General Questions

1What is Quality Assurance?
Quality Assurance is the program by which the laboratory ensures that all standards and procedures are adhered to and that all data delivered to the client meets performance requirements.  Because environmental testing laboratories are regulated by both the State and Federal Government, an active Quality Assurance program is an integral part of a  laboratory’s operations.
2What is the difference between Quality Assurance and Quality Control?
Quality Control is the statistical measurement of the laboratory’s ability to meet analytical specifications for precision and accuracy.  These specifications may be pre-determined by method requirements, or developed within the laboratory as a function to monitor trends and deviations from norms.
3What tools are used in the monitoring of Quality Control?
Torrent’s primary Quality Control tool involves control charts that set acceptance criteria for each method performed at the laboratory.  Some limits are mandated by specific methods but most are a function of collecting and managing data points to determine accuracy and precision criteria.  Quality control charts are updated quarterly.
4What tools are used in the Quality Assurance Program?
Tools include Standard Operating Procedure documents for each method performed, maintenance logs for instrument performance, a wide variety of bench sheets, log books and forms used to track all functions occurring within the laboratory, quarterly internal audits of adherence to standards and procedures, on-going training records for all analysts, and participation in yearly double-blind performance evaluation studies.
5What Quality Control is included in my Report?
There are several levels of QC packages available.  A standard or Level II QC package includes results QC samples associated with your specific samples: Method Blank, a laboratory Control Sample (LCS), a laboratory control sample duplicate (LCSD) and, if applicable to your samples, the results for a matrix spike (MS) and matrix spike duplicate (MSD). An extended or Level III package would include all of the above as well as any QC that is associated with the instrumentation used during the analysis of your samples including calibration data and on-going instrument verification data. With both QC levels, any volatile or semi-volatile analysis performed also includes surrogate data for each sample analyzed.
6What is the purpose of a Method Blank?  And why is it important to me?
A method blank (MB) is an analyte-free matrix such as DI Water for liquids or cleaned sand for solids and/or soils that is processed in exactly the same manner as the samples.  The main function of the MB is to document contamination resulting from the analytical process.  For an MB to be acceptable it must be no higher than the highest of the following:

The method detection limit (MDL)
5% of the regulatory limit of the analyte being tested
5% of the measured concentration in the sample

The importance of the method blank is the confidence it provides in assuring the reported values found in your samples are “real” and not the result of laboratory contamination.
7What are Laboratory Control Samples and why are they important to me?
Laboratory Control Samples and Sample Duplicates (LCS/LCSD) are samples prepared in the laboratory that contain analytes that are representative of the analytes of interest in client submitted samples.  Known concentrations of analytes are added to either DI Water or Sand and are processed in the same manner as the client samples.  The results of the LCS are used to demonstrate that the laboratory is in control of the processes involved in the preparation and analysis of specific tests.  Control charts are maintained with acceptance criteria for each of the LCS analytes.  These acceptance criteria must be met before results for client samples can be reported.  The criteria include both accuracy or bias (% recovery) and precision (% RPD – or reproducibility) measurements.  It is critical that the laboratory be able to not only accurately recover what is present but to be able to reproduce that action as well.  A laboratory control sample duplicate (LCSD) is used to demonstrate reproducibility.

The importance of the LCS/LCSD is to provide confidence to you that what the laboratory claims it can recover and reproduce is actual and not hypothetical.
8What are Matrix Spike and Matrix Spike Duplicates? Why are the significant to me?
A Matrix Spike and Spike Duplicate (MS/MSD) are representative but randomly chosen client samples that have known concentrations of analytes of interest added to the samples prior to sample preparation and analysis.  They are processed along with the same un-spiked sample.  The purpose of the MS/MSD is to document the accuracy and precision of the method for that specific sample.  Control charts are maintained that are indicative of typical MS/MSD recoveries of ‘real’ samples rather than laboratory controlled samples.

