Cloud is not really such a problematic factor:

https://scijinks.gov/solar-energy-and-clouds/

And the nice thing about nights is that they are quite predictable. So not a very difficult problem to solve.

Tesla power wall (I'd argue overpriced...): $10k [1]

Cost of an average nuclear plant today: $20bn [2] (+operating costs)

So you could buy and install 2Mio powerwalls for thr price of one nuclear plant. How many powerwalls (or more cost-effective solutions) would you need to balance the demand vs availability of electricity at night?

1. https://solarmetric.com/learn/tesla-powerwall-review-costs-s...

2. https://thebulletin.org/2019/06/why-nuclear-power-plants-cos...

Nothing in the article you linked to supports your claim that "Cloud is not really such a problematic factor". In fact, the article explicitly states:

> However, clouds can block light from the sun. So, do clouds affect the creation of energy by solar panels? Yes, but it depends on the types of clouds and where those clouds are in the atmosphere.

> When sunlight hits low clouds, a lot of that light – and heat – is reflected back into space. […] So, if you live in a place that commonly has a lot of low clouds, solar panels might not be able to produce as much energy as they would somewhere else.

——

> And the nice thing about nights is that they are quite predictable. So not a very difficult problem to solve.

Non-sequitur. The fact that it's predictable doesn't mean it's not a difficult problem.

Clouds reduce solar about 85% in Seattle.

What is the lifetime of a Tesla power wall? Do we even know?

New nuclear reactors are expected to run for 60-100 years.

This is one of the neglected parameters of such calculations. CAPEX for nuclear is high, but the resulting structure is long-lived, much more so than solar panels (25-30 years) or wind turbines (20 years).