Three numbers, three different failures
A granular media tank is described by three quantities that people routinely collapse into one, and each of them fails differently.
Capacity is a total: how many gallons the bed can treat before the media is spent. It scales with media volume and with nothing else. Empty bed contact time is how long the water spends in the bed, and it is the bed volume divided by the flow. Service loading is the velocity through the media, flow divided by the cross-sectional area of the bed.
EBCT (min) = media volume (cu ft) × 7.48 ÷ flow (gpm)
Service loading (gpm/sq ft) = flow (gpm) ÷ bed area (sq ft)
Two cubic feet of media is 14.96 gallons of bed. At 5 gallons a minute that is just under three minutes of contact. In a 13 inch tank — 0.92 square feet of bed area — the same 5 gpm is a loading of 5.4 gpm per square foot. Change the flow and the second and third numbers both move; change the tank diameter at constant media volume and the loading moves while the contact time does not; change the media volume and the capacity and contact time move while the loading does not. They are genuinely independent and a tank can pass one test and fail another.
The teaching point: the invisible failure is contact time
Running out of capacity is at least an event. Not having enough contact time is a permanent, silent, partial failure that looks identical to the tank working. Adsorption is not instantaneous — the water has to spend long enough among the media for the target to diffuse into the pore structure — and at half the contact time the bed does not remove half as much for twice as long. It removes less of everything, all the time, and the pressure gauge reads exactly the same.
Which is why the flow figure matters more than it looks. A bed sized at three minutes of contact for average household use has under a minute when a shower, a toilet and a kitchen tap are running together, because contact time is inversely proportional to flow just as UV dose is. That peak is not a rare event; it happens every morning. The whole house filter flow rate calculator gives you the number, and it is usually higher than people assume.
The corollary is that the fix for a contact time problem is media volume, and only media volume. A finer grade does not buy contact time. A taller tank with the same media in it does not either. Neither does slowing down the service flow, unless you are prepared to slow down the household, which nobody is.
Bed volumes, and why a capacity figure needs a target attached
Dividing the gallons a bed treats by the volume of the bed gives bed volumes: how many times the bed processes its own volume before it is spent. Two cubic feet treating 80,000 gallons is about 5,300 bed volumes. It is a useful way of comparing media, because it strips out the tank size and leaves the property of the material.
What it does not do is exist in the abstract. A capacity figure is meaningless without three things attached: what the target is, what concentration it is at, and what contact time it was measured at. The same carbon can show wildly different capacities against a strongly adsorbed compound and a weakly adsorbed one, and capacity generally falls as the concentration falls, which is counter-intuitive until you think about it as a competition for a fixed number of adsorption sites. That is why the capacity field on this page is empty of any opinion: it comes from your media supplier, for your target, at the concentration a laboratory measured in your water.
| Tank | Bed area | 2 cu ft bed depth | Loading at 5 gpm | EBCT at 5 gpm |
|---|---|---|---|---|
| 9 in | 0.44 sq ft | 54.3 in | 11.3 gpm/sq ft | 2.99 min |
| 10 in | 0.55 sq ft | 44.0 in | 9.2 gpm/sq ft | 2.99 min |
| 12 in | 0.79 sq ft | 30.6 in | 6.4 gpm/sq ft | 2.99 min |
| 13 in | 0.92 sq ft | 26.0 in | 5.4 gpm/sq ft | 2.99 min |
| 16 in | 1.40 sq ft | 17.2 in | 3.6 gpm/sq ft | 2.99 min |
Every row holds the same two cubic feet of media and therefore the same contact time and the same total capacity. What changes down the column is the velocity through the bed and the depth of it. Narrow tanks give deep beds at high velocity; wide tanks give shallow beds at low velocity. Both extremes have problems — a very shallow bed is more prone to channelling and to the mass transfer zone reaching the outlet early, a very high velocity compacts the bed and raises pressure loss — and the sensible middle is why 10, 12 and 13 inch tanks dominate residential installations.
