6.2 Quality Assurance Project Plan for Phase II

6.2.1 Purpose of the Quality Assurance Project Plan

6.2.1.1 This Quality Assurance Project Plan (QAPP) has been prepared to document the quality assurance protocols for execution of Phase II of the Treatability Studies. The purpose of this QAPP is to define the field and laboratory data requirements for the Treatability Studies as specified in the Field Sampling Plan (FSP) and to ensure that the data are of sufficient quality to support the end use of the data. The QAPP defines the policy, organization, functional activities, and quality assurance (QA) and quality control (QC) protocols that will be used to meet the data quality objectives (DQOs) of this investigation. Descriptions of all of the DQOs and procedures associated with the field programs, including sample collection, sample custody, laboratory analysis, and QA/QC for this project are described in this document. Adherence to the procedures described in this QAPP should generate data that are scientifically sound, valid, defensible, and of known, acceptable, and documented quality.

6.2.1.2 Field samples were collected during the Phase I activities for site characterization and cell characterization. Samples were submitted for laboratory analysis to a contract laboratory. Site characterization data were used to identify locations for placement of the technology demonstration cells. Cell characterization data were used to characterize the constituents in the soil and ground water prior to performing the Treatability Studies. The field program for both Phase I and Phase II Treatability Studies is described in Table B-l. As part of Phase II of the Treatability Studies soil and ground-water samples will be collected from each cell after treatment, and compared to the initial data.

6.2.1.3 The QAPP is organized as follows:

• Section 2.0 Project Organization and Responsibilities

• Section 3.0 Quality Assurance Objectives for Measurement

• Section 4.0 Sampling Procedures

• Section 5.0 Sample Custody, Handling, and Shipping Procedures

• Section 6.0 Calibration Procedures

• Section 7.0 Analytical Procedures and Detection Limits

• Section 8.0 Data Reduction, Validation, and Reporting

• Section 9.0 Internal Quality Control

• Section 10.0 Performance Systems Audits

• Section 11.0 Preventative Maintenance

• Section 12.0 Data Assessment Procedures

• Section 13.0 Corrective Actions

• Section 14.0 Quality Assurance Reports.

6.2.2 PROJECT ORGANIZATION AND RESPONSIBILITIES

6.2.2.0.1. Under the direction of Hill Air Force Base Environmental Management Directorate, Restoration Division (EMR), Montgomery Watson will assist the researchers in conducting the site and cell characterization activities as described in the FSP. The remainder of this section discusses the EPA/University organization and Montgomery Watson's organization. Montgomery Watson is included in this project because they are providing field support to the researchers and other general support services.

6.2.2.1. EPA/University Project Personnel and Organization

6.2.2.1.1. A project management organization chart is shown in Figure 1-1. The cells constructed during Phase I will ultimately be used to field demonstrate and evaluate several innovative technologies. Personnel from the USEPA Robert S. Kerr Environmental Research Laboratory (RSKERL) will oversee development and demonstration of the innovative technologies. The specific innovative technologies are combined into groups in Figure B-1. Each technology group shows the personnel from USEPA RSKERL with oversight responsibilities and the researchers working on the specific technologies. The technologies are being developed and bench scale tests will be conducted by researchers located at different universities including: Rice University, University of Oklahoma, University of Arizona, University of Michigan, Michigan Tech, Clemson University, University of Florida, Tufts University and MIT. Technologies are also being developed by personnel at Tyndall AFB in conjunction with Applied Research Associates and Praxis. In addition to identifying the most viable technology for future implementation at Hill AFB, the results of each technology evaluation will be combined into a single report referred to as the Design Manual. Although the field demonstrations are intended to be independent evaluations of the specific technologies, the researchers will freely exchange information through a centralized data base (see Section 4 of the Work Plan).

6.2.2.1.2. Dr. Dorothy Bertino is the QA/QC coordinator representing the research group and EPA. She is responsible for evaluating the QC for all deliverables.

6.2.2.2. Montgomery Watson Project Staff

6.2.2.2.1. All Montgomery Watson project staff will be from Montgomery Watson's Salt Lake City, Utah, office. Ms. Sue Ann Spencer, P.G., the Montgomery Watson QC Coordinator. Ms. Spencer will review all deliverables and direct quality assurance audits of the field work, and will assist with project development and coordination.

6.2.2.2.2. Ms. Deborah Drain, Senior Soil Scientist, is the Operable Unit 1 Project Manager in this investigation. She will be responsible for assisting Dr. Jon Ginn of Hill AFB in setting project goals and directing technical resources for the satisfactory completion of the project. Her responsibilities include:

• Evaluating all deliverables, including interim and final reports

• Representing the project team at meetings and public hearings

6.2.2.2.3. Ms. Deb Luper, Supervising Engineer, is the Project Manager for the Treatability Studies. She will be responsible for assisting Dr. Jon Ginn of Hill AFB in implementing the goals for the Studies identified by the OU 1 Project Manager, and coordinating with the researchers to oversee all of Montgomery Watson's roles for the project. Her responsibilities include:

• Review deliverables from the researchers, and providing input where

requested

6.2.2.2.4. Mr. Steve Glaser, the Health and Safety Officer for Montgomery Watson's Salt Lake City office, will be the Montgomery Watson Health and Safety Coordinator for this project. Mr. Glaser will be responsible for monitoring all Montgomery Watson health and safety programs that relate to this investigation, and providing on-call assistance to the Montgomery Watson field team members.

6.2.2.2.5. Ms. Jennifer He is the Analytical Coordinator for Montgomery Watson. Her responsibilities include:

• Ensuring that the laboratory implements the requirements of the project Work Plans

• Coordinating with the laboratory on all QA/QC matters

• Providing updates to the Project Manager regarding QA/QC data

6.2.2.3. Mr. Kevin Bourne is the Hill AFB project manager. He will be assisted by Dr. Jon Ginn in the management of the overall project.

6.2.3 QUALITY ASSURANCE OBJECTIVES

6.2.3.1. Data Quality Objectives

6.2.3.1.1. The overall quality assurance objective for this investigation is to develop and implement sampling, sample handling, and analytical procedures that will provide data that can be used to fulfill the data quality objectives (DQOs) as stated in the Wok Plan. DQOs are qualitative and quantitative statements developed by data users to specify the quality of data from field and laboratory data collection activities that is needed to support specific decisions or regulatory actions. The DQOs describe which data are needed, why the data are needed, and how the data will be used to meet the needs of the project. DQOs also establish numeric limits for the data to allow the data user (or reviewers) to determine whether the data collected are of sufficient quality for their intended use.

6.2.3.1.2. DQO development as described in USEPA guidance is based on:

• Identifying project objectives

• Specifying the data necessary to meet project objectives

The project objectives and data specifications are described in the Field Sampling Plan. Analytical and testing methods are described in this QAPP. Table B-2 lists the DQOs for this sampling program.

6.2.3.2. Analytical Quality Control Levels

6.2.3.2.1. Five levels of analytical quality control are identified by CERCLA and are described in Data Quality Objectives for Remedial Response Activities Development Process (USEPA. 1987). These levels are based on the type of site under investigation, the required precision and accuracy, the end use of the analytical data, and the level of documentation. Three levels of analytical data will be collected during this investigation. The analytical levels are listed in Table B-2 and include the following:

6.2.3.2.2. Practical quantitation limits (PQLs) are based on the extent to which the equipment, laboratory or field. or analytical process can provide accurate measurements of a reliable quality for specific constituents in field samples. The PQL for a given analysis will vary depending on instrument sensitivity and matrix effects. PQLs are discussed in Section 6.2.7.

