American Society of Civil Engineers.

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

. (page 64 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 64 of 167)
Font size
QR-code for this ebook

requirements of a filtered water supply.

Plain sand filters, such as are used at Philadelphia, Pittsburgh,
Washington, and other places in the United States, and used exten-
sively abroad, are constructed so that they may be under constant obser-
vation as to the performance of each particular unit, and when the loss
of head has reached a maximum — or other conditions suggest — they
can be taken out of service, the sand beds can be scraped and rehabil-
itated, and the filters restored to their original condition and efficiency.
The Smith filters, however, can be examined and repaired only at long
intervals, and, if it should occur that the quality of the water obtained
from them was unsatisfactory during a period when the stage in the
river was considerably higher than the sand bars in which the filters
are placed, there would be no relief from the unsatisfactory quality

* Cincinnati, Ohio.

discussion: water supply of parkeksburg, w. va. 789
of the water except that which might be obtained from back-flushing Mr

the collecting pipes, and that, in the writer's opinion, is always a
doubtful method of restoring filters to normal condition.

It was also thought, in connection with the Portsmouth proposition,
that, in view of experience with water purification systems somewhat
similar to the Smith filter along the Ohio River, there was no assur-
ance, at any time, that the quality of the water would satisfy strict
sanitary requirements, and, for this reason, objection was then made
by the Ohio State Board of Health to the adoption of this system by
Ohio cities.

At several places in the paper Mr. Hall mentions a possible source
of water supply to wells from the Ohio River, on the assumption that
the water will pass through the silted bed of the river and reach the
wells through the sand and gravel in which the latter are driven.

From long experience, with wells along the Ohio River and else-
'where, it is the writer's conviction that wells or galleries sunk below
the bed of the adjacent river seldom, if ever, receive water from it.
Many tests for hardness, chlorine, and iron, in water from the wells
and from the river or stream, have always shown a distinct difference,
and, in the writer's experience, this has been sufficient to indicate that
the water from the wells is not naturally filtered river water. This
experience has been so uniform over a wide range of territory, and
throughout a long period of time, that it has led him to conclude that
water obtained from wells or infiltration galleries along a stream is
always land-water or ground-water percolating through, or pocketed
in, the pervious strata, some of which in due time reaches the stream
in the form of springs. In a case where a surface connection is made
from the river to the porous material in which wells are driven, the
water may percolate through the porous material and reach the wells
in this manner, but the writer can recall no considerable success, either
from experience in America or abroad, in obtaining large volumes of
satisfactory water in this way.

Considering the Parkersburg wells proposed by Messrs. Knowles,
Fuller, and Fuertes, the writer would assume that all the supply fur-
nished by these would be obtained from the water moving toward the
river from the land side, or pocketed in the porous material along and
under the river bed; and though the stages of the river will have an
influence on the elevation of this ground-water, it would not imply
any flow from the river to the wells.

Several years ago the wrriter had occasion to test the influence of
continuous pumping for 2 months from a gallery or well sunk in a
gravel bed along the Iowa River, and, for the purpose of defining the
extent of the water field, several observation wells were driven at
distances ranging from 100 to 8 325 ft. from the gallery. A level

J. W. Hill

790 DISCUSSION : water supply of parkersbueg, w. va.

J. W. Hill,

base and bench-marks were established, and the levels of the water
in the gallery and in the observation wells before and during pumping
were noted from time to time and carefully referenced to the base.
The pumping was continuous, day and night, for 62 days, and the ele-
vations of the water in the gallery and in the observation wells, at
the beginning, during, and at the end of this period are shown on
Fig. 15.




W 80



§ , , Water Level Sept. 26tli




Based on notes taken July 21tU

to Sept. 25th, 1899.



■3000 4000 5000 COOO

Histance from Pump Well, in Feet,

Fig. 15.


