Subject: problem with reservoirs operation using previous time step discharge/volume Posted: 7/4/2016 Viewed: 15372 times
Hi community, I'm writting you to ask for a problem I have had using WEAP's functions to distribute water using Previous Time Step rivers discharge or reservoirs volume.
In a case, the water demands (hydropower and agriculture) depend on the water level/volume in a reservoir. In real life, those demands (and delivery) are almost on time, but in WEAP it depends on a previous time step so the observed and modeled deliveries have a lag of a month.
In another case, a natural lake discharge depends on its volumen elevation curve and on its maximun hydraulic flow. But in this case, WEAP always give me an error message saying that on several months it didn't use the curve due to computation unstability. So I had to use a Flow Requirement downstream, using the same elevation-curve. Also in this case, the modeled discharge has a lag of one month with respect the observed one.
Do you have an idea how we can solve these two problems? thanks a lot.
Ms. Stephanie Galaitsi
Subject: Re: problem with reservoirs operation using previous time step discharge/volume Posted: 7/8/2016 Viewed: 15327 times
If I understand correctly, you have used an expression (in the expression builder) to tie the system demands to the reservoir volume. How does this work in the actual system? Does the farm wait to find out how much water is in the reservoir before determining its demand?
I'm wondering if it might be better for you to plug in information about reservoir operation levels, hydropower production priorities/buffer coefficient, and agricultural demand priorities. WEAP has many functions built in so you can program the system's actual operations rather than writing expressions about the previous time steps that only approximate what's happening. Can you try to plug the demand priorities into the model?
Can you explain the error message about volume-elevation curve instability? Or you can also send us a report, which is one of the prompts in the window you'll see when you receive the error message.
Mr. David Poblete
Subject: Re: problem with reservoirs operation using previous time step discharge/volume Posted: 7/11/2016 Viewed: 15313 times
Hi Stephanie, thanks for your reply. I'll try to explain better my problem.
Actually the demand from the agriculture/hydropower is quite fixed (monthly values), but the demand from the reservoir depends on the discharge of the main river and the water rights they possess (in the same timestep). So yes, we have written a series of rules in the expression builder (of Flow Requirements) with the use of key assumptions, so each timestep the comparison is made and the instruction to release or not water from the reservoir depends on the difference between the actual discharge (modeled by WEAP) and the discharge they need every month. So the problem is that the actual discharge is always taken from the previous timestep and not from the present.
And for the other problem (other model), we used a cubic relation between the storage elevation in a natural lake (no intervention) for the Maximum Hydraulic Flow. The error message after running the model was: "The Maximum Hydraulic Flow was removed because it causes LP infeasibility...". And the results gave us a lake with no fluctuations in its water level.
Thanks for your help.
Mr. David Poblete
Subject: Re: problem with reservoirs operation using previous time step discharge/volume Posted: 11/28/2016 Viewed: 14772 times
Hi, I'm writing again if anyone knows how to deal with the LP Infeasibility problems associated with reservoirs Maximum Hydraulic Flow.
As I said in the post above, everytime I run WEAP with some curve of MHF I got this message: "The Maximum Hydraulic Flow was removed because it causes LP infeasibility...".
Thanks in advance.
Mr. Jack Sieber
Subject: Re: problem with reservoirs operation using previous time step discharge/volume Posted: 12/9/2016 Viewed: 14632 times
If the flow into a reservoir is much larger than the total capacity of the reservoir, this can cause problems for the LP solver in WEAP in conjunction with the Maximum Hydraulic Outflow for the reservoir. (There is no problem if you do not have an MHO on the reservoir.)
Could that be the case in your model? Do you only get this message in month with high flows into the reservoir?
Mr. David Poblete
Subject: Re: problem with reservoirs operation using previous time step discharge/volume Posted: 7/24/2017 Viewed: 12652 times
Hi Jack, with the latest updates to WEAP, the LP infeasibility problem has been solved. But the original question remains the same: basically I don't know how to mimic the natural behavior of a lake, where its outflow depends on its current volume (or stage) on the same timestep.
Currently I'm using the previous time step function in the MHF tag, but this gives me results with 1-ts lag, compared to the observed discharge in a downstream gauge station.
Ideally, (for me) the MHF should be a elevation-discharge table, similarly to the elevation-volume curve, so WEAP could calculate the discharge of a natural lake in the current time step, as it happends in real life.
Please let me know if this is possible using other WEAP functions that I have missed. Thanks!
Topic "problem with reservoirs operation using previous time step discharge/volume"