GAC Sandwich Layer 2

[tipp79]

Hi,

I'm currently working on a project whee we are looking to refurbh some existing slow sand filters. These slow sand filters have an existing "GAC sandwich layer" incorporated in them. I'm struggling fnd a design basis for depth of GAC and depth of sand above and below this layer. Anyone worked on similar projects or have experience of GAC layers. There's been a few turbidity failures in teh past few years- will refurbishing the SSFs prevent any more? Any good reference articles?

Cheers,

James

 

[waterexpert]

From memory, SSFs are usually filled with a coarser grade of sand than rapid gravity filters, and solids removal takes place within a short distance of the surface in what is called a "schmutzdecke" or dirt layer. This layer develops over time, and is in some designs periodically scraped off, to allow a new schmitzdecke to develop. The sand removed is washed and returned.

GAC layers would be designed on an "empty bed contact time" similar to hydraulic residence time.

Sean

http://www.seanmoran.co.uk

 

[ch81pc]

SSF's ideally cntain sand with a uniformity coefficient of 1.5-2 and an effective size range of 0.1-0.35mm.

The depth of your GAC layer should be based on the necessary hydraulic residence time required. The layer of sand above that should normally be no less than 500mm before the bed is resanded. If you take of 100-150mm per skim then I would suggest a sand layer of at least 1500mm so that you dont have to re-sand your bed every year. I guess that your beds were built taking into account the GAC layer? If the GAC was retro-fitted you will loose that additional head (and hence filter run time) as a reslt of increasing the total bed depth.

I dealt with dsevral filters that have been perfroming poorly with regards turbidity (which often means your failing on coliform count). These were due to:

-Collapsed under-drains, letting through fines
-Heavey plant used to skim the beds compressing and breaking up the GAC layer leading ot carry over
-Operators backfilling beds too rapidly after a skim and blowing holes in the bed leading to hydraulic short circuiting.
- Bed being skimmed below 500mm minimum sand depth
- Bed having too much flow put through it (even though it was removing coliforms) before an adequate schumtzdecke had developed
- If the bed has recently been refurbished, check and double check that if new sand was put in, it had been washed at least 3 times before being putting in.

Hope this helps