Water Tank Overflow Pipe Design

[duk1]

I have been tasked to design an overflow pipe for an elevated potable water tank. I have reviewed several sources to make sure that my assumptions regarding inlet and exit losses, pipe flowing full v. venting through pipe assumptions are correct. To date, I have not found a good comprehensive reference, including AWWA and API. Does anyone know of a good reference for the proper design of a water tank overflow pipe?

 

[BRIS]

No I don't know a good reference. I assume that your problem is that you want the pipe to discharge its design flow before it is submerged and before it is flowing full. i.e when it is also drawing air through a vortex.?? When flowing full the solution is simple. When flowing part full and drawing air the solution is more complex.

The questions are 1) to can you not allow sufficient head above the pipe to assume it flows full (1.5 x D) or 2) can you not add an anti vortex plate or turn the inlet down etc to reduce the amount of air drawn into the pipe.

There are plenty of methods of increasing the flow into an overflow at minimum head such as adding an overflow weir.

 

[JStephen]

Use a weir box and make the top of the box big enough to get the design flow rate with 6" rise over the weir (per AWWA D100). For sizing the pipe, assume that water just dumps out of the weir box into the atmosphere, and use Bernouli's equation to calculate this "orifice flow". This will give you flow rate as a function of pipe area and depth of the weir box. (Try to hold dimensions and shape of the weir box so it is paintable inside.) Once you have the pipe size, go ahead and that pressure losses from pipe friction and fittings don't exceed the gain from the vertical drop. This method is generally conservative.

Typically, the overflow pipe size would be about the same size as the inlet pipes to the tank. If you're a consultant bidding out a job, overflow pipe sizing can be included with the detailed design of the tank (ie, done by the tank contractor).

 

[duk1]

Thank you for the comments. I thought I was making the right assumptions, but just wanted some reinforcement. My main concern was assuming that the pipe would flow full, and forcing the pipe to flow full. I haven't all the conditions from the customer, but I think they are near 250-300 gpm peak inlet rate. I was thinking of using a horizontal plate at top of the pipe to prevent vortexing and force full flow. However, this would increase the entrance losses into the pipe under the orifice condition. Any thoughts on this approach?

 

[JStephen]

I've seen details using the horizontal plate, but have no way of knowing if it's truly effective.

If you check flow rates with a large vertical drop, check absolute pressures along the route. With Bernouli's equation, you can derive physically impossible results.