American Society of Civil Engineers.

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

. (page 28 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 28 of 167)
Font size
QR-code for this ebook

The bank at that point is on the convex side of the river, and is
swept by the main current during the greater part of the year. It
consists of a ledge of conglomerate which apparently presents some
resistance to the boring aaid scouring action of the river. It will
require careful soundings and investigation before its suitability can
be determined definitely.

If the result should be found to be satisfactory, a heading can
be constructed in a manner that will prevent heavy silt from entering
the canal. The writer believes that even in the absence of special
settling basins, the quantity of silt admitted to the canal can be
reduced materially, and that this would produce the effect of scouring
in the Alamo Canal. The present grade of this canal is practically
stationary, and is the result of the discharge of waters carrying
extremely high percentages of silt and sand. If operated under the


Mr. same grade, but with comparatively clear water, a scouring action is
Sonderegger. ^^^^ ^^ occur. It is a phenomenon often observed in the canals in
Imperial Valley.

Probably, as far as the average low-water discharge is concerned,
there would be no difficulty in diverting the quantity necessary for
irrigation; for extreme low waters, as they have occurred in 1912,
1913, 1914, and 1915, when at times practically the entire flow was
diverted, there would still be the necessity for the construction of
temporary dams — whether they are hydraulic-fill or rock-fill. It should
be remembered that increasing quantities of water will be diverted
for the Yuma Project as all its lands are brought under cultivation.
Similarly, there is still a constant increase in the acreage in Imperial
Valley, so that within a few years the entire low-water flow of the river
will be required. This is exemplified in the report* by E. C. LaRue,
Assoc. M. Am. Soc. C. E.

It may be stated, therefore, that the reconstruction of the heading
on the banks of the river, IJ miles above Hanlon, even if found
practicable, will present but a partial solution of the problem of
diverting the low-water flow, and, likewise, only a partial solution as
regards the elimination of silt, provided the river maintains its present
channel. The successful remo,val of all silt by settling basins — in the
opinion of the writer — depends entirely on the operation of the pro-
posed new heading.

Another plan for securing a sufficient supply of clear water for
Imperial Valley is the proposed construction of a canal from Hanlon
Heading to Laguna Dam. There will be no engineering difficulties
in the construction of the connecting canal. The proposed route
crosses several large washes which, at long intervals, are subject to
sudden floods resulting from cloudbursts. However, there is sufficient
grade available, so that short portions of the canal can be constructed
as underground conduits with heavy grades and correspondingly
reduced cross-sections; thus, the cost of construction is not likely
to be excessive.

There is one feature of this plan which will require careful con-
sideration, namely, the necessity for the continuous operation of the
Imperial Canals. The Laguna Project at present diverts probably
not in excess of 750 sec-ft. during the summer. In order to have the
benefit of the settling basin above the dam, it is necessary to flush
the settling pond immediately above the intake one day per week, so
that the canals are operated 6 days out of the 7. The distance from
Laguna Dam to Yuma is about 12 miles.

It is doubtful whether this mode of operation could be applied
to the Imperial system. The summer diversions would be increased
from 750 to about 6 000 sec-ft., and, during the low stages of the

* Water -Siipply Paper No. 395.


river, every drop would be diverted. The distance from Laguna Dam Mr.
to Imperial Valley is about 75 miles, and, in view of this, it is
doubtful whether intermittent operation would be successful. In
other words, though the connection with the Laguna Dam would
undoubtedly secure a permanent supply, it would not solve the silt
problem completely.

Whatever solution will finally be chosen, it will necessitate the
expenditure of large sums of money, and on the success of the
undertaking will depend the future of Imperial Valley. No decision
should be made, therefore, except after full and mature consideration
of all questions involved.

J. A. OoKERSON,* Past-President, Am. Soc. C. E. (by letter). — In Mr.
this paper, Mr. Allison presents some very interesting and novel fea-
tures. His long experience in connection with the irrigation of
Imperial Valley gives exceptional weight to his views as to the proper
solution of the vexing problems involved.

