Surge Tank Headloss Coefficient

Question

Hello good morning; 
 I am modeling a surge tank to protect a pipe for transient event. 
 When I dont consider headloss coefficient ( Headloss coefficient = 0) tthe system works normally but when I consider a headloss coefficient different to 0 (always positive), HAMMMER dont compute.

Additional Surge Tank input data: 
 
 Operating Range type: elevation 
 Elevation (Base): 726.91 
 Elevation (Minimum): 726.91 
 Elevation initial:731 
 Elevation (Maximum):734

Elevation (physical): 726.46 
 Volume(inactive): 0 
 Section: Circular 
 Diameter (m): 0.10 (it is a 4" stand pipe) 
 Treat as Juction: True

Surge Tank Type: Simple 
 Has check valve: False 
 Weir coefficient:1.84 
 Weir Length: 0.5 m 
 Diamter orifice (mm): 100 
 Ratios of losses:2.5 
 Headloss coefficient: ?

The pipe is a 450 mm diameter ductil Iron (lenght 6100m) with a flow of 100 l/s 
 Reservoir elevation: 731.23m 
 Valve close time:120 seg.

Where do you think is the error? Maybe the little diameter of the onlet and surge tank (4"=0.1m = 100mm) 
 Thank you very much for your support 
 Kind regards.

Jesse Dringoli · Answer

Hi Gustavo, 
 Ok, this type of issue can indeed occur when the solver encounter extreme conditions which can be due to unusual model input. It may be a combination of the very small orifice size, small tank diameter, and possibly other parameters. It may be having trouble converging if the water level is changing quickly due to the small size. However, I experimented with one of the example models and was not able to reproduce the issue. 
 Have you tried a smaller calculation timestep? If that does not help, please send the model and we will take a look. 
 In general if you do have a 4" standpipe that "floats" on the system during normal operating conditions and can supply some volume of water during a surge event, then it does sound like your approach here with the surge tank makes sense.

Sushma Choure · Answer

Hello Gustavo, 
 As suggested by Jesse in his reply, the problem here is with small surge tank diameter, in your case diameter is 0.1 m i.e. 100 mm, is that correct diameter ? If you change the diameter of surge tanks to say 0.5 m or 1 m, model runs fine. 
 How is your system ending in real, is it discharge into a water body i.e. open to atmosphere, there is a junction in your model at end. You could model it as discharge to atmosphere element i.e. D2A element. 
 Here is information about it, for your reference. 
 Modeling Reference - Discharge To Atmosphere

Scott Kampa · Answer

Hello Gustavo, 
 You would want to make sure your system is modeled as accurately as possible (and that the model is calibrated). The reason for this is that you want to make sure that the initial conditions at the start of the transient simulation is as accurate as possible. 
 Regarding the demand at the end of the pipe, the link that Sushma provided as steps for setting up the typical flow and typical pressure drop for the discharge to atmosphere element. These values will define the demand at the open end of the pipe. The link further describes how to derive these and which equations are used to do that. See the section called "Common applications of the D2A acting as an Orifice" in the link provided for more information on that. 
 Regards, 
 Scott

Jesse Dringoli · Answer

Gustavo Montero said: Now I am realizing that it drains into a secundary distribution storage tank that it is open to atmosphere preasure. 
 Are you saying that the 4" discharges water to the atmosphere, which then spills into a tank? If so, the D2A approach may work well if there is always water discharge. If upon downsurge, the pipe drains down, then the surge tank approach may still be best. (otherwise with the D2A, air will enter when the pressure drops to zero, but air tracking has limitations . 
 If you are saying that the 4" pipe connects to the bottom of a tank, and that the tank is open to the atmosphere, then this would be modeled with the surge tank element as well (4" pipe connected to surge tank element). 
 If you need further help, a clear diagram of what you are trying to model may help.