Working Volume Percentage of Atmospheric Storage Tanks

[BamaGump]

Hello all,

Boss is gone for the day and I'm still plugging away. Wanted to ask those with experience in tank sizing - what percentage of the nominal capacity do you typically assume for a working volume? I'm currently working with an atmospheric water storage tank, but on a previous project's design was instructed to assume 50% - I'm now thinking that seems pretty conservative.

I understand there's a minimum tank level where you wouldn't want to run a suction pump, as well as space reserved for high / high-high level alarms, but am curious if there's a common rule of thumb out there.

TIA!

"To someone who has seen a dozen people frown, scowl or turn their faces away, your smile is like the sun breaking through the clouds."

 

[fel3]

There is no fixed percentage of or rule of thumb for "dead space" in a water storage tank, but I can tell you from experience that 50% is far too conservative as you surmised.

Are you working with AWWA D-100 (welded steel) or D-103 (bolted steel) and will the water storage tank in question be located in an area prone to seismic activity? If so, I suggest looking at the applicable standard (both standards have the same requirements for freeboard for seismic-induced sloshing) and at ASCE 7, which also has the same requirements (see ASCE 7-22, §15.7.6). The required freeboard is a function of the tank diameter, the maximum depth of water in the tank, and several seismic parameters that you can get from

https://ascehazardtool.org/

. In a high seismic area, the required freeboard might be as little as about 2 feet to more than 6 feet.

You also have to consider the water in the bottom of the tank that is below the outlet. Generally, shell penetrations are at least 12 inches above the tank floor (D-100 discusses this in §14.3.2.7, but there are caveats), so the bottom 12 inches of a water storage tank with an outlet in the shell is usually considered to be "dead space" even if the outlet has a fitting that allows siphoning water from below that level. If the outlet is attached to the tank floor, then there is no "dead space" in the bottom, even though you might want to assume some so as to keep from introducing too much sand into the distribution system (systems fed by groundwater wells often end up with significant sand in the bottoms of their tanks.)

Regarding pump suction, you definitely need to pay close attention to suction head with horizontal pumps. When I have designed horizontal pumps to pull water from a tank, I use a large suction header and I put the pumps as close to the tank as possible to minimize suction losses. I also do what I can to increase the vertical distance between the outlet nozzle and the pump suction. However, when I can, I use vertical turbine pumps for this application because the pump suction is well below ground and this gives me more flexibility with suction head and with locating the pumping station.

============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

 

[JStephen]

One issue is that "nominal capacity" is not always well-defined for tanks.
In some cases, it is the capacity to the top of the shell, regardless of where the actual water levels are. I think this comes from the bolted-tank market and also the older API welded tanks.
In many cases, it is the capacity from tank bottom to the lip of the overflow, regardless of how low you can actually draw water.
In some cases, tanks have an effective range that is only a fraction of the total volume. This is normally seen with standpipes, where you may have a tank 100' high, and only the top 30' is really needed.
On the freeboard- in addition to seismic sloshing, there are some cases where they want a specific air-gap between the inlet pipe and the overflow level, and this can throw a couple of extra feet in.

 

[LittleInch]

As an initial guess I would normally think you should get at least 80% working volume from a nominal volume of a vertical cylindrical tank.

This could reduce based on min head required from the tank, margins for alarms or setting of an overflow.

Each tank is different, so it's difficult to have a fixed view. Maybe the it's a requirement to have a reserve for a water tank so you don't run out of water on a really high flow day?


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[BamaGump]

Wow, thank you all for the input.

The existing is a welded vertical cylindrical tank. The Client is planning to add more water end users at the site, so we're performing a study/evaluation to determine impacts and make recommendations. This has involved more high-level water balance and storage time calculations, so was curious here particularly in the event that we recommend adding/increasing storage capacity. Regarding the seismic considerations, the site is located within the lowest category for peak ground acceleration (PGA) per the linked ASCE Hazard Tool. For the purposes of this report, it sounds like we can be a little less conservative than 50% while still including some design margin.

"To someone who has seen a dozen people frown, scowl or turn their faces away, your smile is like the sun breaking through the clouds."

 

[LittleInch]

The issue the more I think of it is your peak day / hour. Basically you really, really don't want to run out of water so if your "normal" range is 50% it gives you possibly another 30% for your peak day without going mad or being overly conservative.

It's all about the impact of running out of water that you need to think about. saving $ by not adding more tankage, but then the impact on running out for reputation, compensation, flushing lines which now have air in them etc etc is $$$, even if it only happens once in 5 years needs to be considered.

How much margin you need is for you to decide and what your other options are on the supply side ( turn all your pumps on?) when your levels get dangerously low.

but don't cut it too fine...

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.