1) Pump what water up what hill? You need vast quantities of water and terrain capable of being dammed at reasonable cost. Few sources of water can be pumped

1) Pump what water up what hill? You need vast quantities of water and terrain capable of being dammed at reasonable cost. Few sources of water can be pumped at that rate without causing considerable trouble. And places with lots of water tend to be rather sparse in suitable hills. (If the terrain isn't pretty flat the water runs fast and doesn't stick around to be vast quantities of water.) 2) Not even in the ballpark of economic. 3) You'll really crater our economy, you'll get lynched and people will go back to the old way. 4) You realize the low efficiency of the loop you are proposing and big storage headaches it causes? And nuclear power isn't "just that expensive". Rather, US nuclear power is by regulation defined as too expensive. There is a horrible provision in the nuclear world: "as low as reasonably achievable." Sounds good, and probably is good in the medical side. But on the power side it inherently defines nuclear as too expensive because if it wasn't too expensive then additional "safety" (which I find questionable, there comes a point where additional "safety" means more to break and thus doesn't really work) would be reasonably achievable. The Republicans keep crusading about "too much regulation" but because they're not actually interested in the best possible outcome they miss the biggie: We should define that which is say 2x as safe as the status quo is deemed safe enough. And the flip side of this, that which is 2x as dangerous as the status quo is deemed unsafe. (I'd be open to different ratios, I just need to put something down.) Let's look at the reality. Nuclear safety? It's about 10x as good as natural gas. (5x if you count Fukushima--but all of those deaths are from the evacuation. Staying put had an expected death toll of zero.) Natural gas is about 10x as safe as oil. Oil is about 10x as safe as coal. Yes--coal is 1000x as dangerous as nuclear. (And note that these numbers do not include any harms from climate effects and thus are actually an understatement.) Waste? There are two basic types: Low-level: stuff that might have been contaminated. Compare it to ambient (things which aren't hotter than ambient shouldn't be treated as nuclear waste) to see if you need to care, usually you don't. High-level: Yeah, it's hot. Very hot. But we are handling it wrong. The problem is that in the name of preventing proliferation we made reprocessing a dirty word. Plutonium is plutonium, isn't it? No. Bombs need Pu-239 with low amounts of Pu-240. It's extremely hard to make a bomb from reactor plutonium because it's got gobs of Pu-240. Yes, they can be separated--but anybody who can separate them can also separate U-235 from U-238. Pretty much the same thing, it's just the plutonium is 3x harder to separate. Reprocess the spent fuel. 90% of it goes back into the reactor, even more if you're using a breeder design. Of what's left there are some commercially useful isotopes. Cobalt-60 would be pretty nasty spread over the environment but it's pretty darn good at killing things you really want dead. Say, to make shelf stable meat and dairy products. Once you get done with that you have some actual waste. Which will decay to ambient in 10,000 years and note that most of that decay is in the early part. You simply don't need elaborate precautions.