American Society of Civil Engineers.

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

. (page 67 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 67 of 167)
Font size
QR-code for this ebook


table largely to the thoroughness with which the preliminary tests
were conducted to determine, first, the feasibility of putting in such
a system, and, finally, to determine the best location for the plant.

Briefly stated, the method of conducting the preliminary tests in-
cluded the sampling of the underflow of the river, as obtained by
driving small well points to depths ranging from 3^ to 5 ft. beneath
the surface of the bar. Water was pumped from these well points for
such a time as was required to obtain a clear sample which was
finally analyzed for the purpose of determining its chemical constit-
uents. It was recognized that two imjwrtant features must be con-
sidered in selecting the location for the system, namely, that the
underground water must not contain an excessive quantity of iron, and
that it must not be excessively hard, as compared with the river water.

The first preliminary tests were made at Parkersburg on August
6th and 7th, 1909. At this time only four tests were made, at stations
extending from about 5 000 ft. to 1 000 ft. above the city pumping
station. The results of the analyses of the samples collected at this
time are shown in Table 1.



discussion: water supply of parkersbueg, w. va.



815



TABLE 1. — Eesults op Tests of Water Obtained from Beneath Mr.
THE Bottom op the Ohio Eiver^ by Driving Well Points, at ^"'■^^®^-
Parkersburg, W. Ya.

(Eesults in parts per million.)



Station No

Depth of sample below surface J
of bar I

Sample No

Date of collection

Time

Temperature, in degrees, centi-
grade

Turbidity

Free CO2

Dissolved oxygen

Chlorine

Iron (Fe)

Alkalinity

Incrustants

Total hardness

Station No

Depth of sample below surface
of bar

Sample No

Date of collection

Time

Temperature, in degrees, centi-
grade

Turbidity

Free CO2

Dissolved oxygen

Chlorine

Iron (Fe)

Alkalinity

Incrustants

Total hardness



1

Surface of

river

1

8/6

3 P. M.

28.0
11



9.0



1

!• 2.5 ft. ]
2

8/6
4.80 P. M.

23.0

5
23

2.0

5!6
174



Surface of

river

3

8/7
8.30 a. m.

26.0
17


7.7

31

0.5

17

100

117



2

3.0 ft.

4

8/7
8.45 A. M.

23.0

11



2

4.0 ft.

5

8/7
9.30 a. m.

23.0


18

1.8
14

6.0
58
20
78



2.0 ft.
6

8/7
11.45 A. M.

23.0


30

1.0

8

0.5
44
15
59



3.0 ft.
7

8/7
12.30 A. M.

23.0


30

1.0

7

0.9
42

5
47



2.0 ft.

8

8/7

1.55 P. M.

25.0


26

2.4
10

1.8
75


75



5
2.0 ft.



8/7
ZM P. M.



5
9

if
1.0

136
25
161



The analyses indicated considerable differences in the quality of
the water beneath the bar at the different stations. At Station 1,
which was near the foot of Neals Island and between the two channels
of the river, the ground-water was found to be comparatively hard
and of high iron content. This was true also at Station 4, which was
1 000 ft. above the city pumping station. At Station 3, the ground-
water obtained from beneath the river was softer than the river water
and of low iron content. Near this station, on August 8th, a hole,
about 15 ft. in diameter, was dredged in the river bar to a depth of
5 ft. and, in this hole, was placed an 8-in. brass strainer, 8 ft. long,
in a horizontal position. The strainer pipe was covered with 18 in.
of gravel and about 42 in. of sand. Water was pumped from the
strainer by a 2-in. centrifugal pump for the purpose of determining
the quality of the water collected in the strainer pipe. Eepresentative
analyses of the river water and that obtained from the strainer are
shown in Table 2.



Mr.
Burgess,



816 discussion: water supply of parkersburg, w. va.

TABLE 2. — Preliminary Tests of Smith System at Parkersburg,

W. Va., August 10th-15th, 1909.

(Results ill parts per million.)



Source of samples

Sample No .

Date of collection

Time

Temperature, in degrees,
centigrade

Turbidity

FreeCOa

Dissolved oxygen

Chlorine

Iron (Fe)

Alkalinity

Incrustants

Total hardness

Total solids

Loss on ignition

Nitrogen, as nitrites

" as nitrates

Oxygen consumed

Bacteria, per cubic centi-
meter

Coli present in



River

10

8/10
10.30 A. M.

