Hi,

I have a GAMS model of an energy network and am trying to setup a simple equation to add load shifting. I want a maximum of 1000MW to be shifted and the shifting time is limited to two time steps (which in my model is the difference to energy storages). LS = Load Shifting, SL_LS = storage level load shifting, LS_in/out = load shifting input/output from network.

My equations and restrictions are:

LS_level(t,c)..

SL_LS(t,c)=e=SL_LS(t-1,c)+LS_in(t,c)-LS_out(t,c);

LS_balance(t,c)..

LS_in(t,c)=e=LS_out(t+1,c)+LS_out(t+2,c);

SL_LS.up(t,c)=1000;

*SL_LS.fx('t1',c)=0;

SL_LS.lo(t,c)=0;

LS_in.up(t,c)=1000;

LS_in.lo(t,c)=0

My problem is that in the results LS_out(t+1,c) and LS_out(t+2,c) are often higher than LS_in(t,c). Does anyone know why this can happen, or how I can formulate that what is stored from the system has to be put back into the system in the next two time steps?

I would be super thankful for any help !

## Simple Load Shifting Equation

### Re: Simple Load Shifting Equation

Actually it seems like I made a mistake analyzing the results so the problem above is resolved but I figured out another problem with this equation which I have not been able to resolve...

The equation LS_in(t,c)=e=LS_out(t+1,c)+LS_out(t+2,c) leads to the problem that the values LS_out can be counted double, in this and in the next time step.

So for example:

LS_in(t,c)=100

LS_in(t+1,c)=100 (Total input=200)

Then to solve the equation it would be enough for LS_out(t+1) to be 100. Because in the next time step LS_out(t+1) is LS_out(t+2) and the equation is solved again (Total output=100). This leads to the total input being higher than the total output which I would like to avoid. Any ideas how to make sure that the outputs are only counted in one time step??

The equation LS_in(t,c)=e=LS_out(t+1,c)+LS_out(t+2,c) leads to the problem that the values LS_out can be counted double, in this and in the next time step.

So for example:

LS_in(t,c)=100

LS_in(t+1,c)=100 (Total input=200)

Then to solve the equation it would be enough for LS_out(t+1) to be 100. Because in the next time step LS_out(t+1) is LS_out(t+2) and the equation is solved again (Total output=100). This leads to the total input being higher than the total output which I would like to avoid. Any ideas how to make sure that the outputs are only counted in one time step??