Sample Results >Lc but <Ld

[Cambria]

Greetings,

I'm trying to get a handle on L_c vs L_d and their practical application.

I've read the sticky post (https://www.nukeworker.com/forum/index.php?topic=18295.0) and found the info very useful but still have a question. I understand that L_d is a function of the counting system not the sample, but it would seem that if the sample result is >L_c but <L_d then it would not be valid. How can we say that we can say the sample result has meaning if it is less that the amount that we're saying that the counting system can accurately see?

Thanks for any info

[GLW]

read deeper:

read the source document for the sticky post, it is only 16 pages long, one concentrated hour of reading and you would have known without asking,...

and you would have known better than just about anybody has the time to delve into on an internet blog,...

PLUS,...you would have all the good references to go after and read which are referenced in the sticky post's source document,...

in a nutshell from the reference material(s),...

...Two very important statistical concepts, the decision level (DL; a.k.a. critical level [LC]) and the
minimum detectable amount (MDA; a.k.a. detection level [LD], lower limit of detection [LLD], ...)
are based on the standard deviation of the net count rate when an appropriate blank is being
counted. DL and MDA are covered mathematically later. Suffice it to say at this point that one
can determine, in advance of receiving a sample, how small an amount of radioactive material is
likely to be distinguishable from background with a giving counting system and choice of counting
times: this amount is the MDA. The MDA is the value that one can legitimately advertise that one
can measure with reasonable assurance....

take notes, write the hierarchy of thresholds down in a linear fashion, and illustrate for yourself which each progressive threshold of assurance is communicating,...

and then you will be able to conceptualize and remember it,...

it's not overly difficult, but you have to do the work to make it work for you,...

and it's just statistics, statistics are as good as the assumptions behind them,...

airline travel is the safest form of mass transit,...

from the assumption you are not going to die in an impact vaporized fuel cloud of fiery incineration,...

unless that happens to your airliner,...

then you're 100% dead and the assumption is 100% dead wrong,...

and there you go,...

[Cambria]

Thanks for the response, but unfortunately your assumptions are in error. I did read the attached reference. The problem is that I did not understand parts of it. As a newly minted Jr RCT with no statistics background I don't think this is too unexpected. I was hoping that someone might assist me in understand what is (to me at least) a complex subject.

As was stated in the reference material,
"Suffice it to say at this point that one can determine, in advance of receiving a sample, how small an amount of radioactive material is likely to be distinguishable from background with a giving counting system and choice of counting times: this amount is the MDA. The MDA is the value that one can legitimately advertise that one can measure with reasonable assurance...."
It then goes on to say,
"Once one has made a measurement on a sample, one may wish to decide whether there is indeed any activity above background in the sample. This is done by comparing the counting result to the DL, a value typically about half the MDA. Yes, it is possible and not even infrequent to be sure one has detected activity less than the MDA but more than the DL."


This is the part that's tripping me up and I understand that it's likely due to my limited understanding of the underlying statistics. However, the second statement seems to contradict the first. How can we say that we can say a sample result that is >DL (L_c) has activity above background if it is less that the amount that we're saying that the counting system can distinguish from background (MDA or L_d)? I  accept that we can, I'm trying to understand why.

Thanks again for any help

[GLW]

Thanks for the response, but unfortunately your assumptions are in error.....

True that, as I mentioned, statistics are as good as the assumptions,...

When the presenter does not disclose the background facts then the assumptions can be erroneous,...

....The problem is that I did not understand parts of it. As a newly minted Jr RCT with no statistics background I don't think this is too unexpected.....

well, that helps with the scenario since "newly minted Jr RCT" is unlikely to be the source of statistics questions,...

now that you have begun disclosure, we need to need to determine what "newly minted Jr RCT" definition applies in your case,...

(1) - former and erstwhile commercial power plant deconner with enough time under their belt to land a junior job?

(2) - some sorta classroom based commcollege program?

(3) - tech school certificate?

(4) - former DnD, job specific trained, FSS tech?

(5) - off the street direct hire?

(6) - something else?

we need to define and establish your background to give foundation to your statistics questions,...

which is similar to the narrowly applied statistics you are inquiring about,...

your turn,...

almost forgot,....(sic) for beercourt,...

[Cambria]

I'm working at a DOE D&D site and recently transferred into the RP department, they ran us through a 6 week training course. This is more of a quantity over quality operation, but I'm hoping to be a little better than that... 

[SloGlo]

How can we say that we can say the sample result has meaning if it is less that the amount that we're saying that the counting system can accurately see?

yore saying that yew knead an udder counting system witch can measure the sample wit accuracy.

[GLW]

I'm working at a DOE D&D site and recently transferred into the RP department, they ran us through a 6 week training course. This is more of a quantity over quality operation, but I'm hoping to be a little better than that...  

are you trained, tested, qualified and still familiar with this information?

https://www.nukeworker.com/study/hp/rct/pdf/2_16_Radiation_Survey_Instrumentation_sg.pdf

https://www.nukeworker.com/study/hp/rct/pdf/2_17_Contamination_Monitoring_Instrumentation_sg.pdf

https://www.nukeworker.com/study/hp/rct/pdf/2_03_Counting_Errors_and_Statistics_sg.pdf

https://www.nukeworker.com/study/hp/rct/pdf/2_19_Counting_Room_Equipment_sg.pdf

[Cambria]

I won't claim to know it off the top of my head, but I can understand everything there.

