Betavoltaic devives have high energy density but very poor power density. Generating the 10s of kilowatts a car would need would requires acres of surface area squeezed into a few cubic feet. Nanotech may be able to pull it off. It could be a significant source of energy but time to ramp up to quantities to make that impact would take 10 to 20 years which we may not have. A prime candidate for fuel would be potassium 40 which is the most abundant radioactive isotope in the world. It has a half life of about 29 years.

A prime candidate for fuel would be potassium 40 which is the most abundant radioactive isotope in the world.

Sure Tom, sure, no problem.
Just need to mine it, separate from the non radioactive, prevent it to burn off the factory while it's not yet packed in the batteries, etc...

Ahhh.. no.

Potassium 40 has a half-life of 1,250,000,000 years, not 29 years. It is, indeed the most abundant radioisotope (or pretty close to it), and is responsible for a sizable amount of the internal heat generated in the Earth.

But its concentration in natural potassium is .0001

Extraction costs would be high (centrifugal separation of isotopes differing by a few percent), and the energy payback time would be a significant fraction of the lifetime of the Universe.

Tritium is no good either. Too expensive, too dangerous in useful amounts.

Carbon14 is better, easier to make, but still astronomically expensive, and far, far more dangerous. Imagine a car powered by Carbon14 decay... any crash that breaches the containment vessel makes the entire neighborhood uninhabitable for 50,000 years or more.

Forget all betavoltaic applications except for space probes and such.

Oh that's a great plan. Lets use elements that are used/expected in biological reactions as radioactive sources.