6.2.3.3. Data Quality Definition and Measurement

The effectiveness of a QA program is measured by the quality of data generated in the field and by the laboratory. Data quality is judged in terms of its precision accuracy, representativeness, completeness, and comparability. These terms are described in the following sections.

6.2.3.3.1. Accuracy. Accuracy is the degree of agreement of a measurement or an average of measurements with an accepted reference or "true" value, and is a measure of bias in the system. The accuracy of a measurement system is impacted by errors introduced through the sampling process, field contamination, preservation, handling, " sample matrix. sample preparation, and analytical techniques.

Accuracy is evaluated by the following equation:

where: A is the concentration of analyte in a spiked sample

B is the concentration of analyte in an unspiked sample

C is the concentration of spike added.

For this project, accuracy will be assessed and controlled by the results of the following QC samples, which contain known concentrations of specific analytes (spiked):

• Matrix spike (MS) and matrix spike duplicates (MSD)

• Laboratory control samples (LCS) and LCS duplicates (LCSD)

• Surrogate spikes.

As these samples are analyzed, spike recoveries will be calculated and compared to pre-established acceptance limits, as listed in Attachment A. Acceptance limits are based on previously established laboratory performance or specified by the analytical methods. The control limits reflect the minimum and maximum recoveries expected for individual measurements for an in-control system. Recoveries outside the established limits indicate error in addition to normal measurement error and the possible need for corrective action. Corrective action may include re-calibrating the instrument, reanalyzing the QC samples, re-analyzing the sample batch, re-preparation of the sample batch, or flagging the data (if problems can not be resolved). For contaminated samples, matrix spike recoveries may be dependent upon sample homogeneity, matrix interference, and dilution requirements.

6.2.3.3.2. Precision. Precision is the reproducibility of measurements under a given set of conditions. For large data sets, precision is expressed as the variability of a group of measurements compared to their average value (i.e., standard deviation). For duplicate measurements, precision is expressed as the relative percent difference (RPD) of a data pair and is calculated using the following equation:

where: A and B are the reported concentrations for sample duplicate analyses.

For this project, precision will be assessed by calculating the RPD of the MS/MSD sample pairs and the blind duplicate and replicate sample pairs and comparing the results to laboratory-established RPD control limits, which are listed in Attachment A. Precision of blind duplicate samples is dependent upon sample homogeneity.

The analyst, group leader, or technical advisor is responsible for investigating data outside the QC limits. Corrective action may include re-calibrating the instrument, re-analyzing QC samples, re-analyzing samples, or flagging the data.

6.2.3.3.3. Representativeness. Representativeness is a qualitative expression of the degree to which sample data accurately and precisely represent a characteristic of a population, a sampling point, or an environmental condition. Representativeness is maximized by ensuring that, for a given project, the number and location of sampling points and the sample collection and analysis techniques are appropriate for the specific investigation, and that the sampling and analysis program will provide information that reflects "true" site conditions. Results for blind duplicate sample analysis are also used to evaluate representativeness.

6.2.3.3.4. Comparability. Comparability is a qualitative parameter that expresses the confidence that one data set may be compared to another. Comparability of data is achieved through the use of standardized methods for sample collection and analysis, and the use of standardized units of measure.

6.2.3.3.5. Completeness. Completeness is defined as the percentage of valid data relative to the total number of analytes and is evaluated using precision, accuracy, and holding time criteria. Completeness will be calculated using the following equation:
 
 

Project completeness is determined at the conclusion of the data validation and is calculated by dividing the number of valid sample results by the total number of sample analyses listed in the QAPP. The completeness objective for this project is 90 percent for all data and is based on USEPA guidelines (USEPA, 1988a).

6.2.4 SAMPLING PROCEDURES

6.2.4.0.1. All of the sampling locations and procedures to be used for environmental sample collection are presented in the FSP. The FSP describes in detail the procedures that will be followed during sampling to ensure that the data are representative of environmental conditions. The remainder of this section describes the sampling procedures that will be used to collect QC samples in the field.

6.2.4.0.2. Sample Containers. All sample bottles to be used for volatile organic compound (VOC) samples will be chilled and stored in an iced cooler prior to sample collection. The types of sample containers and preservation required for each matrix and analysis are outlined in the FSP for the respective sampling programs.

6.2.4.1. QC Sample Collection

6.2.4.1.1. As discussed above, the sampling procedures for all of the environmental samples are described in the FSP. The following sections outline the procedures to be used to collect QC samples in the field.

6.2.4.1.2. Source Water. A sample of the source water used for equipment decontamination will not be collected during this investigation. Source water data have been collected for other on-going sampling programs at Hill AF6.2. These data will be used to assess source water quality.

6.2.4.1.3. Equipment Blanks. Equipment blanks will be collected at a rate of one per day when non-dedicated or non-disposable equipment is used for sampling. Equipment blanks will be collected for each analytical parameter for which the associated environmental sample was collected. Equipment blanks will be collected immediately after decontaminating sampling equipment by pouring the source water over the sampling equipment, then collecting it in the appropriate sample containers. The samples will be labeled, handled, and shipped following the procedures outlined in Section 6.2.5 of this QAPP.

6.2.4.1.4. Blind Duplicates. During this sampling program. blind duplicate samples will be collected for 10 percent of the total number of ground-water samples for each scheduled analytical method. A blind duplicate sample pair is a single grab sample that is split into two samples during collection. Blind duplicate ground-water samples will be collected by alternately pouring a bailer-volume or discharging a pump-volume of water into the original and duplicate sample containers. One of the samples from the blind duplicate sample set will be labeled with the correct sample identification and the other sample will be labeled with a false sample identification. Both samples will be sent to the same laboratory for analysis. The samples will be labeled, handled, and shipped following the procedures outlined in Section 6.2.5 of this QAPP.

6.2.4.1.5. Blind Replicates. One blind replicate will be collected for every 10 soil samples for each scheduled analysis, except for those samples collected only for soil cutting disposal. A replicate is a single grab sample that is split into two equal parts for the same analysis. One of the samples from the blind replicate sample set will be labeled with the correct sample identification and the other sample will be labeled with a false sample identification. Both samples will be sent to the same laboratory for analysis. The samples will be labeled, handled, and shipped following the procedures outlined in Section 6.2.5 of this QAPP.

6.2.4.1.6. During this investigation, all soil samples are scheduled to be collected in split spoon samplers. The blind replicate soil sample for VOC analysis will be collected by removing soil from the sampler adjoining the soil collected for the environmental samples and placing it directly into the appropriate sample container. Each container will be filled so that there is no head space. For PCB analyses, the soil remaining in the split spoon sampler will be removed, placed in a stainless steel bowl, and thoroughly homogenized. The soil in the bowl will be evenly split between the two sets of sample containers.

6.2.4.1.7. Matrix Spike and Matrix Spike Duplicate Samples. Samples for MS/MSD analysis will be collected for five percent of the total number of samples for each matrix and scheduled analytical method, except for those samples collected only for soil cutting disposal. The Montgomery Watson Project Manager, Field Team Leader, or designee will identify samples to be used for MS/MSD analysis. The same procedures used to collect blind duplicate samples during sampling will be used to collect samples for MS/MSD analysis.

