Surge vessel volume keep on increasing & not getting stable curve on volume, pressure & and also water inflow at surge vessel

Question

5355.LIFT STATION.rar 
 
 Dear Bentley Team, 
 we are facing some problem related to surge vessel, we have attached Hammer native file for for your review. 
 Problem is Gas volume, Gas pressure & Water inflow graphs for the surge vessel is not getting stable whatever run duration (sec) if we provide also, volume is keep on increasing, but in other simulations we are getting stable graph (like a straight line) if we increase run duration (sec), see refer below screenshot taken for other simulation. But in this simulation we are not getting any straight line on graph, request you to review & provide any recommendation/suggestion. 
 
 Thank you

Yashodhan Joshi · Answer

Hello Jayakanthan, 
 I checked your model and I can see the results getting stabilized around 312 seconds with a longer run duration (of 600 seconds) as compared to what you set at 120 seconds. See the snapshot of inflow at the surge vessel when the run duration with 600 seconds; 
 
 Can you try with a lower volume for the surge vessel?

Yashodhan Joshi · Answer

Hello Jayakanthan, 
 Is the model changed from what you shared before? In the response you mentioned that you are getting a stable graph now but you are now getting a user notification "The bladder-type hydropneumatic tank fills its containing tank.". This message usually means that the tank is getting empty during the transient simulation. See this article below which explains how this happens and what you can do to fix this; 
 Hydropneumatic tank user notifications: "gas volume exceeds volume of tank" or "fills its containing tank" 
 Let me know if this helps.

Yashodhan Joshi · Answer

Hello Jayakanthan, 
 If you have updated the model please share your model with us. If you are seeing the same error " The bladder-type hydropneumatic tank fills its containing tank" from the first day it could be an issue with the output files which shows the same user notification. 
 
 Try compacting the database: Go to File > Database Utilities > Compact Database. Choose the database ( .wtg.sqlite file) of your model and compact it. Then reopen the model and check if the crash occurs. More information on this can be found in this article: Compacting a database for optimal performance and general troubleshooting

Try separating the model files: Copy and Paste the model files ( .wtg and .wtg.sqlite ) into a separate folder and try to open the model from there (double click the .wtg file). See if this helps. 
 
 Here is your model which I tested: 3872.LIFT STATION.rar 
 However, this is the same model which I tested on my side and not seeing any user notification of " The bladder-type hydropneumatic tank fills its containing tank". See the image below from HAMMER 10.03.01.08 where I tested this; 
 
 If you have changed the model and are now seeing the " The bladder-type hydropneumatic tank fills its containing tank" notification please share the updated model with us so that we can see the same on our side. See the article on this too in my earlier response.

Yashodhan Joshi · Answer

Hello Jayakanthan, 
 I am now seeing the " The bladder-type hydropneumatic tank fills its containing tank" with a simulation time of 400 or 600 seconds. However, this indicates that the hydropneumatic tank is emptying at this point. When the hydropneumatic tank water inflow level is reaching 0 and stabilizing, it means that the entire tank had emptied. 
 Do you want the entire surge vessel to empty? Because a completely empty surge vessel creates situations which are difficult to model in HAMMER. See this article I shared before for details. 
 I closely went through your model and you have setup the pumps to shut down after a delay of 10 seconds. So initially, for 10 seconds the pumps are on and after 10 seconds (i.e. after shutdown) the hydropneumatic tank starts supplying water downstream. You can see the volume decreasing slowly over the duration of 120 seconds. 
 What you can look at is the profile animation of your model. In your transient calculation options you have set the "Generate Animation Data?" to "False". Set this to "True" and run the model. In the Transient Results check out the profile "Pump Discharge" to see your maximum and minimum hydraulic grades. See this article for details on how you can do this: Using Transient Results Viewer animations for visualizing a transient simulation 
 Additionally see this article which can help you understand how to read the transient animation data: What do the different colored lines in a Transient Results Viewer profile represent? 
 In your model, the minimum pressure is going till -0.65 bars which is above the vapor pressure of -0.98 bars. Also, air / vapor volume generated is 0 L. Without protection (i.e. without the surge vessel) the transient pressures are falling to the vapor pressure (-0.98 bars) and the maximum air / vapor volume is around 0.5 L. 
 See if these results are acceptable for your system. The surge vessel is acting as it is setup and mitigating the transients in the system (as evidenced from the results produced in the "without protection" situation).

