Chapter V: Different Types of Earth Dams

There are several types of earth dams, which may be described as follows:

1. Homogeneous earth dams, either with or without a puddle
trench.

2. Earth dams with a puddle core or puddle face.

3. Earth dams with a core wall of brick, rubble or concrete
masonry.

4. New types, composite structures.

5. Rock-fill dams with earth inner slope.

6. Hydraulic-fill dams of earth and gravel.

The writer proposes to give an example of each type, with such remarks upon their distinctive features and relative merits as he thinks may be instructive.

Earth Dams with Puddle Core Wall or Face.

YARROW DAM.–The Yarrow dam of the Liverpool Water-Works is a notable example of the second type, (a section of which is shown in Fig. 2.) An excavation 97 ft. in depth was made to bed rock through different strata of varying thickness, and a trench 24 ft. wide was cut with side slopes 1 on 1 for the first 10 ft. in depth below the surface. The trench was then carried down through sand, gravel and boulders with sides sloping 1 in 12. The upper surface of the shale bed rock was found to be soft, seamy and water-bearing. Pumps were installed to keep the water out of the trench while it was being cut 4 or 5 ft. deeper into the shale. The lower portion was then walled up on either side with brickwork 14 ins. in thickness, and the trench between the walls was filled with concrete, made in the proportion of 1 of cement, 1 of sand and 2 of gravel or broken stone. By so doing a dry bed was secured for the foundation of the puddle wall. Two lines of 6-in. pipes were laid on the bed rock, outside of the walls, and pipes 9 ins. in diameter extended vertically above the top of the brickwork some 27 ft. These pipes were filled with concrete, after disconnecting the pumps. After refilling the trench with puddle to the original surface, a puddle wall was carried up simultaneous with the embankment, having a decreasing batter of 1 in 12, which gave a width of 6 ft. at the top. This form of construction is very common in England and Figs. 14 and 15 show two California dams, the Pilarcitos and San Andres, of the same general type.

ASHTI EMBANKMENT.–This is not a very high embankment, but being typical of modern dams in British India, where the puddle is generally carried only to the top of the original surface of the ground, and not up through the body of the dam, it is thought worthy of mention. Fig. 16 shows a section of this embankment, which is located in the Sholapur District, India.

The central portion of this dam above the puddle trench is made of “selected black soil;” then on either side is placed “Brown Soil,” finishing on the outer slopes with “Murum.” Trap rock decomposes first into a friable stony material, known in India as “Murum” or “Murham.” This material further decomposes into various argillaceous earths, the most common being the “black cotton soil” mentioned above.

This particular dam has been adversely criticised on account of the lack of uniformity in the character of the materials composing the bank. It is claimed that the materials being of different density and weight, unequal settlement will result, and lines of separation will form between the different kinds of materials.

Earth materials do not unite or combine with timber or masonry, but there are no such distinct lines of transition and separation between different earth materials themselves as Fig. 16 would seem to indicate.

Puddle Trench.

In the last three dams mentioned (Figs. 14, 15, 16) the puddle trenches are made with vertical sides or vertical steps and offsets. A wedge-shaped trench certainly has many advantages over this form. Puddle being plastic, consolidates as the dam settles, filling the lowest parts by sliding on its bed. It thus has a tendency to break away from the portion supported by the step, and a further tendency to leave the vertical side, thus forming cracks and fissures for water to enter. The argument advanced by those holding a different view, namely, that it is difficult to dress the sides of a trench to a steep batter and to timber it substantially, has in reality little weight when put to practical test. Mr. F. P. Stearns, in describing the recent work of excavating the cut-off trench of the North Dike of the Wachusett reservoir, Boston, said it was found to be both better and cheaper to excavate a trench with slopes than with vertical sides protected by sheeting. He favored this shape in case of pile-work and for the purpose also of wedging materials together.

Mr. Wm. J. McAlpine’s “Specifications for Earth Dams,” representing the best practice of 25 years ago, which are frequently cited, contain the following description of how to prepare the up-stream floor of the dam:

Remove the pervious and decaying matter by breaking up the natural
soil and by stepping up the sides of the ravine; also by several
toothed trenches across the bottom and up the sides.

One of Mr. McAlpine’s well known axioms was, “water abhors an angle.” The “stepping” and “toothed” trenches above specified need not necessarily be made with vertical planes, but should be made by means of inclined and horizontal planes. The writer’s experience and observation leads him to think that all excavations in connection with earth dams requiring a refill should be made wedge-shaped so that the pressure of the superincumbent materials in settling will wedge the material tighter and tighter together and fill every cavity. A paper by Mr. Wm. L. Strange, C. E., on “Reservoirs with high Earthen Dams in Western India,” published in the Proceedings of the Institution of Civil Engineers, Vol. 132, (1898), is one of the best contributions to the literature on this subject, known to the writer. Mr. Strange states that

the rate of filtration of a soil depends upon its porosity,
which governs the frictional resistance to flow, and the
slope and length of the filamentary channels along which the
water may be considered to pass. It is evident, therefore,
that the direct rate of infiltration in a homogeneous soil
must decrease from the top to the bottom of the puddle
trench. The best section for a puddle trench is thus a
wedge, such as an open excavation would give. It is true
that the uppermost infiltrating filaments when stopped by
the puddle, will endeavor to get under it, but a depth will
eventually be reached when the frictional resistance along
the natural passages will be greater than that due to the
transverse passage of the puddle trench, and it is when
this occurs that the latter may be stopped without danger,
as the _filtration to it_ will be less than that
_through_ it. This depth requires to be determined in
each case, but in fairly compact Indian soils 30 feet will
be a fair limit.