6.2.4.1.8. Trip Blanks. Trip blanks will be prepared by the laboratory prior to sampling and will consist of two 40 milliliter amber glass bottles filled with preserved reagent grade water. The bottles will be filled so that there is no head space and will be capped with a Teflon septum. Trip blanks will accompany all samples scheduled for VOC analysis, except for those samples collected for cutting disposal only.

6.2.4.1.9. Temperature Blanks. A water temperature blank will accompany each cooler of samples shipped to the laboratory, except for air samples (temperature blanks are not required for air samples). A temperature blank consists of a 40 milliliter amber glass bottle filled with reagent-grade water. The temperature of the blank will be measured upon arrival at the laboratory. If the temperature of the blank is outside the 2E to 6E Centigrade (C) temperature criterion, both the laboratory and Montgomery Watson Project Managers will be notified and the appropriate corrective actions will be taken.

6.2.5 SAMPLE CUSTODY, HANDLING, AND SHIPPING PROCEDURES

6.2.5.1. Sample Custody

6.2.5.1.1. To ensure that samples are identified correctly and remain representative of the environment, the documentation and sample custody procedures specified in this section will be followed during sample collection and analysis. Standard sample documentation and custody procedures, as outlined below, will be used during each sampling program to maintain and document sample integrity during collection, transportation, storage, and analysis. The Field Team Leader, to be designated at the time of the investigation, will be responsible for ensuring proper documentation and custody procedures are initiated at the time of sample collection, and that individual samples can be tracked from the time of sample collection until the samples are relinquished to the laboratory. The laboratory will be responsible for maintaining sample custody and documentation from the time the samples are relinquished to the lab until final sample disposition.

6.2.5.1.2. Chain of Custody. Chain of custody (COC) procedures provide an accurate written record of the possession of each sample from the time of collection in the field through laboratory analysis. A sample is considered in custody if one of the following applies:

• It is in an authorized person's immediate possession

• It is in view of an authorized person after being in physical possession

• It is in a designated secure area, restricted to authorized personnel only.

6.2.5.1.3. Field Procedures. The sample custody and documentation procedures will be initiated at the time of sample collection. Sample collection details will be documented on the ground-water sampling forms or on the lithologic logs (see Section 3.0 of the FSP). Samples will be labeled and the appropriate information will be recorded on the COC form using indelible ink. Any errors will be corrected by drawing a single line through the incorrect entry, entering the correct information, and then initialing and dating the change.

6.2.5.1.4. Sample Labels. Sample labels will be completed and attached to sample containers at the time of sample collection. An example of the type of label that will be used is shown in Section 3.0 of the FSP. The following information will be included on the sample label:

• Project name/location

• Sample location

• Field sample identification

• Date and time of sample collection

• Type of analyses to be performed

• Preservative (if applicable)

• Sampler's initials.

6.2.5.1.5. Chain of Custody Record. Properly completed COC forms will ensure that sample custody is documented, appropriate sample fractions have been collected, and scheduled analyses are properly assigned. An example of the type of COC record that will be used is shown in Section 3.0. of the FSP.

6.2.5.1.6. Unused portions of the COC form will be crossed out and initialed. A completed COC record will be included with each sample cooler. The sampler will retain a copy of the COC. When shipping the sample cooler to the laboratory by a commercial carrier, the COC will be signed, placed in a plastic bag, and taped to the inside of the shipping container used for sample transport. Signed air bills will serve as evidence of custody transfer between the field sampler and courier, and courier and laboratory. The sampler will retain and file copies of the COC record and the air bill after the samples are shipped. The samples are relinquished to the laboratory upon arrival and the laboratory personnel then will complete the COC.

6.2.5.1.7. Custody Seals. Custody seals, as shown in Figure B-2, will be placed in two locations across the cooler closure to ensure that any tampering is detected. The date and initials of the sampler will be written on the custody seal.

6.2.5.1.8. A field sample lot number will be assigned to each sample cooler each day of sampling and will be recorded on the COC form. The field lot numbers are used to identify the field samples that are associated with specific quality control samples, i.e., the field lot control number identifies the environmental samples and QC samples that were collected or submitted for analysis on the same day. The field lot numbers are required for the IRPIMS BCHSAMP file and will be formatted as outlined by the IRPIMS Data Loading Handbook (Version 2.3). The field lot numbers have four characters: the first character identifies whether there is an ambient conditions blank associated with the sample; the second character indicates whether there is an associated equipment blank; the third character indicates whether there is an associated trip blank; and the fourth character identifies the cooler that was used for sample shipping. The following procedure will be used for assigning the field lot control numbers:

6.2.5.1.9. All lot control numbers for this program will begin with 0. An example of a lot control number for one day of sampling is: 011A, where 0 indicates that an ambient conditions blank was not collected, 1 indicates that the first equipment blank collected that day is in the cooler, 1 indicates the first trip blank used that day is in the cooler, and A indicates that this is the first cooler used that day. It is important to note that the same characters are used for the lot control number each day of sampling, and therefore it is quite likely that the same lot number will be assigned on different days of sampling.

6.2.5.1.10. Laboratory Custody Procedures. Upon receipt in the laboratory, the integrity of the shipping container will be checked by verifying that the custody seal is not broken. The cooler will be opened and the temperature blank will be measured to determine the temperature inside the cooler. The sample containers will then be checked for breakage, leakage, damage, and the contents of the shipping container will be verified against the COC. Custody seal integrity, cooler temperature, and sample preservation will be documented on the sample control worksheet.

6.2.5.1.11. A permanent logbook will be maintained in the sample control area to document the following:

• Date of sample receipt

• Sample accession number

• Number of samples

• Source of samples

6.2.5.1.12. All insufficiencies and/or discrepancies will be immediately reported to the Laboratory Project Manager and an Anomaly Form will be completed. The Laboratory Project Manager will either resolve the problem internally or contact EPA or Hill AFB Project Manager for resolution. If the samples and documentation are acceptable, each sample container will be assigned a unique laboratory identification number from the Laboratory Information Management Systems (LIMS) database. One of the functions of the LIMS is to assist in tracking samples while they are in the custody of the laboratory. Other information that will be recorded includes date and time of sampling, sample description, due dates, and required analytical tests.

6.2.5.1.13. When the LIMS log-in has been completed, the samples will be transferred to the appropriate refrigerators in the Sample Control area. Separate refrigerators will be used for samples suspected to contain high levels of organic compounds and for samples for VOC analyses. The sample refrigerators will be kept at 4EC± 2EC and their temperatures will be recorded daily with thermometers calibrated against NlST thermometers. The cleanliness of refrigerators storing samples for VOC analyses will be monitored using refrigerator blanks.

6.2.5.1.14. Samples will be distributed for analysis from Sample Control by either a sample custodian or laboratory chemist. Sample tracking will be documented on the Sample Control Form. After all samples and documentation have been reviewed and appropriately annotated, the Sample Custodian will sign the logsheet and submit it to the Information Services Department for processing. Any marks or notes made on the chain of-custody document by the Sample Custodian will be clearly distinguishable from original field notations.