Freeboard is not wasted tank
A backwashing media tank is deliberately filled well short of the top. During backwash the bed lifts and expands, and it needs somewhere to go; too little room and media leaves through the top of the tank and ends up in the house. How much expansion a given media needs at a given backwash rate and water temperature is a property of that media, published by whoever supplies it, and it is not a number a general calculator has. This page shows you the freeboard you have so you can check it against the figure your supplier gives.
The backwash itself is a substantial water demand in its own right, and on a well it is frequently the largest single flow the system ever sees. The backwash water calculator works out what that costs in gallons, and it is a bigger number than most people expect.
Where this sits
Carbon goes after sediment and before UV. Sediment ahead of it, because silt in a carbon bed causes channelling and pressure loss and the cartridge life calculator covers that stage. UV after it, because carbon removes some of what shades the light, and the UV dose calculator shows how sensitive that stage is to transmittance. Hardness is a separate mechanism entirely and belongs to the softener calculation, which counts grains rather than gallons.
Every water quality number on this page — hardness, iron, solids, transmittance, a concentration, a target dose — is something you enter from a laboratory test of your own water. Use a state-certified laboratory; your state or county health department maintains the list and will usually tell you what a well in your area is commonly tested for. This calculator has no idea what is in your water, will not tell you whether a result is acceptable, and does not select equipment. Treatment follows the report, and a device that removes one thing does not remove another.
Questions people ask
What is EBCT and why does it matter more than tank size?
Empty bed contact time is how long water spends in the media, calculated as the bed volume divided by the flow rate. Two cubic feet of media is 14.96 gallons of bed, so at 5 gallons a minute the contact time is just under three minutes. It matters because adsorption takes time — the target has to diffuse into the pore structure of the media, and below the contact time the media needs, the bed removes less of everything all of the time. That failure is silent. The water looks and tastes the same, the pressure is unchanged, and nothing indicates it except a laboratory test on the outlet.
How long does carbon media last in a whole house filter?
It depends entirely on what it is being asked to remove and at what concentration, which is why the capacity is a field on this page rather than a number in it. If your media supplier gives 40,000 gallons per cubic foot for your application, then two cubic feet treats 80,000 gallons, and a household using 300 gallons a day gets about 267 days out of it. Change either the target or its concentration and that figure moves a long way. Media also ages in the tank regardless of throughput, which is why most installations carry a calendar limit alongside the capacity one and change on whichever arrives first.
Is a taller tank better than a wider one?
They trade different things and neither wins outright. For the same media volume, a narrow tank gives a deeper bed at a higher velocity, and a wide tank gives a shallow bed at a lower velocity — the contact time and the total capacity are identical either way, because both depend only on how much media is in there. Deep and fast raises pressure loss and can compact the bed; shallow and slow is more prone to channelling and to the working zone of the bed reaching the outlet earlier than the arithmetic suggests. The usual residential compromise sits around 10 to 13 inches of diameter, which is why those sizes are everywhere.
Why is my contact time so much shorter than the calculation says?
Because the calculation was almost certainly run at average use and the bed spends its important minutes at peak. Contact time is inversely proportional to flow, so a bed giving three minutes at 5 gpm gives 1.5 at 10 gpm and one minute at 15. A house with a shower running while a toilet refills and the kitchen tap is on is easily at 7 or 8 gpm, and that is the flow the bed sees during most of the water it actually treats. Size on the peak simultaneous flow rather than the daily average, and if a peak cannot be avoided, the only lever that helps is more media.
Can I backwash carbon to make it last longer?
Backwashing lifts and reclassifies the bed, clears out accumulated fines and breaks up channels that have formed, and it does keep a bed working properly. What it does not do is regenerate the carbon. The adsorption sites that have been used are used, and no amount of backwashing at household temperatures and pressures gives them back — reactivation is a high temperature industrial process. So backwashing maintains the hydraulics and the capacity clock keeps running underneath it. Treat the two as separate: backwash to keep the bed physically healthy, and replace the media on capacity or calendar, whichever arrives first.