Hi,
I am trying to balance the flow going into a reservoir. However, there is a snapshot showing a spike which has made me very hard to see the tolerance. I realized the spike happened when the reservoir level reach to the highest level i set at the control or maximum level. Could anyone suggest what is the best way to set a control at the reservoir?
Mark
Azizah,
I'm still looking into your issue. I see the spike that you are talking about, but I'm having a hard time locating the reason it's happening. One thing that I can tell you right now is that it looks like the spikes you are seeing are a product of the pumped flow that you have upstream from pump BP-00006. If you graph the flow out of that pump you'll see a similar spikey pattern. In the meantime while I look into the issue more I have seen a few things in your model that don't quite look correct to me and could help resolve the spikey flow issue. The first thing is the controls for pipes P-08656 and pipe P-11050. Both the open and closed statuses for these controls reference the level of 6.00m. The first thing to note her is that normally you want to stop a tank from filling all the way so the control for your pipe closing should probably be some number just below 6.00m. For example, your closing control should probably read something like "IF R-00014 Level >= 5.99 m THEN P-11050 Pipe Status = Closed". This will prevent the built in altitude valve that is located on the pipe from closing. You can read more about what happens when a tank becomes full or empty on this wiki technote . In this case there are multiple outlets for the water to leave the tank, so this may not be all that important, but it will at least help to eliminate some the user notifications and clean your model up. Another issue that I saw was there are no controls on any of the pipes leading to tank R-00016. If there are control one of those inlet pipes then you should create them in a similar fashion as I've just described above. You could always use a similar control for these situation like you have on pipe P-10813 that reads: IF R-00017 Level >= 6.00 m THEN P-10813 Pipe Status = Closed ELSE P-10813 Pipe Status = Open.
As far as I can see from your model the "node is disconnected" messages that you are getting don't have anything to do with the negative demands, but are rather causes by incidents were there are pumps in the model that are not on, which prevents flow from being provided to downstream elements. One example of this is at junction J-01031, which is disconnected from the system because pumps BP-00003 and BP-00004 are both set to a status of Off for most of the simulation. In order to determine this all I did was right click on the user notification that indicated the problem and choose "Zoom to", which brings me to that element in the drawing pane. Then what I did was trace upstream until I found the source of water. We have a tool for doing this automatically called Network Navigator (View > Network Navigator), where you can use the Trace Upstream to element of specified type function, but many times it's just easier to just do this manually by eye. There are good uses for network navigator though and you can find more information about it here: https://communities.bentley.com/products/hydraulics___hydrology/w/hydraulics_and_hydrology__wiki/3149
Regards,
I've found a solution for your issue. It looks like I was definitely on the right track with what I had referenced above about your controls that reference the same levels on your tanks, but now I can explain a little more about it. The first that you need to do in order to fully see the issue is to go into your calculation options (Analysis > Calculation Options) and set the 'Reporting time step' to " <All> " and then compute your model. Now graph pipe P-08795 and also notice it's location in the model, which is directly upstream of a tank for those who are following this thread and don't have the model to look at. The graph should like similar to what I am showing below:
There are sharp oscillations between about 1.8 hours and 5.2 hours. Now if you examine the control on that pipe you'll notice that they are set as I described in my post above based on the level of the tank, but they both use the same level value to open and close the pipe. In this case the controls read as follows:
IF R-00003 Level >= 6.80 m THEN P-08795 Pipe Status = Closed
IF R-00003 Level < 6.80 m THEN P-08795 Pipe Status = Open
Essentially, what is happening here is pipe P-08795 is opening and closing rapidly because of the controls that are set on it, which causes the spikey looking flow patterns to appear on your graph. An example of the the rapid opening and closing of the pipe by the controls can be observed if you browse to time 2.73 hours (Closed) in your time browser (Analysis > TIme Browser) then browse back to time 2.72 hours (Open) or ahead in time to 2.74 hours (Open). What this tells us is that controls too tightly constrain the model, so the small changes in flow are triggering the controls that you have set to open and close the pipe, which causes the sharp spikey oscillation in the graph. What you need to do in order to resolve this is spread out your control level values. For example, in reservoir R-00003 the minimum level for the tank is 0.1m and the maximum level is 7.0 m., so you if you have controls that read like this:
IF R-00003 Level < 4.5 m m THEN P-08795 Pipe Status = Open
As stated above this is just an example and you should use whatever the real control ranges are in the field on this pipe to resolve this issue. Doing this will get rid of the spikey flows from 0 L/s to 100 L/s. To get rid of the spikey flow from 100 L/s to about 168L/s you will need to look at your controls on the other pipes or pumps that you might have in the model. Make sure the controls on them are not set using exactly the same values to provide some space in between the opening and closing of the pipe.