American Society of Civil Engineers.

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

. (page 66 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 66 of 167)
Font size
QR-code for this ebook

periodical floods, by the water washing and receding from the streets,
cellars, houses, stables, barnyards, chicken coops, etc., of many towns
and cities, Pittsburgh included. In fact, one would imagine that the
water would be much better for fertilizing crops than as a domestic
• supply.

The speaker thinks the Parkersburg authorities were wise not to
try to pump 3 000 000 or 4 000 000 gal. of water a day from fourteen
wells placed some distance from the shore; for he does not think it
possible to pump such a volume of water continuously without remov-
ing more or less sand, perhaps in almost imperceptible quantities, but
still sufficient to form cavities.

These cavities would probably cause no damage in the open fields,
but, sooner or later, these fields will be built on by the city; and
though the subterranean cavities might even then give no trouble for
many years, they might sometime, suddenly, cause very serious col-
lapses, similar to those which have happened to cities built over coal
mines. The experience gained by Parkersburg through its adopted
plan would seem to prove the correctness of this assertion.

Some day all the towns along the Ohio (and this applies also to
other rivers) will combine to dam the water up in the mountains, where
it is comparatively pure and free from sediment, and pipe it to the
different towns, as needed. This will not only give an adequate and
dependable supply of good water, but will also regulate the flow in the
rivers, doing away with the disastrous floods of periodical occurrence.
As an example of many towns along the Ohio : when the speaker first
saw Catletsburg, Ky., the town stood on a bluff some 20 ft. above the
river. The next spring he saw rowboats landing at the second-story
windows of the houses. These floods occur so quickly that, even with
the excellent modern Weather Bureau service, the inhabitants do not
always have time to protect their property; yet they come back, clean
their houses, make the best of their losses, and start over again; they
very seldom move away or build on high ground.

The Court decisions quoted by Mr. Hill would seem to indicate
that the small towns may some day be able to force the towns above
them to provide pure water or to stop polluting the rivers.

Mr. H. F, Dunham,* M. Am. Soc. C. E. (by letter).— A paragraph on

' page 780 giving the dimensions of the intake pipe could be illuminated

by mention of the connections to, and the character of, the pumping

machinery. It would appear from the text that the large diameters

of the intake or "suction pipe" were expected to lessen the otherwise

• New York City.

discussion: watee supply of parkersbueg, w. va. 807

"undesirably rapid rate" of flow through the filtering material. This Mr.
implies the use of intake pipes as reservoirs from which a certain ^^°*^*™-
quantity of water could be taken without marked or legitimate eiiect
at the strainers. Is this view correct? Ordinary experience would
tend to show quite the reverse; that the friction losses would be less
and the draft at the wells greater with large pipes under a vacuum
produced by pumps of any description.

Strainers with V-shaped openings have been used by the writer,
not always in vertical positions, but always at some angle and with
connections that would allow of access to each well or to each one in
a small group when shut off from the others in a system. Any other
method seems to be open to distrust, although it is true that conditions
vary greatly, and each case is a problem by itself.

Many years ago, the writer examined sand bars in the Ohio River
in the vicinity of Wheeling, W. Ya., for the Wheeling Water Board,
but found little encouragement in quality or volume, although the
beds were continuous to a depth of 40 ft. below low-water levels.
Besides hardness, the figures for iron were too high, and filters were

Theodore S. Johnson,* Assoc. M. Am. See. C. E. (by letter). — In Mr.
reviewing the problem of water supply, the same questions that arise •^°'^°^°°-
at the beginning of the study are still to be considered after some
one solution is adopted; namely, is the water obtained of satisfactory
quality from chemical, physical, and bacteriological standpoints, and,
is the requisite quality obtainable at a fair cost of production?
Usually, these questions can be answered after some investigation.
It is to be regretted that rarely are two solutions of a water supply
problem provided so that they may be compared in actual performance
under the same conditions of demand and production and at the same
time. In this discussion the operation of one type of supply is
presented, and an attempt is made to compare it in results with one
not in operation. This is a handicap in solving the problem as to
the scheme which has not been adopted.

