Subject: Natural channels with different zones Posted: 2/27/2018 Viewed: 5541 times
A water management model setup we are trying replicate in WEAP has different zones in natural channels: an ideal zone and one or more relaxation zones. So if the model cannot satisfy the flow required for the ideal zone, it will progressively try to satisfy the flow in the relaxation zones. The ideal zone has no penalty associated with it whereas the penalties incurred for the relaxation zones become progressively higher. How could one implement this policy in WEAP? Would you have a flow requirement for each zone, with the ideal zone minimum flow given the highest priority, and the first relaxation zone minimum flow given the second highest priority etc?
Ms. Stephanie Galaitsi
Subject: Re: Natural channels with different zones Posted: 4/16/2018 Viewed: 5526 times
By zones, do you mean different volume levels in the river? So when the river is very full, that is the ideal zone, and then if the volume decrases, those are the relaxation zones? And then, if I understand correctly, you want the withdrawals in the river to have a penalty if the river is in one of the relaxation zones (and penalty will be according to the zone).
So the volume of the river is a calculated result, meaning that it cannot be referred to as part of the calculations for the same time step. WEAP wants to avoid any infinite loops in calculation. I do understand that there is a downstream element here, which would avoid the loops, but the WEAP calculations unfortunately don't recognize that.
What you can do is link the demand site withdrawals to the original data generating the river flow - in this case the data for the river inflow (or the precipitation generating the inflow, if you're using catchments). So you can say something like "If(lower threshold<input parameter<upper threshold, Maximum volume of transmission link = penalized volume, non-penalized volume). And you can nest if statements. Tying the flow to one of the input parameters will control the system responses in the same time step.
Otherwise, you could have the exact value for streamflow in the respective reach that you're examining, but only for the previous timestep. That would be appropriate only if there is high autocorrelation in your system.