Deep Bed Filter Capacity 1

[DeltaCascade]

Can anyone provide insight into typical total suspended solids (TSS) loading capacity of deep bed filters (lb/ft2 or lb/ft3)?

Realizing that values will range widely with particle size and density and filter media used (eg: 0.1 lb/ft2 for sand filters filtering river water to 20 lb/ft2 for multimedia filters filtering steel mill waste waters), and that pilot testing is the ultimate answer, any data on expected ranges would be valuable for general filter sizing work with:
- dual media filters (eg: sand, anthracite),
- multi-media filters (eg: garnet, sand, anthracite),
- coarse mono-media filters and/or,
- upflow sand filters,
would be very useful. Throughputs rates (gpm/ft2), backwash rates and % solids removal would also be related, but any info appreciated. The only reference found to date states that this data is hard to find. Go figure. Any general trend information would be useful (realizing that pilot test is only way to get true answer).

Thank you.

 

[rbcoulter]

I bet the reason that data is hard to find is that they vary over a wide range. Pilot testing is really the only way to do it. Some solids have a tendency to "blind" the top area of the media rendering the depth capacity nearly useless. You also have to figure in air or water scouring of the media .

I once rented a pilot sand filter unit for the purpose of doing a similar study. The test is straight forward and easily translates to the full scale unit.

 

[ConsultUS]

Welcome to the world of the Unknown.
I been asking the same questions for about 2 months, with no help , my question is posted on the water treatment area.
So far I can share this with you.
1.- Calcule the open area, there are some data on the literature related to materials and size.
2.- Calculate your passing trought velocity , should be below 3 ft/sec wich is the carry over velocity generaly accepted.
3.- Calculate your equivalent FLOW RATES in GPM/SF , my references and Thumb number values are 3to5 GPM/SF for normal flow rates, 5to8 for very high poor separation.
4.- Estimate the void area, and use 10 % for material collection before backwash, since this will reduce your open area and you do not want to carry over.
5,- Once you know this you have the carry over velocity for backwash.

IF YOU FIND SOMETHING BETTER PLEASE SHARE IT WITH ME.
GOOD LOOK.
Javier.

PD.-Look into old, old .......Chemical Enginnering Text Books on Filtration.

 

[DeltaCascade]

It's difficult to pilot test, 'cause we are working with a slipstream of recirculating cooling tower water and the plant isn't built yet ... at least we are working towards doing settling tests and perhaps, jar tests and deep bed filter piloting on spring-run off quality river water, prior to detailed design. Tx for describing the piloting as straight forward ... that helps to build confidence to &quot;do the riht thing&quot;, and pilot (if we can ju$tify it). Filter backwashing shouldn't be a problem, especially with air scour (approx 6 scfm at 7 or so pisg, per ft2 filter area, etc). &quot;Rapid&quot; sand filters are pretty much out of our consideration, due to rapid blinding (yeah, particles collecting at top inch or so of sand bed and &quot;plugging&quot... we believe that these filters are only good for filtering (&quot;polishing&quot low turbidity water (<5 ntu). We understand that even deep bed filters (eg: anthracite coal on top of sand on top of garnet) are only good for filtering up to 50 ppm TSS influent water (about 25 NTU turbidity units)... this is the worst case cooling water quality that we want to design for. Any really high turbidity water ( > 50 ppm TSS) would surely &quot;prefer&quot; sedimentation treatment (eg: solids contacting clarifier or lamella settler, etc). Checking old books for deep-bed filter loading is a good idea ... we continue to &quot;dig&quot;. Tx

We can specify suspended solids particle size distrubution as 2 to 10 micron to narrow it down ... that type of thing ... to get general idea of filter loading capacity (lb solids/ft2 filter area or lb/ft3 filter bed volume) to &quot;bracket&quot; the problem (and estimate backwash frequency to &quot;nail down&quot; design). In our application, for example, we aren't really concerned with filter effluent quality, we just need the filter to remove about 80% or so of solids in a single pass. In the meantime, estimating void area should at least give an estimate for a maximum solids holding capacity that the filter could have. And Consultus, yes, our specific throughputs (3-5 up to 5-8 usgpm/ft2, and the higher the throughput rate, the lower the effluent quality (hihger the TSS), etc) agree with yours, too. Tx
I'll work to share what we've compiled thus far .. a litle later. Hopefully others may be able to contribute.
... here's wishing you all a very merry Christmas