Puddle Wall vs. Puddle Trench.

There is a diversity of opinion among engineers in regard to the proper place for the puddle in dam construction. Theoretically, the inner face would be preferable to the center, for the purpose of preventing any water from penetrating the embankment. It is well known that all materials immersed in water lose weight in proportion to the volume of water they displace. If the upper half of the dam becomes saturated it must neccesarily lose both weight and stability. Its full cohesive strength can only be maintained by making it impervious in some way. The strength of an earth dam depends upon three factors:

1. Weight.
2. Frictional resistance against sliding.
3. Cohesiveness of its materials.

These can be known only so long as no water penetrates the body of the dam. When once saturated the resultant line of pressure is no longer normal to the inner slope, for the reason that there is now a force tending to slide the dam horizontally and another due to the hydrostatic head tending to lift it vertically. When the water slope is impervious the horizontal thrust is sustained by the whole dam and not by the lower half alone. When once a passage is made into the body of the dam, the infiltration water will escape along the line of least resistance, and if there be a fissure it may become a cavity and the cavity a breach.

For practical reasons, mainly on account of the difficulty of maintaining a puddle face on the inner slope of a dam, which would require a very flat slope, puddle is generally placed at the center as a core wall.

It was thought possible at the Tabeaud dam to counteract the tendency of the face puddle to slough off into the reservoir by use of a broken stone facing of riprap. This covering will protect the puddle from the deteriorating effects of air and sun whenever the water is drawn low and also resists the pressure at the inner toe of the dam.

Percolation and Infiltration.

The earlier authorities on the subject of percolation and infiltration of water are somewhat conflicting in their statements, if not confused in their ideas. We are again impressed with the importance of a clearly defined and definite use of terms. The temptation and tendency to use language synonymously is very great, but it is unscientific and must result in confusion of thought. Let it be observed that _filtration_ is the process of mechanically separating and removing the undissolved particles floating in a liquid. That _infiltration_ is the process by which water (or other liquid) enters the interstices of porous material. That _percolation_ is the action of a liquid passing through small interstices; and, finally, that _seepage_ is the amount of fluid which has percolated through porous materials.

Many recent authorities are guilty of confusion in thought or expression, as will appear from the following:

One says, for instance, that a

rock is water-tight when non-absorbent of water, but that a soil
is not water-tight unless it will absorb an enormous quantity of
water.

This would seem to indicate that super-saturation and not pressure is necessary to increase the water-tightness of earth materials.

Again, in a recent discussion regarding the saturation and percolation of water through the lower half of a reservoir embankment, it was remarked, that

the more compact the material of which the bank is built,
the steeper will be the slope of saturation.

Exception was taken to this, and the statement made, that

_with compact material_, the sectional area of flow
is larger below a given level with porous material, and as
the bank slope is one determining factor of the line of
saturation, this line tends to approach the slope line;
while with porous material in a down-stream bank, the slope
of saturation is steeper and the area of the flow less.

In reply to this, it was said,

that it is obvious that if the embankment below the core
wall is built of material so compact as to be impervious
to water, no water passing through the wall will enter it,
and the slope of saturation will be vertical. If it be
less compact, water will enter more or less according to
the head or pressure, and according to its compactness or
porosity, producing a slope of saturation whose inclination
is dependent on the frictional resistance encountered by
the water. And the bank will be tight whenever the slope of
saturation remains within the figure of the embankment.

Further,

that it was necessary to distinguish between the slope assumed
by water _retained in_ an embankment and that taken by water
_passing through_ an embankment made of material too porous to
retain it; where the rule is clearly reversed and where the more
porous the material the steeper the slope at which water will run
through it at a given rate.

These citations are sufficient to emphasize the importance of exact definition of terms and clear statement of principles.

The latest experiments relating to the percolation of water through earth materials and tests determining the stability of soils are those made during the investigations at the New Croton Dam and Jerome Park Reservoir, New York, and those relating to the North Dike of the Wachusett Reservoir, Boston. These are very interesting and instructive, and it is here proposed to discuss the results and conclusions reached in these cases, after some introductory remarks reciting the order of events.

NEW CROTON DAM.–In June, 1901, the Board of Croton Aqueduct Commissioners of New York requested a board of expert engineers, consisting of Messrs. J. J. R. Croes, E. F. Smith and E. Sweet, to examine the plans for the construction of the earth portion of the New Croton Dam, and also the core wall and embankment of the Jerome Park reservoir.

This report was published in full in Engineering News for Nov. 28, 1901. It was followed in subsequent issues of the said journal by supplemental and individual reports from each member of the board of experts, and by articles from Messrs. A. Fteley, who originally designed the works, A. Craven, formerly division engineer on this work, and W. R. Hill, at that time chief engineer of the Croton Aqueduct Commission.

After describing the New Croton Dam, the board of experts preface their remarks on the earth embankment by saying that

it has been abundantly proven that up to a height of 60
or 70 ft. an embankment founded on solid material and
constructed of well-selected earth, properly put in place,
is fully as durable and safe as a masonry wall and far less
costly.

There are, in fact, no less than 22 earth dams in use to-day exceeding 90 ft. in height, and twice that number over 70 ft. in height. Five of the former are in California, and several of these have been in use over 25 years. The writer fails to appreciate the reason for limiting the safe height of earth dams to 60 or 70 ft.