Jesse Dringoli · Answer

Jayakanthan, 
 As you change the hydropneumatic tank total volume, you also change the pressure vs. volume relationship because you have chosen to use the bladder-type tank and likely have kept the preset pressure the same. To understand how this works, please see the article previously provided, plus the sections titled "Bladder" and "Transient Simulation Behavior" > "With a Bladder" in this article: Modeling Reference - Hydropneumatic Tanks 
 Also, this appears to be a pump shutdown event with no downstream storage. So, as long as the pumps remain off, no matter how large the tank size is, it will eventually become empty and cause the problem in question. You may need to include any missing boundary conditions (tanks or reservoirs that are in the actual system) or determine how long the pumps will need to be off, in order to size the tank. The following article goes into more detail on this situation: Negative pressures occur during a transient simulation no matter what protection is used 
 If you turn on profile animations as Yashodhan mentioned, you can visualize how the tank eventually drains if there is nothing else to provide head when the pumps are off.

Jesse Dringoli · Answer

Hi Jayakanthan, 
 If the reservoir elevation is too low, the hydropneumatic tank could certainly still drain, and the results should be accurate. Reviewing the animated profile results as previously suggested should help you visualize this. You would need to use the advice from the previous article along with your engineering judgement to decide what needs to be done to prevent this from happening (again using an example of determining how long the pump will actually remain off). 
 Note that a reservoir would simulate the boundary condition of a fixed hydraulic grade and water supply. If the pipe is connected to the bottom of the "Atmospheric tank", a reservoir may be a good approximation, but if it connects to the top, above the water surface elevation in the tank, then a D2A node would better represent the boundary condition as it can model air inflow.

Jesse Dringoli · Answer

Hi Jayakanthan, 
 Jayakanthan said: The problem is with surge vessel elevation, we have decreased the surge vessel elevation & verified, now we are getting stable curves. 
 Yes, as noted in a previous reply, the elevation of the tank needs to be lower than the downstream storage in order to have a chance of not emptying if the pump remains off. 
 Jayakanthan said: One more final query: Is it really necessary to get the stable curve (Gas volume, Gas pressure & waterinflow) after some time ? 
 In general, a transient simulation will usually settle on a final "steady state" where results like flow and pressure become constant/steady (a transient is a transition from one steady state to another), but there are some factors to consider: 
 
 If extreme conditions occur such as tanks becoming completely empty or large amounts of vapor or air entering and staying in the system, it may be very difficult (or take a long time) for the model to reach a true final steady state. Care must be taken to understand the assumptions in the software and how it is handling certain conditions occurring in the model, and adjust the model as needed to make sure it is an accurate reflection of the real system. In some cases if the model does not settle to a final steady state, it may still provide you with the answer you need - for example if a high or low pressure occurs before reaching a final steady state, that may tell you that something needs to be done to prevent that from occurring, such as a change in the transient protection strategy or configuration. 
 If the transient conditions are such that a zero flow is expected to be reached at the end of the transient simulation in some or all pipes, the transients may not settle down or will take a very long time to settle down because of the absence of friction. The following article explains this in more detail: Transient pressure wave not dampening or unexpected lack of headloss 
 In rare circumstances, the transient numerical solver may have difficulty converging, resulting in unstable results that are skewed by numerical error rather than hydraulic instability. In such cases you will typically see user notifications indicating such problems. These cases can usually be resolved by fixing problems with data entry, but can sometimes be related to extreme conditions that are expected to occur (see point above). In some situations the information in the following article can help: Changing transient solver accuracy and max trials