26.0
15



56
1.2

20
110
130



Strainer
11

8/10
10.30 A. M.

25.0




24

0.5
24

85
109



River

12

8/15
9.30 A. M.

24.0
1750

2.5

7.0
21

15.0
42
50
92
749
58

0.010



9.3

8 700

0.05 c.c.



Strainer
13

8/15

9.30 A. M.

23.0


23

3.2
17

1.2
34
15
49
167
28

0.008



1.8

82
0.3 c.c.



In general, the water obtained from the strainer pipe was entirely
clear, even when the river water had a turbidity of nearly 2 000 parts
per million; moreover, the total hardness of the strainer water was
less than that of the river water. It was undoubtedly true that the
success of this preliminary test interested a large number of citizens
and officials at Parkersburg, and showed them the possibility of ob-
taining a satisfactory water supply in this manner.

At the request of Mr. Gray, still further tests were made on May
7th, 1910, to determine the quality of the water in the bar beneath
the river and the feasibility of obtaining a satisfactory supply from
horizontal strainers placed beneath the bed of the river. At this time,
also, by means of a dredgeboat, the material composing the bar was
excavated and examined thoroughly to determine its characteristics
at numerous points in the river.

As was true during the first preliminary tests, considerable dif-
ferences were noted in the character of the water obtained from beneath
the river at different depths beneath the surface of the bar and at
different locations in the bar. The results of the analyses of the
samples collected at this time are shown in Table 3.

After the City had entered into a contract with Mr. Smith to build
his system, still further tests were made by the writer in June and
July, 1910, for the purpose of determining the best location. More
than seventy-five well tests were made at intervals of about 100 ft.
over the river bottom, and a small pile-driver was rigged up on a
small flatboat to assist in driving the well points. The sampling
stations were located by stadia measurements from a base line on



DISCUSSION : WATER SUPPLY OF PARKERSBUEG, W. VA. 817



TABLE 3. — Eesults op Tests of Water Obtained from Beneath Mr.
THE Ohio Eiver, by Driving Well Points, at Parkersburg, ^"'■^®^^-
W. Va., May 7th, 1910.

(Results in parts per million.)



Station

Depth of sample below (^

surface of bar j

Sample No

Time of collection

Temperature, in de- (

?rees, centigrade (

Turbidity

Free CO2

Dissolved oxygen

Chlorine

Iron (total Fe)

Alkalinity

Incrustants

Total hardness

Total solids

Loss on ignition



1


1


1


1


2


Surface
of river


[ 2.0 ft.


3.5 ft.


5.0 ft.


j Surface
1 of river


1


2


3


4





H A.M.


11 A. M.


12.20 P. M.


1 P. M.


4 p. M.


15.5
35


14.0

50 ±


13.3
10 ±


13.3

5 ±




40


1

9.1
15


4


4

2.6

9


5

0.7
10






9


13


4.0


2.0


0.2


0.1


1.5


24


128


142


223


20


23


20


15


10


27


47


148


157


233


47


176


226


203


289


136


68


42


43


17


29



2

[ 4 ft.

6

4:30 p. M

18.3

^^
2.2
13

0.3
40
12

52
169
46



the bank of the river. By this method of testing, it was learned that
the best location for the system proposed was not opposite the Camden
Farm, but several thousand feet below this farm at a point where
the material composing the river bar was very satisfactory and the
underlying water was of low iron content and softer than the river
water. Consequently, this location was selected as best adapted for
the system.

Further studies of the design of the proposed strainer system
indicated the desirability of dividing it into a number of units with
independent suction lines, rather than to make only two units, as
was contemplated in the original plan. Moreover, on account of the
large losses of head which would develop during back-flushing, it was
realized that the pumping station should be at the nearest possible
point to the strainer system. As shown by the author, the strainer
system, as constructed, was composed of five individual units, each
connected by separate 18-in. suction lines to the main suction of the
pumps, and the latter were near the river bank opposite the center
of the strainer system. ,

It is the writer's belief that two important features of construction
have contributed especially to the success of the strainer system,
namely, the provision of a deep bed of clean, coarse gravel over the
entire area covered by the strainer system, and the reduction in the
area of the strainer pipes at their connections to the main distributing
pipes to a diameter of 2^ in. The deep gravel layers have increased
the capacity of the strainer system materially, and also the reduction
in the area of the connections of the strainer pipes has also increased
the . efficiency of back-flushing the system, by introducing a loss of



818



discussion: watek supply of parkersbueg, w. va.