The importance of the MS/MSD is specific in nature.  The MS/MSD is only significant to the client whose sample was chosen for spiking.  The MS/MSD data serves as an indication of the problems that may be associated with a specific sample or sample site.  For example, if a known concentration of Lead is added to a sample but only 20% is recovered, it may be an indication of matrix interference resulting in suppressed recovery of Lead.  This is important when evaluating all of the lead results that may be associated with a group of samples taken from the same general location. The MS/MSD is not useful to a client whose sample was not chosen and thus, is not reported to that client unless specifically requested.
9What is a surrogate?
A surrogate is an organic compound similar to the analytes of interest in both chemical composition and behavior in the analytical process, but not normally found in environmental samples.
10Why are surrogates used and what do they mean to me?
Surrogates are used in most organic analysis to demonstrate matrix compatibility with the chosen method of analysis.  Each sample is spiked with a known concentration of surrogate compound(s) prior to the preparation and analysis of the sample.  Control charts are maintained that are indicative of typical surrogate recoveries of ‘real’ samples rather than laboratory controlled samples.  Samples that fail surrogate recovery criteria are re-spiked and re-analyzed to determine if poor recovery is due to laboratory spiking error or matrix interference.

Reported surrogate recoveries are important because low recoveries indicate matrix interference in the sample resulting in suppressed recovery of analytes of interest.  This becomes important when evaluating the data.  For example, if TCE is an analyte of interest but the associated surrogate is only recovered at 28%, the reported value for TCE may be biased low as well and the client may want to consider the possibility that more TCE is present than is being reported.
11What are Method Detection Limits?
Method Detection Limits (MDLs) are the minimum concentrations of substances that can be measured with 99% confidence that the analyte concentrations are greater than zero. MDL studies are performed  for all analytes of interest for each method at a minimum of once a year.  MDLs are matrix specific as well and have very specific analytical procedures and acceptance criteria.
12What is the difference between an MDL and an RL/PQL?
MDLs are the starting point within a laboratory of defining normal Reporting Limits (RLs), also called Practical Quantitation Limits (PQLs).  An MDL is a value that is statistically determined and represents what can be identified above the “noise” level of an instrument as being present but not necessarily accurate.  The RL/PQL is a laboratory determined value at 2 to 5 times above the MDL and can be reproduced in a manner that results in a 99% confidence level that the result is both accurate and precise.
13Does Torrent report results below the RL/PQL?
Although Torrent would prefer to report data to the laboratory determined RLs, Torrent will report results between the MDL and the RL/PQL under very specific circumstances or upon client request.  Whenever results are reported to the MDL, the sample will have a specific comment associated with it indicating the reason MDLs are reported.  Additionally, any detected values between the MDL and RL will be qualified as ‘estimated’ rather than ‘quantitative’ concentrations and will be flagged with a’J’ qualifier in the case of organic analysis or a ‘tr’ in the case of inorganic analysis.
14Does Torrent report data that is out of control limits?
Torrent makes every effort to report only qualified data to the client.  Occasionally, there will be QC failures that can not be corrected prior to the release of data.  An example of this would be a MB or LCS failure for an extractable analysis that requires a minimum volume of sample.  Normal corrective action requires the re-extraction and re-analysis of all samples associated with the failing QC.  However, sometimes the client is not able to collect the extra volume of sample originally requested due to low flow wells or high level of sediment, and there is no possibility of performing the re-extraction.  In such cases, the data is reported and a case narrative is included with the report describing the nature of the QC failure and why corrective action was not possible.

A second circumstance where data is reported despite QC failure is when the QC limits are exceeded on the high side of the acceptable range but the samples are Non Detect (ND) for any associated compounds.  Both situations are rare but do occur.
15What do the terms ‘RL’ and ‘MRL’ on your report stand for?
RL refers to the initial “Reporting Limit”.  This is the lowest quantifiable reporting limit that can be achieved when an analysis is performed under ordinary conditions.