The New Croton Dam was designed as a composite structure of masonry and earth, crossing the Croton Valley at a point three miles from the Hudson River. The earth portion was to join the masonry portion at a point where the latter was 195 ft. high from the bed rock. The Board thought there was no precedent for such a design and no necessity for this form of construction. The point to be considered here was whether a dam like this can be made sufficiently impermeable to water to prevent the outer slope from becoming saturated and thus liable to slide and be washed out.

The design of the embankment portion was similar to all the earth dams of the Croton Valley. In the center is built a wall of rubble masonry, generally founded upon solid rock, and “intended to prevent the free seepage of water, but not heavy enough to act alone as a retaining wall for either water or earth.”

Fig. 17 shows a section which is typical of most New England earth dams; and Fig. 18, the sections of two of the Croton Valley dams, New York water supply. These dams all have masonry core walls, illustrating the third type of dams given on page 33.

The board of experts made numerous tests by means of borings into the Croton Valley dams to determine the slope of saturation. The hydraulic laboratory of Cornell University also made tests of the permeability of several samples of materials taken from pits. All the materials examined were found to be permeable and when exposed to water to disintegrate and assume a flat slope, the surface of which was described as “slimy.”

Pipe wells were driven at different places into the dams and the line of saturation was determined by noting the elevations at which the water stood in them. In all the dams the entire bank on the water side of the core wall appeared to be completely saturated. Water was also found to be standing in the embankment on the down-stream side of the core wall. The extent of saturation of the outer bank varied greatly, due to the difference in materials, the care taken in building them, and their ages. Fig. 19 gives the average slopes of saturation as determined by these borings.

The experts stated

that the slope of the surface of the saturation in the bank
is determined by the solidity of the embankment: The more
compact the material of which the bank is built, the steeper
will be the slope of saturation.

As a result of their investigations, the experts were of the opinion that the slope of saturation in the best embankments made of the material found in the Croton Valley is about 35 ft. per 100 ft., and that with materials less carefully selected and placed the slope may be 20 ft. per 100 ft.

Further, that taking the loss of head in passing through the core wall, and the slope assumed by the plane of saturation, the maximum safe height of an earth dam with its top 20 ft. above water level in the reservoir and its outside slope 2 on 1, is 63 to 102.5 ft. This is a remarkable finding in view of the fact that the Titicus Dam, one of the Croton Valley dams examined, has a maximum height above bed rock of 110 ft. and has been in use seven years. This dam is not a fair example to cite in proof of their conclusion, because its _effective head_ is only about 46 ft.[2]

FIG. 18.–CROSS-SECTION OF TWO CROTON VALLEY DAMS, SHOWING SATURATION.]

Mr. Fteley gave as a reason for the elevation of the water slope found in the outer bank of the Croton dams the fact of their being constructed of fine materials and stated that with comparatively porous materials they would have shown steeper slopes of saturation.

Mr. Craven argued that all dams will absorb more or less water, and that porosity is merely a degree of compactness; that slope implies motion in water, and that there is no absolute retention of water in the outer bank of a dam having its base below the plane indicated by the loss of head in passing through the inner bank and then through a further obstruction of either masonry or puddle; that there is simply a partial retention, with motion through the bank governed by the degree of porosity of the material.

Fig. 19 is a graphical interpretation of the conclusion reached by the board of experts, as already given on page 41. “A” is an ideal profile of a homogeneous dam with the inner slope 3 on 1 and the outer slope 2 on 1. The top width is made 25 ft. for a dam having 90 ft. effective head, the high-water surface in the reservoir being 10 ft. below the crest of the dam. This ideal profile is a fair average of all the earth dams of the world. Not having a core wall to augment the loss of head, it fairly represents what might be expected of such a dam built of Croton Valley material, compacted in the usual way. It should be noted that the intersection of the plane of saturation with the rear slope of the dam at such high elevation as shown indicates an excessive seepage and a dangerously unstable condition.

Preliminary Study of Profile for Dam.

The preliminary calculations for designing a profile for an earth dam are simple and will here be illustrated by an example. Let us assume the following values:

a. Central height of dam, 100 ft.

b. Maximum depth of water, 90 ft., with surface 10 ft. below
crest of dam.

c. Effective head, 90 ft.

d. Weight of water, 62.5 lbs. per cu. ft.

e. Weight of material, 125 lbs. per cu. ft.

f. Coefficient of friction, 1.00, or equal to the weight.

g. Factor of safety against sliding, 10.

The width corresponding to the vertical pressure of 1 ft. is,
(62.5 × 10)/125 = 5 ft.

The hydrostatic pressure per square foot at 90 ft. depth is, 62.5 × 90 = 5,625 lbs.

The dam, having a factor of safety of 10, must present a resistance of, 5,625 × 10 = 56,250 lbs., or 28 tons per square foot.

The theoretical width of bank corresponding to 90 ft. head and a factor of 10 is shown by the dotted triangle (A-B-B) to be 450 ft., (B, Fig. 19) with slopes 2½ on 1.

To this must be added the width due to the height of crest above the water surface in the reservoir and the width of crest.

The former would be, 2 (2½ × 10) = 50 ft., and the latter by Trautwine’s rule, 2 + 2√100 = 22 ft., giving a total base width of 522 ft.

