Hydropneumatic Tank Gas Volume

Question

I have a model including two main line at east and west side of a pumping station. I have included two Hydropneumatic Tanks at west and east side. The problem is that for the tank in the west side, the gas volume is increasing no matter how the size is (figure below). If we put 2000 L as the size of the Tank, the gas volume is more than 2000 and if the tank size is changed to 5000L, the gas volume would be 5000 L (as increasing line) and so on. For the east side , the gas volume increases to the tank size and then decreases, if we put 20000L as the tank size the maximum gas volume will be 20000 and if we change the size to 50000 the maximum gas volume would be 50000L and so on. 
 Could you help me to find out the issue? 
 Thanks,

Yashodhan Joshi · Answer

Hello Vali, 
 Try and use a value like 10 for Report Period (Transient) in the Transient Calculation Options. See if this helps with the linear increase of gas volume. See this article which provides more details on this: Hydropneumatic tank Extended Node Data graph shows linear continuous increase 
 Let me know if this helps.

Scott Kampa · Answer

Hello Vali, 
 Are you seeing any user notifications related to the hydropneumatic tank? If so, information on these may be useful. Otherwise, I would recommend reviewing this wiki with general information on hydropneumatic tanks: Modeling Reference - Hydropneumatic Tanks . This includes information that may be useful for what you are seeing. Overall a change to the properties of the tank will impact the results. A bigger tank will allow for a greater gas volume. Overall, you would general size the tank so that it does not become empty.

Jesse Dringoli · Answer

To add to Scott's response - 
 
 As mentioned in the technote , an empty tank condition cannot be modeled and HAMMER simply continues to use the gas law equation to calculate the increase in air/gas volume (beyond the total tank volume) as it cannot accurately determine what will happen otherwise in this condition. Hence, the importance of sizing the tank such that this does not occur 
 as you change the total tank volume while keeping other parameters the same, you are changing the relationship between pressure and flow in the gas law equation. This is also mentioned in the article and explains why the curve is different (and still becomes empty) when you change total tank volume. You may need to consider the other factors involved such as the orifice diameter and losses, initial gas/liquid ratio, and (if applicable), the bladder preset pressure as these (along with the total size) can all impact the performance of the tank. See the article for more details. 
 There are some situations and topographies in which a hydropneumatic tank will always drain eventually no matter how large you make it. For example if the downstream tank or reservoir is at a lower elevation than the hydropneumatic tank, the tank will eventually drain as long as the pump remains off. Read more about it here . 
 I recommend animating your transient results to get a better visualization of what is happening during the transient event, and potentially better understand how the tank is responding. Read more here .

Jesse Dringoli · Answer

Vali, there is a balance between the different parameters - you may need to adjust the tank size at the same time that you adjust the preset pressure and the inlet orifice (headloss). As seen in the article, the pair of initial conditions pressure (P) and volume (V) change the gas law relationship so they are interrelated. The orifice size and ratio of losses can impact the rate of flow and thus the rate of volume change. You want the PV relationship and headloss to be such that the tank drains quickly on downsurge (to keep the water column moving and improve the pressure drop from the transient) but does not drain too fast so as to cause it to empty completely. You could also consider trying a sealed tank instead of a bladder. 
 You may also want to take a closer look at your pump shutdown + startup configuration. Use this article to ensure that you are properly capturing what actually happens to the pump, using the operating rule pattern and control type (torque or speed). 
 Also check the inertia value for your pump if you are using torque as the Control Type, as it can have a large impact on the severity of the downsurge (and resulting emptying of the tank). See this: 
 How does pump inertia effect the pump calculations during a transient simulation? 
 If the tank is at a high point that could be a major contributor to the problem, since the tank will continue draining as long as the pump is off, since no other sources are high enough to prevent that from happening. So, if the pump turns back on after 200 seconds, the tank will need to be large enough to prevent a full draining within those 200 seconds. This may not be feasible in which case you may need to consider the cost trade-off between a very large tank at the high point (or a tank that is not able to provide protection because it cannot be allowed to drain fast enough on downsurge) vs. a smaller tank closer to the pump and at a lower elevation, but with potentially large costs to install in what I assume would not be an ideal location for such a facility. HAMMER can help you present these options, by using scenarios and alternatives to show the difference in performance between the options (big tank at high point vs small tank closer to the pump). 
 You could also consider other protection strategies, such as the use of air valves along with a higher pump inertia ( flywheel ), as just one example.