The creationists who quote Kieth and Anderson never tell you this, however.
Question: A sample that is more than fifty thousand years old shouldn't have any measurable C-14. Radiocarbon dating doesn't work well on objects much older than twenty thousand years, because such objects have so little C-14 left that their beta radiation is swamped out by the background radiation of cosmic rays and potassium-40 (K-40) decay. this isotope [K-40] accounts for a large part of the normal background radiation that can be detected on the earth's surface" (p. This radiation cannot be totally eliminated from the laboratory, so one could probably get a "radiocarbon" date of fifty thousand years from a pure carbon-free piece of tin.
In the new study using samples taken from Xingkai Lake near the Sino-Russian border in Heilongjiang province, the scientists used both radiocarbon dating and another method known as optically stimulated luminescence.
Using light to measure the amount of free electrons trapped in quartz, the team was able to tell how long the samples had been kept away from sunlight, and therefore estimate when it was that they first fell in the lake.
But the new study suggests that the sediment might be over 80,000 years old, possibly formed during an ice age.