Let us now assume that the slope of saturation may be 35 ft. per 100 ft. We observe that this intersects the base 40 ft. within the outer toe of the bank slope. If the plane of saturation was 33 ft. per 100, it would just reach the outer toe. It would be advisable to enlarge this section by adding a 10-ft. berm at the 50-ft. level, having a slope not less than 3 on 1 for the up-stream face, and two 15-ft. berms on the down-stream face, having slopes 2½ on 1. The additional width of base due to these modifications in our profile amounts to 65 ft., giving a total base width of 587 ft., and increasing the factor of safety from 10 to 13. It should be remembered that if the bank becomes saturated this factor of safety may be reduced 50%, the coefficient of moist clay being 0.50.

The loss of head due to a core wall of masonry, as designed for the New Croton Dam, was assumed by the board of experts to be 21 ft., or 17% of the depth of water in full reservoir. It has been stated by several authorities that the primary object of a masonry core wall is to afford a water-tight cut-off to any water of percolation which may reach it through the upper half of the embankment. It appears that absolute water-tightness in the core wall is not obtained, although the core walls of the Croton dams are said to be “the very best quality of rubble masonry that can be made.”

Mr. W. W. Follett, who is reported to have had considerable experience in building earth dams, and who has made some valuable suggestions thereupon, is emphatic in saying,

that the junction of earth and masonry forms a weak point, that
either a puddle or masonry core in an earthen dam is an element
of weakness rather than strength.

He also thinks the usual manner of segregating and depositing materials different in density and weight, and thus subject to different amounts of settlement, as bad a form of construction as could be devised.

Core walls may prevent “free passage of water” and “excessive seepage,” but are nevertheless of doubtful expediency.

Earthwork Slips and Drainage.

Mr. John Newman, in his admirable treatise on “Earthwork Slips and Subsidences upon Public Works,” classifies and enumerates slips as follows:

Natural causes, 7.
Artificial causes, 31.
Additional causes due to impounded water, 7.

After describing each cause he presents 39 different means used to prevent such slips and describes methods of making repairs.

Mr. Wm. L. Strange has had such a large and valuable experience and has set forth so carefully and lucidly both the principles and practice of earth dam construction, that the writer takes pleasure in again quoting him on the subject of _drainage_, of which he is an ardent advocate. He says that,

thorough drainage of the base of a dam is a matter of vital
necessity, for notwithstanding all precautions, some water will
certainly pass through the puddle.

It is at the junction of the dam with the ground that the maximum amount of leakage may be expected. The percolating water should be gotten out as quickly as possible. The whole method of dealing with slips may be summed up in one word–_drainage_.

The proper presentation of these two phases of our subject would in itself require a volume. The interested reader is therefore referred to the different authorities and writers cited in Appendix II.

Jerome Park Reservoir Embankments.

The Jerome Park reservoir is an artificial basin involving the excavation and removal of large quantities of soil, and the erection of long embankments with masonry core walls, partly founded on rock and partly on sand. The plan and specifications call for an embankment 20 ft. wide on top, with both slopes 2 on 1, and provide for lining the inner slope with brick or stone laid in concrete, and for covering the bottom with concrete laid on good earth compacted by rolling.

FIG. 20.–GRAPHICAL EXHIBIT OF STUDIES OF JEROME PARK RESERVOIR EMBANKMENT.]

Wherever bed rock was not considered too deep below the surface the core walls were built upon it. In other places the foundation was placed 8 to 10 ft. below the bottom of the reservoir and rested upon the sand.

It appears that the plans of the Jerome Park embankment were changed from their original design, prior to the report of the board of experts, on account of two alleged defects, namely, “cracks in the core wall” and “foundation of quicksand,” and incidentally on account of the supposed instability of the inner bank.

In describing the materials on which these embankments rest the experts remarked

that all these fine sands are unstable when mechanically
agitated in an excess of water, and that they all settle in
a firm and compact mass under the water when the agitation
ceases. That they are quite unlike the true quicksands whose
particles are of impalpable fineness and which are “quick”
or unstable under water.

Fig. 20 is a graphic exhibit of the results of tests made at “Station 76 + 20,” and at “Station 99,” to determine the flow line of water in the sand strata underlying the embankment and bottom of the Jerome Park reservoir.

The experts reported that there was no possible danger of sliding or sloughing of the bank; that the utmost that could be expected would be the percolation of a small amount of water through the embankment and the earth; and that this would be carried off by the sewers in the adjacent avenues; that a large expenditure to prevent such seepage would not be warranted nor advisable.

In concluding their report, however, they recommended changing the inner slope of 2 on 1 to 2½ on 1, and doubling the thickness of the concrete lining at the foot of the slope to preclude all possibility of the sliding or the slipping of the inner bank in case of the water being lowered rapidly in the reservoir.

Mr. W. R. Hill, then chief engineer of the Croton Aqueduct Commission, favored extending the core walls to solid rock. He took exception to the manner of obtaining samples of sand by means of pipe and force-jet of water, claiming that only the coarsest sand was obtained for examination. He did not consider fine sand through which three men could run a ¾-in. rod 19 and 20 ft. to rock without use of a hammer, very stable material upon which to build a wall.

North Dike of the Wachusett Reservoir, Boston.

The North Dike of the Wachusett Reservoir is another large public work in progress at the present time. It is of somewhat unusual design and the preliminary investigations and experiments which led to its adoption are interesting in the extreme.[3]

The area to be explored in determining the best location for the dike was great, and the preliminary investigations conducted by means of wash drill borings, very extensive. A total of 1,131 borings were made to an average depth of 83 ft., the maximum depth being 286 ft. The materials were classified largely by the appearance of the samples, though chemical and filtration tests were also made. The plane of the ground water was from 35 to 50 ft. below the surface, and the action of the water-jet indicated in a measure the degree of permeability of the strata.

