Tank never runs out of water (Hammer V8i)

Hello Anthimos,

Please provide the model files, so we can take a closer look. Instructions for uploading the models are provided here.

See: Sharing Hydraulic Model Files on the Haestad Forum

Here are the requested files: 

Note that purpose of the tank is to model a pressure breaker chamber.

Hello Anthimos,

HAMMER is not intended to model a tank becoming empty. The flow you see reported downstream of the empty tank is the movement of the water column away from the tank as the air pocket becomes larger. This is essentially the pipe draining, but the air pocket simulated in HAMMER (when the HGL drops below the tank base elevation) is limited to the location of the tank, so the impact of the draining of the pipe is not simulated.

See wiki article (I just added a section for HAMMER):

What happens when a tank becomes empty or full?

I'm not familiar with "pressure breaker chamber", but you may need to consider whether the impact of the drained tank is of interest to the overall transient results, or if you only need to focus on the transient that occurs just after closure of the valve. (in which case you could either accept the results you see from the limitation mentioned above, or omit/simplify the area around the tank).

First of, thanks for the quick response.
The pressure breaker chamber is the closest translation of what we are doing.
It is essentially a concrete chamber (around 2x3x3) that has the purpose of "resetting" the hydraulic grade line of the network, because it is getting too high.

As you stated, if the pipe is draining after the tank, is of no interest to me. The results didn't make sense and I thought about asking, to get a better understanding of how the tank modelling works.

What interests me, is if the tank lets the wave travel through, or if the network is kinda separated (upstream and downstream of the tank). Because after the closing of the TCV, I was expecting to see some some negative pressures travelling, or for the pipe to drain, which I didn't. Should I leave it as it is, or should I model it as a surge tank?

Can I use the tank for my simulation, or should I use it only as a boundary condition?

First of, thanks for the quick response.
The pressure breaker chamber is the closest translation of what we are doing.
It is essentially a concrete chamber (around 2x3x3) that has the purpose of "resetting" the hydraulic grade line of the network, because it is getting too high.

As you stated, if the pipe is draining after the tank, is of no interest to me. The results didn't make sense and I thought about asking, to get a better understanding of how the tank modelling works.

What interests me, is if the tank lets the wave travel through, or if the network is kinda separated (upstream and downstream of the tank). Because after the closing of the TCV, I was expecting to see some some negative pressures travelling, or for the pipe to drain, which I didn't. Should I leave it as it is, or should I model it as a surge tank?

Can I use the tank for my simulation, or should I use it only as a boundary condition?

Hi Anthimos,

The networks on either side of a tank are indeed separated; the tank's water surface elevation provides a boundary condition. You can observe this by animating a profile path in the Transient Results viewer. It is difficult to recommend the correct approach since it depends on the details of how this device works.

If the tank in your actual system does not fill from the bottom but rather spills into the tank from the top, then it would behave as an open discharge point, which you could model using the Discharge To Atmosphere (D2A) element (see item #2 in this article, under "Common applications of the D2A acting as an Orifice"). This would "end" the system at this point, since the transient waves would not propagate downstream past the open discharge point (the spilling down into the tank).

If the device is actually pressurized, then the transient waves would propagate through it, in which case it may be better modeled as a PRV element. This simply induces a headloss through the valve.

Regarding surge tanks - by default, the surge tank element will operate exactly the same way as the tank node element, during a transient simulation. The main difference between the elements is that the surge tank has some additional options such as a check valve (one-way surge tank) and the option to model a complex "differential" tank.