10.7 LABORATORY QA/QC

An accurate estimate of the precision or accuracy of analytical results is only possible if sample results are derived within laboratory reporting limits (RL) required by the DQO described in the SAP or QAPP. The RL represents the concentration of a specific analyte the laboratory can detect to a high degree of confidence for a particular sample.

The QAPP identifies DQO for the project; the laboratory report indicates RLs for each result. Variables that affect the laboratory's ability to achieve the RL conforming to the QAPP include: the sample matrix, naturally occurring background concentrations, and laboratory instrumentation. QA/QC requirements include following the referenced analytical method for each chemical of concern.

Most analytical data from laboratories are documented in computer records or on printouts generated by the instrument data-handling computer and transferred to a centralized acquisition server. Standard logs are maintained to document the preparation of standards. The identity and number of the parent material is recorded and each prepared standard is assigned a number that is traceable to the parent material. All data from analytical laboratories should be collected and documented in such a manner that allows the generation of data packages that can be used by an external data auditor to reconstruct the analytical process.

10.7.1 Method Blanks

The laboratory analyzes method blanks for each analytical batch and uses results to assess laboratory background or reagent contamination. An aliquot (extraction blank) equal in mass to the sample and known to be free of the COPCs is used for method blank analysis. The matrix of the method blank is selected to represent the sample matrix as closely as possible. The method blank is taken through the whole analytical process and is analyzed exactly like the calibration standards, field samples, and field replicate samples. Method blank analytical results are included in the analytical report. Method blanks should be prepared and analyzed at a frequency of at least 1 per every 20 field samples (5%) of the same matrix (USEPA, 2003a).

10.7.2 Laboratory Control Samples (LCS)

The laboratory analyzes an LCS to assess overall method performance; it is the primary indicator of laboratory performance. The LCS is commonly accompanied by an LCSD. The LCS and LCSD pairs should be prepared and analyzed at a frequency of at least 1 per every 20 field samples (i.e., 5%) of the same matrix (USEPA, 2003a). The LCS and LCSD are typically similar in composition to the primary samples, contain known concentrations of all analytes of interest, and undergo the same preparatory and determinative procedures as the primary samples. LCS and LCSD pairs are used to assess laboratory specific precision and accuracy or to assess the performance of an analytical method. Laboratories should have established internal QC RPD and Percent Recovery limits as defined in Section 10.3.1 for each method. The parameters should be developed in accordance with guidelines established in USEPA SW-846 (USEPA, 2008a). In the absence of established guidelines, RPD goals of 20% and Percent Recovery goal ranges of 70 to 130% should be used as default objectives (USEPA, 2003a).

When both an LCS and an LCSD are processed for a batch of samples, there is no significant physical distinction between the LCS and LCSD. Both the LCS and LCSD must satisfy the same recovery acceptance criteria, which is usually based on the laboratory specific control limits.

The LCS and LCSD are prepared by spiking an uncontaminated sample matrix with known amounts of analytes from a source independent from the calibration standards. Should the LCS and LCSD fail the acceptance criteria, the entire analytical batch must be re-analyzed with another LCS and LCSD pair.

10.7.3 Matrix Spikes (MS)

An MS sample is evaluated to assess the accuracy and precision of an analytical method with respect to the sample matrix. The MS is commonly accompanied by an MSD sample. The MS and MSD samples are prepared by adding known concentrations of analytes to the sample matrix prior to sample preparation. The MS/MSD pairs should be prepared and analyzed at a frequency of at least 1 per every 20 field samples (i.e., 5%) of the same matrix (USEPA, 2003a). The concentrations of analytes in the sample matrix are known prior to the addition of matrix spike analytes.

The MS and MSD are used to identify matrix interference peaks that may co-elute with target analytes. The MS and MSD are taken through the whole analytical process. Following the analytical process, the recoveries of the spike analytes are calculated and reported for assessment of accuracy. When an MSD is analyzed, the relative percent differences between the MS and the MSD results will also be calculated and reported. The percent recoveries and the relative percent difference are used to evaluate the effect of the sample matrix on the accuracy and precision of the analysis. Matrix interference effects may result in the MS and MSD failing the acceptance criteria. However, the MS and MSD pair must satisfy their acceptance criteria for the analytical batch to be considered in control and acceptable.

