Size of inlet and outlet holes

[Galantino]

-function of flow and depth -bigger holes on bottom layer for backwash -flow characteristics and loss calculations (probably unnecessary to some extent

[eak24]

Hmm... what is the constraint here? Normally, we size filter holes to ensure proper flow distribution during backwash. Then we use that same total orifice area for each entrance/exit to ensure good distribution amongst the different entrances/exits. Can we do the same with the horizontal filter? Here are my first thoughts about designing hole size/ plate settler size:

1. Pick a desired filter height (filter_h) and calculate the maximum width (the length shouldn't matter, as that changes with flow) where filter_w = (cos(plate_settler_angle)*plate_settler_L x 2)+sand_w . The width and height have the following relationship to ensure the forward filtration velocity (FFV) and backwash velocity (BWV) are achieved with the same flow rate:

FFVfilter_Lsand_h = BWV sand_w filter_L => sand_h = (BWV/FFV) sand_w => filter_h = (BWV/FFV) sand_w * sand_h_bw_factor

2. Size holes for backwash condition. For backwash, we block all layers of holes except for the bottom, otherwise the bottom-most sand wouldn't fluidize, and water would only come out of the upper-most unblocked holes. Now we size the total orifice area so that it's between the two competing constraints, on the one side we can't make it too small or the headloss grows out of control, and if we make it to large, we'll need to grow all the plate settlers so that we achieve a low enough capture velocity in the rest of the plate settlers. I think we have a bit of room here. Maybe we just say we want 5 cm of headloss through the holes during backwash? and we can enlarge the holes later if necessary?

3. Is that it? Are there other constraints I'm missing?

Oooof, I can't wait to see a hydrogen design sheet and a Fusion design!

[monroews]

The inlet and outlet system (outside of the sand filter) can be easily designed to have no significant head loss (meaning head loss is very small compared with head loss through the sand). With slotted pipes it was challenging to get the open area required for low head loss. With the tube settler grids on both influent and effluent it may be possible to have the head loss kept small enough that it is the sand that determines uniformity of flow during filtration.

The next challenge is to design for uniformity of flow during backwash to ensure that there isn't a channeling of high velocity water up through one side of the sand bed. I assume that this requires the head loss through the orifices be large relative to the head loss through the sand at the fluidization velocity in the path between two orifices. The idea is that if only part of the sand bed is fluidized, that the head loss through the inlet orifices will be high enough to force enough water through the orifice that is connected to the settled bed of sand that is closest to the fluidized bed such that the next orifice will force the settled sand to fluidize. If we meet that constraint, then more and more of the sand will fluidize. I think this means that the head loss through the orifices must be greater than the horizontal spacing of the orifices.

I've never thought about the failure mode of fluidizing a bed at this level of detail before, so please critique this!