search

vicelab / cen-sierra-pywr

Code base for modeling the central Sierra Nevada hydropower systems
4 stars 7 forks source link

ramping rate IFR below Collierville PH logic #46

Open drheinheimer opened 5 years ago

drheinheimer commented 5 years ago

This is in reference to #43. However, unlike below diversion dams, this is for below where Collieville PH discharges into the Stanislaus. According to @alanccai, this is from the license (please confirm and provide more context):

image

drheinheimer commented 5 years ago

And for reference, here is how Collierville is simulated currently:

image

alanccai commented 5 years ago

The language is a little ambiguous as is typical of older license documents. Article 50 of this document says the ramping rate requirement applies to each "project discharge facility". They don't list the facilities it applies to. So if you define that literally as a facility from which water is discharged, that would include Collierville Powerplant, since water is discharged from it into the mainstem Stanislaus River. The full list of project components is copied below:

image

alanccai commented 5 years ago

1982 document which those screenshots are taken from: P-2409_NorthForkStanislaus_1982_OrderIssuingMajorLicense.pdf

alanccai commented 5 years ago

I will contact the operator of this project to clarify which places the ramping rate restriction applies to.

drheinheimer commented 5 years ago

That's useful. It remains somewhat unclear whether this applies to the PP or to the river. There doesn't seem to be a gauge in the river itself, but there is a guage for the PP.

I added a min & max constraint to the river:

The method is to use a piecewise IFR: the first block (min flow) has a high negative cost (benefit), the second block (in between min & max) has zero cost, and anything above the max has a high positive cost (penalty). I applied an initial value of 75% and 125% of previous flow for min & max (i.e., 25% up & down ramp rate), using Python functions.

Result is in first fig below. The river is generally managed to stay between the min & max, which is great. The PP, on the other hand, doesn't quire resemble actual operations (second figure below). It's a little hard to see, but the daily variability is way too high for the PP vs. observed. So this makes me wonder if the ramping rate constraint is actually applied to the PP instead of the river. Note the scales on these figs are completely different.

The good news here is that the general approach for applying an up/down ramp constraint does appear to work.

image

image