Dr Martin Elvis, from the Harvard-Smithsonian Center for Astrophysics in Cambridge, US, has developed an equation to estimate the number of asteroids in the solar system that could be exploited in a cost-effective way.
Elvis evaluated the factors that would make an asteroid commercially viable to mine and what fraction of known space rocks met these requirements.
He emphasised there were large uncertainties in the values and called for more thorough surveys of what's out there, 'BBC News' reported.
These are rare in the Earth's crust because they dissolve in molten iron, instead being mainly concentrated in the planet's core.
More From This Section
However, according to the analysis, just 1 per cent of near-Earth asteroids are rich in these elements.
Suitable asteroids also need to be relatively easy to reach, further narrowing the pool by ruling out all but the nearest objects to Earth.
The operative parameter here is delta-v - the change in velocity needed to send mining equipment to the target and return with a larger mass of ore.
Elvis pointed out that the ore values in his analysis range from a low of USD 800m to a high of USD 8.8bn.
"Such a large range of values could greatly change the profitability of a venture, making more accurate assays necessary," he said.
Eric Anderson, co-founder of asteroid mining company Planetary Resources, said there were key errors in the study.
"Number one, the author points to an assumption of only wanting to go to M-type asteroids. Assuming we were only going after platinum-group metals, the most platinum-rich asteroids are the C-class ones," he told the BBC.
Also, Planetary Resources' engineers were prepared to include objects that required a delta-v of 7km/s, a more ambitious limit than the 4.5km/s used in the study.
"I want to stress that my paper does not mean that there is no commercial future for asteroid mining. It does mean that gold mines are rare, which shouldn't be too surprising," Elvis said.