ANALYTES OF INTEREST Soil gas, soil and groundwater samples will be collected from the sites of interest. The target compounds analyzed for during this project are cis-1,2-dichloroethene (cis-1,2-DCE), trichloroethene (TCE), tetrachloroethene (PCE), and carbon tetrachloride (CCL4). For optimum detection of the analytes of interest, the gas chromatograph (GC) will be equipped with a photoionization detector (PID) and an electron capture detector (ECD) in series. The ECD is extremely sensitive to TCE and PCE and will be used as a confirmation back-up tool, while the PID will be calibrated to and used for actual analysis of cis-1,2-DCE, TCE, and PCE. The ECD will be used to quantify CCL4.

SAMPLE COLLECTION PROCEDURESTo minimize cross-contamination of samples, each test hole is probed with pre-cleaned sampling rods. Pre-cleaning consists of steam-cleaning, washing with Alconox soap, and rinsing with deionized water. Enough probe rods are available so that they are used only once between decon periods. In addition, disposable gloves, polyethylene tubing, and syringes are used for sample collection and analysis to minimize cross-contamination. All disposable equipment is used only once.
Procedures specific to the collection of soil gas, soil and groundwater are detailed in the following paragraphs.

Soil Gas Sampling ProcedureSoil gas sampling procedures include advancing 1.25-inch outer diameter (O.D.) probe rods with a threaded point holder and disposable point to the desired sampling depth on each test hole. Polyethylene tubing (1/4" dia.) is then lowered down the inner diameter of the probe rods and threaded onto the threaded point holder. An o-ring is utilized to ensure a tight seal surrounding this threaded fitting. The probe unit is then used to retract the rods approximately 2- to 4-inches, thereby disengaging the disposable drive point attached to the threaded point holder. Once a space or void has been made for soil gas extraction, the tubing is purged, utilizing the on-board vacuum system, to remove any extraneous air present in the polyethylene tubing. Once purged, a soil gas sample is collected using a disposable syringe. The soil gas sample is collected by first pinching off the upper sample tubing, puncturing the tubing directly above the upper probe rod with the syringe, and evacuating 0.20 cc of soil gas from within the tubing into the syringe. The sample, 0.20 cc of soil gas, is then immediately injected into the GC for analysis.

Soil Sampling Procedure Soil sampling procedures include vertically profiling the soils to the desired sampling depths or probe refusal. Continuous profiling procedures include attaching a 4-ft. in length Geoprobe macro-core soil sampler to the leading probe rod, and advancing the sampler in 4-ft. increments from the ground surface to the desired sampling depths. Discreet profiling procedures include attaching a closed piston-assembled 4-ft. in length Geoprobe macro-core soil sampler to the leading probe rod, and advancing the sampler in 4-ft. increments from the ground surface to the desired sampling depth. Once at depth, small diameter (1/8") extension rods are lowered down the I.D. of the probe rods until they make contact with the piston rod and the reverse-threaded, piston rod stop pin. The stop pin is unthreaded freeing up the inner piston rod and driving point. The 4-ft. sampler is then advanced filling the sample tube with the discreet sample. The sampler is equipped with an inner PVC, transparent liner, useful for logging purposes. The filled liners are placed on visqueen and placed on the ground surface for visual observation. Soil samples are then collected and logged by Client personnel. Once a selected sample has been determined, EPS will conduct the on-site analysis of the soil samples. Selected duplicate samples will be sent from the Client to a certified laboratory for analysis.

Groundwater Sampling Procedure Once it has been determined that depth to groundwater is within sampling range of the hydraulic probe and that sufficient water is available for sampling, groundwater samples will be collected for analysis. A 1-inch by 2-foot length of mill-slotted groundwater sampling tool, which is attached to the leading probe rod, is advanced in 3-foot sections to the desired sampling depth at each test hole. A larger diameter (1.5-inch) pre-probe point is connected to the bottom end of the mill-slotted tool. This pre-probe point is used to prevent the mill-slots from plugging with sediment as the rods are advanced to depth. If conditions do not warrant mill-slot usage due to the presence of too much clay or too small fines, then a Geoprobe SP-15 Screen Point Sampler will be used. This sampler is equipped with an inner 4-foot length of mesh-wire screen. Once the rods and SP-15 sampler are advanced to depth, the rods are retracted approximately 4 feet, which exposes the inner screen. Once groundwater has infiltrated the sampler, a groundwater sample can be collected.
Prior to sample collection, a groundwater depth measurement will be taken using a Solinst water level indicator. Static water levels are taken by lowering the water level indicator inside the probe rods until contact with water occurs. The water level is then measured to within 1/100th of an inch.
To collect the groundwater sample, the on board vacuum/volume system, a peristaltic pump, or a ball-and-seat check valve system is used. Whichever system is utilized, depending on hydrologic conditions, 3/8" polyethylene tubing will be used to collect the groundwater samples. The 3/8" tubing is lowered down the inside of the probe rod until it reaches the bottom of the mill-slotted section or SP-15 sampler. An up-down motion of the tubing allows groundwater to infiltrate up through the tubing to the ground surface if the ball-and-seat check valve system is used. Utilizing the vacuum/volume system, groundwater is extracted through the tubing by using a vacuum of no more than 21 inches of mercury. Once a sufficient amount of the groundwater has been purged from the sampling string, the tubing is pinched off at the ground surface and pulled from the probe rods. Groundwater is then immediately transferred from the lower end of the tubing directly into 40-ml vials. The vials are filled to minimize the headspace, and the samples are placed on ice and set aside awaiting analysis. Depending upon groundwater availability, a minimum of two 40-mL vials is typically collected at each sample location.