6.2.5.1.15. Shipping receipts will be stapled on chain-of-custody logsheets and stored in the project file. Samples will be placed in appropriate storage areas in the laboratory depending on storage requirements. The Department Managers or their designee will be notified that the samples have arrived through the distribution of arrival notices. The majority of the samples are stored in the main coldroom with the temperature maintained at 4±2EC. The Sample Custodian will log the samples delivered into the coldroom in the Cold Room Sample Arrival Logbook. The coldroom will be kept locked when not in use. The water samples for metals analysis will be stored in a separate air-conditioned storage room located near the metals sample preparation area. This room will be kept locked if not being used by the analyst. The samples in these storage areas will be assigned to labeled shelves by field group. A sample location list is posted at the door of each storage room. Access to samples will be limited to authorized personnel, and a Sample Check In/Out Log is maintained.

6.2.5.1.16. Sample Handling and Shipping. After each water or soil sample is collected, it will be placed in a cooler containing ice, and the cooler will be shipped by overnight courier to a contract laboratory. The samples will be placed upright in the cooler, and secured with inert cushioning material to prevent breakage. A completed COC form will accompany all samples. Complete packaging and shipping procedures are as follows:

• The lid will be secured with strapping tape by wrapping it completely around

the cooler.

• Signed and dated custody seals will be placed on the cooler in two locations

across the opening of the cooler lid.

6.2.5.1.17. Sample Disposal. Thirty days after a laboratory report has been generated and submitted to EPA or Hill AFB, the samples are transferred to the sample disposal area. This transfer is also documented on the Sample Control Form. Samples will be disposed according to each laboratory's SOP, which is based on both State and Federal guidelines. Samples taken to the university laboratories will be disposed of by the Universities in accordance with their standard operation procedures for Treatability Study materials originating from a CERCLA site.

6.2.6 CALIBRATION PROCEDURES

6.2.6.0.1. This section discusses general requirements for field equipment and laboratory instrument calibration and standards preparation. Instrument calibration is necessary for accurate sample quantitation, and establishes the dynamic range of an instrument. Criteria for calibration are specific to each method and instrument manufacturer. The following paragraphs outline the calibration procedures for the field equipment and laboratory instrumentation.

6.2.6.1. Field Equipment

6.2.6.1.1. The field equipment to be used during the ground-water sampling program include a water-level sounder; an Eh meter; and a specific conductance, dissolved oxygen, temperature, turbidity meter, and organic vapor meter. The meters will be calibrated according to the procedures outlined below.

6.2.6.1.2. Water-Level Sounder. Electric water-level sounders will be checked before the beginning of field activities by comparing the scale on the water-level tape against an engineering measurement tape.

6.2.6.1.3. Eh Meter. An Orion model 250A or equivalent will be used to measure Eh. In accordance with the manufacturer's instructions the meter is calibrated with an Eh calibration kit that includes a 90 to 100 millivolt (mV) and a 250 to 270 mV standard.

6.2.6.1.4. pH, Turbidity, Dissolved Oxygen, Temperature, and Specific Conductivity Meter. A Horiba U-10 Water Quality Instrument or equivalent will be used for pH, salinity, turbidity, dissolved oxygen, temperature and specific conductivity measurement. The instrument will be calibrated daily prior to use according to the manufacturer's instructions. The Horiba meter follows an automatic calibration routine in which a single standardizing solution, supplied by the manufacturer, is used to calibrate the meter for pH, specific conductivity, salinity, turbidity, and dissolved oxygen. Periodically, the automatic calibration of the Horiba will be checked manually using independent reference solutions including two pH buffers that bracket the expected pH (generally pH 7 and pH 10), a 1.000 Fhos/cm standard conductivity solution, and turbidity-free distilled water. The temperature probe will be checked periodically against an NIST-traceable thermometer to confirm measurements.

6.2.6.1.5. Organic Vapor Meter. Any organic vapor detectors including flame ionization detectors (FIDs) and photoionization detectors (PIDs) will be calibrated daily prior to use and any time that instrument drift is suspected. In addition, calibration will be checked at the conclusion of each day of use in order to evaluate instrument performance. Instruments will not be adjusted before the final calibration check has been performed and recorded. Calibration procedures will be documented in the log book or B-17 on the appropriate field form. Calibration gases that have a shelf life will not be used past the expiration date.
 
 

6.2.6.2. Laboratory Instruments

A subcontract laboratory will provide analytical services for all Level III data. The following paragraphs describe procedures for standard preparation and instrument calibration for SW-846 methods. Attachment B summarizes instrument calibration and corrective action procedures.

6.2.6.2.1. Standard/Reagent Preparation. Data accuracy is dependent upon the accuracy of the standards used for instrument calibration. To ensure the highest quality standard, primary reference standards used by the contract laboratory are obtained from the National Institute of Standards and Technology (NIST), EPA CRADA vendors, or other reliable commercial sources. When standards are received at the laboratory, the date received, supplier, lot number, purity, concentration, and expiration date are recorded in a standards log book. Vendor certification for the standards are retained in the files.

6.2.6.2.1.2. Standards are obtained either in their pure form, or in stock or working standard solutions. Dilutions are made from vendor standards. All standards are given a standard identification number and the following information is recorded in the standards log book; source of the standard, the initial concentration of the standard, the final concentration of the standard, the volume of the standard that was diluted, the volume of the final solution, the solvent and the source and lot number of the solvent used for standard preparation, and the preparer's initials. All standards are validated prior to use.

6.2.6.2.1.3. Validation procedures for standards include a check for chromatographic purity and verification of the standard's concentration by comparing its response to a standard of the same analyte prepared at a different time or obtained from a different source. Reagents also are analyzed for purity; for example, every lot of dichloromethane (used for organic extraction) is analyzed for contaminants prior to use in the laboratory. Standards are checked routinely for signs of deterioration (e.g. discoloration, formation of precipitates, and changes in concentration) and are discarded if deterioration is suspected or the expiration date has passed. Expiration dates are based on vendor recommendation, the analytical method, or internal research. Stock solutions for VOCs are not be held for more than 30 days. Fresh working calibration standards shall be prepared every week. Stock solutions for semi-volatile organic compounds shall not be held for more than 90 days. Dilutions below 1 ppm shall not be held more than 30 days.

6.2.6.2.2. Calibration of Organic Methods. Calibration of instrumentation is required to ensure that the analytical system is operating correctly and functioning at the sensitivity necessary to meet established reporting limits (i.e., PQLs). Each instrument will be calibrated with standard solutions appropriate to the type of instrument and the linear range established for the analytical method.

6.2.6.2.2.1. Analytical instruments will be calibrated using standards in accordance with the specified analytical methods and manufacturer's procedures. At a minimum, written calibration procedures include the equipment to be calibrated, the reference standards used for calibration, the calibration techniques, actions, acceptable performance tolerances, frequency of calibration, and calibration documentation format. Records of standard preparation and instrument calibration will be maintained. Instrument calibration will include daily checks using standards prepared independently of the calibration standards and instrument response will be evaluated against established criteria. The analysis logbook, maintained for each analytical instrument, will include at a minimum: the date and time of calibration, the initials of the person performing the calibration, the calibrator reference number and concentration. Attachment B summarizes the calibration procedures. Instrument calibration procedures for specific instruments used for organic analyses are discussed in the following paragraphs.

6.2.6.2.2.2. Gas Chromatography/Mass Spectrometry (GC/MS). Each day prior to analysis of samples for VOCs, the instrument will be tuned with bromofluorobenzene (BFB) (according to the tuning criteria specified in the USEPA Contract Laboratory Program [CLP]). No samples will be analyzed until the instrument has met tuning criteria.