This paper recites the preliminary investigation for a well supply
for Parkersburg, and then describes the construction and operation
of a system not capable of such preliminary investigation, concerning
which little or no previous research or study had been made. It would
have been desirable to have answered in the paper the following
questions :

1- — How does the present supply compare, in turbidity, total
hardness, iron and bacterial content, in color, and in other factors
affecting the desirability of a water supply, with the same factors
and qualities of the water taken from the well supply as investigated

* Granville, Ohio.

808 discussion: watee supply of parkeksbueg, w. va.

Mr. by Mr. Knowles and by Messrs. Fuertes and Fuller, which supply was
Johnson, rejected?

2. — How does the cost of pumping from the present beds compare
with the estimated cost of pumping from the line of wells? Speaking
more generally, how does the cost of operation, including capital
charges, compare with the same cost as estimated for the rejected
supply ?

3. — What may be the probable life of the present plant, as com-
pared with that of the plant for well pumping? The object of this
question would be to show the effect of depreciation and maintenance
charges on the cost of pumping.

The data in the paper are not sufl&cient to enable the reader to
determine the answer to the first of these questions. The reports
published in the paper, showing turbidity, alkalinity, and total hard-
ness, are given only for one month, and therefore cannot be
regarded as showing very fully the characteristics of the new supply.
The figures in the table on page 785 are interesting. For example,
the iron content of the raw river water varies from 1.06 parts per
million, on January 20th, to 7.04 parts per million on January
24th, 1914. This is a very wide range for iron content in the raw
river water in such a short period. Again, the average iron content
for the month is 3.33 parts per million. Having been familiar with
the untreated Ohio River water from his earliest remembrance, and
being at present a user of water in which the iron content is slightly
more than 1 part per million, the writer is surprised to find the Ohio
River water so contaminated with iron. He is still more surprised
to find, on page 783, a report of an investigation of the supply by
Mr. I. L. Birner, which states: "The total iron in solution, of the
filtered water, was 0.19 of one part per million, against an average
of 5 to 16 parts in the river." From the wording of Mr. Birner's
report, one is led to believe that he refers to parts per million, and
this would seem to indicate a river water extremely polluted with
iron. The only data which the writer has immediately at hand
are those given in the Report of the Ohio State Board of Health for
1908, on page 206, where results are given for tests made at Marietta,
Ohio, a few miles up stream from the site of the present filtration
works. These tests show an average iron content for 1906 of 0.62
part per million, and the maxima for the year are found to be 1.00
and 1.23 parts per million for August and November, respectively.
It is difficult to reconcile these two statements.

In the same report the bacterial results show the untreated water
at Parkersburg to have from 100 to 3 100 bacteria per cu. cm., and
the filtered water from 24 to 1, with averages for the month of 900
and 9.3 for the untreated and filtered waters, respectively. Tests

discussion: watek supply of parkeesburg, w. va. 809

by the State Board of Health at Marietta show that the average of Mr.
a large number of determinations indicates a bacterial count of 8 200
per cu. cm. in the raw water during 1906, and 7 350 during 1907.
The same investigators, in tests made on July 24th, 1906, at Pomeroy,
Ohio, show 19 100 bacteria per cu. cm. in the raw river water and
8 200 on November 19th, 1906. Although these are only scattered
results of tests, they serve to cast some comparative light on the
published figures. That the safe condition of the water, as indicated
by the tests of Dr. Rose and Mr. Birner, is not generally recognized,
is shown by the suggestion that a report prepared by the City of
Wheeling and tests by the State Bacteriologist of West Virginia both
agree that at times the water is quite unsafe.

Engineers may then reasonably inquire whether experience with
other well supplies, or tests on the wells as driven at Parkersburg,
would not indicate that the quality of the water from the well supply
is better than that furnished by the present infiltration system. If
there are published or available reports on the well supply, or addi-
tional tests on the present supply, which will throw more light on
this question, it would seem highly desirable to have them published.