VERIFICATION OF REPORTING LIMITSThe reporting limit for each compound is determined by the sensitivity of each detector that is utilized. For example, based upon instrument performance using heated headspace the PID is typically sensitive to cis-1,2 DCE, TCE and PCE at or below 2 parts per billion (ppb). Similarly, the ECD is sensitive to CCL4 at or below 2 parts per billion. Therefore, the reporting limits for cis-1,2 DCE, TCE and PCE would be anticipated to be at 2 ppb in non-diluted samples. When running a different detector, or dual detectors, the same idea would apply to the ECD or FID pertaining to compounds to be screened. Prior to field activities, the reporting limits will be verified as described in the following paragraphs.

Verification of Soil Gas Reporting LimitsTo verify the detection limit, a standard is made for each analyte of interest, or a commercially prepared standard is utilized. The gas standard is made from stock solutions of the target compounds preserved in methanol. The distributor (Chem Service) guarantees the concentrations of the stock solutions (including DCE, TCE, PCE and CCL4) are 100ug/ml plus or minus 5% in methanol. To produce a 2 ppb gas standard, 20 microliters ( l) of each stock solution is injected into a 1000-ml glass bulb. To produce a 5 ppb gas standard, 50 l of the stock solution is injected into a 1000-ml glass bulb, and so on.

The verification of the detection limit is made by analyzing three (3) injections of the standard prepared at or below the chosen reporting limit (i.e., 2 ppb). If the instruments signal to noise ratio is 10.1 or greater, then the detection limit has been verified. If this criterion is not met, then the detection limit should be increased accordingly.

In order to meet data input requirements for the Shimadzu and P/E Nelson Analytical Data Processors, the average retention time and peak area is calculated for each analyte of interest during the verification process. The result of an internal 3-point calibration curve automatically determines a response factor, which enters into the Identification Table of the Data Processor. The response factor, as defined/required by the Data Processor, is determined by dividing the known concentration of each individual compound within the standard by its average peak area. Appropriate retention time windows are selected for the identification of each compound, based on degree of separation.

Verification of Groundwater/Soil Reporting Limits To verify a 2 g/L reporting limit, a standard is made for each analyte of interest. The liquid standard is made from stock solutions of the target compounds preserved in methanol. The stock solutions contain concentrations of the stock solutions (including DCE, TCE, PCE, and CCL4) at 100ug/ml plus or minus 5% in methanol.

A 1:5 Working Standard is prepared using the stock solutions. Standards are prepared in a 50cc volumetric flask for each calibration level using this compound mixture. Three 40-ml vials are then prepared by transferring 20-ml of the Working Standard from the volumetric glassware. Standards may also be prepared by spiking a half full 40-ml vial with the compound mixture. The vials are placed in a temperature-controlled oven at approximately 90°C for approximately 15 minutes. Once a vial has been thoroughly heated, 0.20 cc of headspace from within the vial is collected with a syringe and immediately injected into the GC for analysis.

The verification of the reporting limit is made by analyzing three (3) injections of a standard prepared at or below the chosen reporting limit (i.e., 2 g/L). If the instruments signal to noise ratio is 10.1 or greater, then the reporting limit has been verified. If this criterion is not met, then the reporting limit should be increased accordingly.

In order to meet data input requirements for the Shimadzu and P/E Nelson Analytical Data Processors, the average retention time and peak area is calculated for each analyte of interest during the verification process. The result of an internal 3-point calibration curve automatically determines a response factor, which enters into the Identification Table of the Data Processor. The response factor, as defined/required by the Data Processor, is determined by dividing the known concentration of each individual compound within the standard by its average peak area. Appropriate retention time windows are selected for the identification of each compound, based on degree of separation.

CALIBRATION OF GAS CHROMATOGRAPH Prior to the start of field activities, the GC will be calibrated using a 3-point calibration. Concentrations of 2, 10, and 20 ppb are typically used for the initial calibration for fuel compounds and chlorinated solvents. Broader calibration ranges may need to be utilized if high concentrations of the analytes of interest are detected within the field. This may be accomplished by a calibration check of a concentration range equivalent to the high concentration sample. It is anticipated that the initial calibration may be sufficient for the entirety of this field activity. Calibrations will be re-established if necessary. The stability of the calibration will be assessed during the field program through the use of calibration checks.

A chromatograph and area report that was generated by the chromatography software will be provided for each standard that is analyzed. Calibration reports that are generated by the chromatography software will also be provided.

