I have a transient model of pump failure, and I put a surge valve as a protective device. According to the following image and from the black line did not get back to the initial state of the system like it is failed.
The hammer file is also attached.
File Download Link: https://we.tl/t-40VvFlgNKn
Hello,
The results you are seeing are accurate for a pump shut down. Without the pump adding head to the system, the system will tend to settle on an equilibrium with a lower head, like what you are seeing. If you need to see what happens when the pump shuts down and then turns back on, such as to see the timing of when it a normal state, you would want to model a Variable Speed/Torque transient pump type. This link has information on this: Modeling a pump startup and shutdown transient event in the same simulation.
The surge anticipator valve would not add head to the system. It opens when a certain pressure threshold has been reached in anticipation of a high pressure surge from the pump shutting down. I would recommend reviewing this wiki for more information: Modeling Reference - Surge Valves. If you are trying to maintain some head in the system when the pump is shut down, something like a surge tank or hydropneumatic tank may work better.
Regards,
Scott
Ok, but why in the program sample case of the surge valve, the system did not fail like my case? Just a fluctuation around the initial state.
Why the two cases are different ?
Most likely a check valve is closing in that case, preventing reverse flow (and the resulting downward sloping hydraulic grade from the headloss from that flow). Are you using a check valve at your pump station?
Jesse DringoliTechnical Support Manager, OpenFlowsBentley Communities Site AdministratorBentley Systems, Inc.
I used the check valve at the transient operational settings of the pump.
The downstream end of the surge valve sample model uses a reservoir. This is a source of head in the system. With the check valve on the pump, the system is balanced on a hydraulic grade around the elevation of the reservoir.
In your model, you are two discharge to atmosphere elements (which is a discharge point, and not a source of head), where an air volume is continuous forming throughout the simulation (since there is no head to expel the the air pocket that is forming without the pumps operating). This is the difference between the sample model and yours. In order for the profile to balance on a new equilibrium, additional head would be needed, like a surge tank. The pumps turning back on after some amount of time would also add head to the system.
But I changed the two D2A points to two reservoirs, and the problem still exists. I do not know where is the problem.
Please send us an updated copy of the model files and we can take a look.