Storage Trajectory Non-anticipativity
Units: | $/MWh |
Mode: | Input Only |
Multi-band: | True |
Default Value: | -1 |
Validation Rule: | Any Value |
Key Property: | No |
Description: | Price for violating non-anticipativity constraints in scenario-wise decomposition mode |
Detail: |
Storage Trajectory Non-anticipativity is the penalty applied to violation of the 'non-anticipativity' constraints - those that force the storage release decisions across multiple samples to be identical i.e. that prevent the optimization from anticipating the Natural Inflow. The default value of -1 means the constraints are hard i.e. the same storage trajectory decision must be made for all samples.
For all stochastic optimization problems, except SDDP, defining the property adds the following constraints to the formulation:
End Volume(s,t) = End Volume(s+1,t) ∀s < S
where:
s is the sample number
S = Stochastic
Risk Sample Count
- Non-anticipative (value = -1)
- There is no recourse on the decisions for the release decisions from the storage inside the defined time. The non-anticipativity constraints (as above). The value of -1 in effect means a penalty of 'infinity' on the non-anticipativity constraints.
- Recourse (value = 0)
- There is full recourse on the decisions. The non-anticipativity constraints are relaxed and the optimization is free to set the releases in each stochastic sample separately.
For SDDP, setting this property to-1 simply indicates that this is a stochastic storage.
This property can be defined in combination with Trajectory Non-anticipativity Volume to define a multi-band function.
Property | Value | Units | Band |
---|---|---|---|
Trajectory Non-anticipativity | 0 | $/MW | 1 |
Trajectory Non-anticipativity | 1000 | $/MW | 2 |
Trajectory Non-anticipativity Volume | 1 | GWh | 1 |
Trajectory Non-anticipativity Volume | 1000 | GWh | 2 |
Note that the units of Trajectory Non-anticipativity are $/MWh
while Trajectory
Non-anticipativity Volume is in units of storage (e.g.GWh, CMD, AF) and thus internally a
conversion is made for this penalty to units of storage and this is
done by using either Downstream
Efficiency or if that is not defined then by computing the Potential
Energy.
See also: