
coming into contact with the concrete. Several hundred bolts of this
type were used, with entire satisfaction, for fastening the Holophane
globes to the ceiling of the tunnels.

The difference in the movement of the two walls, to which Mr.
Lynch calls attention, was undoubtedly due to a change in the design



296 DISCUSSION" ox the yale bowl

Mr. of the structure, rather than to any difference in the foundation, as
^"^' that is of the same general character throughout the field.

The material at the field consisted of strata, of sand and gravel,
of varying thicknesses. Among these were three or four layers of a very
fine, compact sand — material which would be quicksand if in contact
with water. One of these strata, about 6 in. thick, occurred about
3i ft. below the base of the high retaining walls, and was uncovered
in excavating for the foundation of the adjacent tunnels. The con-
struction manager expressed grave fears that, unless the foundation
of the wall was carried down through this stratum and built on the
coarse sand beneath it, the wall would slip. As this stratum was about
20 ft. above the level of ground-water, and as no water could reach
it from above after the embankment was concreted, there was no
danger of the sand becoming "quick"; therefore the writer believed
that, to support the wall, this material was equal to, if not better than,
the coarse but less compact sand beneath it, and he strongly objected
to the expenditure of the several hundred dollars necessary to extend
the footing as proposed. In spite of his protest, however, the extra
footing was built under one wall, the forms for the other wall having
been erected and the steel placed in position before Mr. Atwood became
connected with the work. This extra footing, under the outer edge
of the base of the T-shaped wall, was built in the form of the frustum
of a wedge, with the narrow edge down, thus reducing the effective
bearing area of the sand about one-half, the sides of the wedge having
comparatively little supporting value. This is the wall which Mr.
Lynch refers to as having moved | in. The companion wall, at the
other end of the gate-house, which he says has moved less than i in.
is the same in all respects, except that it is 1 ft. higher, thereby sus-
taining that much additional pressure, but is built with a flat base,
as planned originally.



AMERICAN SOCIETY OF CIVIL ENGINEERS

INSTITUTED 1852



TRANSACTIONS



This Society is not responsible for any statement made or opinion expressed
in its publications.



Paper No. 1387

CONTROL OF THE COLORADO RIVER

AS RELATED TO THE PROTECTION

OF IMPERIAL VALLEY*

By J. C. Allison, Assoc. M. Am. Soc. C. E.



With Discussion by Messrs. A. L. Sonderegger, J. A. Ockerson, and

J. C. Allison.



Synopsis.

The general problems involved in the control of the Colorado River
and the conservation of its waters for irrigation have already been
treated exhaustively in papers and discussions before the Society, and
it is not the intent of the writer to reopen the subject for purposes
of debate. In this paper the direct relation of river control to the
maintenance and ultimate perfection of levee systems protecting Im-
perial Valley, California, is the fundamental topic, as it has been
the phase of the subject with which the writer necessarily has been
most concerned.

In a broad sense, the problem of excluding the Colorado's overflow
from the Valley has been solved. The maintenance of existing lines
of defense, and the establishment of exterior auxiliary or supporting
lines, may be considered as matters of detail. Final and complete
protection, making possible the reclamation and beneficial use of
all the arable lands of the delta, is to be secured, in the opinion which
the writer holds in common with other engineers acquainted with the

* Presented at the meeting of September 6th, 1916.



298 CONTROL OF THE COLOEADO RIVER

situation, most satisfactorily by forcing the river back into the bed
that it abandoned in 1910, and keeping it there with impregnable
retaining works.

Methods of accomplishing that feat have been discussed, and esti-
mates of cost have been made by several engineers, but methods and
estimates have been based on emergency closures of breaks of the Colo-
rado which, the writer believes, do not necessarily govern the case.
Immediate protection being afforded by lines of defense which may be
maintained with no difficulty and at small cost, the restoration of
the river to its old channel is no longer an urgent need, and does
not call for emergency methods. A favorable time may be chosen
for the work, and methods less expensive than those heretofore con-
sidered may be applied.

It is safe to assume that engineers generally know the situation
of Imperial Valley in relation to the Colorado, and only a brief state-
ment of the problems presented to operators of the irrigation system
will be necessary. The slope of the Valley from sea level at the Mexican
boundary line is northward to the Salton Sea, 280 ft. below sea level.
About 25 miles south of the line is the highest land in the delta, the
elevation of which, before the change in the river's course, was 34 ft.
above sea level at the westward end of the delta cone. The normal
elevation of the river bed at the canal intake is about 100 ft. above
sea level.

