Conceptual - why is UC_Growth limit LO not UP in Part IV?

[ach]

Hello,
In 3.7.4.4 of Part IV of the documentation, where UC_Cap is introduced, why is LimType LO? I initially (mis)interpreted it as an enforced lower limit and was expecting UP for an upper limit on growth rate of capacity.
Thanks.

[Antti-L]

As explained in the documentation (Part II), dynamic constraints can be of type (t, t+1) or (t, t−1). The example uses (t, t+1), because GROWTH is specified on the LHS side.  So, denoting by Growth(t) the growth multiplier, we are defining a constraint with the following formulation (constraint type = LO):

   VAR_CAP(t) × Growth(t)  ≥  VAR_CAP(t+1) − 1

Swapping the sides we would get an equivalent formulation:

   VAR_CAP(t+1)  ≤  VAR_CAP(t)×Growth(t)  + 1

As you can see, the constraint is as intended. Therefore, you should readily see why the LimType in UC_RHSRTS has to be LO. However, you could also define GROWTH on the RHS side, and then the constraints would be of type (t, t−1), and you would use LimType=UP.