8000 9000

Although the test wells shown on Fig. 15 range from 100 to more
than 8 300 ft. in a direct line from the pumping well or gallery, they
were distributed east, west, and northwest from it, so as to include the
water field which it was thought would be affected by the test. Fig. 15
shows some curious results, with reference especially to observation
wells Nos. 10, 11, and 12, all of which were open and penetrated the
sand and gravel water-bearing strata, and in which the water levels
fell from time to time during the test.

discussion: water supply of parkersbueg, w. va. 791

Though the water in No. 12 dropped 1.20 ft, during the 62 days Mr.
of pumping, at the end of the period the surface of the water therein
was considerably above the general elevation of other wells nearest to
it in point of distance and in the same general direction from the
pump well (northwest).

On the other hand, the water level in IsTos. 10 and 11, which were
east of the pump well, was at no time as high as in the nearest wells
in point of distance, although not lying in the same direction from the
pump well. The only explanation for this discrepancy that could be
given at the time of the test was that the wells were on lower ground
and nearer the river, and might have been supplying the river rather
than the test pump well. In attempting to draw smooth curves on
the diagram for the elevations at the beginning and end of the test,
Nos. 10, 11, and 12 were omitted, all the others lying sufficiently near
to a smooth curve to indicate the probable water level at the beginning
and end of the test.

"Reference to the diagram, upon which have been plotted the
elevations of water in the pump well and observation wells, will show
the mean curve of water level which existed in the known area affected
by pumping on September 18th. All water in the field above this
curve has been pumped away or lost by natural causes, and for the
rate of pumpage from the test pump well at that time [693 000 gal.
per day] this curve represents the average condition throughout the

"The facts obtained from the experimental work [in Iowa] con-
ducted during the summer [1899] are as follows:

"The total amount of water pumped from the test well from
July 24th to September 21st, inclusive, was 38 796 662 gal., an average
of 646 600 gal. per day. During this interval of time, from the data
submitted, the water level in the test pump well was lowered from
Elevation 87.78 to Elevation 73.93, or 13.85 ft.

"The total rainfall reported after July 24th aggregates 6.03 in.,
and, applying this to the area embraced in the investigations, viz.,

1 367.4 acres, and allowing an absorption of 40% by the sand and
gravel overlying the rock and under the top soil and clay, the absorbed
rainfall within the field of observation amounts to 89 555 837 gal.,
or more than twice the pumpage from the test pump well for the inter-
val of time during which the rainfall was reported.

"The loss of head or lowering of water level from all causes within
the field of observation during 57 days of pumping was approximately

2 ft., equivalent to a loss in the field of 235 251 147 gal., to which
add the estimated absorbed rainfall, making the total estimated loss
from all causes 324 806 984 gal., of which the pumpage from the test
well was about 12 per cent."

An interesting part of Mr. Hall's paper relates to the cost of the
Smith system of filtration. This is stated as $80 700 for a capacity
of 170 000 gal. per hour, or 4 000 000 gal. per 24 hours, or more than
$20 000 per million gallons of daily capacity. This is a very high cost

792 discussion: water supply of paekersbueg, w. va.

Mr. for a system of filtration of this kind. Elaborate slow sand filters of
J. w. Hill. ^]^g ^ypg a(jopted by Philadelphia, Pittsburgh, Washington, etc., cost no
more, and are of such a character that the operation may be observed
from day to day, and controlled in capacity and efiiciency; and rapid
mechanical sand filters would have cost at Parkersburg (in 1911)
about one-half as much per million gallons of capacity as the Smith

The performance of slow and rapid sand filters can be fully con-
trolled, and maintained at high efficiency, regardless of river stages
and other unavoidable conditions which are calculated to operate
against sand filters of the Smith type.

The guaranty of 4 000 000 gal. per day for a total area of 1.65
acres is equivalent to 2 426 000 gal. per acre per day, or about the
capacity of plain sand filters operating without any preliminary

The manner of making payments for the Parkersburg filter was
unusual. A retention of 55% of the contract price for one year
after completing the construction indicates that there must have been
a sublime faith, on the part of the contractor, in the final performance
of the filter system.