In flood season, which reaches the maximum in June, the river
overflows the delta south of the crest of the delta cone, and fills a
shallow sink at the west end known as Volcano Lake. The most rapid
fall is southward, and the flood-waters find their way to the Gulf of
California through Hardy's Colorado, which drains Volcano Lake;
but the cone is so flat that some of the flood-waters normally spill
from the lake northward into what was a shallow drainage channel
known as the New River.

In 1905 the river broke through its bank north of the delta cone,
or "Paredones" crest, and flooded the north slope of the delta. In
draining off the inundated land to the Salton Sink, the water cut
deep gorges in New River and in another drainage channel on the
east side of the Valley, known as the Alamo. In the event of another
incursion of the river, the gorges would carry the water, and there
would be no inundation of farm lands on the American side. If per-



CONTROL OF THE COLORADO RIVER 299

mitted to flow for any great length of time, the flood in the New River
gorge would cut back the grade, and in time the recession would reach
the old channel and lower the bed below the level of the intake, so
that no water could be diverted into the main canal. Therefore the
menace to Imperial Valley of an unchecked break would be drought,
not inundation.

The problem presented by the change of course of the river through
the delta in 1910, following the drainage channel known as Bee River
to Volcano Lake, and thence through the Hardy, was the prevention
of a northward flow into New River. That problem was solved by
building the Volcano Lake levee, and by extending the Paredones levee
of the California Development Company, parallel with the Bee River.

A deposit of silt raises the bed of Volcano Lake and consequently
the surface elevation of the overflow water, and that has been met
by increasing the height and cross-section of the levee. A small break
in the levee in 1914 allowed some of the overflow water to go north-
ward, but it was diverted easily from the head of New River gorge and
turned into New River at a point where it did not produce any dan-
gerous recession of grade.

Interior Lines of Defense.

North of the levee paralleling Bee River, which is now the main
Colorado, interior lines of defense exist, and may be strengthened and
extended very readily. The Inter-California Railroad grade serves
as a barrier, and it may be connected with the Volcano levee. The
function of interior lines would be to intercept and divert to the
west a shallow body of overflow water having no deflned channel and
no eroding velocities. Numerous canal banks also serv^e as auxiliaries
to check velocities and turn the water westward, where it may be
taken care of readily.

The effect of one of these canal banks on the overflow of 1914
suggests a third line of defense and a method of diverting any water
that might pass the Paredones and Volcano levees. This canal is
the New Alamo Mocho, constructed by a rancher in Mexico to irrigate
a large tract lying on both sides of New River channel. The water
that came through the levee break, about 6 000 sec-ft., spread over
the ground in a wide, shallow sheet, and was held by the canal bank,
along which it flowed slowly to the west.



300 CONTEOL OF THE COLORADO EIVER

Assuming a breach in the Volcano Lake levee to be closed
promptly, the inflow northward would be cut off before the slowly
spreading sheet of water reached the west end of the proposed Alamo
Mocho Canal bank, and it might not be necessary to let the water into
New River at all. The overflow could be impounded and used as a
reservoir for the West Side Main iintil drained off.

Each year prior to the break of 1905-06, which eroded the bed
of ISTew River, the overflow water from the Colorado entered on the
northern slope of the divide. Each year this flow to the north sub-
sided of its own accord, before the flow of the water to the south. The
interior line of defense cuts off the water from entering New River
gorge and so reinstates former conditions.

As an interior line of protection in emergency arising through pos-
sible failure of the primary lines, this proposed extension of the Alamo
Mocho bank as a diverting levee seems entirely feasible and compara-
tively inexpensive to construct and maintain. It is important that the
run-off channels leading southward from Volcano Lake be kept open
and free from obstruction. Bee River has now cut a defined channel
for itself through the silted Volcano Lake, and may be expected to
keep to that approximate course for a long time.

Primary Line of Defense.

Only through negligence and failure to maintain the old California
Development Company levee from Hanlon Heading south could the
river again get into the region that was flooded in 1905-06, and be a
menace to the Alamo Canal and delivery structures. The existing
levee, properly maintained and protected from destruction by erosion
of the banks of the river, is sufficient protection to the canal system
as far as it extends. In 1914 a breach at the point where the levee
bends to the southward resulted from negligence. Erosion of the
river bank threatened the levee two years before the breach occurred.