[GLW]

I won't claim to know it off the top of my head, but I can understand everything there.

fair enough,...

the question was "are you trained, tested, qualified and still familiar with this information?",...

to make the appropriate assessment which will satisfy the assumptions, the devil is always in the details,...

[Cambria]

Yes...

[GLW]

okay good,...

to answer this short form:

...I'm trying to get a handle on L_c vs L_d and their practical application.

Lc -

you have something greater than your known background which is most likely your target ROC,...

Ld -

you have something greater than your known background which is your target ROC,...


for the long form answer follow the threshold hierarchy instructions in the earlier post, read the Brodsky, Tritch, Sommers, Strom, Turner, et al, papers and you should be good to go,...

and keep in mind that distributions can be applied to isotopes, grasshoppers and buttocks,...

the term analyte is used a lot, it's a chemist's term,...

nuke are concerned with isotopes, nukes get to toss in cool qualifiers such as Secular Equilibrium, chemists get to toss in cool qualifiers such as London Forces, when you are ready to sink your teeth in a bit deeper to the marriage from hell aka radiochemistry, read this one (cover to cover);

Handbook of Radioactivity Analysis (3rd Edition),...

page 177 starts your "critical level et al" education beyond the "practical application" stage,...

you know they have degrees in this stuff if it really trips your trigger,....

almost forgot,....(sic) for beercourt,...

[Chimera]

unlike our "newly minted" RCT who started this thread, I'm carrying 45+ years experience in DOD, DOE and NRC and I have pretty much the same issues.  Trying to relate Lc to Ld just ain't right.  First off, Lc is in cpm and Ld is in dpm.  If you look at equations 16 and 17 in module 2.03 of the DOE study materials, Ld is just 2 time Lc plus a constant (in this case 2.71 although we use a different constant at my site).  Since one is in counts and the other is in activity, I'm not even sure they belong on the same graph.  On our surveys, we only calculate the activity on a given smear or direct reading if it's greater than or equal to Lc.  Ld is just something nice to know for the biennial requal exam from an RCT's point of view.

[GLW]

unlike our "newly minted" RCT who started this thread, I'm carrying 45+ years experience in DOD, DOE and NRC and I have pretty much the same issues.  Trying to relate Lc to Ld just ain't right.  First off, Lc is in cpm and Ld is in dpm.  If you look at equations 16 and 17 in module 2.03 of the DOE study materials, Ld is just 2 time Lc plus a constant (in this case 2.71 although we use a different constant at my site).  Since one is in counts and the other is in activity, I'm not even sure they belong on the same graph.  On our surveys, we only calculate the activity on a given smear or direct reading if it's greater than or equal to Lc.  Ld is just something nice to know for the biennial requal exam from an RCT's point of view.

Lc has risen to prominence with the "no detectable activity" or the "indistinguishable from background" mantra for free release acceptance criteria,...


ergo, Lc is the level at which there is a statistical probability with a predetermined confidence of correctly identifying a measurement value as "greater than background",...

compliance with the acceptance criteria (i.e., indistinguishable from background) is demonstrated when the "net" result for a measurement is less than or equal to Lc, any response above Lc will be considered as above background (i.e., a net positive result),...

your constant is a Poisson probability sum,...

Ld can be calculated in counts, refer to page 6-34 of NUREG 1575,...

so much for practicality,...


almost forgot,....(sic) for beercourt,...

[Chimera]

I understand about the Poisson probability sum and will even accept that Ld can be recalculated in cpm.  All I'm saying is that, given the study materials availabe to most RCTs, background, Lc and MDC (minimum detectable concentration) are the salient points they need to know.  Module 2.03 equates Ld with MDC which tends to just confuse the issue for them.  On their surveys, background, Lc and MDA are all that's recorded.  The rest is a rather esoteric discussion for the breakroom and has little practical value for the working RCT.

[GLW]

I understand about the Poisson probability sum and will even accept that Ld can be recalculated in cpm.  All I'm saying is that, given the study materials availabe to most RCTs, background, Lc and MDC (minimum detectable concentration) are the salient points they need to know.  Module 2.03 equates Ld with MDC which tends to just confuse the issue for them.  On their surveys, background, Lc and MDA are all that's recorded.  The rest is a rather esoteric discussion for the breakroom and has little practical value for the working RCT.

true all that,...

the kid ( I'll assume kid being a newly minted Jr RCT and all) came in asking,...

it's been noted that it's uncommon for a newly minted Jr RCT to delve into the vagaries of the statistics behind their procedures, and vagaries can be the perception of it all from the practical application viewpoint,...

but the kid asked, so we point them in the directions where they can begin to expand what they know, and we try to shine some light on applications such as advanced degrees, regulatory compliance frameworks, et al,...

when I was an ex-ELT / RCM / RCSS, newly minted shipyard 0288 IF I had not wondered a bit more and delved a bit deeper, I'd still be working out of a green memorandum book with nineteen Co60 equivalent thumbrules, and be self deluded that I was all that and I was one kick ass SRCM et al,...

in reality I'd be a spoon fed tool,...

a smart, ethical, well versed and valuable tool,...

but still a tool,...

the kid asks, we shine a little light, the rest is up to them,....

[Rennhack]

in reality I'd be a spoon fed tool,...

Must resist....

[Chimera]

GLW - point well taken!

[tolstoy]

Yes...

Hilarious thread! See what you get for asking?