Intermittent Sand Filters...
The intermittent sand filter is a specially prepared bed of sand on which effluents from primary treatment or from trickling filters or secondary settling tanks may be applied intermittently by using troughs or perforated pip distributors. The effluent from the filter is removed by an underdrainage system.
The filter bed should have a depth of clean sand of at least 24 inches overlying clean, graded gravel. The gravel should
be placed in at least three layers around the underdrains and to a depth of at least six inches over the top of the underdrains. The sand itself should have an effective size of 0.3 to 0.6 millimeters and have a uniformity coefficient
of not more than 3.5. The center to center spacing of underdrains should not be more than 10 feet.
When treating a primary effluent from normal sewage, the rate of application on the filter should not exceed 125,000 gallons per acre per day, which should be reduced if the sewage is strong. With trickling filter and secondary settling tank effluent, the loading should not exceed 500,000 gallons per acre per day.
The intermittent sand filter is a true filter which strains out and retains fine suspended solids and also acts an an
oxidizing unit. The major portion of the straining and oxidation is effected at or near the surface of the sand. Straining results from the fine nature of the sand medium with small voids and from a biological slime growth of organisms which develops on the surface of the sand. Oxidation is effected, as in all secondary treatment devices, by the living aerobic
microorganisms which develop primarily at the surface, forming a slime layer, but also extending into the sand medium.
It is important that the filter be allowed to empty itself and obtain a fresh air supply at intervals. This is accomplished by intermittent dosing of the sewage onto the filter. Sewage is applied from two to six times a day in quantities sufficient to cover the surface down through the sand, air is drawn in from the surface. Sand filters are constructed in two or more units which are used in rotation. Eventually the slime layer on the surface causes the top layer of sand to become clogged and necessitates the removal of the top layer of sand in order to put the unit back into efficient operation.
Should not be allowed to develop on the beds as this tends to produce septic action, obnoxious odors and an effluent of poor quality. Pooling indicates that cleaning is necessary. The surface of the beds should be kept level to afford uniform distribution of the sewage and weeds, grass, etc., should not be allowed to grow on the beds.
For winter use, the open air beds in the northern part of the State should be ridged every two feet in order to hold the ice off the main body of sand. An alternate method is that of raking up small four to six inch piles of the top surface every three feet over the surface of the bed. The bed should be cleaned and leveled as early as possible in the spring. Where natural sand percolation beds are used it is sometimes advisable to shallow harrow the beds after careful removal of all organic deposits. Such harrowing is not advisable with underdrained beds because of possible damage to the underdrainage system.
Efficiency and Use...
A well operated intermittent sand filter will give a clear, sparkling stable effluent almost completely oxidized and
nitrified. Over-all plant removals of 95 percent or more of the BOD and suspended solids in the raw sewage can be
expected. This exceeds other accepted secondary treatment processes.
When compared with other sewage treatment processes, large areas of land are needed, construction costs per unit of sewage treated are high and maintenance in keeping the filters clean is considerable. The use of filters of this type is restricted to situations where the volume of sewage to be treated is small, or where an exceptionally high grade of plant effluent is necessary. They have been used effectively for the additional treatment of secondary treatment effluents. Modification of
the sand filters have been proposed in the past but, in general, have not proven practical for treatment of large volumes