Complicated hy International Boundary Line. — A comprehensive
plan for the reclamation of the Colorado River delta is complicated
by the presence of the International Boundary Line between the United
States and Mexico.

If that was eliminated, the restoration of the river to its former
bed and the limitation of its floods thereto by suitable levees or
embankments, together with such revetment as might be necessary to
check and prevent excessive bank erosion, would solve the problem of
regulation definitely, so far as the river itself is concerned. "Interior
lines of defense", which, at best, are temporary makeshifts, would then
be unnecessary.

The water of the river, however, is necessary to supply moisture
to about 500 000 acres in Imperial Valley, California, and about double
that area on the Mexican side of the line. This is more than double
the total area of land reclaimed by the United States Reclamation
Service, and Imperial Valley alone has a cultivated area measuring
more than two-thirds of the total of all the Reclamation projects.
This gives a fair measure of the importance of the projects under

It is the diversion, control, and distribution of this water in
Imperial Valley which has largely claimed Mr. Allison's attention for
several years, the development of the Mexican side being compara-
tively small.

The New River Gorge. — The development of the New River gorge
is commonly attributed to the break of 1905, when the Colorado River
left its channel and emptied its waters into Salton Sea. An investi-
gation of the earliest records left by the stream itself in its wander-

* St. Louis, Mo.


Mr. ings shows plainly that this diversion to the Salton Sea had occurred
Ockerson. j^^^^ times in the distant past.

Early Mexican maps, dating back to a time before any attempt had
been made to divert and control the waters, show the New River gorge
to have been developed to a considerable extent even then. The rapid
enlargement of the gorge during 1905-06 was most impressive as an
example of excavation on a gigantic scale, such as would make the
Panama Canal work seem insignificant, and has rarely been equalled
elsewhere in Nature.*

Break in the Levee at "House 7". — The author speaks of the breach
in the levee at "House 7", and suggests that the use of "jetties" to
deflect the current away from the bank would have prevented the
damage until rock revetment could have been applied. He also states
that there was "no danger of destructive flooding, as the river did not
overflow the banks". If the records of the Reclamation Service are
correct, the river banks at the point named have never been overflowed
by the highest known floods, and why "the simple expedient of building
a levee aroiuid the break", or why a levee 9 ft. high was ever built at
all on a bank which is 4 ft. above the highest knoAvn flood, has always
been a mystery to the writer. It is fair to say that the author was not
responsible for either of these.

Deflecting Dikes or Jetties. — The use of dikes or jetties to deflect
the current away from the banks is a common delusion. R. Jasmund,
in his review of the regulation of the Rhine, covering some 150 years,
recites the "countless failures" in which deflecting dikes had a very
large share. He says that, in spite of such dikes inclining down stream,
the bank receded more and more and, in many cases, the erosion was
strongest just behind the dikes, where protection had been looked for,
and the dikes themselves were cut out and disappeared. After being
thoroughly tried out, the deflecting type of dike was abandoned, and
dikes of suitable length perpendicular to the bank were substituted,
which resulted in deposits along the shore below them, which method
has had a large part in the successful regulation of the Rhine and
other streams.

Like experience on the Mississippi and Missouri Rivers has also
demonstrated the futility of current deflectors in the protection of
river banks from erosion.

Levees Along the West Bank. — When the break of 1905 was closed,
a levee was built to the southward for a few miles, and then turned
westward toward the old overflow channel known as the Paredones. It
was well known at that time that a break in the near future was
imminent by way of Bee River. If, instead of carrying the levee
back away from the river, it had been extended down along the river,
across and well below the Bee River, which was dry except at flood

* Transactions, Am. Soc. C. E., Vol. LIX, Plates V and VI, and p. 37.


stages, the break of 1910 with its attendant losses never would have Mr.
occurred. It would also have been in line with "final and complete"
regulation as advocated by "engineers best acquainted with the

The writer discussed this matter at the time with those in charge
of closing the break, and understood that the plan was to extend the
levee down along the west bank of the river to what was reported to
be high land well down toward the lower end of the delta. There are,
in fact, evidences on the ground, in the nature of a partly constructed
levee for a short distance below the point where the levee turns to the
westward. It would be interesting to know what caused the change
in the plans.