To answer the second question — how does the cost of pumping
from the present beds compare with the estimated cost of pumping
from the line of wells? — the necessary information is not available
in the paper and, by all means, should be presented by the author or by
the engineers concerned. The figures on page 781, showing the cost
to be $11.03 per million gallons, are very evidently only for the
cost of operation, and include no capital or depreciation charges.
The data on pages 780 and 781 do not separate items in construction
cost in such a way that direct comparisons can be made.

The service life of the present plant is necessarily one of doubt.
It is not the time in which the present construction will furnish the
proper capacity, with which we are concerned, but rather the life of
any one filter unit. Relative to this it is noted that an examination
of the beds in 1916 showed that, at several places, there were holes
in the sand and gravel covering. These holes were from 5 to 10 ft.
in diameter, and were formed like craters, extending to within 1 or 2
ft. of the brass strainer pipes. The writer has been long familiar
with the locality in which the filter beds have been placed, and was
engaged on some surveying work, in connection with the quantities
of excavation under the contract, at a time very near that at which
the photograph shown in Fig. 13 was taken. At that time, he was
able to walk to the coffer across the slightly raised level of the bar,
and, on every side, were to be seen similar crater-like holes in the
natural surface of the sand-bar. In some of these there were pools
of water in which children were bathing. It has been the writer's

810 DISCUSSION : watee supply of parkersburg, w. va.

Mr. opinion that if these craters would form in the natural bed, they would
Johnson. -^^ ^^ likely to occiir over the filter bed, and, if so, would of course
greatly impair its efficiency. Though these holes may be refilled, it
would seem hardly possible to conduct the refilling so as to secure
uniform stratification in the bed, and the reasonable life of the
infiltration plant would be considerably shortened by this cause alone.
The efficient life of a rapid sand filter or a slow sand filter is a
question which can be answered with fair accuracy. It would not
seem that the conditions providing stability in the plant were to be
found in the shifting variable sand bottom of the Ohio River.

In a general way, the comparison of this plant with either a
filtered water supply from the Ohio River or a supply drawn from
wells in the gravel beds of the adjacent bottom-lands can be put on
a satisfactory basis only when the Smith infiltration system is
regarded as being of the caisson type or as a type of sand filter. If
of the latter type, it would seem that the same conditions which have
been regarded as necessary to secure a satisfactory supply from a
rapid sand or slow sand filter would have to be applied to the Smith
system. Among the first requisites for either of the usual filter types
is a careful control of the operating head, as affecting the rate of
filtration, and both types are considered as sensitive to small changes
in the effective heads. With a range of 53.1 ft. in the stage of the
river, it would hardly seem probable that the filtering conditions
remain the same.

Another requisite of proper sand filtration is a careful control
of the cleaning processes. The filters are constantly inspected and
watched, and the conditions carefully maintained for securing equal
fitnesses and thorough washing of all their parts. If the beds of the
Smith system are pitted with crater-like holes extending to within a
short distance of the strainer pipes, the efficiency of the back-flushing
would certainly be reduced. If, as a matter of experience, such back-
flushing does cleanse all parts of the filter bed very thoroughly and
uniformly, the data which prove this fact would be interesting.

With a supply of extremely varying nature, such as that found
in the Ohio River, sand filtration plants are accustomed to have con-
stant and rigorous tests at least daily, and often more frequently, on
the basis of which the rate and character of their operation is altered.
Although the author recommends systematic tests, these tests will be
lessened in value by the fact that little or no change could be made
in the operating conditions of the filter system. It may then be
concluded, in the writer's opinion, that the Smith infiltration system
may not be regarded as a purification plant, capable of providing at
all times a safe and wholesome supply of water such as is obtainable
from rapid sand filtration plants, and, therefore, must be considered

discussion: watek supply of parkersburg, w. va. 811
as an underground caisson or ground-water collection system, of rather ^ Mr.

, . 1 . 1 . . , - , . Johnson.

high construction cost, drawing water either from the river above
or the ground-water below, and confining its purification processes
to mere straining.

Philip Burgess,* M. Am. Soc. C. E. (by letter). — The writer Mr.
has read this paper with much interest. As stated in the paper, it has "'■^'^''^•
been the author's purpose principally to discuss the investigations made
by the engineers who were retained by the City to determine the best
method of obtaining a satisfactory water supply. The writer was
retained by Mr. Smith, the contractor for the system which was finally
built, principally to conduct the preliminary investigations which Mr.
Smith had made before the construction of the plant, and, finally,
to prepare the plans from which the system was actually constructed.
The following discussion is offered principally for the purpose of
describing the preliminary tests and reports made prior to actual

From his position as an outsider, the writer was particularly in- »
terested in the opinions of the comparatively large number of engineers
obtained before the City authorities finally agreed as to the system
which was considered best adapted to serve the needs of the city. The
writer was led to believe that perhaps this condition was attributable
largely to the fact that the reports of the engineers as submitted were
not sufficiently conclusive and convincing, and that they contained
serious differences of opinion.

The first report submitted by Messrs. Chapin and Knowles recom-
mended "that the ultimate object to be looked forward to is the
establishing of new water-works at the Camden Farm, with the build-
ing of either wells or a mechanical filter plant; the former, if in-
vestigations show the feasibility, and the latter, if not so found."

Again, on page 16, of that report, it is stated: "It cannot be too
strongly urged that, while the well proposition is attractive, further
investigations should be made, perhaps in other localities nearer the
proposed reservoir site."

It is apparent that, although this report seemed to favor the con-
struction of a well system, it showed that the engineers were not in
a position to make a definite recommendation for building such a
system without further investigations.

Under date of February 21st, 1910, Messrs. James H. Fuertes and
Greorge W. Fuller submitted to the city 'officials a joint report, in which
appears the following statement:

"We find, as will be -seen from the estimates of cost of construction
and operation, that a first-class well system and a first-class filter system
at the Camden Farm are substantially on a parity as to the quality

* Columbus, Ohio.

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

Mr. of the water, and as to cost, both of construction and of operation

Burgess, (-^yhen a suitable charge for the required land is made against the

well system), and that either plant could be depended upon to supply

such quantity of water as would be needed by the city for some

years. * * *."

The conclusions and recommendations of Messrs. Fuertes and
Fuller were very definite, but an examination of their report indicates
that these conclusions and recommendations were based on certain
assumptions, as shown on page 13 of that report, as follows :

"During wet weather, perhaps when the water in the Ohio River is
high, it appears to us that the source of water contained in the pores
of the extensive sand and gravel layers is largely river water, but to
some extent water from upland sources, the high river stage damming
up some of the upland water.

"We are firmly of the opinion that, during low-water stages in the
Ohio River, the water which can be drawn from the strata is that
which has been stored within the pores of the material at times of
previous high water in the river. As this stored water is removed, or
naturally flows away, we are certain that, during low stages in the
river, whatever water enters these sub-surface strata of sand and
gravel is largely of river origin.

"Our conclusion, therefore, is that, as a source of supply for a
municipality, the water from the underground sources in question, at
the end of protracted droughts, must come from the Ohio River through
the bottom or sides of the present river bed along the stretch of the
bottom lands below Briscoe."

The conclusions of the engineers as to the source of the gravel-
water modified their report, as will be shown in the following discus-
sion. On pages 14 and 15 of the report appear the following state-
ments :

"In the first place, it was found that the water in the porous sand
and gravel layers in question did not rise to such high levels as found
in the river itself. This is true as indicated by readings in test wells
almost at the bank of the river; and shows that, even during fairly
high velocities in the river, there is still so much mvid and silt de-
posited on the sides and bottom of the river that a considerable head
is required to overcome the friction sufficiently to allow the water,
in substantial quantities, to pass through this thin and more or less
impervious deposit of mud."