Burgess.



Mr. head at this point, thus causing a better distribution of the wash-
water.

In view of the possible differences of opinion as to the source,
or sources, of the water in the adjacent gravel bed along the Ohio
River, and also in the gravel beneath the river, it may be noted, first,
that the entire river bottom at Parkersburg is covered to a depth of
approximately 18 in. with very hard impervious gravel layers firmly
cemented together. In fact, in some places, during the preliminary
testing, the gravel layers were so tight that it was found impossible
to pump water from the well points after they were driven. This
feature is of interest in considering the recommendations contained
in one of the preliminary reports to use a hydraulic dredge of such
design that it would remove material from the river bar to a depth
of 1 or 2 in. Experience in dredging material from the bottom of
the Ohio shows that it is impossible to use hydraulic dredges in this
river above Portsmouth, Ohio, on account of the compactness of the
surface material.

Moreover, some light as to the probable source, or sources, of
water obtained at shallow depths from beneath the river may appear
from Table 4, showing results of analyses of samples of water collected
by the writer at Sistersville, W. Va., on August 1st, 1910.

TABLE 4. — Result op Analyses of Samples of Water from the

Ohio River at Sistersville, W. Va., August 1st, 1910.

(Results in parts per million.)



stations

Depth of water, in feet

Temperature, in degrees, centigrade

Dissolved oxygen

Total solids

Iron (as Fe)

Incrustants

Alkalinity*

Total hardness

Chlorine



1


2


3


4.7


3.9


3.3


23.5


23.5


23.0











7 300


6 400


6 100


1 100


1 300


900


475


482


475


147


150


156


622


632


631


4 500


4 250


4 500



Ohio River
'26'.0



• Alkalinity. August 3d, 1910, Sample No. 1 = 28 ; No. 2 = 5; No. 3 — 44.
•Samples of ground-water were obtained 4.0 ft. beneath the surface of the bar.

The samples in Table 4 were obtained from the bar at different
stations near the center of the river by exactly the same method as
that used at Parkersburg, and it is of interest to note that, even at
shallow depths beneath the surface of the bar, the ground-water was
of remarkably high mineral content. The samples were clear when
collected, but in 15 min. they looked like tomato ketchup.

The writer has made a large number of other investigations, along
similar lines, to determine the character of the water obtained at
shallow depths beneath river bars, principally at Owensboro, Ky.,
Gallipolis, Ironton, Zanesville, Portsmouth, and Newark, Ohio, . and



DISCUSSIOX : WATER SUPPLY OF PAEKERSBURG, W. VA. 819

Wheeling, W. Ya., and, in all cases, the results obtained have been Mr.
generally similar to those at Parkersburg. The tests have indicated
great variations in the quality of the water at different locations
and at different depths of sampling. The writer's conclusion is that
the underground water, even at shallow depths, is always of very
different quality from the river water, especially during low stages of
the streams; also, that during low stages it cannot have its source
in the river, but must come from adjacent higher territory and flow
toward the river.

An interesting feature was developed at Parkersburg where, in
some cases, the water obtained from the well points was of very high
iron content and contained only 5 parts of incrustants when analyzed
by the usual methods. Its alkalinity was 55 parts. Two or three
such points were included within the area covered by the strainer
system, and excavations disclosed the fact that a log or trunk of
an old tree had been deposited and covered within the river bar.
Immediately around these logs, for distances of about . 10 ft., the
surrounding gravel and sand was filled with iron. When the log and
iron deposits were removed, the quality of the water obtained from
the ground changed, and resembled that flowing from adjacent areas.
Its iron content was reduced to about 0.3 part per million, the incrust-
ants increased to about 70 parts, and the alkalinity was reduced to
about 40 parts. The writer has no explanation for this phenomenon.

In that part of the report of Messrs. Fuertes and Fuller which
discusses the quality of the ground-water obtained from the test
wells it is stated that, in some instances, the ground-water contained
objectionable quantities of iron "due to the water dissolving iron
during the long intervals of standing in contact with the metal;"
and, again, that the large quantity of iron in the well water at the
steel plant is explained "partly by the flow of water through a long
line of pipe." The writer believes that these statements should not
go unquestioned, because undoubtedly the high iron content of the
ground-water was not attributable to its contact with iron pipes, but
the iron was present in the water in the gravel strata. It is significant
that the test wells producing water of high iron content were opposite
those points in the river where the well point tests indicated that
the underground water beneath the river, also, was of high iron
content. It is a common popular fallacy to attribute the iron found
in some well waters to theii" contact with the pipes.

