Water-Works at South Haven, Mich.*
The city of South Haven lies on the shore of Lake Michigan at the mouth of the Black river. It has about 2,500 inhabitants. The larger portion of the city, including most of the business blocks, lies on a plateau south of the river and close by the lake. About one-third of the city occupies a similar plateau north of the river. The portion of these plateaus occupied by buildings lies at an elevation of thirty or forty feet above the level of the lake. South Haven has within a few years suffered two disastrous fires, and about a year ago voted bonds to the amount of $35,000 for system of water-works. This amount was subsequently increased to $40,000, a portion of which was expended in the equipment of a fire department and the extension of lines of water-pipe not originally contemplated. The water supply is derived from tubular wells on the lake front and is forced into a stand-pipe located near the business portion of the city, and standing on ground about thirty-five feet above the level of the lake. Aside from the method of obtaining a water supply there is nothing about the system particularly interesting to engineers, the details being, in the main, similar to those in other works of this class. Attempts had previously been made to secure water from the strata bordering the black river at points above the city, but without success. A suitable supply could be obtained from the lake by a long intake pipe extending out beyond the liars, but the estimates for this intake pipe, protected from the action of storms, which are severe at this point, were about $10,000. This sum, added to the cost of the stand pipe and pumping plant, left very little of the appropriation to be applied to the construction of pipe lines. The problem, then, was to secure at less cost a supply of water sufficient for present needs which could be supplemented in future as the demand arose, at reasonable cost and with certainty. An examination of the lake revealed the following conditions : The bed and shore of the lake is sand, rather fine, but well washed and permitting the passage of water with reasonable freedom. The shore rises from the lake at a slope of about three feet in 100 to the foot of the bluff, which is about 400 feet distant. The bluff, with the exception of a surface covering of about three feet of sand, is of tenacious clay as deep as it has been explored. The same clay formation extends under the beach of the lake at a depth of eight or ten feet below the surface of the sand and, continuing, extends under the bed of the lake at about the same depth.
Test wells were sunk at various points in the beach to the level of the stratum of clay, and front their behavior the following conclusions were drawn :
1. There was a slight upward inclination of the water surface from the level of the lake toward the foot of the bluff and consequently a movement of water from the sub-surface underlying the city towards the lake. Chemical analyses of the water from a series of wells arranged at right angles to the lake shore indicated that the water in the upper portion of the beach was polluted, and that the degree of pollution was less as the lake shore was approached.
2. The sand was sufficiently free to admit the passage of water in considerable quantities.
3. Vertical tubular wells of the ordinary form sunk in the shallow layer of sand above the clay must necessarily have short screens placed just above the clay in order to prevent the taking of air by the pumps when the water stratum is lowered by continuous pumping ; and since a gang of wells in order to supply a sufficient quantity of water from this stratum must be at considerable intervals, it follows that the relative proportion of connecting pipes to screening surface must be large.
4. The farther from the bluff a supply of water could be secured the purer would the water be.
The following method of obtaining the supply was adopted : At points thirty feet back from the water line of the lake pits were dug to water. Starting from these pits six-inch tubular wells were pushed out under the lake nearly horizontal, the inclination being about one inch per foot downward. This is about the inclination of the stratum of clay underlying the bed of the lake. Where the tubular wells pass under the shore of the lake they are about three feet deep, and the depth below the sand gradually increases to five or six feet at a distance of about 125 feet. At about this distance from the shore line screens, thirty feet long and as large as can be inserted in the six-inch casing, are set. There are three of these wells at present, and they are twenty-five feet apart and are connected by pipe lines parallel with the beach into a twelve inch suction line leading to the pumps.
♦Reprinted from the Michigan Engineers Annual.
The history of subterraneous water supplies records many failures. The natural conditions essential to success are 1
1. A sufficient water shed from which the sub-surface water naturally flows to the point of supply or may be drawn to it by pumping.
2. I’orous strata, both at the point of supply and remote from it, through which the sub-surface water may pass to the point of supply in sufficient volume.