6.2.6.2.2.3. After the instrument has met tuning criteria, it will then be calibrated for all target compounds. An initial calibration curve will be produced, and certain compounds referred to as System Performance Calibration Compounds (SPCC) and Continuing Calibration Compounds (CCC) will be evaluated to ensure that the system is within calibration. If the daily SPCCs and CCCs do not meet the established criteria, the system will be recalibrated.

6.2.6.2.2.4. Calibration standards at a minimum of five concentrations will be prepared by secondary dilution of stock standards. All or a subset of the compounds listed in EPA Methods 8240 can be used as calibration standards.

6.2.6.2.2.5. Each calibration solution including internal standards and surrogates will be introduced according to EPA Method 5030 for volatile compounds. A relative response factor (RF) will be calculated for each compound relative to the internal standard whose retention time is closest to the compound being measured. The RF is calculated as follows:

Where:

A_x = Area of characteristic ion for the compound being measured

A_is = Area of characteristic ion for the specific internal standard

C_is = Concentration of the specific internal standard

C_x = Concentration of the compound being measured.

6.2.6.2.2.6. The average relative response factor (RF₁) will be calculated for each compound using the values from the five-point calibration. A system performance check must be made before the calibration is accepted as valid. The SPCCs are checked for a minimum average relative response factor. The five volatile SPCCs are chloromethane, 1,1-dichloroethane, bromoform, 1,1,2,2-tetrachloroethane, and chlorobenzene. The minimum acceptable average relative response factor for volatile compounds is 0.300 (0.250 for bromoform).

6.2.6.2.2.7. The percent relative standard deviation (% RSD) for the CCCs will be calculated from the RFs in the initial calibration and must meet specified criteria. The volatile CCCs are 1,1 -dichloroethane, chloroform, 1,2-dichloropropane, toluene, ethylbenzene, and vinyl chloride. The formula used to calculate % RSD is:

%RSD=D x 100

Where:

RSD = Relative standard deviation

= Mean of 5 initial RFs for a compound

SD = Standard deviation of the RFs for a compound

6.2.6.2.2.8. Every 12-hour shift, each GC/MS must be tuned by purging or injecting 4-bromofluorobenzene (BFB) for volatile compounds. Also, initial calibration of the GC/MS will be checked by analyzing a calibration standard (usually the mid level standard) and checking the SPCC and CCC performance. If the minimum relative response factors for SPCCs are not met, corrective action must be taken before samples are analyzed. The percent difference of relative response factor compared to the average relative response factor from the initial calibration is calculated as follows:

Where:

RF₁ = Average relative response factor from initial calibration

RF_c = Relative response factor from current calibration check standard.

If the percent difference criterion for each CCC compound is met, the initial calibration is assumed to be valid. If the criterion is not met for any CCC, corrective action must be taken. A new five-point calibration must be generated if the source of the problem cannot be found and corrected.

6.2.6.2.2.9. The internal standard responses and retention times in the CCC must be evaluated. If any internal standard retention time changes by more than 30 seconds from the last calibration check (12 hours), the system must be checked for malfunctions and corrected as necessary. If the extracted ion current profile (EICP) area for any of the internal standards changes by a factor of two from the last daily calibration standard check, the system must be checked for malfunctions and corrections made as necessary. All samples analyzed during the time the system was malfunctioning must be re-analyzed.

6.2.6.2.2.10. Gas Chromatography. Initial calibration consists of determining the linear range, establishing detection limits, and establishing retention time windows. The calibration will then be checked daily to ensure that the system calibration remains within specifications. If the daily calibration check does not meet established criteria, the system will be recalibrated.

6.2.6.2.2.11. Calibration standards will be prepared according to the standard operating procedure for the method. For the SW 846 8000 series methods, a calibration standards will be prepared for each analyte of interest at five concentration levels. One of these standards will be slightly above the method detection limit. The other standards will bracket the concentration range expected in the environmental samples, but not exceed the working range of the detector.

6.2.6.2.2.12. A reagent water blank will be run prior to calibration to show the absence of interferences. The calibration standards then will be introduced into the system and a calibration curve will be generated for each analyte. The response factor for each analyte at each concentration will be calculated as follows:

6.2.6.2.2.13. Acceptance criteria for instrument response linearity checks are based upon the correlation coefficient (r) of the best fit line for the calibration data points, or on the percent relative standard deviation (% RSD) for response factors calculated for each analyte at each level over the working range. The correlation coefficient is calculated as:

Where:

x = Calibration concentrations

y = Instrument response (peak area)

n = Number of calibration points (x,y data pairs).

The percent RSD is calculated as:

Where:

% RSD = Relative standard deviation

= Means of 5 initial RFs for a compound

SD = Standard deviation of the RFs for a compound.

6.2.6.2.2.14. If the coefficient of correlation, r, is greater than or equal to 0.995, or the % RSD is less than or equal to 20 percent, the calibration is considered valid. The use of r or % RSD is instrument specific, and only one of these criteria will be used on each instrument.

6.2.6.2.2.15. The calibration curve and response factors will be checked daily by injecting at least one calibration standard, usually the mid-range standard. The percent difference between initial and continuing response factors will be calculated using the following equation:

Where

RF₁ = Average relative response factor from initial calibration

RF₂ = Response factor from continuing calibration.

An acceptable percent difference will be within plus or minus 15 percent.

6.2.6.5.2.16. Retention time windows must be established for each analyte during initial calibration per SW 846, Method 8000. The retention time window must be checked prior to sample analysis using the calibration check standard. A warning limit specific to the method will be used. If the standard fails to meet the retention time window, the instrument will be recalibrated.

6.2.7 ANALYTICAL PROCEDURES AND DETECTION LIMITS

6.2.7.0.1. All samples collected for Level III data will be analyzed by a subcontract analytical laboratory. All samples will be prepared and analyzed using the Test Methods for Evaluating Solid Waste, Physical/Chemical Methods, SW-846, 3rd Edition (USEPA, 1986). The methods and holding times are listed in Table B-3. The units of measure and typical practical quantitation limits for each analyte are listed in Attachment C. These are laboratory-specific target reporting limits that can be met in the absence of matrix interferences or high contaminant concentrations, and are at least as stringent as the reporting limits specified for the individual analytical methods. The laboratory-specific PQLs will be determined when the subcontract laboratory is identified.

6.2.8 DATA REDUCTION, VALIDATION, AND REPORTING

6.2.8.1. Field Measurements

6.2.8.1.1. Raw data from field measurements and sample collection activities will be documented in the field log book and on the appropriate forms, as described in the a FSP. The field measurements and data collected during sampling will be presented in the Summary Report scheduled for this project. All field data generated during this investigation will be evaluated under the direction of the Montgomery Watson Quality Control Coordinator before it is incorporated in the report.

6.2.8.2. Laboratory Measurements

6.2.8.2.1. Data Reduction Calculations. Data will be reduced as specified by the analytical methods. These calculations are specific to the analytical instruments that are used for the analysis, the level of automation, and the type of software used to reduce the data. The procedures used for data reduction for each analytical method are described in the laboratory's SOPs.

