Positive flow is observed in a pipe downstream of a closed element (ie. pump or valve) during a transient simulation, and the flow continues to be positive regardless of the simulation duration.
Or, positive flow continues to move away from an air valve that opens during a transient simulation, and never stops as if the pipe never drains.
How does HAMMER's assumptions regarding vapor and air pocket tracking effect these situations?
If this is happening next to a closed valve, first make sure that the transient operational rule for your valve has the correct times and units.
Most likely, a vapor or air pocket has formed, and the separated water column is moving away from that pocket, which is expanding due to low pressure.
By default, HAMMER cannot track the air-liquid interface and assumes the pocket is concentrated at the location of formation, which is why this flow never stops. Animating the profile path of the area in question in the transient results viewer may be very useful in visualizing this behavior. Observe the air/vapor volume at the top of the profile.
In reality if there is a downstream reservoir at a lower elevation (lower than the HGL at the vapor/air pocket), the pocket might remain until the valve reopens, but would balance on a certain volume as the pocket travels down the length of the pipe and the pipe becomes partially drained. However, by default HAMMER does not track the liquid/vapor interface (see article below for more on air/vapor tracking limitations), so it would not be able to simulate this. If the downstream reservoir is at a higher elevation, then it is expected that the HGL will eventually settle on a positive value at the node in question, thereby diminishing and collapsing the vapor pocket (which HAMMER should be able to simulate).
You may want to focus on the impact of the collapse of the air or vapor pocket. For example for a pump shutdown, consider modeling a shutdown followed by startup, to see the transient impact when the air or vapor pocket is fully expelled.
For cases where an air pocket forms at an air valve at a local high point, you could try using the Extended CAV calculation option to track the air/liquid interface in the two adjacent pipes. If the ends of the adjacent pipes are at a lower elevation that the boundary conditions, then the HGL and flow may settle.
Modeling Reference - Air Valves
Assumptions and limitations of tracking air or vapor pockets in HAMMER
Surge mitigation for systems with intermediate high points experiencing negative pressure or spikes in pressure
Does the Extended CAV option apply to vapor pockets in the system or only air?