Hmm, major reversals of conflict-of-interest policy? (ie. now chairman of two YC companies and personally investing in them)
Large claims with no hard data comparisons? (eg. atomic power has major cost, density, and predictability advantages)
This(?):
>There will only be one cheapest source of energy
Really? There can't be two sources at or near equilibrium?
These systems are not deployed in isolated, designer environments, but instead are deployed in complex environments. Transportation and project logistics will prefer some sources over others.
The lack of any real metric for cost or "cheapness" is a red flag. Is cost being measured in nominal dollars? Will such a thing even exist in the 22nd century?
I still think there are huge issues with nuclear that need to be solved but nothing Sam said was controversial;
Cost: http://www.eia.gov/forecasts/aeo/electricity_generation.cfm -- on an LCOE basis, Advanced Nuclear is cheaper than everything but natural gas and that's entirely due to CapEx costs to build new plants -- which are what the modular nuclear companies that Sam's joining are working on fixing.
So, comparing the report in your link for cost, Solar PV capital cost is a bit more expensive than "Advanced Nuclear" capital cost. However, the ongoing maintenance costs for solar PV are less than 50% of the ongoing maintenance costs for Advanced Nuclear. In fact, the additional (over Solar PV), annual maintenance costs for Advanced Nuclear are about 17% of the entire capital costs. After just a few years, Solar PV will be saving massive amounts of cost over Advanced Nuclear, just in maintenance.
This seems to at least be partially taken into account in the "Total LCOE" numbers (sorry, but this report doesn't seem to have much as far as exactly how these numbers are calculated). Without subsidies, Solar PV is about 30% higher. However, when you look more closely at the actual ranges behind the averages that are the primarily cited data, you'll find that the Total LCOE estimate ranges for Advanced Nuclear and Solar PV actually overlap. The "minimum" Solar PV estimate is 97.8 and the Advanced Nuclear "maximum" is 101.0.
The claim that atomic energy has major advantages over solar is not borne out in this EIA report. In fact, the report seems to show that solar is highly competitive with almost every generation source and actually has a huge advantage in ongoing maintenance costs, being practically at the bottom.
I haven't investigated the other links yet, since the first one didn't seem to pan out.
That makes sense given those numbers tabulated, but the real question is cost of solar plus backup or storage. One means nothing without the other in terms of both cost and carbon.
Above I illustrated my calculations on just one hour of battery backup for US grid with lithium ion- would require 10x annual global production of lithium. Generally cost of battery backup looks like 35c/kWh and has limitations on duration so it must come back to fossils.
Solar makes a lot of sense if backup/storage/smoothing weren't needed since pv has gone down in cost. It still makes a lot of sense in many off grid applications, and for instance in my home we have solar concentrator lights and I would love to have solar water pre-heaters in the summer. But for on grid the capital requirements are much more complicated than generally illustrated, and ultimately the environmental impacts are greater than apparent at first glance.
Lithium ion makes no sense for storage.
Something like large-scale flywheels would probably be much better. Using gravity for storage is even more sustainable. Flywheels for night use and stored water for long term use (storms, fog, prolonged winter at poles [would need to keep from freezing]) all powered by solar or other low-ongoing-input sources should provide a pretty nice environment.
Flywheels have proved extremely expensive with massive engineering challenges. When you have to account for precession due to Earth's rotation, things get tough.
For short term very high flux power conditioning they've got uses. For long-term storage not so much.
Seconds to minutes is likely their effective range. Not hours to days.
Yes but solar doesn't last as long as nuclear plants. The current nuclear generation, like the AP-1000, are certified for 60 years with an option to extend for another 60.
Solar has the advantage of trivial decommissioning, but a huge disadvantage in requiring things like storage, overproduction, long distance transmission, and smart grids, all of which raise its effective cost if you're not putting it on top of a fossil or nuclear-based power grid.
Large claims with no hard data comparisons? (eg. atomic power has major cost, density, and predictability advantages)
This(?):
>There will only be one cheapest source of energy
Really? There can't be two sources at or near equilibrium?
These systems are not deployed in isolated, designer environments, but instead are deployed in complex environments. Transportation and project logistics will prefer some sources over others.
The lack of any real metric for cost or "cheapness" is a red flag. Is cost being measured in nominal dollars? Will such a thing even exist in the 22nd century?