Physically, what kind of wells will be considered to be not alive?

[GitPaean]

From the code, it looks like if a well does not inject, and also no flow towards well-bore from reservoir happens, the well will be considered not alive.

But in the actual simulation, what kind of set up will make a well not alive? Thanks.

[totto82]

For instance setting the injection rate to zero will as I understands it result in a not alive well. You may still have crossflow in a not alive well in cases where the total flow towards the well-bore from the reservoir equals the total flow out of the well-bore into the reservoir. For not alive wells the control equations will be trivial and special treatment is therefore needed.

[GitPaean]

The formulation is as follows, (compi.col(pos) * injectingPhase_selector.select(q_s,ADB::constant(V::Zero(nw)))) - q_ps;

The first term only handles the injecting phase, which will be positive in values (>=0). And the second term only handles ProducingPerforations, which means the flow towards the well-bore, and will be negative in values (<=0).

It looks like if there are some flow towards the well-bore happen for a well, it will be an alive well. It does not matter if the sum of the in-flow and out-flow is zero.

Please let me know if there some misunderstanding here.

[totto82]

You are right. The total wellbore volumetric rates (wbqt) consider the total inflow to the well-bore and is used to calculate the mixture of the fluids inside the wellbore. I.e. It means that a well is not alive if there is nothing flowing into it neither from the reservoir or from the top side.

[GitPaean]

Thanks.