Nuclear power is the clear and obvious winner. No other technology comes close.
Cost per kWh is a seriously contentious question. Nuclear power is fairly expensive, but mostly that's because powerful lobbies have spent six decades making it as expensive as possible, particularly in the USA. The cost per kWh is, however, still competitive with other generation technologies to a sufficient extent that plants remain open, and new plants are being built.
The OECD's Nuclear Energy Agency gives a levelised cost in the USA of $77.71/MWh for nuclear, based on a 7% interest rate, with a cost on the same basis of $65.95/MWh for gas, and $93.75 for coal; The cheapest Solar PV option (large ground based panels) in the US has an LCOE of $79.84/MWh, and the cheapest wind power (onshore wind, 49% capacity factor) $42.85/MWh but this includes nothing to offset the intermittency of this power (see below). If the Capacity factor of the same onshore wind generator is 35%, the LCOE rises to $65.32/MWh.
By comparison, in Korea, the LCOE of nuclear power is just US$40.42/MWh, making it the cheapest even before the cost of intermittency is considered. The inherent cost of nuclear power is very low, and that shouldn't be a surprise - nuclear power is very simple technology, once you have produced a suitable fuel. You put the fuel in a big lump wrapped in concrete and steel, and it gets really hot. You needn't do much else, and you don't even need to put new fuel in (or remove the "waste") for at least several months. Plants last for many many decades, largely because of this simplicity and lack of moving parts.
The anti-nuclear lobby love to quote low costs for wind power, which really does have low LCOE, but this is somewhat misleading. The viability of a power plant is not solely determined by the cost of its output - it's also a function of the value of that output. A kWh of electricity at 3pm on a sunny and breezy afternoon in California can be worth less than nothing - that is, you have to pay the grid to take it off your hands. The same kWh on a still, hot evening, just after sunset, can be vastly more valuable - but wind and solar generators cannot take advantage of that high wholesale market price, unless they stored electricity generated when prices were low. But of course, storage is expensive. An LCOE calculated without including the cost of such storage results in figures that appear artificially low, and give a misleading idea of viability and profitability.
Deaths and environmental damage are perhaps less contentious. Deaths are difficult to fudge the numbers on for any technology, and the deaths per TWh for various sources were calculated by ExternE, a European project that set out in 2005 to calculate the various 'external' costs of energy, including public health and environmental impacts, and their data on deaths per TWh have been widely reported since that study (P. Bickel and R. Friedrich, Externalities of Energy, European Union Report EUR 21951, Luxembourg (2005); I haven't got a link to the original paper online, but reports such as
James Conca's and
Brian Wang's give the figures) was published. Their initial figure of 0.04 deaths/TWh for nuclear power was later revised down to 0.01/TWh, on the back of lower than projected actual cancer deaths amongst Chernobyl survivors, and the absence of a detectable increase in such deaths across Europe; James Conca's article linked above has been updated with this new information. Whichever dataset you choose, nuclear power is by far the least deadly way to make electricity, with only onshore wind coming close.
Environmental damage is not as clear cut as deaths, and depends on what you consider 'damage', and on how much of it you are willing to tolerate. Typically, the tolerance of environmental activists varies by technology, with a nuclear plant that raises river water temperatures by a couple of degrees being 'bad', while the mining of rare earths for wind turbines is simply ignored. On a like-for-like basis, nuclear power is one of the least damaging to the environment of any electricity generating technology. The materials used are small in quantity per kWh generated, and waste streams are completely managed and segregated from the environment indefinitely - a claim no other industry, power generation or otherwise, can match. Iida Ruishalme's Thoughtscapism blog is an excellent place to find hard data on the environmental impacts of nuclear power; She is a cell biologist and always backs her arguments with solid references.
https://thoughtscapism.com/2017/11/04/nuclear-waste-ideas-vs-reality/ is a good starting point;
https://thoughtscapism.com/2017/11/...astest-and-lowest-cost-clean-energy-solution/ is another article in which she addresses these questions.
The last four paragraphs above are a copy/paste of a post I made in a previous discussion of nuclear power; I saved a copy because the same basic slurs just keep being repeated over and over by the anti-nuclear lobby
en.wikipedia.org