The danger could have been averted by inexpensive means, but
responsibility for the maintenance of the levee was a matter of con-
troversy, and nothing was done until the bank and a point of the
levee had been cut away by the river. ]\[uch money was wasted in
an attempt to close the breach by the trestle and rock fill method on
the line of the original location. The attempt was a complete failure,
and the simple expedient of building a levee aroxmd the break was



CONTEOL OF THE COLORADO RIVER



301




303 CONTROL OF THE COLORADO RIVER

adopted, and the damage was readily repaired. In this case there
was no danger of destructive flooding, as the river did not overflow
the banks, and there was no head of water against the levee. The
breach was caused by erosion of the bank and levee which could have
been prevented by temporarily diverting the current away from the
bank with jetties, until rock revetment at the toe of the levee could
be placed.

The breach at "House 7", as the point is designated, was not due
to any defect in the location or construction of the primary line of
defense, but solely to failure to maintain the line. It was an adminis-
trative, not an engineering failure.

During the flood of 1914, the long stretch of bank between the levee
and the river was cut away, up to the toe of the levee, but the latter
was saved intact by dumping large quantities of rock at the foot of
the slope, which filled the trench scoured by the current, and formed
permanent revetment. A railroad track on the levee, and an inex-
haustible supply of rock in a quarry at Hanlon Heading, near by,
facilitated this method of revetment.

To make this line of defense impregnable, it is necessary only to
reinforce the levee wherever bank cutting occurs, doing the work from
year to year as conditions call for it. Assuming that reasonable
attention is given to the matter, and that ordinary diligence in main-
tenance is observed, the protection afforded by the levee south of
Hanlon Heading is sufiicient. When thoroughly revetted down to
the scour line, the levee cannot be broken by flood.

Head-Gate Foundation.

The only weak point in the line is the head-gate in the Alamo, or
Imperial Canal, at Hanlon Heading; and the weakness there is due
to the character of the rock fomidation, which is not what it was
assumed to be by the builders of the gate.

Some have held that the location of the gate in a rock spur of
Pilot Knob was not necessarily of strategic importance, and that such
a structure might be built safely and satisfactorily on alluvial soil,
on the Mexican side of the line, using wooden caissons as a foundation.

Without going into the reasons for this proposed change of location
of the head-works, which are political and, in the opinion of the
writer, not based on engineering considerations, it seems almost self-



CON'TEOL OF THE COLOKADO EIVER 303

evident that rock foundation is a vital factor in the construction of
a head-gate on the Colorado. This is demonstrated, not only by the
failure of the structures lacking such foundations, but by the weakness
developed in the foimdation of the concrete gate at Hanlon Heading.

The Rockwood gate, constructed of timber in 1905, was' built in
the alluvial bank of the Colorado Eiver in Mexico, some distance from
the channel, to serve the double purpose of first controlling the run-
away river, and later becoming the head-gate for the Alamo, or
Imperial Canal. The writer is familiar with the design and con-
struction, and all problems connected with that gate, and was on the
ground at the very moment when it failed. It is his opinion that it
would be difficult to devise, for a structure resting on alluvial soil, a
more substantial foundation than was designed for this gate. In other
words, the Rockwood gate represents a type of construction as nearly
perfect as can be devised, at any reasonable expense, for a structure
resting on alluvial silt deposits, and yet it failed.

In building structures, whether of wood or concrete, on alluvial
soil, the problem presented to the engineer does not concern strength
of material and design to meet the stresses to be put on it, so much
as it involves provision to meet the boring action of water in front
of and back of the structure, wherever large volumes are dealt with.
For several months after it is built, the structure must be watched
constantly because of the boring action and the resulting settlement
or displacement of filling under and aroiind the wings and curtains.
After the settlement of the filling, the structure may be reasonably
safe for the head under which it was "cured", but an increase or
decrease of the head tends to develop new conditions which may cause
damage.

This applies to Imperial Valley generally, but the difficulty is
even more pronounced along the banks of the Colorado, where the
land has been built up in strata of sand and silt containing more or
less decayed vegetable matter, and where the structure is subject to
extreme variations of pressure induced by a rapidly fluctuating river.

The Rockwood gate failed, not from the pressure of earth or
water it was required to withstand, but from the boring which took
place in front of the front curtain. In spite of the extended upper
curtain, and the extensive apron built in front of the structure, this
boring continued to develop with the increased volume of water put



304 COXTEOL OF THE COLORADO EIVEE

through the gate, until it had entered under the front end of the
structure and lifted it bodily from its anchorage.

