Demand load in Water Hammer

Mohamad Azzam said:

How is Bentley Water Hammer deal with Demand?

Demands during a transient simulation are always handled as pressure-dependent. HAMMER takes the initial demand and corresponding pressure and calculates a discharge coefficient from that, just like it does with the D2A element. So, as the pressure drops at the demand node, the outflow drops accordingly. If the pressure drops to zero, the flow drops to zero. From a search in our Wiki, you can find the following related article: How are demands treated during the transient simulation in HAMMER?

Mohamad Azzam said:

if I am modelling a transient for force main pipe from pump to the wet well (or Manhole) should I add the D2A element to the end of the pipe?

If there is a free-discharge between the end of the force main and where it empties into the wetwell, then the D2A element would make sense, as it may be easier for you to determine the pair of "typical pressure" and "typical flow" than it would be for you to estimate the outflow as a demand. Also, the D2A would permit air inflow if the pressure drops below zero, which would make sense in the case of a free-air discharge.

If the force main connects below the water level in the wetwell, then it would probably make more sense to model the wetwell as a tank. Different boundary conditions (reservoir vs. demand vs. D2A) will have a different transient response (different wave reflection behavior, etc)

Mohamad Azzam said:

when usually use D2A element? what are the cases to use that element (I know one case to model hydrant).

See the section "When to use it" in this article in our Wiki: Modeling Reference - Discharge To Atmosphere

The D2A element is primarily just a different way of establishing a pressure-dependent outflow. If you know the exact demand you want in the initial conditions, then it is often easier to enter a demand on a junction, but if you have some type of free-air discharge where you may not know exactly what the outflow is, the D2A element is helpful because you can calculate the pair of "typical flow" and "typical pressure drop" for example using the orifice equation. For example you could pick an estimated flow and use the orifice equation to calculate the corresponding pressure drop based on the size of the opening to the atmosphere, or vice-versa.

The other differentiating factor with the D2A element is how it models air inflow upon sub-zero pressure. This is why it is best when modeling any discharge to the atmosphere (as the name implies)

Jesse Dringoli said:

If there is a free-discharge between the end of the force main and where it empties into the wetwell, then the D2A element would make sense, as it may be easier for you to determine the pair of "typical pressure" and "typical flow" than it would be for you to estimate the outflow as a demand

If I am modelling a force main connected to a tank and the pipe invert below the water level; you recommend to use D2A element; do you mean to replace a tank with the D2A or placing the D2A before the tank? 

the pressure here at the tank is atmospheric pressure; what is your recommendation; may you please elaborate a little about your approach?

Jesse Dringoli said:

If you know the exact demand you want in the initial conditions, then it is often easier to enter a demand on a junction

meaning if I am modelling a water network and I place a demand on junctions; no need to used the D2A element since the junction will act the same here no in air would enter the network once connected to a meter. am I right?

Mohamed, if the end of the pipe has a free discharge and the water spills out and into the tank from above, then use the D2A since that is a better representation of a free discharge and can model air inflow. If the end of the pipe is submerged, model it as a tank. If in doubt, try a sensitivity analysis (try different approaches and see if the change in wave reflection at the boundary has a significant impact on the transient envelope). 

Mohamad Azzam said:

meaning if I am modelling a water network and I place a demand on junctions; no need to used the D2A element since the junction will act the same here no in air would enter the network once connected to a meter. am I right?

The junction-with-demand will indeed act the same way as the D2A during the transient simulation, but they offer two different ways for the data input to establish the outflow. Meaning, with the junction demand, you enter the exact outflow you want and the initial conditions solver computes the pressure around that (and the pair is used during the transient simulation to model the pressure-flow relationship). With the D2A, you enter a known pair of flow/pressure, and the initial conditions solver calculates the pressure and flow based on that relationship (and just like the demand, the pair is used during the transient simulation to model the pressure-flow relationship).

The other difference is that the D2A can model air inflow upon sub-zero pressure while the junction demand does not.

Here is another related article for your reference: Options for modeling an outflow that varies with pressure

Thank you Jesse 

Everything is clear; however, in the model presentation and layout D2A for free discharge do I have to remove the downstream tank and provide D2A element in water Hammer? assume we are modelling the force main in watercad and now need to do the surge analysis in water hammer. please clarify.

Yes, if you're modeling a top-fill tank in HAMMER (top diagram in my previous reply labeled "D2A"), the D2A would replace the tank element. See the bottom of this article: Modeling top fill tanks and throttling inlet valves