Mr. Alexander Potter,* Assoc. M. Am. See. C. E. — The City of

Potter. Pa^j.ijgj.g]3urg ig to be commended for its courage in putting in opera-
tion the so-called natural slow sand filtration method of procuring a
supply of pure water. This method has been advocated so vigorously
by its proponents and has been opposed so bitterly by its antagonists
that a plant of this kind actually constructed should prove of a distinct
advantage in demonstrating what is of scientific value in the plan
and what of a distinctively personal bias on both sides of the much
controverted subject.

The speaker has watched with much interest the attempts which
have been made from time to time to solve successfully the problem
of so-called natural filtration. There can be no doubt that the suc-
cessful establishment of such a plant will meet with popular favor,
and therefore such a project should be given a fair and impartial trial.

The speaker recalls that on an invitation for competitive plans
for a new water supply from the City of Evansville, Ind., in 1893,
fourteen or fifteen engineers submitted plans. In this competition
George S. Davison and W. G. Wilkins, Members, Am. Soc. C. E.,
were awarded the first prize, the speaker received the second prize, and
Arthur S. Tuttle, M. Am. Soc. C. E., the third prize. The plan
recommended by the speaker was submerged filter galleries on a sand-
bar in the bed of the Ohio River, and, as he recalls it, most of the
fourteen competitive plans contemplated similar treatment.

* New York City.

discussion: water supply of parkersburg, w. va. 793

Though the speaker's project was not developed by the city, the Mr.
study made at that time convinced him that, given proper river con- ° ^^'
ditions, infiltration galleries can be designed to give a uniform and
pure effluent.

The uniformly good results secured in temperate climates, how-
ever, by rapid or slow sand filtration, clearly limit the field of opera-
tion of the so-called natural slow sand filtration process, or infiltration
galleries, to places where the cost of the latter — the first cost plus
operating charges — is less than mechanical or slow sand filtration.
In the tropical zone, however, a strong prejudice exists — and rightly
so — against the use of stored water, even when filtered; consequently,
the development of a supply from infiltration galleries warrants a
higher cost than is involved in the construction of mechanical filtra-
tion with adequate storage.

Though the analyses of the water included in the paper show a
satisfactory effluent, they represent, unfortunately, only a limited
period. From the statement made by the author, it appears that
radical changes must be introduced in the details of the system in
order to prevent a recurrence of the condition found when the beds
were examined after having been in service about 3 years. The author
states that conical holes were discovered in the sand, having diameters
of from 5 to 10 ft. and from 3 to 3^ ft. deep.

It is evident that at all stages of the river, except extreme low
water, difficulties will be experienced in maintaining a uniform depth
of sand over the filters, and that, under certain conditions beyond
the control of the operator, raw water in large quantities may pass
directly into the city supply, thus breaking down the barrier against
typhoid fever. If no adequate remedy can be provided, either to pre-
vent blow-outs due to back-flushing, or to restore positively and
promptly the sand over the strainers when such blow-outs occur, the
system particularly described by the author must be considered a
failure, no matter how successful it may be most of the time.

With Messrs. Gray, Hall, and others, standing sponsor for the
success of the Parkersburg system, it is to be hoped that the best and
most intelligent care will be given to its proper maintenance, and
that every effort will be made to correct the defects already apparent,
which prevent it from producing a uniformly dependable effluent at
all times.

A water supply is not built for a few years, or for a decade, but for
a lifetime, and because there are those who prefer some other system,
no matter what the local conditions may be, engineers should not
blind their eyes to the importance of giving a fair trial to this method
of water purification.

Passing to the question of infiltration galleries proper, there is no
good reason why they cannot be designed to maintain a constant flow

794 discussion: watee supply of paekeksbukg, w. va.

Mr. and not gradually decrease and ultimately become obsolete, as is gen-
^^' erally the case. They should be as dependable as the identical natural
process of ground-water seepage into surface streams. An infiltra-
tion gallery, to be permanently successful, must never be forced to
such an extent as to allow it to be partly emptied. When it is partly
empty, high velocities are set up in its vicinity, which, in time, cause
the silting up of the filter media immediately surrounding the gallery,
which condition slowly but inevitably results in the breaking down of
the system.