Moreover, the writer would question the judgment which dictated
placing the wells along the river bank 400 ft. apart. Such a spacing
would seem to be too great, and would not make available all the water
contained in the gravel strata. Although the water levels in the test
wells indicated that the contours were parallel to the river bank, it



820 DISCUSSION : water supply of parkersburg, w. va.

Mr. should be noted that these levels were taken during comparatively
urgess. jjjgj^ stages of the river. It is the writer's opinion that if levels had
been taken during low stages, it would have been found that the
ground-water table had a different slope, and that the flow of water
would have been slightly down stream toward the river bank. This
would conform more nearly to the conclusions of Mr. Clapp, as stated
in his report.

It is believed that experience at Parkersburg warrants the conclu-
sion that each of the following features is of importance and should
be considered in the construction of strainer systems or infiltration
plants of this type:

1. — The presence of a permanent bar of sand and gravel of suit-
able size within a reasonable distance of the pumping station
to which the water is to be carried;

2.— The constant submergence of the bar from which the supply
is to be obtained;

3. — The velocity of flow in the stream, which velocity must be
sufficient to maintain the bar clean and free from silt and
other drift;

4. — The depth at which the collecting units are placed with rela-
tion to the bottom of the stream; and

5. — The character of the water which is found beneath the bottom
of the stream.

Experience at Parkersburg has shown, also, that it is very important
that there should be a suitable meter or measuring device on each
suction line from each section of the system. Without such meters,
it is impossible to operate the different sections of the system at the
same rate. Immediately after back-flushing a section, the loss of head
is very much decreased, and it is difficult to avoid excessive rates of
filtration from the newly washed unit without a suitable meter.

Although this paper was presented primarily with a view of dis-
cussing the preliminary investigations by engineers to determine the
merits of the various water supplies available at Parkersburg, it may
be well to mention briefly the merits of the system actually built. Its
obvious advantages are its low cost of operation and its simplicity
and ease of operation. The annual cost of operation is only one-half
of the estimated cost of a suitable well system or rapid sand filter
plant, as discussed in the preliminary reports of the consulting engi-
neers. Moreover, the operation of the plant is accomplished success-
fully without expert control.

Though it may be true that the results accomplished by the plant
are not within certain arbitrary standards adopted by Federal author-
ities, it is to be noted that the plant was constructed under more severe



DISCUSSION': WATER SUPPLY OF PARKERSBURG, W. VA. 821

guaranties of efficiency than have ever been enforced at any other Mr.
water purification plant in the United States. Moreover, the plant "^^®®^-
was accepted on the basis of these guaranties and on the basis of
thorough tests of operation by a competent representative of the City.
It is the writer's belief that the City should erect a suitable chlori-
nating plant as an additional precaution in securing always an entirely
satisfactory supply. In doing so, however, it should be recognized
that the City is taking no additional precautions other than those now
adopted by very nearly all cities which obtain their water supplies
from either rapid or slow sand filters. There is no doubt that the results
obtained by the present plant are comparable with those from rapid
sand filters, and that the cost of operation is so low that it makes the
type now in use at Parkersburg of very great advantage to any city
where conditions are such as to permit of its satisfactory installation.

James H. Fuertes,* M. Am. Soc. C. E. — Messrs. Fuller, Hill and Mr.

FuGrtcs

Potter, having discussed quite fully the principal and more important
matters dealt with by Mr. Hall, a few remarks concerning experiences
with somewhat similar methods of securing water from river beds
may be of interest.

A number of water companies and cities on the Allegheny and
Ohio Rivers in the vicinity of Pittsburgh take water from these rivers
by a plan having general features in common with those of the plan
adopted at Parkersburg.