10.7.4 Matrix Cleanup

Matrix cleanup methods are applied to the extracts prepared by one of the extraction methods to eliminate sample matrix interferences. Several cleanup methods may be employed depending upon the target analytes of interest. USEPA Method 3600 from SW-846 provides general guidance on selecting cleanup methods (USEPA, 2008a).

As indicated in USEPA Method 3600, the purpose of applying cleanup methods to extracts is to remove interferences and high boiling point material that may result in the following:

• Errors in quantitation [data may be biased low because of analyte adsorption in the injection port or front of the gas chromatograph (GC) column, or biased high because of overlap with an interference peak]
• False positives because of interference peaks falling within the analyte retention time window
• False negatives caused by shifting the analyte outside the retention time window

Most extracts of soil require some degree of cleanup. Highly contaminated extracts (e.g., soil containing oily residue) often require a combination of cleanup methods. Following extraction and cleanup, the extract is analyzed by one of the determinative methods. If interferences still preclude analysis for the analytes of interest, additional cleanup may be required.

10.7.5 Surrogates

Surrogate spikes involve the addition of a known concentration of a non-target analyte prior to sample preparation and analysis. The surrogate is chemically similar to the target analyte(s) and behaves similarly during extraction and analysis. The surrogate spike recovery must meet the established acceptance criteria, and measures the efficiency of the steps of the analytical method in recovering the non-target analytes.

10.7.6 Laboratory Sub-sampling Replicates

Laboratory sub-sampling replicate QA/QC samples are generally employed for all soil, sediment, or other particulate samples analyzed for non-volatile contaminants (from Multi-Increment or discrete samples). The HEER Office recommends triplicate sub-sampling and determination of the RSD. Due to the typically smaller mass of discrete soil samples, there may be situations where only duplicate lab sub-samples may be feasible. This issue should be considered during the systematic planning phase of the investigation when determining DQO and coordinating with the laboratory.

The HEER Office recommends collecting laboratory sub-sampling replicate QA/QC samples at a frequency of at least one per 20 samples, or at least one if there are less than 20 samples. Replicate sub-samples are collected from the entire mass of sample available (e.g. the entire mass of sample available after drying and sieving to project-specific particle size, typically < 2mm). Sub-sampling should be performed using a sectorial splitter or by hand Multi-Increment sampling. The USEPA lab sub-sampling guidance (USEPA, 2003b) provides detailed information on sub-sampling procedures.

10.7.7 QA/QC Reports

The investigation team generating the data should include an experienced data reviewer or a third party data validator to review the analytical data to determine its validity and therefore usability.

The data reviewer or validator should review all QC-related information provided in the data package and project-specific laboratory report provided by the analytical laboratory. As part of the process of selecting the project analytical laboratory, the investigation team will ensure that the laboratory assigns a data analyst. The analyst should review the data to assess that:

• Sample preparation information is correct and complete.
• Analysis information is correct and complete.
• The appropriate SOPs were followed.
• Analytical results are correct and complete.
• Quality control samples were within established control limits.
• Documentation, including the case narrative is complete.

The analyst will then review the analytical data package to verify that:

• Calibration data are scientifically sound and method compliant.
• QC samples were within established guidelines.
• Qualitative and quantitative results are correct.
• Documentation and the case narrative are complete.
• The data package is complete and ready for document archiving.

The laboratory report must provide the following QA/QC information:

• Sample temperature at time of receipt
• Whether sample hold times were within method limits.
• Whether samples were received in good condition.
• Whether bubbles were present in volatile organic analysis (VOA) vials at time of receipt and size of bubbles if any.
• Description of corrective measures taken, if any QA/QC sample results were out of laboratory control limits.