CSP power generation into Current Notebooks

[GbotemiB]

This PR includes integration of the CSP technology into the build_renewable_profiles and the landuse_availabilty notebooks.

[GbotemiB]

While I reran the notebooks, some of the analysis plots changed. I don't know if that's ok.

[ekatef]

While I reran the notebooks, some of the analysis plots changed. I don't know if that's ok.

Could you please elaborate a bit? 🙂 Notebooks are not exactly friendly in terms of reproducibility, that is always best to restart, when generating the results.

[ekatef]

My general impression is that the results are quite insightful. In particular, I haven't expected that CSP resources will be so close to PV ones.

To have some cross-validation: this resource which can be convenient to inter-compare PV vs CSP. Note that the colorbare scale is different for both resources, which make it difficult to compare them. My general impression is that spatial patterns are pretty much the same in both cases. Values seem to be also in the same range, with CSP potential being only slightly lower.

@GbotemiB could you please post results of the quantitative comparison you have done for solar and csp?

[GbotemiB]

My general impression is that the results are quite insightful. In particular, I haven't expected that CSP resources will be so close to PV ones.

To have some cross-validation: this resource which can be convenient to inter-compare PV vs CSP. Note that the colorbare scale is different for both resources, which make it difficult to compare them. My general impression is that spatial patterns are pretty much the same in both cases. Values seem to be also in the same range, with CSP potential being only slightly lower.

@GbotemiB could you please post results of the quantitative comparison you have done for solar and csp?

This is the result obtained using build_renewable_profile notebook

Also, the time series output for solar and csp looks similar

[GbotemiB]

While I reran the notebooks, some analysis plots changed. I don't know if that's ok.

Could you please elaborate a bit? 🙂 Notebooks are not exactly friendly in terms of reproducibility, that is always best to restart, when generating the results.

Discussing the difference in results, Here is the plot from landuse_availability notebook This is from the main notebook

This is from rerunning the notebook

Considering the technical results too Here are the results from the initial notebook

While looking at the results obtained from rerunning the notebook

[ekatef]

Hello @GbotemiB!

A great addition to the documentation 😄

A couple of comments:

1. It looks like CSP has pretty same landuse availability as PV. At least, that seems to be the case in the current implementation, for which we do not account for the slope effect. So, not sure it makes sense to include CSP into the final version of the landuse-availability notebook. [Thogh, that is definitely a great idea to have both plots on the testing stage]
2. Regarding the renewable profiles, I think we need to replace csp["p_nom_max"] * csp["profile"] with the generator output, as you have done when analysing the dispatch plot.
3. A technical point: "solar-power-profile.pdf" should be replaced with "csp-power-profile.pdf" under plt.savefig.

[GbotemiB]

Hello @GbotemiB!

A great addition to the documentation 😄

A couple of comments:

1. It looks like CSP has pretty same landuse availability as PV. At least, that seems to be the case in the current implementation, for which we do not account for the slope effect. So, not sure it makes sense to include CSP into the final version of the landuse-availability notebook. [Thogh, that is definitely a great idea to have both plots on the testing stage]
2. Regarding the renewable profiles, I think we need to replace csp["p_nom_max"] * csp["profile"] with the generator output, as you have done when analysing the dispatch plot.
3. A technical point: "solar-power-profile.pdf" should be replaced with "csp-power-profile.pdf" under plt.savefig.

Thanks @ekatef for the review.

Regarding 2, the generator output that was used for analysing the dispatch plot was done from the network profile. I am not sure if we can achieve the same with just the renewable_profile object file. Also, that would also translate to changing the formula for wind and solar pv too.

[ekatef]

3. sp-power-profile

Hello @GbotemiB,

As a second thought, it's also highly meaningful information that CSP has the same land availability as PV for the considered case. Apart of that, the notebook can also be used to investigate the areas where that is not the case. So, I'd leave the landuse plot as is.

Agree on the point regarding the generation output for CSP. Let's just fix the axis name changing csp power output into something like CSP input solar power to highlight that it's not yet the electric power. After adding this fix and changing the name of the output file (point 3), I think the notebook is ready to merge 🙂

[GbotemiB]

@ekatef Thanks for the review. I have implemented the suggestions. I have also reverted the landuse notebook to the original version. The CI is also successful.

[ekatef]

Merged @GbotemiB 🎉 🎉 🎉 Thanks a lot for the great contribution! 😄

[GbotemiB]

@ekatef Thanks for the amazing support.