Slow Sand Filter Operation.
The filtration plant at Camp Perry, O., designed by and constructed under the supervision of R. Winthrop Pratt, was operated for 11 weeks at an average rate of 3,000,000 gallons per acre daily. The plant consists of two slow sand filters, kilter 1 was cleaned alter being in service an average of 5.6 days, and filter No 2, 6.2 days. Neither filter was operated continuously, the time varying from 6.6 to 22 3 hours during each day to supply an average consumption of 108,700 gallons. The plant was built in 1908 to provide a supply for the members of the State National Guard and the U. S. Army and Navy, which attend the shooting contests at the rifle range. There was an average of 1500 persons in attendance during 1911. The water supply and filter plant comprises two 22×45-ft. open filters, two settling reservoirs holding 2 hours’ supply, a control chamber and a circular clcar-watcr basin holding about 4 hours’ supply. Water is obtained 800 feet from the shore. At this point the water is only 6 feet deep and is constantly stirred up by the waves. In 1911 the turbidity varied from 19 to 65, averaging 48. The filters remove all of thq suspended matter. Sand in filter 1 was taken from the Lake Erie beach. It had an effective size of 0.21 mm. and a uniformity coefficient of 2.5. For the second filter a larger sand was obtained from Michigan, having an effective size of 0.26 mm, and a uniformity coefficient of 2.0. Both filters originally had 3 feet of sand supported by a 12-inch underdrain system of graded gravel or limestone. Unlike most small plants, the Camp Perry filters arc equipped with loss-of-head gauges and Venturi meters, so the operator has a way of knowing how the plant is operating and when scraping is necessary. Accompanying tables show results of cleaning the filters, which is accomplished by two men in about two hours. As filter 1 has the finer sand, the runs are shorter. The comparatively small amount of water passing both filters between cleanings is said to be due to the small sand size, high turbidity and small amount of settling-basin capacity. In 1909 filter 1 passed 12,900,000 gallons per acre between scrapings and filter 3, 21,300,000 gallons. Colony counts were made in 1911 every two or three days. The supply varied from 70 to 1600, but averaged about 600 bacteria per cubic centimeter. Filter 1 turned out an efficient averaging 66 in the 7 samples taken. The maximum and minimum counts were 150 and 12, respectively. In filter 2 effluent the counts varied from 11 to 75. averaging 41. B. Coli tests were made on 33 days.