Surface release levels and NORM

["Hampshire, John" <John.Hampshire@Jacobs.com>]

I offered the Reg Guide 1.86 release limits in reply to a specific
question asked by Mark Winslow on 12/16.  In reply, I noted that Chris
Alston stated radium release levels might apply independently of
"uranium and associate decay products" (the terminology of Reg Guide
1.86 and 10 CFR 835), and suggested that "associated decay products
would be limited to the short-lived early daughters (I assume Chris
meant Th-234, Pa-234m, and Pa-234 for U-238, and Th-231 for U-235).
Chris' observation that field probes are unable to distinguish alphas
from uranium or radium is well taken, and is the basis for the common
practice of assuming the worst-case condition until it can be proven
otherwise.  This would lead to using the radium release levels in this
case, if they are appropriate to the situation.  Which leads to my first
question.

I have always assumed that the associated decay products included the
full gambit of daughters for materials such as uranium ores.  I looked
over Reg Guide 1.86 again, as well as 10 CFR 835, and could find nothing
to confirm nor deny either interpretation.  Is there anyone else with
experience in this area, preferably with a reference to further guidance
clarifying this point?

I have been under the impression that UMTRA sites, as well as non-UMTRA
uranium mills, have been using the release levels assuming a full
sequence of daughter products; today's e-mail from Bob Morris seems to
imply this as well.  I also thought an exception was made for particular
mill circuit elements that would be reasonably expected to
preferentially concentrate either radium or thorium in excess of
equilibrium with the uranium; anyone (Bob?) know different?

As an aside, UMTRA does deal with naturally occurring radionuclides, but
I thought this was somewhat distinct from "NORM",  a term which carries
an implication of radioactivity being "accidentally" concentrated as a
byproduct of a process not specifically meant to do so (i.e., without a
radioactive materials license, so in an uncontrolled manner).  Even so,
there are some process which would concentrate radium preferentially
over uranium (a tin smelter in Texas comes to mind, but my involvement
with it was somewhat limited), which in effect generates a radiological
problem similar to mill tailings.  So much so that the proposed (now
final?) Texas NORM regulations in effect adopt UMTRA (40 CFR 192)
standards.

Also posted was a reference to Steve Green's paper in the HP Journal
regarding statistical methods for surveying.  Anyone reading this
article may be interested to note that the method worked very well for
items that were reasonably expected to be free releasable, but were not
as successful elsewhere.  (Any WSSRAP-specific Radsafers know better?).
One of his key assumptions was the existence of a normal distribution of
contamination levels, leading to a test to confirm measurements fit that
expectation.  In retrospect, this may have been unnecessarily limiting.
I have considered modifying the approach to test a sample set for
goodness of fit to other distributions (log-normal comes to mind) in
addition to normal, and use the best-fitting distribution to guide
acceptance levels.  I would also appreciate feedback on this aspect.

I'm just full of all sorts of questions today...


Just my own thoughts (or even ignorance), not my employer's...

John Hampshire
John.Hampshire@jacobs.com