In addition to these tests experimental dikes of different materials, and deposited in different ways, were made in a wooden tank 6 ft. wide, 8 ft. high and 60 ft. long. The stability of soils when in contact with water was experimented with, as shown in Fig. 21, in the following manner:

An embankment (Fig. 21) was constructed in the tank of the material to be experimented with, 2 ft. wide on top, 6 ft. high, with slopes 2 on 1, and water admitted on both sides to a depth of 5 ft. The top was covered with 4-in. planks 2 ft. long and pressure applied by means of two jack screws resting upon a cross beam on top of the planks.

With a pressure of three tons per square foot, the 4-in. planks were forced down into the embankment a little more than 6 ins., resulting in a very slight bulging of the slopes a little below the water level. Immediately under the planks the soil became hard and compact. A man’s weight pushed a sharp steel rod, ¾-in. in diameter, only 6 to 8 ins. into the embankment where the pressure was applied, while outside of this area the rod was easily pushed to the bottom of the tank.

These results corroborate in a general way the practical experience of the author, both in compressed embankments, where he found it necessary to use a pick vigorously to loosen the material of which they were composed, and in embankments made by merely dumping the material from a track, in which case the earth is so slightly compressed that an excavation is easily made with a shovel.

FIGS. 21 TO 24.–EXPERIMENTAL DIKES AND CYLINDER EMPLOYED IN STUDIES FOR THE NORTH DIKE OF THE WACHUSETT RESERVOIR; AND (FIG. 25) CROSS-SECTION OF THE DIKE.]

The difference in the coefficient of friction of the same material when dry and when wet greatly modifies the form of slope. The harder and looser the particles, the _straighter_ will be the slope line in excavation and slips. The greater the cohesion of the earth, the _more curved_ will be the slope, assuming a parabolic curve near the top–the true form of equilibrium.

RATE OF FILTRATION.–The rate of filtration through different soils was experimented with by forming a dike in the tank previously mentioned, as shown in Fig. 22.

The dike was made full 8 ft. high, 7 ft. wide on top, with a slope on the up-stream side of 2 on 1, and on the down-stream side 4 on 1. This gave a base width of 55 ft. Immediately over the top of the dike there was placed 3 ft. of soil to slightly consolidate the top of the bank and permit the filling of the tank to the top without overflowing the dike. The water pressure in different parts of the dike was determined by placing horizontal pipes through the soil crosswise of the tank. These pipes were perforated and covered with wire gauze, being connected to vertical glass tubes at their ends. The end of the slope on the down-stream side terminated in a box having perforated sides and filled with gravel, thus enabling the water to percolate and filter out of the bank without carrying the soil with it.

When the soil was shoveled loosely into the tank, without consolidation of any kind, it settled on becoming saturated and became quite compact. It took five days for the water to appear in the sixth gauge pipe near the lower end of the tank. After the pressure, which was maintained constant, had been on for several weeks, the seepage amounted to one gallon in 22 minutes. When the soil was deposited by shoveling into the water, the seepage amounted to one gallon in 34 minutes.

The relative filtering capacities of soils and sands were thought to be better determined by the use of galvanized iron cylinders of known areas.

Fig. 23 shows one of the cylinders. These latter experiments confirmed those previously made at Lawrence, by Mr. Allen Hazen, for the Massachusetts State Board of Health. They showed that the loss of head was directly proportioned to the quantity of water filtered and that the quantity filtered will vary as the square of the diameter of the _effective size_ of the grains of the filtering material.[4]

The material classed as “permeable” at the North Dike of the Wachusett Reservoir has an effective diameter of about 0.20 mm. A few results are given in the following table:

Amount of Filtration in Gallons per Day, Through an Area of 10,000 Sq. Ft., With a Loss of Head or Slope of 1 ft. in 10 ft.

Material. Unit ratios. U. S. gallons.

(1) Soil 1 510
(2) Very fine sand 14 7,200
(3) Fine sand 176 90,000
(4) Medium sand 784 400,000
(5) Coarse sand 4,353 2,200,000

To be sure that the accumulation of air in the small interstices of the _soil_ was not the cause of the greatly reduced filtration through it, another series of experiments was conducted in the wooden tank, as shown in Fig. 24.

A pair of screens was placed near each end of the tank, filled with porous material, sand and gravel, and the 50-ft. space between filled with soil. The soil was rammed in 3-in. layers, and special care taken to prevent water from following along the sides and bottom of the tank. One end was filled with water to near the top, while the other end gave a free outlet.

After this experiment had been continued for more than a month, the amount of seepage averaged 1.7 gallons per 24 hours, or about 32 drops per minute.

Filtration tests were also made through soil under 150 ft. head, or 5 lbs. per sq. in., with results not materially different, it is stated, from those already given. The soil used in all these tests contained from 4 to 8% by weight of organic matter. This was burned and similar tests made with the incinerated soil, resulting in an increase of about 20% more seepage water.

PERMANENCE OF SOILS.–This last material experimented with suggests the subject of _permanence_ of soils. This was reported upon separately and independently by Mr. Allen Hazen and Prof. W. O. Crosby. These experts agreed in their conclusion, stating

that the process of oxidation below the line of saturation would be
extremely slow, requiring many thousands of years for the complete
removal of all the organic matter, and that the tightness of the
bank would not be materially affected by any changes which are
likely to occur.

