mbakker7 / ttim

MIT License
34 stars 23 forks source link

second head specified element effects discharge result #16

Open PilbaraPanther opened 5 years ago

PilbaraPanther commented 5 years ago

When the head of the second element is fixed at 0 for t>0 and t<t1, the discharge of the second (and more elements) is compensated Y adding to the system to keep the head at 0. This effects the discharge results of other elements. for example, 2 head specified wells, running at different times, the second well adds to the system to maintain the head at 0. The first well increases the discharge to maintain the specified head.

The issue is experienced HeadWell and HeadLineSinkString

mbakker7 commented 5 years ago

Can you submit a small scripts that shows the problem?

PilbaraPanther commented 5 years ago

It seems that because the headinside the second HeadWell (w2) is fixed at 0 for t>0 and t<t1, the model needs to add water to keep it at zero. This addition of water is included into the model water balance and thus the discharge of w1 needs to be increased to get to the head specified in tsandh of the HeadWell. Example below

`ml = ModelMaq(kaq=10,z=[20,0],Saq=1e-5,tmin=1e-3,tmax=1000) w1 = HeadWell(ml, xw=0,yw=0,tsandh=[(0.1,-5),(100,-10),(200,-15)], rw=1) ml.solve()

results

print(w1.discharge(t=120))

ml = ModelMaq(kaq=10,z=[20,0],Saq=1e-5,tmin=1e-3,tmax=1000) w1 = HeadWell(ml, xw=0,yw=0,tsandh=[(0.1,-5),(100,-10),(200,-15)], rw=1) w2 = HeadWell(ml, xw=700,yw=0,tsandh=[(150,-5),(250,-10),(350,-15)], rw=1) ml.solve()

results

print(w1.discharge(t=120)) print(w2.headinside(t=120))`

Another example with plots Two HeadWell example with different discharge results.zip

PilbaraPanther commented 5 years ago

This also effects the output of the contours as the head is fixed at 0 for the second (and more) elements.

mbakker7 commented 5 years ago

The problem is that fixed head elements can not be turned on or off (contrary to fixed discharge elements, which can have a discharge of zero). What can be done in TTim is turn a fixed head element on, by not including it in the solution before it is turned on, then compute a new solution as soon as it turns on (after which you can not turn it off anymore). I thought this was done, but apparently not correct or not for all head specified elements. This needs to be investigated and fixed. Thanks for your report.

On Fri, May 24, 2019 at 5:24 AM PilbaraPanther notifications@github.com wrote:

This also effects the output of the contours as the head is fixed at 0 for the second (and more) elements.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/mbakker7/ttim/issues/16?email_source=notifications&email_token=AAKM3SEZMH3HPQUTYWN54D3PW5NWDA5CNFSM4HO2FXAKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWEBUGY#issuecomment-495458843, or mute the thread https://github.com/notifications/unsubscribe-auth/AAKM3SE6I7ICAPT6NG2LI5DPW5NWDANCNFSM4HO2FXAA .