6.2.8.2.2. Data Validation. The laboratory will perform in-house analytical data reduction and review under the direction of the Laboratory Project Manager and the Laboratory QA Officer before the data are released to Montgomery Watson. The Laboratory Project Manager and Laboratory QA Officer also are responsible for assessing the data quality and qualifying any data that may be unreliable. The laboratory will prepare and retain full analytical and QC documentation. The data reduction and review will be conducted as follows:

The laboratory review of the data includes assessing compliance with the control limits in QAPP. Accuracy an precision are the primary data parameters that can be used to calculate control limits. Data to evaluate accuracy are obtained primarily from separately prepared laboratory QC samples or from spiked field samples. Data used to evaluate precision are QC sample analyses or the replicate analysis of field samples. The calculations that are used to evaluate precision and accuracy are defined in the laboratory's SOP and/or QA/QC manual. Precision and accuracy quality control limits are generated from the statistical analysis of QC sample results. The quality control limits that will be used to evaluate the data are listed in Attachment A.

6.2.8.2.3. Data Reporting. The analytical data will be reported in a format organized to facilitate data evaluation. All of the data, including QC data. will be reported in the chronological order in which they were produced. The following information will be included in each data package:

• A list of diluted samples including their dilution factors.

6.2.9 INTERNAL QUALITY CONTROL

6.2.9.1. Field Program

6.2.9.1.1. Internal quality control evaluates whether a method is performing within acceptable limits of precision and accuracy. On the sampling level, quality control samples used to assess field sampling techniques and environmental conditions during sample collection and transportation include blind duplicates, trip blanks, source water blanks, and equipment blanks.

6.2.9.1.2. Blind duplicate or replicate samples will be used to assess variability in the sample matrix and to assess sampling precision. The sampling procedures will be evaluated by comparing the analytical results of blind duplicate or replicate sample pairs. If the reported values for the sample pair are similar, the samples are considered to be representative of the environment. A large difference (greater than 40 percent) between the reported values for the sample pair indicates that there may have been a problem during sampling or analysis. Blind duplicate analyses will be used to evaluate precision by calculating the RPD between a blind duplicate sample and its associated environmental sample. The RPD will be compared to the MS/MSD QC limits for precision. (QC limits for precision for blind duplicate or replicate samples have not been established.) Relative percent difference values within the QC guidelines indicate that good sampling and analytical procedures were followed. Relative percent difference values outside the QC limits indicate that sample may be heterogeneous, or that there may have been a problem during sampling and/or analysis. Section 6.2.4 outlines the procedures for collecting blind duplicate samples.

6.2.9.1.3. Trip blanks will be used to evaluate representativeness by assessing whether VOCs were introduced into samples during handling, shipping, or storage at the laboratory. Trip blanks prepared by the laboratory (see Section 6.2.4) will be included with each sample shipment that contains ground-water or surface water samples for VOC analysis

6.2.9.1.4. Source water data from ongoing investigations at OU 1 will be used to assess representativeness, by evaluating whether contaminants present in the samples are representative of the environment or are attributable to the water used for equipment decontamination. A source water blank will not be collected during this investigation (See Section 6.2.4).

6.2.9.1.5. Equipment blanks will be used to assess the equipment decontamination procedures and evaluate whether the samples are representative of the environment. The results of each equipment-blank analysis will be reviewed for the presence of target analytes. If target analytes are found, the data from the associated environmental samples will be evaluated to determine if they are representative of environmental conditions or the result of incomplete equipment decontamination. Equipment-blank samples will be collected as outlined in Section 6.2.4.

6.2.9.2. Laboratory Analysis

6.2.9.2.1. The general objectives of a laboratory QC program are to:

• Ensure that all data are properly archived.

6.2.9.2.2. Internal quality control for analytical services will be conducted by the laboratory in accordance to their standard operating procedures, the individual method requirements, and this QAPP. Before making significant changes to the QAPP or analytical methodology, the laboratory will notify the Montgomery Watson Project Manager in writing.

6.2.9.2.3. Laboratory quality control consists of two distinct components: a laboratory and matrix component. The laboratory component measures the performance of the laboratory analytical process during the sample analyses, while the matrix component measures the effects on the method performance of a specific matrix. Method blanks and laboratory control samples uniquely measure the laboratory component of method performance, while matrix spikes, matrix spike duplicates, laboratory sample duplicates, and surrogate spikes measure the matrix component of method performance, but also reflect laboratory performance. The following paragraphs discuss the QC samples and parameters to be evaluated to assess the overall laboratory data quality.

6.2.9.2.4. Holding Time. Holding time reflects the length of time that a sample or sample extract remains representative of the environmental conditions. Depending on the analysis, either one or two holding times will be evaluated. For those analyses that do not include sample extraction, one holding time will be evaluated: the amount of time between sampling and analysis. For analyses that have an extraction prior to analysis (e.g., herbicides), two holding times will be evaluated: 1) the length of time from sampling until extraction, and 2) the length of time from extraction to analysis. Holding times for each analytical method are listed in Table B-3 of this QAPP. Analytical results of samples whose holding times are exceeded are considered quantitatively questionable and may be biased low.

6.2.9.2.5. Blind Duplicate and Blind Replicate Samples. Like the field procedures, the analytical procedures will be evaluated by comparing the analytical results of blind duplicate or replicate sample pairs. If the reported values for the sample pair are similar, the samples are considered to be representative of the environment. A large difference (greater than 40 percent) between the reported values for the sample pair indicates that there may have been a problem during sampling or analysis. Blind duplicate analyses will be used to evaluate precision by calculating the RPD between a blind duplicate sample and its associated environmental sample. The RPD will be compared to the MS/MSV QC limits for precision. (QC limits for precision for blind duplicate or replicate samples have not been established.) Relative percent difference values within the QC guidelines indicate that good sampling and analytical procedures were followed. Relative percent difference values outside the QC limits indicate that sample may be heterogeneous, or that there may have been a problem during sampling and/or analysis. Section 4.0 outlines the procedures for collecting blind duplicate samples.

6.2.9.2.6. Method Blanks. Method blanks will be used to evaluate representativeness by identifying any contaminants that have been introduced during analysis. Method blanks are generated in the laboratory and consist of ultra-pure water. Method blanks are carried through each processing step necessary for an analytical procedure and are analyzed at frequency of one per 20 samples or daily, whichever is more frequent. These blanks measure contamination originating from the laboratory (i.e., water, air, reagents. equipment, and instruments used for analysis), and help in distinguishing low-level field contamination from laboratory contamination. If analytes of interest are found in both the method blank and in associated environmental samples, the environmental data will be qualified as per USEPA guidelines (USEPA, 1988b). The data from the associated samples may be considered quantitatively questionable depending on the relative concentrations of contaminants in the method blank and the environmental sample.

6.2.9.2.7. Surrogate Spikes. Surrogate spikes will be used to evaluate the accuracy of method and analytical instrument performance for VOC and PCB analyses. During analysis surrogate spike compounds behave similarly to the analytes of interest. Surrogates are added to each sample and method blanks, including QC samples, prior to extraction or analysis. The surrogate spikes specified in the SW-846 Test Methods for Evaluating Solid Waste (USEPA, 1986) will be used for analysis. After the analysis has been completed, the percent recovery of each surrogate spike will be calculated and compared to the QC limits established by the method of analysis (see Attachment A). Percent recoveries within the QC limits indicate acceptable accuracy during analysis. Recoveries outside the QC limits indicate that there may have been a problem during analysis (matrix or non-matrix) and that the data may be of questionable value; the data will be qualified accordingly.

