Some localities get assigned technologies with higher levelized cost

[shakythesherpa]

If we look at the output table with the Levelized Cost Comparison with Localities it appears that for some localities (mostly those assigned to mini-grid) have higher levelized cost than solar home system yet they get proposed for mini-grid. Is this comparison at LCOE at the localities level even valid? Previously LCOE was reported only at the aggregate level by technology for all localities similar to what you have in the executive summary.
I

Does it make more sense to replace this table of Localities with the discounted cost and population rather than LCOE and connection order (which would be more appropriate in the Proposed Grid Medium Voltage Line Table) as in the old NP system instead?

[invisibleroads]

Currently, the system chooses the standalone technology based on lower discounted cost.

How do the discounted costs compare?

If you want, we can change the system to use levelized cost instead.

[shakythesherpa]

I think this should be discussed among the group as to what the the least-cost option should be based on? I believe the discounted cost is what we previously used and should continue to do so..

[jea98]

(sorry, forgot to address this at our meeting this afternoon)

On the question of how best to choose the lower cost of the two non-grid options (i.e whether LCOE or discounted cost should be the decisive factor):

My first question is why do they produce different selection for the least-cost technology? It seems like the discounted cost divided by discounted demand IS the levelized cost. And if the two metrics are calculated for the same location, then they should address the same total nodal demand. So.... if X1 < Y1, then dividing both by Z to create X2 and Y2 doesn't seem like it should make Y2 < X2. Am I missing something?

[invisibleroads]

Actually, the issue was that there were empty columns in the spreadsheet so that when Shaky sorted the table, only a subset of the table was being sorted.

The solution, we discussed, is to add column names to the empty columns.

[shakythesherpa]

Still seeing some cases in some of my test scenarios where a node is assigned a technology with higher LCOE and total discounted cost. Here is sample output from a test data in Myanmar

[shakythesherpa]

Also, seeing some results where a node is assigned a technology with lower total discounted cost but LCOE reported is higher than other technologies: Shown below is a screenshot of the results.