It has been remarked,

that of all the materials used in the construction of
dams, _earth_ is physically the least destructible of
any. The other materials are all subject to more or less
disintegration, or change in one form or another, and in
earth they reach their ultimate and most lasting form.

In speaking of the North Dike of the Wachusett Reservoir, Mr. Stearns remarked that,

it was evident by the application of Mr. Hazen’s formula for
the flow of water through sands and gravels, that the very
fine sands found at a considerable depth below the surface
would not permit enough water to pass through them if a dike
of great width were constructed, to cause a serious loss of
water, and it was also found that the soil, which contained
not only the fine particles or organic matter, but also a
very considerable amount of fine comminuted particles, which
the geologist has termed “rock flour,” would be sufficiently
impermeable to be used as a substitute for clay puddle.

Fig. 25 shows the maximum section of the North Dike with its cut-off trench. The quantities and estimated cost of the completed structure are given in the table herewith:

|––– Cost –––|
Per cent.
Work. Quantities. Unit Actual. total.
(cu. yds.) Price.
Soil 5,250,000 $0.05 $262,500 34.7
Cut-off trench 542,000 .20 108,400 19.3
Borrowed
earth and gravel 200,000 .20 40,000
Slope paving 50,000 2.20 110,000 14.6
Sheet-piling,
pumping, etc. 117,000 15.5
Engineering and
preliminary investigations 120,000 15.9
––––––– –––––
Total cost $757,900 100.0

Druid Lake Dam, Baltimore, Md.

Another very interesting and instructive example of high earth dam construction is that of the Druid Lake Reservoir embankment, Baltimore, Md.

This dam was built under the supervision of Mr. Robt. K. Martin. Construction was begun in 1864, and the dam was finished in 1870. Mr. Alfred M. Quick, present chief engineer of the water-works of the City of Baltimore has given a very lucid description of this work in Engineering News of Feb. 20, 1902.

Fig. 26 is a cross-section of this dam, showing the method of construction so clearly as to scarcely need further description. The banks D-D on either side of the central puddle wall were carried up in 6-in. layers with horses and carts, and kept about 2 ft. higher than the puddle trench, which always contained water. The banks E-E were made of dumped material, after which the basins F-F were first filled with water and finally filled by dumping material into the water from tracks being moved in toward the center.

After reaching the top of this fill, banks B-B-B were built up in layers similar to D-D. The second set of basins C-C were then filled in a manner similar to F-F. The remaining portion A-A was constructed in layers like D-D and B-B, with the addition of compacting each layer with a heavy roller.

Finally the inner face slope was carried up in 3-in. layers and thoroughly rolled, after which 2 ft. of “good puddle” was put upon the inner slope the latter was rip-rapped, the crown covered with gravel and the rear slope sodded.

Some years after completion, a driveway was built along the outer slope, as shown, which had a tendency to strengthen the dam, though not designed expressly for that purpose.

It is of interest to know that the influent, effluent and drain pipes were originally constructed through or under the embankment. These pipes were laid upon solid earth, and where they passed through the puddle wall were supported upon stone piers 6 ft. apart. As might be expected, they soon cracked badly and were finally abandoned, new ones being placed in the original ground at the south side of the lake. Mr. Quick states that so far as is known there has never been any evidence of a leak through the embankment during these 32 years of service.

New Types of Dams; Bohio, Panama Canal.

A brief description will now be given of three different dams designed for Bohio, on the proposed Panama Canal. Mr. George S. Morison’s paper before the American Society of Civil Engineers, on “The Bohio Dam,” and the discussion thereon, especially that by Mr. F. P. Stearns, were quite fully reported in Engineering News for March 13 and May 8, 1902. In constructing the Panama Canal it will be necessary to impound the waters of the Chagres River, near Bohio, to maintain the summit level of this canal and supply water for lockage.

THE FRENCH DESIGN.–Fig. 27 is an enlarged section of the original design of the new French Co. This design has no core wall, but at the up-stream toe a concrete wall was to be built across the river between the two lines of sheet-piling. At the down-stream toe a large amount of riprap was to be placed to prevent destruction of the dam during construction. In this case it would be necessary to construct a temporary dam above and also to use the excavation for the locks as a flood spillway. This method would involve considerable risk to the work, on account of the large volume of flood waters it might be necessary to take care of during construction.

ISTHMIAN CANAL COMMISSION.–The dam proposed by the Isthmian Canal Commission is shown by Fig. 28. This was designed to be an absolutely water-tight closure of the geological valley, by using a masonry core wall carried down to bed rock. The maximum depth being 129 ft., it was planned to rest the concrete wall on a series of pneumatic caissons reaching to rock. The spaces between the caissons would be closed and made water-tight. Both slopes of the earth embankment were to have horizontal benches and be revetted with loose rock.

MR. MORISON’S DESIGN.–To appreciate fully the object and aim of the third design, Fig. 29, which may be called a new type, although similar in many respects to the North Dike of the Wachusett reservoir already illustrated and described, it should be stated that the equalized flow of the Chagres River is put at 1,000 cu. ft. per sec. Of this quantity it is estimated that 500 cu. ft. would be needed for lockage and 200 cu. ft. for evaporation. This leaves 300 cu. ft. per sec. available for seepage and other losses or to be wasted.

It will thus be seen that a scarcity of water is not in this instance a condition demanding an absolutely water-tight dam. The amount of seepage permissible without endangering the stability of the structure is the real point now to be discussed.

