PDA

View Full Version : Flow Control Valve Irregular limiting capability



FCJR
August 14, 2016, 03:37 AM
Hi All,

As again I've stumbled to a confusing FCV behavior. Hope someone can give me some light on the situation. anyhow this is how it goes... I have this complex water network going on, but I've chopped down the elements concerned, so that it would be easier to be analyzed. this chopped network that I have consists of Reservoir-pump-tank-pump-node with demand. (I'll attach illustration later on to make it easier to visualize). I have digitized a FCV in between my first pump and tank so that it will regulate the refilling rate on my tank. Without any FCV to limit it, makes my entire refilling finish in just 2 to 3 hours, which is not fine by me because it makes the flow on my first pump really high which gives as well the warning flow exceeded on the first pump. I was intending to regulate the flow so that pumping would be a continuous and gradual operation and flow on first pump won't exceed its pump curve. Hence, my short network now goes like "reservoir-pump-FCV-tank-pump-node".

Now this is where it becomes tricky... Whenever I set my FCV to a certain setting, (e.g. 130 lps,140 lps... etc), it does regulate the flow following the specified setting, on certain time periods of the 24hr EPS while at some of the other periods it exceeds. I have attached some illustrations below to give you an idea.
319320321
Do you guys have any idea what might possibly cause the FCV to not follow the specified setting??? all flow average of the end node of my network is accurately the same so it means that my flow average is constant as dictated on the pattern I loaded the node.

Please help.

regards,
Felipe

Patrick Moore
August 15, 2016, 12:25 PM
Felipe,

I would guess part of the problem is that you may have a control on the FCV and when you are turning it "ON" with a statements like "If Tank Z level < X Then FCV "Y" is OPEN" instead of "If Tank Z level < X then FCV setting = 130 lps" This can also occur in Initial Status statements or in Rule Based or simple controls. To turn a valve "ON" you need to assign it a setting not tell it to be "OPEN" as this will make the model treat it like an open pipe rather than a control valve until another control statement changes the status or setting.

It looks like the initial status is set to "OPEN", so please check that for sure. To get a valve to run initially use an initial status of "None" and it will use the setting seen in the model explorer. If you want it initially closed set the initial status as closed.

I would recommend you first check the initial status and any rule based or simple controls associated with the valve.

One other thing to consider is the following



Does the real system have a control valve here?
Is your pump definition off so that you are trying to compensate with a control valve?


Generally a real world system would in most cases not have a control valve before the tank and would simply control the flow into the tank by the pump operation and pump sizing. In addition, the mathematics of the situation you are setting up with a pump then a FCV, can get a bit tricky for the model and should only be used if the real system had this arrangement.

You may wish to review the actual system to see how it operates and verify the following items as you may simply have something off that is making the pumps run differently than expected.


Are the pumps VFD's (variable speed) - Pump Speed control would likely be used to control pump flow or discharge pressure
Is the reservoir supply HGL accurate? This will play a huge role on where the pump operates
Is the pump curve or pump definition accurate for both pumps? The characteristics assigned to the pump will dictate how the pump operates.

The higher the head gain of the pump, generally the lower the pump flow, the smaller the head gain, generally the higher the pump flow.
Quick check of expected head gain subtract the overflow HGL of the tank (elevation plus max height) from the HGL of the reservoir. How does this compare to the design head assigned to the pump. If they differ widely from the actual pump in real life, something is wrong.


Is the elevation and max height of the receiving tank accurate? This will also impact where the pumps operate.


If you find an issue with one of the above 4 things you may find you don't need the FCV at all and you can make adjustments to the data that is least reliable whether that is the pump data, or the reservoir head to have the pumps operating as expected to the real world without adding unneeded complexity.

If you find you have checked all this data and found not issue you can contact us at support@innovyze.com for further assistance.

Patrick Moore

FCJR
August 16, 2016, 10:06 PM
Hi Patrick,

It appears that the previous modeler has created a pattern for the pump. If it was for calibration purpose, that I haven't had the chance to confirm. I've removed the FCV and created a new pattern myself for the pump.

Thanks a lot as usual!

Regards,
Felipe