About 10 years ago three citizens of Wilkinsburg, Pa., a suburb of
Pittsburgh, brought suit against the Pennsylvania Water Company
to compel that company to furnish pure water, as required by the
terms of its charter. The suit was rather unique, and was one of
the few recorded in which a decree has been entered directing a com-
pany to secure and provide a sufficient supply of pure and wholesome
water. The supply against which complaint was made was drawn
from gravel-covered cribs in the Allegheny River bottom. The first
supply crib, built in 1897, was 32 ft. wide, 308 ft. long, and 5 ft. high,
made of 2 by 6-in. and 2 by 8-in. planks, placed edgewise and flatwise,
and separated by intervals of 2 in. A 24-in. pipe extended from the
interior of the crib to the pump-well on the shore. The crib was sunk
in a dredged excavation, about 300 ft. from the north shore of the
river, and at a depth placing the top of the crib about 5 ft. below
the natural bed of the river. The excavation was refilled and the crib
covered to a depth of 1 ft., with 2-in. gravel, and then with from 4
to 4 J ft. of sand and fine gravel, such as had passed through a IJ-in.
mesh screen. The depth of the water over the crib was ordinarily
from 8 to 10 ft.

* New York City.



822 discussion: water supply of paekersbukg, w. va.

Mr. It was shown by analysis of the water that the supply was drawn

' partly from the river by filtration through the gravel over the cribs,
and partly from the sub-stratum of ground-water in the 30-ft. bed of
gravel and sand in the river bottom.

In 1902, two additional cribs, each 48 ft. wide, 408 ft. long, and
5 ft. deep, were put in service. They were constructed in the same
general manner as the first, but excluded from the gravel covering
stones larger than 1 in. in diameter. Three cribs were sunk side by
side and 10 ft. apart, and a line of 42-in. pipe, decreasing to 24-in.
in diameter, was laid in the 10-ft. space between the cribs, with four
short pipe connections therefrom to each crib about equally spaced
throughout their length. The main 42-in. pipe was extended under
the river bottom to the pump-well on shore.

The water from the cribs discharged under its own head into the
pump-well, and was pumped thence to a reservoir, from which the
street mains distributed it to the consumers.

After periods of operation of varying lengths, the surfaces of the
gravel over the sunken cribs would become more or less choked with
silt, dirt, and other matters suspended in the river water, and gradu-
ally fail to yield a sufficient supply. To relieve this shortage, the
company had recourse to cleaning or washing the filter areas with
water. The filtering area immediately over the cribs — with all three
cribs in use — was about 1.13 acres; undoubtedly, however, water
entered the cribs from the bottom and from areas outside of the boun-
dary lines of the cribs. It is not possible, therefore, to state the rate
of filtration per unit of area per day, except that it was less than
about 7 000 000 gal. per acre per day by the quantity derived from
sources other than the filtering area over the cribs.

The plan followed in cleaning these filter areas was to drag back
and forth across the area a rake having a 2-in. pipe header, 3 ft.
long with seven teeth, 15 in. long, each perforated with two ^-in.
holes on the front and one in the end, water being forced down through
this rake at a pressure of 125 lb. per sq. in. by a pump on a barge. The
rake was dragged across this area, the wash-water rising up through
the sand and bringing the mud with it, the principle resembling
that by which the Blaisdell filter washer is operated. The river cur-
rent was depended on to carry the washings down stream and away
from the washed areas. The rake was dragged back and forth across
each area three times, by hand windlasses on barges anchored on each
side of the areas being washed, before a new area was attacked.

Raking, or scouring, was practically continuous from May to
November, 1905, and from May to December, 1906, except as inter-
fered with by uncontrollable conditions, such as river floods. In 1906,
the surface was examined several times by a diver, who reported



discussion: water supply of parkersbueg, w. va. 823

considerable disturbances of the sand surface, and, under his guidance, Mr.
depressions were filled up with gravel and sand deposited from a barge
through a tube. Frequently, the diver rode on the rake, as it was
dragged back and forth, to press the teeth down into the sand. It
was stated that the teeth usually penetrated the top of the gravel
bed to a depth of from 8 to 12 in. While the process of raking or
scouring was going on, the cribs were kept in constant use. The dis-
turbance of the surface and the destruction of the silt formation
thereon permitted a largely increased volume of water to pass freely
through to the cribs, through the washed areas, and during and imme-
diately after these periods of washing, the water was easily shown
to be very little different, in character, from the raw river water.

Much evidence was introduced at the trial as to the quality of
the water, both as to its chemical and biological features, and it was
shown that there were times when, for a number of consecutive days,
the quality of the water would be excellent; all of a sudden, however,
some change in conditions of river flow would cause the silt to be
washed from the river-bed over the cribs, or parts of their areas, leaving
places where the river water could penetrate to the interior of the



Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 67 of 167)