The third design, which was proposed by Mr. Morison, is shown by Fig. 29. The topography and configuration of this dam site is not unlike that of the San Leandro Dam, California, soon to be described, while the general design is similar, as has been remarked, to the North Dike of the Wachusett Reservoir.

This third design contemplates a compound structure, formed by two rock-fill dams situated about 2,120 ft. apart, with the intervening space filled with loose rock, earth and other available material. Immediately below the upper and higher rock-fill dam, it is proposed to place across the canyon a puddle wall 50 ft. in width, resting over two lines of sheet-piling 30 ft. apart. This piling would probably not reach farther than 50 ft. below tidewater, the solid rock floor being about 100 ft. deeper.

Mr. Morison made use of Mr. Hazen’s filtration formula for estimating the rate and quantity of seepage through the permeable strata below the dam. This formula is:

h t + 10°
V = cd² –– ––––––
l 60

where

V = rate of flow in meters per day through the whole section.
c = constant varying from 450 to 1,200,
according to cleanness of the sand.
d = “effective size” of sand in mm.
h = head in feet.
l = length or distance water must pass.
t = temperature of the water (Fahr.)

This formula should be used only when the _effective sizes_ of sands are from 0.10 to 3.0 mm. and with _uniformity coefficients_ below 5.0[5].

Mr. Morison used the following values: c = 1,000; d = 1.0 mm.; h = 90 ft.; l = 2,500 ft.; t = 90°; for the solution of this problem, and obtained a velocity of 0.002 ft. per sec. The bed of sand and gravel was assumed to have a sectional area of 20,000 sq. ft. for 2,500 ft. in length. This gives a seepage of 40 cu. ft. per sec.

It is believed that the above rate of 0.002 ft. per sec., equivalent to 1⅜ ins. per minute, or 7 ft. per hour, is not sufficient to move any of the material. The velocity of water percolating through sand is found to vary directly as the head and inversely as the distance.

The value of “c” in the formula is larger for sands of filters favorable for flow, and smaller for compacted materials and dams.

Mr. Morison thought it might be nearer the actual conditions to assume d = 0.50 mm.; c = 500; and l = 5,000 ft.; in which case the seepage would only amount to 2.5 ft. per sec. In this last assumption the “effective size” of sand grains is 2½ times that classed as “permeable material” at the North Dike of the Wachusett Reservoir.

Prof. Philipp Forchheimer, of Gratz, Austria, recommends the use of the formula,

h
––– = a√ + b√²
l

for the percolation through soils between loam and loamy sand. Sellheim, Masoni, Smreker, Kröber and other authorities on filtration use still other formulas, to which the reader and student is referred for further research.

The writer, having had occasion in his professional practice to study quite carefully the subject of ground waters, and their percolation or flow through different classes of materials and under varying conditions, is of the opinion that rarely does the cross-section of a stream-channel, filled with sand, gravel and debris, present, even approximately, a homogeneous or uniform mass; and that there are, almost without exception, strata of material much coarser and more porous than the general average. In other words, that it is extremely difficult to arrive at a uniformity coefficient. It is unwise to place much reliance upon an estimated flow where this is the case. The formula may be used with confidence where the layers are artificially made, and where there is no uncertainty regarding the uniform character of the material. In most natural channels there are distinct lines of flow, and under considerable hydrostatic head or pressure these lines of flow would surely enlarge. There is a wide difference between permissible and dangerously excessive percolation through an earth embankment. The local features, economical considerations and magnitude of the risks, all bear upon this question and must be considered for each particular case.

It is of interest to compare the estimated cost of the three designs proposed for the Bohio Dam, based upon the same unit prices, as follows:

French Engineers’ design $3,500,000
Isthmian Canal Commissioners’ design 8,000,000
Mr. Morison’s design 2,500,000

No comments will be made upon these figures, further than to remark that the successful building of a stable dam, accomplished by the use of an excessive quantity of materials and at a cost beyond reasonable requirements, is mainly instructive as illustrating “how not to do it.” It is creditable to execute substantial works at a reasonable cost, but it reflects no credit upon any one to construct them regardless of expense.

Combined Rock-fill and Earth Dam.

Fig. 30 shows a section of the Upper Pecos River Dam near Eddy, N. M.

This dam is quite fully described by Mr. Jas. D. Schuyler, in his recent book on “Reservoirs for Irrigation, Water-Power and Domestic Water-Supply,” and need not be mentioned in this paper, further than to call attention to the combination of rock-fill and earth which constitutes its particular type of construction. This type of dam is believed to be for many localities a very good one, but up to the present time has only been adopted for dams of moderate height, under 60 ft.

The San Leandro Dam, California.

A section of the San Leandro Dam, near Oakland, Cal., is shown by Fig. 31. This section was supplied by Mr. W. F. Boardman, hydraulic engineer, who superintended the construction of the dam, from his own private notes and data. It differs materially from sections heretofore published, and is 5 ft. higher, thus making it rank as the highest earth dam in the world of which we have an authentic record.

The dam was commenced in 1874, and brought up to a height of 115 ft. above the bed of the creek in 1898. At the present time it is 500 ft. in length on the crest and 28 ft. wide. The original width of the ravine at the base of the dam was 66 ft. The present width of base from toe to toe of slopes is 1,700 ft. The height of embankment above the original surface is 125 ft., with a puddle trench extending 30 ft. below.

All that portion of the dam within a slope of 2½ on 1 at the rear and 3 on 1 at the face is built of choice material, carefully selected and put in with great care. The portion outside of the 2½ on 1 slope line at the down-stream side of the dam, was _sluice in_ from the adjacent hills regardless of its character, and is composed of ordinary soil containing more or less rock.

