Hello Doan, Apologies for the delay in response as it took some time to fully analyze this model. Here are my observations and comments: 1) The hydropneumatic tank (HT) becomes empty at about 5.9 seconds, which you can see by graphing the gas volume in the Transient Results Viewer. This happens because the tank is at a higher elevation than the surrounding elements, and so has a higher chance of becoming empty or experiencing negative pressure. Results are not valid in this condition and you will see a user notification indicating that a larger tank may be required. There are several other parameters that impact a HT's performance, though. 2) The gage pressure at the HT becomes negative, but the gas law calculations use absolute pressure, which is still positive (see extended node data tab of TRV, graph gas pressure at HT-2. Also see: Differences in hydropneumatic tank gas pressure results ) 3) The Surge Tank (ST) becomes empty as well, at about 7.7 seconds. An air pocket forms after this point, simulating air from the empty tank entering the pipeline. Like the HT, the ST is also at a higher elevation than the surrounding elements. 4) There is a large headloss through the opening and the pipe next to the ST, which you can view if you create a profile from a nearby pump, to the ST. This is from the reduction in diameter from 2000 mm to 600 mm. 5) The negative pressure at the surge tank occurs because the pressure at the pipe next to the tank drops to zero before the pressure inside does (due to the headloss from the orifice). You can view this by checking the differences between the "level" and "head" in the table at the bottom of the Transient Analysis Detailed Report. If you change the headloss coefficient of the ST to zero for example, this does not happen because the tank and pipe pressure reach zero at the same time, after which air inflow begins. The main point is that this happens due to the tank becoming empty and an air pocket forming, which you will want to avoid anyways. 6) Both the ST and the HT have the "treat as junction?" option set to false, and the adjacent pipe flow is positive. This means the initial conditions are not at rest because the tanks are still filling/emptying at the start of the simulation. Set this to "true" to model the tanks "floating" on the system HGL, where they are neither filling nor draining in the initial conditions. 7) With tanks for protection, it's a careful balance between volume of water in the tank to supply mass during a transient event to keep the water column moving, vs. restriction through the opening to prevent the tank from becoming empty. With the large flows going through these large pipes, it appears that it would take larger volume of storage to prevent them from becoming empty within the timeframe being analyzed. As an example, if you make the following changes, the tanks will not become empty and not drop below zero pressure, at least for about 45 seconds after the pump shutdown. HT-2: - Treat as junction: True - Volume (tank): 10 m^3 - Volume of gas (initial): 3 m^3 E-ST1: - Treat as junction: True - Diameter: 4 m - Diameter (orifice): 1 m PS-13 (pipe next to E-ST1): - Diameter: 1 m Without a downstream boundary, the tanks will eventually drain out though, as long as the pumps remain off. See more here: Negative pressures occur during a transient simulation no matter what protection is used You may want to consider how long the pumps will be off and what other options you have for surge protection strategies. For example you could try placing the tanks at different locations, or try adding other measures like a higher pump inertia/flywheel.
↧