This might have been avoided had rock and other material been
available in quantities sufficient to prevent this scouring, but the same
condition might have developed at any time under either wing of
the structure, or from leakage through the silt puddle work when
the structure was later suddenly subjected to increased water pressure.
This is especially true because, during a recent seismic disturbance,
the silt puddle work around most of the secondary structures in the
canal system was disturbed.

The concrete gate at Hanlon Heading has been in use since Novem-
ber, 1906. During this time it has served its purpose as a regulator
in the Imperial Canal, but with excessive operating costs, due to its
design, and, as recently discovered, with considerable danger to the
system during high water. The cost of handling drift through the
gate has been excessive, and represents the largest operating expense
of the river division, outside of the dredging of the canal. The
enormous quantity of drift which the river brings into the intake and
lodges against the gate has, in the past, blocked the openings to such
an extent as to cause temporary water shortage in Imperial Valley.

As a partial solution of the drift problem, a sheer boom was built
across the mouth of the intake during the season of 1912, and most
of the surface drift has been handled successfully at that point, rather
than at the gate itself. The submerged drift continued to give
trouble by lodgment in the gate tunnels until a new section of the
gate was built during the spring of 1913. The construction of the
new gate was such as permitted the free passage of submerged drift.
The level of the sill of the Hanlon Gate, as originally constructed,
was calculated to pass all the water required for irrigation in Imperial
Valley without a diversion weir in the river. The width of the struc-
ture was in excess of what was calculated for carrying water in the
canal alone, as it was presumed that the gate would play a part in
turning the water of the Colorado River back to its old channel.
The sill proved to be sufficiently low to allow the desired quantity
of water to pass over it into the Imperial Canal until 1912; but, be-
cause the average bottom of the river was lower during the low-water
period than previously, owing to the new route which the river takes
to the Gulf through the Bee River, and because the demands for



CONTEOL OF THE COLORADO RIVER 305

water during the low stage of the Colorado River were greater than
was calculated, the sill of the gate proved to be too high.

During the summer of 1910 it became necessary to construct a rock
weir across the Colorado River, in order to divert sufficient water
through the gate to meet the demands. The necessity for the weir
arose from the fact that the canal above and below the gate had
silted, but since then it has developed that, even with the intake clear
to the elevation of the former sill of the gate, sufficient water could
not have been diverted at the low stage of the river.

It became evident that a second diverting weir would be necessary,
or that the sill of the gate would have to be lowered. A permanent
weir in the Colorado River was out of the question. The property
of the California Development Company was in the hands of a
receiver, and he could not do permanent work of this nature; neither
were the funds in his hands sufficient to meet the costs of a permanent
weir at this point. A temporary weir, such as was constructed in 1910,
would not have served the purpose, on account of the heavy main-
tenance expense and the danger to the levees and banks on both sides
of the river, caused by its presence during high water.

It was decided, therefore, to lower one 25-ft. section on the
north end of the concrete gate, and in the new opening the radial
gates were replaced with a sliding gate of the Stoney type. After
the coffer-dams were in place and the excavation work had commenced,
a flow of water developed around the end of the coffer-dam connecting
with the original concrete gate, and thence inside the coffer-dam. A
careful investigation disclosed the fact that the original gate did
not rest on solid rock, but on a very inferior grade of decomposed
granite, full of soft seams, all of which were water-bearing. This
decomposed material had been scoxired from under the floor of the
old concrete gate until the piers and more than one-half of the
original floor had scarcely any foundation left. Holes bored by the
water under the original floor were in some places as deep as 8 ft.
below the floor line. It is a marvel that the gate had stood, against a
head of water varying in the simimer from 13 to 16 ft., on the little
remaining foundation, through which there must have been constantly
a great quantity of percolation. Had this condition been allowed to
remain through one, or perhaps two, additional flood seasons of the
Colorado River, the remaining foundation would certainly have been



306 CONTEOL OF THE COLORADO RIVER

scoured out, causing the complete failure of the structure. This con-
dition under the north side of the gate was corrected in that portion
enclosed within the coffer-dam by a series of concrete curtain-walls
running down in front and below the structure to serve as cut-off
walls. The foundation was refilled with concrete, and a strongly-
reinforced concrete iloor was laid on top. This section of the struc-



Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 25 of 167)