It is not a difficult matter to regulate the output from a given
infiltration gallery so that it cannot exceed a predetermined maximum
discharge, and, by so doing, the lowering of the ground-water plane
below the top of the gallery can be prevented, which condition will
keep entrance velocities within safe limits and thus maintain in-
definitely the integrity of the system.

Throughout the country there exist large deposits of sand and
gravel which, if intelligently treated, can be counted on to yield large
supplies, and there are also sand and gravel deposits in which the
depleted water supply can be augmented by the construction of large
storage reservoirs over such a system of infiltration galleries. These
reservoirs function as feeders to the infiltration galleries constructed
within the confines of the subterranean basin.

The maximum rate of downward filtration which can be main-
tained permanently is still a mooted question. In the Morris Canal,
the leakage, after 100 years of service, averages at the present time
1 250 000 gal. per mile, or from about 200 000 to 250 000 gal. per acre
per day. A filter gallery constructed parallel and adjacent to the
Morris Canal could be counted on to derive indefinitely 1 250 000 gal.
per mile per day from the canal proper. Although feasible to draw
off water for short periods at higher rates than this, such an attempt
might prove disastrous, as rates of leakage in excess of the natural
flow might cause the bottom of the canal to silt up, resulting in a
possible break-down in the gallery.

The attention of engineers must be devoted more and more to the
determination of safe yields from such underground supplies and gal-
leries, and to the method of controlling the draft therefrom to pre-
determined and safe quantities.

The speaker is in hearty accord with Messrs. Fuller and Fuertes as to
the wisdom of changing the present method of determining and report-
ing B. colt. When the warning signal rings just as loudly at the
approach of a cow, or a tramp, walking on the railroad track, as it does
for the fast moving express train, one is likely to have contempt for
the signal or demand its change. When the presence of B. coli may
mean evidence of contact with anything, from the human intestine to


the covering of a kernel of wheat, one of two things must necessarily Mr.
happen: contempt for the warning, or unnecessary and unwarranted
alarm. It is difficult to say which is the more damaging.

This unnecessary alarm is exhibiting itself in the laws proposed
for enaction in certain of the State legislatures. Certain legislation
existing or proposed in many of the States indicates clearly that the
authors or the sponsors for such legislation lack the fundamental prin-
ciples of sanitary science. These laws seek to prevent the entry of
sewage, either crude or purified, into the waters of the State. Such
a bill is now before the New Jersey Legislature, submitted under the
very attractive title, "An Act to Prevent the Pollution of Sources of
Potable Water Supply in This State." The principal provision is con-
tained in Paragraph 2, which reads as follows:

"2. It shall be unlawful for any person or persons, private or munici-
pal corporation, to discharge into any water which may constitute or
form part of a source of potable water supply any raw or treated sewage,
or the effluent from any sewage disposal works, after the date fixed
by the State Board of Health when such acts shall cease. Notice to
discontinue the depositing of such sewage or effluent shall be given
by the State Board of Health to such person, or private or municipal
corporation, at least three months before the date fixed by said State
Board of Health as aforesaid."

Other provisions of the Act leave the State Board of Health no
alternative but to enforce the law prohibiting the discharge of the
effluent of any sewage disposal plant, no matter what its degree of
purity may be.

Another bill, almost as drastic, has been made a law by the State
Legislature of Texas, and, in Oklahoma, the Health Officer, 6 years
ago, laid down the principle that neither crude sewage nor purified
sewage would be permitted to enter the streams of the State. Even
New York State is now seriously contemplating the abandonment
of hundreds of thousands of dollars spent on new public institutions,
because the authorities feel that they cannot even trust the engineer
to design a plant which will insure the conveying of the effluent by
pumping from an efficiently designed sewage disposal plant outside
of the water-shed of a portion of New York City's water supply.

This idea, which is gaining momentum, carried to its absurd but
logical conclusion, means that no city in the United States can dis-
charge its sewage into any body of water which is not tide-water. It
is the speaker's opinion that the matter is sufficiently important to
demand the serious thought of this Society, and he would suggest
that the attention of the Board of Direction be especially directed
to this matter, to the end that some intelligent measures may be taken
by the Society to counteract this dangerous proposition.