This process of sluicing was carried on during the rainy season, when there was an abundance of water, and it was intended to be continued until the canyon below the dam had been filled to an average slope of 6.7 on 1 at the rear of the dam. It was thought that the location was particularly favorable for this kind of construction, the original intention being to raise the dam from time to time, not only to increase the storage as the demand for water increased, but to meet the annual loss in capacity caused by the silting up of the reservoir basin. The latter has amounted to about 1 ft. in depth per annum.

METHOD OF CONSTRUCTION.–Under the main body of the dam, the surface was stripped of all sediment, sand, gravel and vegetable matter. Choice material, carefully selected, was then brought by carts and wagons and evenly distributed over the surface in layers about 1 ft. or less in thickness. This was sprinkled with just enough water to make it pack well, not enough to make it like mud. During construction a band of horses was led by a boy on horseback over the entire work, to compact the materials and assist in making the dam one homogeneous mass. No rollers were used on this dam.

The central trench was cut 30 ft. below the original bed of the creek. In the bottom of this trench three secondary trenches, 3 ft. wide by 3 ft. deep, were made and filled with concrete. These concrete walls were carried up 2 ft. above the general floor of the trench, to break the continuity of its surface.

The original wasteway, constructed at the north end of the dam, has been practically abandoned, having been substituted by a tunnel of larger capacity. The original wasteway was excavated in the bed rock of the natural hillside, and although lined with masonry, is not in the best condition. The author considers its location an objectionable feature, as menacing the safety of the dam, and thinks it should be permanently closed.

A wasteway tunnel, 1,487 ft. in length, was constructed in 1888, through a ridge extending north of the dam. This has a sectional area of about 10×10 ft., lined with brick masonry throughout, having a grade of 2½%.

The criticism might be made of the tunnel that it is faulty in design at the entry or reservoir end, where the water must first fall over a high spillway wall, aerating the water before entering the tunnel proper. The water even then has not easy access to the tunnel, and no adequate arrangements have been made for ventilation, so as to insure the utilization of its maximum capacity. The maximum depth of water in the reservoir is about 85 ft., and the full capacity 689,000,000 cu. ft. of water. The catchment area is 43 square miles, and the surface of the reservoir when full 436 acres. The outlet pipes are placed in two tunnels at different elevations through the ridge north of the dam. There are no culverts or pipes extending through the body of the dam itself.

Hydraulic-fill Dams.

No discussion of earth dams would be complete without some reference being made to the novel type of construction developed in western America in recent years, by which railroad embankments and water-tight dams are built up by the sole agency of water. The water for this purpose is usually delivered under high pressure, as it is generally convenient to make it first perform the work of loosening the earth and rock in the borrow pit, as well as subsequently to transport them to the embankment, and there to sort and deposit them and finally part company with them after compacting them solidly in place, even more firmly than if compressed by heavy rollers. Sometimes, however, water is delivered to the borrow pit without pressure, in which event the materials must be loosened by the plow or by pick and shovel by the process called ground sluicing in placer mining parlance.

An abundance of water delivered by gravity under high pressure is usually regarded as one of the essential factors in hydraulic-fill dam building, but it is not essential that there be a large continuous flow. The Lake Frances Dam, recently constructed for the Bay Counties Co., of California, by J. D. Schuyler, is 75 ft. high, 1,340 ft. long on top, and contains 280,000 cu. yds. The dam was built up by materials sluiced by water that was forced by a centrifugal pump through a 12-in. pipe and 3-in. nozzle, against a high bank, whence the materials were torn and conveyed by the water through flumes and pipes to the dam. About 6 cu. ft. per sec. of water was thus used, and at one stage of the work the supply stream was reduced to less than 0.1 ft. per sec., the water being gathered in a pond and pumped over and over again.

The chapter on hydraulic-fill dams in Mr. Schuyler’s book on “Reservoirs for Irrigation” will be found to contain matter on the subject interesting to those who desire to pursue it further, and the reader is again referred to that work.

An Impervious Diaphragm in Earth Dams.

As a result of the recent extended discussion concerning the design of the New Croton Dam and the Jerome Park Reservoir embankments, the Engineering News of Feb. 20, 1902, contained a very suggestive editorial entitled, “Concerning the Design of Earth Dams and Reservoir Embankments.” The opinion is given that no type of structure that man builds to confine water can compare in permanence with earth dams, after which the following pertinent questions are asked:

1. How shall an earth dam be made water-tight?

2. What is the office and purpose of the masonry core wall?

3. Would not a water-proof diaphragm of some kind be better
than a core wall of either masonry or puddle?

The article then suggests a number of designs of diaphragm construction, with a special view of obtaining absolute water-tightness, by use of asphaltum, cement mortar, steel plates, etc. Special emphasis was put upon the _principle_ of constructing a water-proof diaphragm. The matter of relative cost is advanced as an argument in favor of the diaphragm principle as against the usual orthodox method. The saving in cost is to be accomplished by the use of inferior materials and less care in the handling of them, or by both. It is suggested that almost any kind of material available, rock, sand or gravel, will answer every purpose where good earth is not to be found. Further, that this material may be dumped from the carts, cars or cableways, or be placed by the hydraulic-fill method.

The writer believes the diaphragm method of construction may have some merits, but that it is attended by the very great risk of neglecting principles most vitally important to the successful construction of high earth dams, which will now be formulated and advanced, as follows: