Hi Raquel,
Just out of curiosity, did you get resolve this? I know the post is more than a year old but it would be of interest to me to know what the outcome was of this unexpected result. The reason is that, if I understood correctly what you are trying to model, I think the result should be expected.
You say: "the solar multiple that minimized the LCOE with no TES can't give me a lower LCOE if I implement TES, it should provide a greater value." I don't think this is the case... the LCOE should come down because the power unit which has a capital cost is being better utilised.
I built an alternative model to SAM in order to compare current state of the art with a new concept in collector/receiver configuration I'm working on. I wanted to compare like for like so, rather than starting from the power unit and then applying a solar multiple, I targeted a 90% capacity factor and sized the rest of the components to suit (thermal storage, generator etc). What was interesting was not only to see a drop in LCOE across the board but also the difference in drop between the different technologies. in both my concept, and power tower, the drop in LCOE was in the order of 45 - 50% whereas with parabolic trough it was only 22%. The reasons are the balance in costs between the components of the system
- The storage added to the cost of the installation but it meant that the power unit could be significantly smaller while the same overall amount of salable electricity was produced.
- Because, for power tower, thermal storage is cheap relative to the power unit (when compared to parabolic trough), the cost savings from being able to use a smaller power unit are higher so the LCOE comes down more significantly with storage.
- What is interesting was to see how my design compares... I am looking at lower overall efficiency of the system but with cheaper energy capture costs ~$70/m2 collector and higher concentration ratios (~4,000 to ~6000 range). This allows me to target higher temperatures, which allows me to significantly drop the per watt cost of the power unit. It should also allow me to drop the storage costs but I still have to assess that so have kept the same per kWht costs as with power tower. This means that the ratio of the cost of storage to cost of power unit bumps back up to that of the parabolic trough and yet increasing the capacity factor still produces a drop in LCOE similar to that of a power tower... hmm.
- I concluded that, while the LCOE / capacity factor relationship is primarily dependent on the storage/power unit cost ratio, one has to relate it all back to the capture costs: As the capture costs come down, the impact of this storage / power unit ratio becomes more significant in overall economics.
All that said, this leads me to conclude that your observation could be correct and if it is then all CSP plants should be built with storage for 24/7 operation.
Regards,
Andrew