796 discussion: water supply of parkersbueg, w. va,
Mr. George W. Fuller,* M, Am, Soc. C. E. — This is quite an unusual


paper. It deals with local politics and personalities, exploits some of
the general features of the laws of hydraulics pertaining to under-
ground water supplies, and closes with a description of a filtration
project which is unique.

In association with James W. Fuertes, M. Am. Soc. C. E., the
speaker reported on the situation, after Morris Knowles, M. Am. Soc.
C. E., had been called to Parkersburg to advise as to whether a well-
water project was feasible, and preferable to filtration of the polluted
and impure water of the Ohio Eiver.

Mr. Knowles' general viewpoint was that a well-water project was
a fairly feasible one. That report produced some differences of opinion
in the community, with the result that a Board of Engineers was
appointed to make further investigations.

The speaker agreed with Mr. Fuertes that a mechanical filtration
system was to be preferred to a well-water system. Whatever merit
there may be in the Fuertes and Fuller report is not in the detailed
results set forth in the paper, but rather in the careful application of
the fundamental principles involved.

Effort was made to explain, as engineers, to a community that had
its well-defined factions, the xinderlying principles involved. In out-
line, the fundamental propositions involved are: first, that, for an
adequate underground water supply, it is necessary to determine
whether or not there are sufficient porous water-bearing strata; second,
where does the water come from, which supplies the strata; third,
where, under present conditions, does that water go; and, fourth,
what is the best way to collect that water for municipal water supply

Some rather unusual conditions are found above Parkersburg, in
that there is a rocky promontory, a few miles above the city, which
makes a rock dam across the Ohio River. Consequently, there is no
large flow of water in this vicinity beneath the bed of the river. The
old river-bed undoubtedly swung to the south, against the West
Virginia foot-hills. The surface water from those impervious hills
came out above an impervious stratum at the surface of the bottom-
lands, and found its way into the river, not through the underground
porous strata, but as a surface stream flowing over the impervious
clay and silt directly to the river.

Then, it was found necessary to reckon with a partial flow from
a very limited drainage area of the river plain, or bottom-lands, plus
the underground flow of water to the wells from the adjacent Ohio
River. The various hydrographs in the paper show some rather
unusual conditions. They show in this river plain a very marked rela-

• New York City.

discussion: watek supply of parkeksbukg, w. va. 797

tion between the widely fluctuating water level of the river and the Mr.
water level in the deep gravel strata into which the wells were driven.

The obvious deduction was that during high river stages there
was plenty of water which would flow to wells that might be sunk in
the bottom-lands near the river. When the river level was very low,
there was not only a gradually receding water-table in the bottom-lands,
with the water from the land side making its exit into the river, but
there was also the very striking difficiolty of having only small volumes
of river water flowing back in a reverse direction from the river to
the system of wells.

That small increment of naturally filtered river water in the gravel
strata was due, of course, to the blanket of silt and clay which formed
on the bottom and sides of the Ohio, at times when the flow down the
river was insufficient to produce a scouring velocity to push it away.

The ground-water proposition was abandoned for the reason that
there was not water in ample quantities coming from the land side,
and, as to that coming from the river side, it was very uncertain whether
at times of low water the blanket or carpet of clogging materials on
the river-bed could be disturbed so as surely to allow water to reach
those wells. Mechanical filters were recommended.

S. M. Gray, M. Am. Soc. C. E., was called in to review this propo-
sition, after an injunction suit precluded carrying out the recommenda-
tions of the Fuertes and Fuller report. Mr. Gray recommended a
project which, although it may have been adopted in certain respects
in other communities, was new in its entirety, as far as the speaker
knows, in a city of this size. The Smith system appears to be essen-
tially an old-fashioned slow sand filter layout, without any impervious
concrete bottom, without any concrete walls, and with reliance placed
on two things to maintain the supply during the very varying stages

Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 64 of 167)