Today we received the following question about modeling the impacts of learning in LEAP and NEMO:

Is it feasible to configure the learning rate of technical parameters within the LEAP power optimization module and implement endogenous technological change modeling through the application of a learning curve model? Specifically, how does the endogenous technological change model for power technology integrate with the LEAP power system optimization model?

Our response:

You should be able to simulate technological impacts due to learning by creating appropriate expressions in LEAP for transformation input variables such as Process Efficiency and Variable OM Cost. The right variables to target depend on the type of impact you aim to model. You can create a wide variety of expressions in LEAP that could represent learning processes, such as improvements in efficiency or costs as a function of time, Exogenous Capacity, or Minimum Capacity. Once you set up your expressions, LEAP will use them to generate inputs for the optimization calculation (which takes place in NEMO). So, for example, it will tell NEMO that efficiencies or costs improve over the modeling period due to learning.

One thing you can’t do in LEAP optimization models is write a LEAP expression that includes result variables. For instance, you can’t have an expression for Variable OM Cost that links the cost to dispatch (a result variable) in prior time periods. If you need this functionality, you’ll have to consider writing a custom constraint for NEMO. See https://sei-international.github.io/NemoMod.jl/stable/custom_constraints/ for a guide.

I thought this might be helpful for others!