Hydraulic-Fill Dam Across a Flowing River. — The author's suc-
cess in building, across a flowing river, a hydraulic-fill dam out of
material pumped from the lower strata of the bed of the stream is
a striking example of what a resourceful engineer can accomplish
when an emergency arises.

In his official report of 1910 the writer suggested that the whole
operation of the closure of the Bee River break "might be economically
handled in the dry", by operating during a low-water season, using
the temporary rock-fill dam built just below the intake, and raising it
to the required height to divert the whole flow of the river through
the Imperial Canal.

The writer's experience does not agree with the statement that the
depth of scour is limited by the heavier material composing the bottom,
and that "the materials are graded as to weight, and deposited in
direct relation to the velocities". In the Mississippi Eiver, the bars
composed of gravel larger than the shingle described by the author,
extend well above low water, and the deep concave bends opposite the
bars have no gravel in them at all. It is not uncommon to see coarse
gravel distributed over the high-water banks. Take the exposed sec-
tion of an alluvial bank in general, all of which has been deposited
from a state of suspension, and there is little evidence to show that
the materials are graded as to weight.

The scour line of the bed of the river at Yuma is also doubtless
influenced by the bed-rock which extends across the river in that

The construction of the dam described was only possible during
the extremely low stage in the river, amounting to only 4 000 sec-ft.,
all of which could easily be carried through the head-gates to Imperial
Canal. The low cost of the work fully justified the effort, and its
novelty and successful outcome reflect great credit on the builder.

It is stated in the press that there is under construction a new tem-
porary rock-fill dam which will cost several times as much as the


Mr. hydraulic-fill dam tinder discussion, and it would be interesting to
c erson. ^^^^ ^j^y ^j^g latter was not repeated.

Restoration of Colorado River to Its Former Bed. — The restoration
of the Colorado River to its former bed must be realized before a final
solution of the problem of control is reached; and, when it is under-
taken, the experience gained in the construction of the hydraulic-fill
dam can be applied with success. As the author well says, closure
should not be attempted under emergency conditions, as heretofore,
but a favorable time should be selected, when the whole volume can be
diverted down the canal through the head-gates, and then it will be
possible to build "dry-shod" such dam or barrier as may be necessary
to turn the flow back to its old channel. The latter would also probably
need clearing out and rectifying, in order to facilitate the flow.

Railroad on Crown of Levee. — The vsTiter cannot agree that a rail-
road on the crown of the levee is a "vital factor in all levee protection
on the Lower Colorado"; but it may be a convenience in the work of
maintenance, which cannot be neglectfed if the levee is to remain

The writer is of the opinion, also, that, in the use of rock to pro-
tect the exposed face of the levee from erosion, economy as well as
safety dictates that the rock facing shall be laid by hand with some
care, instead of being dumped at random from cars, as a large part of
the rock thus dumped serves no useful purpose; witness the levee
north of Bee River, which was breached in spite of such protection.

Restricting the River to Narrow Limits. — The author states that
"the less room the river is allowed for meandering, the greater its
velocity and scouring action, and the easier it is to keep it under
control". This statement needs qualification. When the Sonora Mesa
is approached, only one bank of the river requires control by revet-
ment; the Mesa itself takes care of the other bank, as do the hills on
the right bank from near Yuma to Andrade, which the author cites as
a narrow confined bed. Increase of velocity necessarily increases bank
erosion, as well as the difficulty of control by bank revetment. Increased
velocity does not mean that the energy will all be expended on the bed
of the stream, for it will also attack the banks. Therefore, why not
give it more lateral room, so as to reduce the flood height?

Location of the Government Levee. — The location of the Govern-
ment levee on the west bank of the river was fixed after a careful
study of the changes in the bank line, as shown on a map defining
several positions of the river dating from the earliest surveys.

This study developed the fact that, had a levee been built 20 years
earlier, and placed at a distance of 3 000 ft. from the westerly bends
of the river, it would have remained intact, as far as river bank erosion
was concerned.


As the project of 1910 could not cover bank revetment, it was Mr.
decided to place the Government levee so that it would be safe from
destruction by river bank erosion for a probable period of about 20
years, the primary object of the levee being to carry the flood-water
so far down stream as to eliminate the possibility of its flowing into
Imperial Valley. With the exception of the detour to connect with
the Bee River crossing, which was regarded as the best available point,
the levee alignment was fixed in accordance with this plan.

In a comprehensive scheme for the regulation of the Colorado
River, including bank revetment and flood control, the levee alignment
might be somewhat different.

River Side Borrow-Pits. — The author seems to object to river side
borrow-pits, which constitute a cardinal principle, followed in all parts
of the world, except in the Colorado River delta; and in most of the
levees built there that practice has been followed, except when pre-
vented by the encroachment of the water on the river side.

An examination of the Government levee after the floods of 1911
disclosed the fact that there were no channels in the pits paralleling
the levee, but, on the contrary, the pits in many cases were obliterated
by deposits of sediment. There was no evidence anywhere that the
river side pits were responsible for damages to the levee.

The imperative necessity for at least temporary protection of the
river slope of the levee was fully realized by the writer, but, when pro-
posed to those '"higher up", was rejected, because there was "no authority
mider the appropriation, or negotiations with Mexico for maintenance".

Hanlon Head-Gates. — Apparently, very little attention had been
given to the hydraulics of the river, its variations in stage, and the
oscillation in the elevation of the bed conforming in a measure to the
stage, until after the concrete head-gates had been placed at Andrade
in 1906. Shortly after they were built, it was realized that the sill
of the intake, which was 5 ft. below the bed of the river at the time,
was too high, and this is what made it necessary to build temporary
dams at low stages below the intake, in order to increase the head so
as to divert enough water to supply the demands of Imperial Valley.

In July, 1910, the supply dropped to 746 sec-ft., only 395 sec-ft.
reaching Imperial Valley. Much of this trouble was due to a silting
up of the intake above the head-gates.

Both the United States and Mexico Interested. — A satisfactory and
comprehensive solution of conserving, controlling, and using the waters
of the Colorado River can only be reached by a joint commission of
engineers from both sides of the International boundary line. It has
already been too long deferred, and it is hoped that Mexican affairs
will before long reach such a peaceful and prosperous state as to per-
mit the consideration of the agricultural and other developments of
the delta of the Colorado.


Mr. J. C. Allison,* Assoc. M. Am. Soc, C. E. (by letter). — A state-

'^°°' meut of developments, in the control work of the Colorado River pro-
tecting the Imperial Valley, since the paper was written in October,

1915, and since the discussion was published in September, 1916, is
of value.

During January, 1916, there occurred the greatest flood ever
recorded on the Lower Colorado. At Yuma the discharge was 210 000
sec-ft., or 60 000 sec-ft. in excess of any previous measurement. The
restricted channel opposite the Imperial Valley intake carried the
flood, without side-scouring and with no meandering of the stream.
The channel in the vicinity of "Houses 5 and 6" continued to widen
and meander to the extent of undermining the primary line of defense.
Bank erosion through the restricted area was not increased in propor-
tion to the stream-bed erosion by the increase in velocity. The writer's
statement that "the less room the river is allowed for meandering, the
greater its velocity and scouring action, and the easier it is to keep
it under control", to which Mr. Ockerson has taken exception, is not
only qualified here, but was again proved during the summer flood of

1916. The theory that the greater energy due to increasing velocity
is expended on the bed of a stream, and not on the bank, is correct,
applied, of course, to locations where the stream alignment is good
and where local deflections do not occur. Local deflections will not
occur where the stream bed is narrow and deep, but they do occur
where it is wide and shallow and the formation of shifting sand-bars
is made easily possible.