This statement is of interest, in view of the fact that the bottom
of the Ohio at Parkersburg is always clean and free from deposits of

In their concluding discussion of the quality of the supplies, the
engineers stated as follows :

"There are some slight differences in quality of the water from
wells and from filters, the principal feature of which would be hardness
on account of the wells yielding water coming largely from the Ohio

discussion: avater supply of parkersburg, w. va. 813

River at times when it is in flood and fairly soft. We do not attach Mr.
much importance to this difference because ultimately during low- burgess,
water periods the well water would also have its source in the Ohio
River. So far as quality is concerned, we put waters from filters and
from wells in the vicinity of the Camden Farm on a parity. * * *."

In describing the well system and its operation, the engineers recom-
mended the following method of increasing the supply during low-
water stages:

"River Plant. To insure the removal of mud and silt from the bot-
tom of the river adjacent to the well plant at the latter end of low-water
periods, we have considered it prudent to provide for a hydraulic
dredge located upon a barge equipped with 25-horse power engines,
steam-driven, and which would operate not only the centrifugal pumps
for dredging purposes, but also some arrangement for propelling the
boat at intervals up and down this distance of 4 000 ft. adjacent to
the wells. The suction of these dredges would be specially designed,
resting upon a shoe and with the end attached to a special casting,
with a view to removing say for a depth of 1 or 2 in. the material
over as wide a strip as possible. It is probable that such a general
device could be arranged so as to clean the bed of the river for a
width of 20 or 25 ft. and for a length of 4 000 ft. in a working day.
This would give an available filtering area of roughly two acres. * * *.

"We estimate that the annual cost of operation and maintenance
of a 4 000 000-gal. well water project, exclusive of fixed charges, would
be as follows :

Fuel, based on gas $11 680

Pay-roll 4 000

Supplies, repairs, and renewals (average) ... 4 000

River dredging 3 680

Total $23 360

"We estimate that the annual cost of operating and maintaining
a 4 000 000-gal. filter plant in a first-class manner, and of pumping the
water to the distributing reservoir, exclusive of fixed charges, would
be as follows :

Fuel, based on gas $11 000

Pay-roll 6 500

Coagulants 2 500

Supplies, repairs, and renewals (average) ... 4 000

Total $24 000"

From these citations, it is apparent that the conclusions of the
engineers as to the cost of operation and as to the comparative merits
of siipplies derived from a filter plant or from a well system were
largely influenced by their assumption as to the source of the well
water. It was this assumption in regard to the source of the well
water and in regard to the method of maintaining and increasing the
available well water supply during dry-weather conditions, which was

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

Mr. not convincing or conclusive to, at least, some of the city officials and
urgess. jjjgj^jjgrs of the Water Commission. If one omits the annual cost
of river dredging, estimated at $3 680, from that of operating the
well system, it is materially less than that of operating mechanical
filters, as determined by the engineers. In other words, omitting such
cost of river dredging, and assuming the filtered river supply and
well water to be equally satisfactory as to quality, the engineers' con-
clusions and recommendations for establishing a mechanical filter
plant were not warranted.

Messrs. Fuertes and Fuller did not agree with the conclusions of
Mr. Clapp, who thought that the ground-water at the wells had its
source partly from the sand plain districts and partly from the bottom-
lands in the main valley of the Ohio.

As stated by the author, the City finally employed Mr. Samuel M.
Gray, who submitted a final report, under date of May 26th, 1910,
which was so convincing and conclusive, especially as to the merits
of the system advocated by Mr. Smith, that the City soon entered
into an agreement with Mr. Smith to build his system.

The author has given a very interesting discussion of the pre-
liminary investigations made at Parkersburg to determine the com-
parative merits of supplies, obtained either by filtering the Ohio
River water or by puting in wells adjacent to the river. However, he
does not mention any of the preliminary studies made to determine
the feasibility of the Smith system. This feature of the work embraced
perhaps a more thorough investigation than was made to determine
the merits of the well system. Moreover, it is the writer's belief
that the final success of the Smith system at Parkersburg is attribu-

Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 66 of 167)