Refurb of Slow Sand Filters

[tipp79]

Anyone have experience of refurbhising slow sand filters? What are the chracteristics of a slow sand filter in good condition- ie. good quality sand (uniformity coefficient), best practice for skimming?

Cheers,

Jim

 

[ccor]

Usually a finer sand is used, with an effective size of 0.25-0.35mm (D10) and a uniformity coefficient of 2 - 3, and specific gravity > 2.6. Depth is usually 1-1.5 m. When maximum allowable headloss occurs (20-30 days) the top 1 - 2 cm is scraped off and stockpiled on site. This is repeated until the filter depth is around 0.5 meter and the whole bed is "resanded". SSF is not recommended for influent waters > 50 NTU, >30 color units, or for waters containing colloidal clay.

 

[ch81pc]

A good slow sandfilter will give you a turbididty of <0.5 NTU with no coliforms. A fully refurbished bed can take weeks to become mature and develope an adequate schmutzdecke. I have managed to speed the process up a little by adding some sand from an established filter (taking sand from the top 20mm only).

For skimming the key is to get the sand depth as uniform as possible, areas with less sand depth or a higher permiabilty (for what ever reason) will obviously take more flow, becoming blind and anearobic long before the rest of the bed which could lead to quality issues.

As mentioned by CCOR, your filter will only be as good as the pre-treated water you are putting on to it (exponential relationship between loading and run time). Is you raw water from a river or a borehole?

I have managed to acheive 0.1m3 (water)/m2 (sand) per hour for well over 18 weeks feeding the filter with water from the Thames pre-treated with pre-chlorination (no Cl residual must reach the SSF's!!), pre-ozonation, rapid filtration (assisted at times of high turbidity with ferric chloride dosing).

Hope this helps.