The filtration of public water-supplies Third edition, revised and enlarged. — Reading Notes

Hydraulic Calculations and Filter Design

Hazen's analysis of the Jewell filter reveals a preoccupation with hydraulic resistance and water distribution. He calculates that strainer necks occupy only 0.07 per cent of the filter area, forcing wash water through at 22 feet per second and creating an estimated 30 vertical feet of head loss. This friction ensures uniform distribution but requires high pressure. Hazen critiques this design, noting that enlarging the necks without enlarging underdrain pipes would cause uneven flow. He advocates for a more rational system with larger waterways throughout the outlet and wash-water system.

Sand Selection and Stratification

Hazen emphasizes the importance of sand uniformity coefficients to prevent particle separation during washing. In the Warren and Jewell filters, crushed quartz and rounded silicious sand from Red Wing, Minnesota, are used—coarser than typical sand filter media. Hazen notes that low uniformity coefficients are essential to avoid stratification, where coarser sand allows most water to pass through the least efficient filter area. He argues that sand must be selected with much greater care than for traditional sand filters, linking material properties directly to filtration performance.

Filtration Rates and Economic Trade-offs

Hazen reports experimental filtration rates ranging from less than 100 to about 130 million gallons per acre daily. He states that higher rates require coarser sand or impractically high water levels, while lower rates increase filter costs without material advantage. This cost-benefit reasoning recurs throughout the text, as Hazen weighs the expense of larger sedimentation basins or additional pumping against improvements in water quality. The Warren filter's hour-long sedimentation basin is contrasted with the Jewell's 15-minute basin, illustrating different design philosophies.

Mechanical vs. Sand Filters: A Comparative Approach

Hazen structures his book by juxtaposing mechanical filters like the Jewell and Warren with traditional slow sand filters. He details the Jewell filter's rake system for mixing sand during washing and its distribution of wash water through strainers, while the Warren filter uses a separate, larger sedimentation basin. The text moves between European precedents and American adaptations, using specific installations—such as the Pittsburg experiments and the Elmira, New York, filter—as case studies. This comparative method allows Hazen to evaluate each system's hydraulic efficiency and practical limitations.

Readers will find that Hazen's work rewards attention to his quantitative details—flow velocities, head losses, and sand grain sizes—as these numbers drive his engineering judgments. The book's value lies not in broad conclusions but in the precise observations that informed early twentieth-century water purification. Approach it as a technical manual shaped by a single engineer's rigorous testing and comparative analysis.

Reading about Allen Hazen’s careful, patient measurements of sand filters brought back an afternoon in my grandfather’s shed, where he showed me his old logbook of well levels. That quiet, precise attention felt familiar. Something similar stirred with Søren Hjorth: Inventor of the Dynamo-electric Principle — Text and Context, where an overlooked life’s work waits, unhurried, for anyone to pause and listen.