Merit order

[Hamed Kouchaki Penchah]

Dear all, 
I'm trying to find out the merit order of large-scale hydrogen sources (BECCS, natural gas, electrolysis) and what key parameters may change the order.  Any ideas on how to determine which technology deploys first and the merit order of the technology mix?
Thanks,
Hamed

[Antti-L]

Merit order is a concept of price formation, usually on the electricity market. It is a static description concept well suited for representing short-term electricity price formation. All generating units, or technologies, can be ordered based on their variable costs. This ordering results in the merit-order curve. It ranges from the lowest variable cost merit-order units such as hydro, renewable and nuclear capacities to the high cost merit-order peak-load units, as illustrated for example on the Kraftwerke website.

You can easily construct the merit order from the model results. Just calculate the unit variable costs of all electricity (or hydrogen) generating technologies, and then rank the technologies accordingly. The Val_Flo attribute gives the purchase costs of all inputs, the Cost_Act attribute gives the activity costs, and the Cost_Flo attribute gives the flow costs. Depending on your modeling approach, you may also need to add/subtract the value flows for by-products and emissions modeled at the technology level. By summing these up and dividing by the generated electricity (or hydrogen) you can easily get the unit variable costs for all technologies.

[Hamed Kouchaki Penchah]

Thank you, Antti. So, to calculate the unit variable costs of a technology (for example, a BECCS technology producing hydrogen, byproduct1, GHG emissions, and negative emissions as outputs and biomass and electricity as inputs), I can use the following formula:
The unit variable costs = Val_Flo (biomass, electricity) - Val_Flo (byproduct1, negative emissions) + Val_Flo (GHG emissions) + Cost_act + Cost_Flo

[Antti-L]

"Unit variable costs" in time T means (total variable costs in time T) divided by (activity/output in time T).  In other words, for hydrogen production, you should divide the (total variable costs in time T) by the (hydrogen output in time T). Your expression appears to be completely missing the "unit" part.

Ok, I realize now that for hydrogen production you may need to have the value flows by timeslice as well (electricity for electrolyzers), to get the merit order by timeslice. And because Val_Flo is reported on an ANNUAL basis only, you would in that case need to use the timeslice-specific commodity balance marginals (which are already unit prices) instead of Val_Flo, and so the calculation will become somewhat more tedious.

[Hamed Kouchaki Penchah]

I see! Thank you. What about using Var_NcapR (Levelized cost (LC) of a process)? Based on the documentation "A process's LC is a life cycle quantity that aggregates all costs attached to a process, whether explicit or implicit. It is a useful quantity for ranking processes. However, such a ranking is dependent upon a particular model run, and may vary from run to run."

[Antti-L]

Well, you cannot use levelized cost for obtaining the merit order, as it is a long-term cost measure.