Non-zero flow in closed pipes next to hydropneumatic tanks

Question

Ref: http://communities.bentley.com/products/hydraulics___hydrology/f/5925/t/123522 
 Thank you for the response. I was able to get the booster pump to turn on and prevent the hydropneumatic tanks from emptying completely. All four tanks are connected to the main waterline by individual pipes (one pipe for each tank) that connect to a common junction on the main line. With the EPS I've been running, the four tanks should be inactive during the day (with one or two booster pumps providing all the needed flows and pressures during the day), and there should therefore be no flow in the pipes that connect to the tanks. However, whenever I graph the flow for any of these four pipes, I get the plot shown below. These spikes and drops in flow are causing the tanks to experience abrupt changes in pressure and liquid volume when these parameters should remain constant during the day (due to the pipe being closed). I have controls set up so that the pipes are set to be closed, so I'm not sure why these spikes and drops in flow are occurring in the first place. Do you know why they may be occurring?

Jesse Dringoli · Answer

Hello Rob, 
 It looks like the zip file only contains the .WTG file. Please include the .WTG.SQLITE file as well, as both are needed to open the model. 
 In the meantime, some general suggestions - if the hydropneumatic tanks are offline or "float" on the system, try setting the "treat as junction" property to "true", for the appropriate scenarios. 
 For the other issue - try a smaller calculation timestep. This tends to be the cause of many EPS problems with hydropneumatic tanks. 
 See also: 
 How are surge tanks and hydropneumatic tanks (hydro tanks) used in WaterGEMs and WaterCAD models?

Jesse Dringoli · Answer

I concur with Mark that if you have multiple hydropneumatic tanks in hydraulically close proximity (and hydraulically connected - not closed off), you should combine them into a single equivalent tank. You may need to do a few calculations with the gas law to ensure that the tank's relationship between pressure and volume is equivalent. 
 
Consider simple pump controls - one to turn off when the pressure gets to the high setpoint and the other to turn off when the pressuer gets to the low set point. If the pressure "over shoots" the control range, try a smaller timestep. 
 
If the tanks are taken offline in a particular scenario, set "treat as junction" to "true".

MPachlhofer · Answer

Robert, 
 
The following information can be found in the HAMMER help documentation for the topic titled, 'Gas Law vs. Constant Area Approximation' and 'Initial Conditions Attributes': The effective volume represents the difference in volume between the HGL 'On' and HGL 'Off'. The HGL 'On' and HGL 'Off' fields represent the maximum and minimum hydraulic grade lines within the hydropneumatic tank (i.e. when an associated booster pump would turn on or off). An approximate diameter is computed based on the effective volume of the hydropneumatic tank so that the tank cross sectional area multiplied by the distance between HGL on and HGL off gives the same volume as the hydropneumatic tank. 
 
In order to help you out with the 'volume of Gas (Initial)' question I'll need to know what version of WaterCAD, WaterGEMS, or HAMMER are you using? The version number can be found by going to HELP > About > <ProductName> inside your program and locating the number in the lower left hand corner and may be in brackets. It should start like this 08.11.XX.XX where X are some numbers. I don't see a reason why you shouldn't be able to enter this value in the field. 
 
The 'pressure (gas-preset)' field is used if there is a bladder in the tank . It's the pressure of the gas prior to exposing the tank to pipeline pressure and should be otherwise, this should be omitted as it is ignored. 
 
To test your theory on changing the inlet orifice diameter and gas law exponent you could perform a sensitivity analysis as outlined in the wiki below. 
 
 communities.bentley.com/.../22932.how-do-i-perform-a-sensitivity-analysis 
 
Regards, 
Mark

MPachlhofer · Answer

Robert, 
 
To add to what I posted yesterday, it's difficult to tell whether you're using WaterCAD, WaterGEMS, or HAMMER for this model. It seems like you're using WaterCAD or WaterGEMS because you were referring to running an EPS simulation. If you are using WaterCAD or WaterGEMS it's not necessary to input any information like the 'pressure (gas-preset)' or gas law exponent under the 'Transient (physical)' section of the hydro tank properties because that information is only used during a transient simulation for HAMMER. 
 
For more information on hydropneumatic tanks see the wiki solution below. The topic is extensive and covers most of what you need to know about the element. 
 
 communities.bentley.com/.../modeling-reference-_2D00_-hydropneumatic-tanks 
 
Regards, 
Mark

Jesse Dringoli · Answer

Robert, 
 
I agree with Mark - the properties of the hydropneumatic tank that you asked about in your previous reply are all related to transient simulations in Bentley HAMMER. HAMMER uses the same file format as WaterCAD and WaterGEMS, so a steady state or EPS model can be later opened in HAMMER for a transient simulation. For a steady state or EPS, the transient attributes are not used. 
 
Regarding the "effective volume" - that only applies to the "constant area approximation" option in the "tank calculation model" attribute. It is a legacy option, mainly for compatibility with the method by which you would model a hydropneumatic tank in versions of the program before the hydropneumatic tank element was introduced. So, in almost all cases, you would use the "gas law" option, in which case the effective volume does not come into play. This is explained further in the "Initial conditions" sections of the hydropneumatic tank technote that Mark linked to.