MRL  refers to “Modified Reporting Limit” and is the final reporting limit that applies to the sample once all sample preparation factors and/or dilution factor have be applied.  For example, if Chromium in soil was reported at a dilution factor of 10 the RL column would read ‘5’ but the MRL column would read ‘50’. Or, for example, if an oil and grease sample was analyzed with only 500 mL rather than 1000 mL , a preparation factor of 2 would be necessary and the RL column would read ‘5’ while the MRL column would read ‘10’.
16Are the values reported in the Results column final or do I need to correct for dilution?
All data in the results column is final.  Any preparation factors and/or dilution factors have already been applied.
17What is a “Case Narrative”?
A Case Narrative is included with final reported data whenever there are any issues regarding sample receipt, preparation, analysis, and/or reporting that are outside of the normal policies and procedures of Torrent Laboratory.  Items included in the case narrative can be as benign as a client request to amend sample IDs, deviations to methods due to limited sample mass or volume, or more serious issues regarding Matrix Spike recoveries or explanations of why data is reported despite failing QC.  Within the case narrative, information regarding specific samples may be present, such as matrix interference problems, instances where the sample required special preparation procedures, and/or information regarding requests made by the client as to how the sample should be treated.  The case narrative is also where Torrent indicates whether any analyses required sub-contracting to another certified laboratory and which laboratory the samples were subcontracted to.  And, finally, if a report is revised or re-issued for any reason, the incident leading to the revision/re-issue would be described and the date of the revision/re-issue noted in the case narrative.
18I was checking my data and some of the values do not seem correct.  What can Torrent do to help me?
Please feel free to contact the QA department at Torrent anytime you have questions or concerns about the presentation of the data or the data itself.  Torrent can open a ‘Data Validation’ request to re-validate the reported data.  While instances are rare, mistakes in data can be made and Torrent will check each step of the analytical process in order to re-assure you that the data presented is correct.  If an error is found, we will re-issue the report to you immediately with a case narrative describing the genesis of the error and the corrective action that has been taken to ensure that the same error is not repeated.
19How do you ensure that your error rate is low?
Torrent operates with a 6 sigma approach to the reporting and presentation of data.  All of our analysts are empowered to make QA decisions within the scope of their expertise and each department manager is encouraged to think of project management as their final client.  In this way, data coming into project management has already passed rigorous analyst review, peer review and department manager review.  Conversely, Project Management is encouraged to think of each department as their vendor which means any special requirements or protocols outside of the normal scope of work must be communicated to the departments prior to the release of samples for analysis.  This ensures that the departments have the information they need to qualify their work to client specifications before submitting final data to Project Management.
20What if I need to change or add tests to samples that have already been submitted to Torrent?
If the samples are in process, a simple request from the client (in writing) will initiate a ‘Change Order’ form and the changes will be communicated to the affected department.  Every effort is made to adhere to the client’s original turn around time request (TAT). Due to capacity limitations, changes may delay the final report by a day or two.

If the report has already been signed and released, any requests for additional analyses will be logged in under a new CoC that will be filled out and forwarded to the client for signatory approval.  A new work order ID and TAT will be assigned to the additional requests and the new CoC will refer to the CoC under which the original samples were received.
21What if I need a report re-issued ore revised?
Contact the Project Management or QA department and explain what you need.  Torrent’s QA procedures include the electronic storage of all generated reports.  A revised report will be issued with a unique name and number identifier and will not replace the original report but be stored alongside it.  This procedure allows for the re-creation of any action that may have taken place during the issuing of an original and/or revised report.
22Does Torrent supply chromatograms to clients?
Yes, chromatograms are available upon client request.  Torrent does not release quant sheets unless a Level IV data package is requested.
23Will Torrent release preliminary data?
Yes and no. Yes, if multiple analyses are requested and some of the data is completed but other data is still not ready, a preliminary report can be issued for the completed data upon client request.  The report is still considered preliminary, and is identified as such, but the results reported are final.

Torrent does not report any data that has not been qualified by the appropriate passing QC samples.  This is to protect the client as well as the laboratory from any liability issues resulting in decisions made with data that may change due to QC failure.
24Does Torrent release data verbally?
For analyses requested on a TAT of less than 8 hours, Torrent will release data verbally if all associated QC has passed acceptable criteria.  Under no circumstances are verbal results given without passing QC.  Verbal results may be given over the phone but a follow up email will be sent containing information shared during the phone conversation and that email is considered as the official delivery of verbal results.