2008-03-10_GENERAL DOCUMENTS - M1977247

• Permit No: M1977247
• IBM Index Class Name: GENERAL DOCUMENTS
• Doc Date: 3/10/2008
• Doc Name: Scope of work for preblast surveys & water well analysis
• From: Holcim
• To: DRMS
• Email Name: ACS
• Media Type: D
• Archive: No

ACZ Laboratories, Inc. August 10, 2007
Quality Assurance Plan Version 12
SOPADO 18.08.07.12 Page 44 of 96
11.1.2.3 Radiological Tracers
Radiological tracers are used for Thorium and Uranium analyses. The tracer reacts in the
same manner as the target isotope and is used to asses analyte recovery. The tracer is added
to client samples, controls, and blanks in accordance with the requirements stipulated in the
test SOP. Because the tracer recovery has a direct impact on the LLD, the recovery must be
high enough to yield LLDs that are within the scope of the project or meet ACZ's acceptance
criteria. Refer to the test SOP for evaluation criteria and corrective action(s) for out-of-
control tracer recovery.
11.1.3 Sample Specific Controls
The effect of different sample matrices on the performance of any method can be profound;
therefore, matrix spikes, duplicates, and surrogate compounds are analyzed to evaluate matrix effects
on data quality. Each SOP includes specific information regarding the usage and evaluation of matrix-
specific QC samples and also states the required corrective action to take if any matrix QC fails.
ACZ provides analytical services to numerous and varied clients; therefore, the possibility of routinely
favoring one client is highly unlikely. Over the course of time, no single matrix type will always be
spiked or duplicated, and no one client will be selected for a high percentage of spiked or duplicated
samples. If either of these occurs, it is due entirely to chance. Samples are selected for a
workgroup by due date or priority - not by client - and are presented in the workgroup in increasing
numerical order according to project number. A client's samples will be grouped together within the
batch - in this way, a single client cannot be selected for a spike or duplicate, unless all of the client
samples in the batch are from the same project. ACZ recommends that the analyst, to the extent
possible, select samples to spike or duplicate that are representative of the workgroup. Analysts are
not to associate QC with a client sample known to be or believed to be any type of blank or
Proficiency Testing sample. Several exceptions exist for selecting samples for spiking or duplicating:
1) A sample is not spiked or duplicated if the volume is inadequate, and the client sample and QC
sample(s) would require dilution; however, if no other option is available then the client sample
and QC sample should be prepared and analyzed on the same dilution whenever possible.
2) Use the same weights (or as similar as possible) to prepare duplicates of solid matrix samples.
3) A client may request that one or more of their samples be spiked or duplicated. A "RUN QC"
comment is added when the sample is logged in to notify the analyst that QC must be
performed for a specific sample or project. If a client requests that their sample(s) be spiked
or duplicated then ACZ is obliged to accommodate the client.
4) If TDS data indicates a sample would require dilution, then the sample should not be selected
for spiking. Performing dilutions increases the likelihood of introducing error due to pipetting,
and it is possible that spike recoveries may be incorrectly influenced by this error. A high TDS
value will not influence whether or not a sample is duplicated.
5) A reactive sample is unpredictable and is a poor choice for spiking or duplicating.
6) A PT sample is not a real-world sample and is a poor choice for spiking or duplicating,
because the data does not provide any useful information about possible matrix effects.
Spike or duplicate a PT sample only when there are no client samples in the workgroup.

?J

r

2773 Downhill Drive 970-879-6590 '
Steamboat Springs, CO 80487 www.acz.com