 

[waterexpert]

For loading,you can use Metcalf and Eddy. Used to be pp241-For prediction of backwashing expansion, there was an article by Amirtharajah and Cleasby in 1972. I've got a photocopy of the Amirtharaja and Cleasby article if you can't get one. Reference is AMIRTHRAJAH,A AND CLEASBY,J.L. &quot;Predicting Expansion of Filters During Backwash&quot; J.AWWA Jan 1972

Seán Moran

 

[waterexpert]

Oh, and there are all sorts of experience based rules of thumb on selection of media size(s), backwash and running flowrates, distribution and type of header and underdrain systems, design of air scour systems, and so on. There is plenty of information of this sort available within specialist process contractors, but it's nearly all proprietary.

You may well be best buying a proprietary design,or getting a specialist to design for you. Otherwise you'll might be reinventing the wheel.

 

[DeltaCascade]

Thank you, folks, for the most helpful suggestions. I will dig into those as well!

As promised, the following table is what I've gathered to date. As expected they are just generalizations. The formatting isn't very good but an interested reader may be able to paste it to a wider word document to make better sense of it. As in many processes, particularly water treatment processes it seems, pilot testing on the water to be filtered is required for precise data...

Thank you. Cheers...

Typical Filter Parameters and Performance Expectations
Filter Type& Media Maximum InfluentSolids Contentppm TSS BedDepthft Solids Removal Capacitylb/ft2/filter-run Effluent TurbidityNTU Service Flowusgpm/ft2 Pressure Drop atEnd of Runft water Backwash Flowusgpm/ft2
Pressure Sand 5 - 20 1 0.1 – 0.125 < 1 1 – 3 5 - 10 13 – 15
Gravity Sand 5 – 20 1 0.1 – 0.125 < 1 1 – 3 5 - 10 12 – 15
PressureGranular Anthracite 5 - 50 n/a n/a < 1 3 – 4.5 10 8 – 12(12-15 for hot waters)
GravityGranular Anthracite 5 - 50 n/a n/a < 1 3 – 4.5 10 6 – 15
PressureMixed MediaAnthracite & Sand 5 – 50 2(1 & 1) 0.75 0.1 – 1 6 - 8 10 6 – 15
GravityMixed MediaAnthracite & Sand 5 – 50 2(1 & 1) 0.75 0.1 - 1 5 - 8 10 6 – 15
PressureMixed MediaAnthracite & Sand 5 – 50 3(2 & 1) 1.2 0.1 – 1 6 - 10 15 6 – 15
GravityMixed MediaAnthracite & Sand 5 – 50 3(2 & 1) 1.2 0.1 - 1 5 - 8 15 6 – 15
PressureMixed MediaAnthracite & Sand 5 – 50 4(3 & 1) 1.3 0.1 – 1 6 - 10 15 – 20 6 – 15
GravityMixed MediaAnthracite & Sand 5 – 50 4(3 & 1) 1.3 0.1 - 1 5 - 8 15 – 20 6 – 15
PressureCoarse Mono-Media 50 - 200 6 2 – 5 < 5 5 – 10 10 – 20 6
GravityCoarse Mono-Media 50 - 200 6 2 - 5 < 5 5 – 10 10 – 20 6
Notes:
· Values given above represent wide ranges anticipated; pilot studies are required to obtain precise data for a given application.
· Increased influent solids concentration leads to filter blinding and short run lengths.
· Increasing service flow rates reduces fraction solids removed in a filter pass.
· Deep bed filter service rates range from 1.5 to 20 usgpm/ft2, the lower rates for chemically treated surface waters, the higher rates for steel mill mill-scale particles filtered on coarse grained media.
· Filter solids loading numbers are general estimates which are best determined by pilot testing.
· Air scouring will lower required backwash flow rates slightly.
· Gravity filtration may not be a feasible option where end-of-run pressure drops are high.
· Filter backwash rates required will vary with backwash water temperature and media properties and are typically determined by vendor.

---------- end -----------