6.2.9.2.8. Laboratory Control Samples. Laboratory control samples (LCS) will be used to evaluate accuracy. These samples are carried through the same analytical procedures as the environmental samples and are used to evaluate method and analytical procedure performance in the absence of matrix interference. Laboratory control samples are prepared in the laboratory and consist of ultra-pure water that is spiked with specific compounds as outlined in the analytical methods. An LCS sample will be prepared and analyzed at a frequency of one per 20 samples, or daily, whichever is more frequent. Accuracy will be evaluated by calculating the percent recovery for each spiked compound and comparing it to the QC limits established by the individual methods. Values within the established QC limits indicate acceptable analytical accuracy. Values outside the QC limits indicate that the data may be questionable.

6.2.9.2.9. Matrix Spike and Matrix Spike Duplicate Samples. Results of MS/MSD sample analysis will be used to evaluate accuracy and precision. Unlike LCSs, MS/MSD samples are used to assess the influence of the sample matrix (matrix interference) on the analysis. Each MS/MSD sample will be spiked with the compounds specified by the analytical method. To evaluate accuracy the percent recovery for each spiked compound will be calculated and compared to the QC limits established by the method. Precision will be evaluated by calculating the RPD between the MS and MSD samples for each spiked analyte. These RPDs will be compared to the QC limits established by laboratory performance. Percent recovery and RPD values within the QC limits indicate acceptable precision and accuracy Values outside the QC limits indicate that there may have been a matrix interference during analysis. The laboratory data validation protocol will be based on precision and accuracy measurements from MS/MSDs. Individual compound recoveries will be compared with acceptance limits. If a matrix spike analyte fails acceptance criteria, the MS/MSD will be reanalyzed and a LCS also will be analyzed. For the method to be considered in control, those compounds that failed the matrix spike criteria must be within acceptance limits in the LCS. If, after re-analysis, analytes that failed acceptance criteria in the MS and MSD pass acceptance criteria in the LCS, these analytes may be considered biased due to sample matrix effects.

6.2.9.2.10. All samples analyzed or prepared in a process batch without an MS and MSD will, at a minimum, have a method blank and LCS. The environmental samples in this batch will be considered in control if more than 80 percent of the target compounds in the LCS are within acceptance limits

6.2.10 PERFORMANCE SYSTEMS AUDITS

6.2.10.1. Field Programs

6.2.10.1.1. Oversight of Montgomery Watson field procedures will be the direct responsibility of the Montgomery Watson Project Manager, who will review all elements of the QAPP to ensure that the objectives of the RI are met. In addition to an initial review, the sampling procedures will be reviewed as the field work progresses so that any necessary modifications can be made

6.2.10.1.2. Internal audits of Montgomery Watson field activities (sampling and measurements) will be conducted by the Montgomery Watson QC Coordinator or the coordinator's designee. The audits will include examining field measurement records, field equipment calibration records, field sampling records, field instrument operation records, sample collection procedures, sample handling and shipping procedures, and chain-of-custody procedures. Field activities will be audited early in the project to verify that all of the procedures outlined in the FSP and QAPP are being followed. Follow-up audits will be conducted to correct deficiencies, and to verify that QA procedures are maintained throughout the project.

6.2.10.2. Laboratory Audits

6.2.10.2.1. In-house and regulatory agency audits of laboratory systems and performance are a regular part of a laboratory QC program and are outlined in the subcontract laboratory's QA/QC plan. The audits consist of a review of the entire laboratory system and at a minimum include: examination of sample receiving, log-in, storage, and chain of-custody documentation procedures; sample preparation and analysis; and instrumentation procedures. An external audit may be performed by Hill AFB or its designee, USEPA, or UDEQ personnel prior to or during the field work, to verify proper implementation of laboratory procedures and adherence to this QAPP.

6.2.11 PREVENTATIVE MAINTENANCE

6.2.11.1. Field Equipment

6.2.11.1.1. The field equipment that will be used during this investigation includes an electronic water-level sounder; an Eh meter; a PID and/or FID; and a pH, specific conductivity, salinity, turbidity, temperature, and dissolved oxygen meter. All meters and instruments will be maintained and used according to the manufacturers' directions. Each piece of equipment will be inspected on a regular basis to ensure that the equipment is operational. Any preventative maintenance or repair conducted in the field will be recorded in the field log book.

6.2.11.2. Laboratory Equipment

6.2.11.2.1. Documentation. All maintenance performed on an instrument is documented analyst performing the maintenance, and the type of maintenance are recorded in this logbook. Receipts from routine maintenance performed by the manufacturer's representative are kept in folders and filed in the department's file cabinets.

6.2.11.2.2. Contingency Plan. In the event of instrument failure, every effort will be made to analyze samples by alternate means within holding times. If the redundancy in equivalent instrumentation is insufficient to handle the affected samples, efforts will be made to secure the same or equivalent analyses at another location. Dr. Jon Ginn, Hill AFB, will be advised of any proposed changes in methodology or location.

6.2.12 DATA ASSESSMENT PROCEDURES

6.2.12.0.1. The quality of the field and analytical data will be evaluated using precision, accuracy, representativeness, completeness, and comparability (PARCC) parameters, which are quantitative and qualitative statements that describe data quality. The PARCC parameters will be used to determine whether the data quality objectives of this investigation have been met by comparing QC sample results and standard procedures with acceptance criteria established for this project. The PARCC parameters that will be used for data evaluation are defined in Section 6.2.3.

6.2.12.1. Field Data

6.2.12.1.1. Field measurement data will be assessed by the EPA and/or Montgomery Watson QC Coordinator or designee. The data quality evaluation, in terms of the PARCC parameters, will focus primarily on the laboratory data. However, the field data will be evaluated qualitatively in terms of the PARCC parameters. The following sections discuss how the PARCC parameters will be used to evaluate the field data and field sampling procedures.

6.2.12.1.2. Precision. Sampling precision is affected by the procedures used for sample collection, handling, and transportation. To reduce the variability that may be introduced during sampling, the FSP outlines the standard sampling, handling, and shipping procedures that will be used for each sampling program. The use of these procedures should minimize variability in the sampling process.

6.2.12.1.3. In addition, the results of blind duplicate and blind replicate sample analyses also will be used to evaluate sampling precision. The RPD will be calculated for each blind duplicate sample pair. Although the results of blind duplicate sample analyses also reflect the variability associated with analytical procedures, low RPD values are an indication that consistent sampling techniques were used for sample collection.

6.2.12.1.4. Accuracy. Although there is no way to quantitatively measure the accuracy of the field program using percent recovery, some aspects of accuracy can be assessed, such as comparing the length of the water-level probe to another measuring tape of known length and proper calibration of the field instruments.

6.2.12.1.5. Representativeness. The representativeness of the field data is determined by the design of the data collection procedures. The sampling and field measurement procedures to be used are based on existing analytical data, hydrogeology, the physical setting of OU 1, and the past land use history. Representativeness of the field sampling procedures and the field measurements will be evaluated by comparing the sampling and measurement procedures used in the field to the procedures outlined in the FSP and this QAPP. In addition, the results of equipment blank samples will be used to evaluate the representativeness of field sampling procedures. Contaminants detected in equipment blanks are indications that the decontamination procedures are not completely effective, and that contaminants detected at specific sites may be attributable to cross-contamination rather than the environment.