The determination of the extent to which these conditions are present is generally a difficult matter. The action of test wells when pumped for a limited time is no guarantee of their performance under continuous pumping, and the fact that one test well supplies a certain quantity of water in a given time is no guarantee that we can duplicate it by others at intervals of 50 or too feet or more which will correspondingly increase the supply even for a limited time ; and if they do so for a limited time it may be that all are in a pocket of limited area which will be exhausted so much the sooner.
In the case in point, however, many of the uncertainties surrounding a work of this class are removed. The stratum in which the screens lie is comparatively accessible for examination. There is every reason to believe that it is uniform in character. Lake Michigan, which determines the static head above the screens, will, so long as it stands above them, insure a constant supply of water to them, limited only by the capacity of the screens and the five feet of sand above them to convey it. Furthermore, wells placed at no great distance from one another horizontally will not sensibly encroach upon each other, being supplied along the lines of least resistance or mainly through the sand above the screens. The supply can, therefore, be increased with certainty by additional wells.
If the lake recedes so as to uncover the screens, or the beach extends so as to force the open water beyond them, the wells can readily be extended, or additional ones can be put in to make up for the increased resistance due to the passage of the water through the sand for a greater distance.
The pumps are placed on foundations only seven feet above the level of the lake, in order to make the lift as low, and the action of the pumps as smooth as possible. The three wells readily furnish one million gallons of water every twenty-four hours under a vacuum of ten inches, as indicated by a gauge set on an air chamber thirteen feet above the level of the lake. The air chamber is taken off the suction line at a point seven feet above the level of the lake. Correcting for the lift of the pumps, and the friction in the suction line, the vacuum necessary to induce the passage of one million gallons of water per day through the sand and screens, is about two or three inches of mercury, or, say, 2.25 to 3.4 feet of water.
The water is uniformly clear, even during severe storms. No deposit of fine sand particles can be detected after the water has stood iira deep vessel for some time.
It was anticipated that possibly the sand about the screens might become so compact by the action of the waves and the ice in winter as to impede the passage of water through it, and a by-pass is provided around the pumps so that the stand-pipe pressure may be put upon the screens, and the current through them reversed. Each well is provided with a gate, so that it can be isolated from any, or all the others at will.
In the pump house are two direct acting duplex pumps of the Hughes pattern, each having a capacity of seven hundred and fifty thousand gallons each twenty-four hours.
There are two boilers fifty-four inches bv twelve feet, with the usual attachments. Coal oil is used as fuel. It is stored in a special tank near the pump house, from which it is brought to the boilers by a pump. On the supply pipe from the oil tank is a meter. The flame from the burning oil strikes against a bridge wall of fire bricks at the front of the boiler, which is laid up without mortar, and the bricks spaced about two inches, so that a portion of the flame passes through and against other similar walls. These biidge walls not only distribute the heat along the boiler with less damage to it, but after the fires go down, have a marked influence in keeping up the steam by the heat which they give off. The pumps are run two hours each day. Morning and evening, oil is burned under the boilers while pumping only, the steam pressure seldom falling below’ thirty pounds at any time during the twenty-four hours.
The arrangement of the distribution system is shown on the accompanying map. The main leading from the pump house is ten inches in diameter, and the main connecting the standpipe with the balance of the distribution is also ten inches in diameter. All hydrants except one are off from lines six inches in diameter or larger, and this one is fed by a four inch line from both directions, which is reinforced at the distance of a block in each direction by a six inch pipe.
Where the water main crosses Black river, it is laid fifteen feet below the surface of the water. It was desirable to test the river crossing before connecting it with the balance of the distribution system, and the followinig method w’as adopted.
The pipe closed at one end above the water by a gate, and into the other end was calked an iron plug. Into this plug a inch water pipe was tapped, having a water meter and water gauge attached. Water was forced into the river crossing through the ¾” inch pipe, and the meter attached to it by a force pump until the desired pressure was obtained, and then the pump was operated enough to preserve this pressure. By this means could be detected the leakage in a given time at the required pressure.