At the breach in the levees between "Houses 5 and 6", which took
place iu January, 1916, the Inter-California Railroad, the second line
of defense, held the flood-waters, and kept them from entering the
Alamo Channel near the site of the original 1906 break. The breach
occurred again from lack of maintenance, and the incident does not
alter the advisability of retaining this line of defense and the present
method of revetment. Where the breach occurred, the levees were
designed with land-side borrow-pits, and though the flood-water
through the breach entered these land-side borrow-pits, it was easily

On the Lower Colorado River protective work, past experience
demonstrates beyond a doubt that the land-side borrow-pits are more
advisable. The writer has even gone so far as to propose in one in-
stance the construction of a drain canal in the borrow-pit parallel
with the protective levee, for the purpose of preventing the saturation
of the lands adjacent to the levees. In no instance has a levee failure
been traced to the presence of a continuous pit on the land side, so
long as the width of the levee and the width of the berm are sufficient.

The irrigation system and the system of protective work are now

* Calexico, Cal.


in the hands of the Imperial Irrigation District. Improvements to Mr.
the flood-protection system of the Valley have been begun, the greater ^ ''^°°
part of which are in accordance with the recommendations set forth
in the paper. An additional control-gate is now being constructed
above the present Hanlon gate, as suggested by Mr. Sonderegger.
The abandonment of the southerly end of the Hanlon Head-gate and
the construction of additional openings in the rock to the north, are
being planned to remedy the defects in that structure, as mentioned
in the discussion.

The second line of levee-protection defense has been improved by
placing a railroad track thereon throughout the whole extent of the
Volcano Lake Levee, in accordance with early suggestions, and it is
planned to open a quarry, in addition to the Hanlon Heading quarry,
at the Volcano Lake end of the system. It is likewise planned to
supply the district with sufficient equipment to facilitate present
methods in maintaining the levee, namely, with rock revetment. In
this regard, the suggestion made by Mr. Ockerson, which was carried
out most efficiently and economically in his reconstruction of the
Volcano Lake Levee, namely, that the rock facing should be laid by
hand with some care, instead of being dumped at random from cars,
is well founded in localities where the levee is not subjected to under-
mining from the river. This method of revetment is satisfactory for
the entire second line of defense for Imperial Valley, but it is not
practical on the primary line of defense and its extension, as proposed
in the paper, on account of the difficulty of placing such rock to scour-
line elevation.

The following is a resume of the arguments and discussions con-
cerning the theory and practices incident to the construction of a
hydraulic-fill dam, as a means of diverting a stream flowing in an
alluvial plain, which forms the principal subject of the paper.

The writer has made the statements that: "a direct relation exists
between the character of the materials forming the bed and the
velocities of the water", that "in refilling the bed, the materials
are graded as to weight and deposited in direct relation to the veloci-
ties", and that in refilling the bed, the materials decrease "in weight
as they approach the elevation corresponding with the lesser velocities."
Mr. Ockerson cannot agree with this theory, and cites, as an example,
the Mississippi River, in which bars composed of gravel larger than
shingle exist well above the lower water, and deep concave bends op-
posite the bars contain no gravel whatever. Mr. Ockerson's citation
is of a stream in which the channel is permitted to deflect from one
side to the other, and in which increased areas are developed by side
scour rather than by bed scour.

Such a stream as that cited by Mr. Ockerson is that section of the
Colorado River at the site of the proposed dam to deflect the Colorado


Mr. River water from its present course into the Bee River, back into the
^^°°' old channel of the Colorado. It is on account of such a condition
that the writer has proposed the construction of a hydraulic-fill dam
up stream some 12 miles, where the channel has been virtually con-
fined to one bed for a great number of years, rather than where it has
been shifting in different locations each year. The Mississippi River
condition and that which makes possible the construction of the

Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 28 of 167)