6.2.12.1.6. Comparability. The comparability of the field sampling procedures and field measurement data will be evaluated by comparing them to previous sampling rounds.

6.2.12.1.7. Completeness. Completeness of the field program will be evaluated to ensure that the appropriate number of samples were collected for analysis, and that field data of the type and quantity outlined in the FSP were collected. Completeness of the field investigations will be evaluated by comparing the actual number of samples and the actual quantity of data that were collected to the requirements outlined in the FSP.

6.2.12.2. Laboratory Data

6.2.12.2.1. The laboratory data will be assessed by EPA project manager or the Montgomery Watson QC Coordinator or designee, and based on the assumption that the sample was collected, handled, and analyzed according to the FSP and this QAPP. The data reviewer will conduct a systematic review of the data for compliance with the QC criteria established in the QAPP, and will identify any data omissions or data that do not meet the quality control criteria. The reviewer also will interact with the laboratory to correct any data deficiencies. Decisions to repeat sample collection or analyses will be made by EPA or Hill AFB based on the extent of the data deficiencies and their importance in the overall context of the project. Results of the data assessment will be presented in an appendix of the report scheduled to summarize the results of this investigation.

6.2.12.2.1.1. Laboratory Data Assessment Procedures. As discussed above, PARCC parameters will be used to evaluate the quality of analytical data and determine whether the data quality objectives of the project have been met. To assess the quality of the analytical data, the results of QC sample analyses will be evaluated using quality control limits established by the analytical methods used for analysis, or by past laboratory performance (see Attachment A). Results of the quality control sample evaluation then will be expressed in terms of the PARCC parameters and used to assess the quality of the analytical data.

6.2.12.2.1.2. The quality control samples that will be used to evaluate the analytical data for this program include trip blanks, equipment blanks, source-water blanks, blind duplicate samples, blind replicate samples, method blanks, surrogate spikes (when applicable), laboratory control samples (when applicable), and matrix spike/matrix spike duplicate samples. The specific types and descriptions of the QC samples that will be collected in the field are presented in Section 6.2.4 of this QAPP. The total number of each type of QC sample that will be collected during each sampling program is listed in Table B-4. The quality control samples that are prepared in the laboratory and the rate at which these samples are analyzed are method-specific (see Section 6.2.9). A summary of the QC sample evaluation of laboratory data in terms of the PARCC parameters is in Table B-5, and the quality control criteria used for evaluating the QC sample data are presented in Attachment A. The acceptance limits for MS/MSD, surrogate spikes and LCS are updated periodically. The laboratory shall inform Montgomery Watson before new limits are implemented. The following sections describe the criteria that will be used to evaluate the laboratory data.

6.2.12.2.1.3. Precision. Analytical precision is determined by analyzing field duplicates or replicates submitted "blind" to the laboratory, and MS/MSD samples. Relative percent difference is calculated between the sample pairs and compared with control limit acceptance criteria. The data quality objectives for precision during this program are based on laboratory established control limits, which are specific to each analyte (see Attachment A).

6.2.12.2.1.4. Accuracy. Accuracy is a quantitative measure of the bias of a method or the level of agreement between a measurement and a known true value. Laboratory accuracy will be evaluated using the results for surrogate spike, MS/MSD, and LCS/LCSD sample analyses. As with precision, the accuracy objectives for the data are based on laboratory established limits, and vary with the specific analyte (see Attachment A).

6.2.12.2.15. Representativeness. Representativeness is a qualitative parameter that evaluates whether or not the data represent the actual environmental conditions. Representativeness will be evaluated by analysis of laboratory method and equipment blanks, and blind duplicate or replicate samples. Laboratory method and equipment blanks will be used to duplicate or replicates will be used to evaluate laboratory performance.

6.2.12.2.1.6. Representativeness also is evaluated using holding-time criteria, which reflect the length of time that a sample or extract remains representative of the environmental conditions after sample collection. Holding times are compared to standard method-specific holding times accepted by the EPA. All holding times within the acceptance criteria are considered representative. Those holding times outside of EPA acceptance criteria are qualitatively evaluated to determine the effect on sample representativeness.

6.2.12.1.1.7. Comparability. Comparability is a qualitative expression of the confidence with which one data set can be compared to another. Comparability is maximized through the use of standard analytical method and units of measurement.

6.2.12.2.1.8. Completeness. Completeness also will be used to assess the data. Completeness is expressed as a percentage and is defined as the number of valid samples relative to the total number of samples gathered during the sampling programs. Completeness will be calculated using the following equation:

Completeness = (Number of Valid Samples) / (Total Number of Samples) x 100

6.2.13 CORRECTIVE ACTIONS

6.2.13.1. Field Programs

6.2.13.1.1. The field staff will be responsible for documenting and reponing all suspected technical and QA non-conformances, and suspected deficiencies during any field activity. The non-conformances and/or deficiencies will be documented in the field log hook and reported to the Hill AFB Project Manager. If the problem is associated with field measurements or sampling equipment, the field staff will take the appropriate steps to correct the problem. Typical field procedures to correct problems include the following:

• Repeating the measurement to check for error

• Checking or replacing batteries

• Recharging batteries

• Recalibrating the instruments

• Replacing the meters or instruments used to measure field parameters

• Stopping work until the problem is corrected (if necessary).

6.2.13.1.2. If a non-conformance or problem requires a major adjustment to the field procedures as outlined in the FSP (e.g., changing sampling methodology), the Project Manager, in conjunction with the Quality Assurance Coordinator, will be responsible for initiating corrective actions. The Project Manager will be responsible for the following:

• Evaluating the reported non-conformance

• Controlling additional work on non-conforming items

• Determining the appropriate corrective actions

• Maintaining a log of all non-conformances and corrective actions

The Hill AFB Project Manager will ensure that no additional work that is dependent on the non-conforming activity is performed until the appropriate corrective actions are completed.

6.2.13.2. Laboratory Analysis

6.2.13.2.1. Corrective actions are required whenever unreliable analytical results prevent the quality control criteria as specified by the method or the laboratory QAPP from being met. The corrective action that is taken depends on the analysis and the non conformance.

6.2.13.2.2. Corrective actions will be undertaken if one of the following occurs:

• QC data are outside the acceptance windows for precision and accuracy

• Blanks contain contaminants above acceptance levels X

• There are unusual changes of detection limits during analysis

• Deficiencies are detected during QA audits

• Inquiries concerning data quality are received from Montgomery Watson.

6.2.13.2.3. Corrective actions are primarily handled at the bench level by the analyst who reviews the sample preparation or extraction procedures, and performs the instrument calibration and analysis. If the problem persists or its cause cannot be identified, the matter will be referred to the department supervisor or QA department for further investigation. Once resolved, full documentation of the corrective action procedure will be filed with the QA department. A summary of the corrective actions will be included in the data package submitted to Montgomery Watson.

6.2.1 QUALITY ASSURANCE REPORTS

6.2.14.0.1. All of the analytical data collected during the investigation will be presented in an appendix to the Summary Report scheduled for this investigation. The data validation for all Level III data will be included in the final Treatability Study Report. The following information will be presented in this report:

• Sampling procedures (planned and implemented, problems, and corrective actions)

• Analytical procedures and detection limits

• Analytical data (environmental and QC sample results)

• Results of the data quality evaluation

• Conclusions and recommendations.

Attachment A

Attachment B

Attachment C