Calibration of Gas Chromatograph for Soil Gas SamplesStandards are commonly made at concentrations of 2, 10, and 20 ppb for each chlorinated analyte of interest. However, in situations in which higher concentrations of target analytes are anticipated, a broader range of standard concentrations will be utilized. The gas standard is made from stock solutions of the target compounds preserved in methanol. Concentrations of the stock solutions (including DCE, TCE, PCE and CCL4) are 100ug/ml plus or minus 5% in methanol. To produce each concentration of the gas standard, the appropriate volume of each stock solution is injected into a 1000-ml glass bulb. Once the gas standard has been made, a known volume (typically 0.20 cc) of the gas standard is injected into the GC for each concentration. A calibration factor is calculated for each concentration of each target compound. The ratio of the detector response (peak area or peak height) to the concentration of analyte in the calibration standard is defined as the calibration factor. The CF is calculated as follows:



To evaluate the linearity of the initial calibration, the P/E Nelson integration software calculates a correlation coefficient (CC) by linear regression, from three calibration injections using a straight-line fit. The mean CF, thestandard deviation (SD), and the relative standard deviation (RSD) can be manually calculated as follows:  
Note: In the preceding calculations n is the number of calibration standards and RSD is expressed as a percentage (%). 

 


If the RSD of the calibration factors is less than or equal to 25% over the calibration range for each compound of interest, then linearity through the origin is assumed. The average calibration factor (mean CF) can then be used to determine sample concentrations by manual calculation.

If the RSD for any target analyte exceeds 25%, then one or more of the following corrective actions will be necessary and will be performed in the following order:

" Check the instrument operating conditions
" Review the results (peak area) and subsequent calculations that do not meet the linearity criteria. If the problem appears to be associated with a single standard or injection, that standard/injection may be reanalyzed and the calibration statistics recalculated.
" Re-calibrate until all compounds meet the linearity criteria.

Calibration of Gas Chromatograph for Groundwater/Soil SamplesStandards are commonly made at concentrations of 2, 10, and 20 g/L for each analyte of interest. Please note that calibration ranges may be increased within the field if high concentrated samples detected cannot be diluted within the 20 ppb calibration range. A photoionization detector (PID) will assist in determining sample locations with elevated concentrations of target analytes. This may be accomplished by a calibration check of a concentration range similar to the high concentration sample detected within the field. The liquid standards are made from stock solutions of the target compounds preserved in methanol. The stock solutions contain concentrations at 100ug/ml plus or minus 5% in methanol. A 1:5 Working Standard stock mixture is prepared using these stocks.

Working Standards for each calibration concentration (i.e. 2, 10, and 20 g/L) are prepared from Intermediate Standards. Following preparation in a 50 cc volumetric flask, 20-ml of each calibration concentration is transferred to a 40-ml vial. The standard can also be prepared by spiking a half full 40-ml vial with the compound mixture. The vials are placed in a temperature-controlled oven at approximately 90°C for approximately 15 minutes. Once a vial has been thoroughly heated, 0.20 cc of headspace from within the vial is collected with a syringe and immediately injected into the GC for analysis.

Linear regression is then used by the Nelson software to evaluate the linearity of the initial 3-point calibration by internal calculation of a correlation coefficient (r) that represents the "fit" of a straight line to the data. A value of r=1.00 indicates a perfect fit. Calibration is determined acceptable when r 0.98. Alternatively, the coefficient of determination (r2) can be determined. Calibration is determined acceptable when r2 0.97. Calibration of certain compounds with low detector response or short shelf life may be considered acceptable at a slightly lower coefficient. The Client will approve any deviance from the r 0.98 (or r2 0.97) criteria prior to analysis of any samples.

While the linear regression is performed internally by the Nelson software, it can also be performed manually using statistical procedures. The linear regression calculations produce the slope and y-intercept of a line of peak area versus compound concentration. The linear equation is as follows:

y = ax + b

Where: y = Peak area or height a = Slope of the line x = Concentration of target compound b = y-intercept Alternatively, linearity of calibration can be manually evaluated using the average CF (CF). Equations for the manual calculation of the mean CF, the standard deviation (SD), and the relative standard deviation (RSD) are used.

Once each standard has been analyzed, a calibration factor is calculated (Shimadzu software) for each concentration of each target compound. The equations for the mean CF, the standard deviation (SD), and the relative standard deviation (RSD) can be done manually as previously presented (See Page 8, Calibration of Gas Chromatograph for Soil Gas Samples).

FEILD SCREENING & ANALYSES

Soil Gas Analysis At the beginning of each day of analysis, a gaseous calibration check and method blank analysis will be performed. (See Page 12, 13) Quality Control/Quality Assurance, for further discussion of calibration checks and method blanks.) Once it has been established that both the calibration check and method blank meets the acceptance criteria, field samples are analyzed. As previously presented, a soil gas sample is collected from the sampling tubing using a disposable syringe. The sample, 0.20 cc of soil gas, is then immediately injected into the GC for analysis. After the analysis of 10 field samples, the method blank analysis is repeated. The calibration check is repeated at the end of each day. Once the calibration is established, or the response factor defined (Shimadzu software) and retention time for each individual compound has been entered into the ID table of the GC, the GC's Data Processor identifies each compound and reports the concentrations directly. The concentration of an analyte in the sample can also be calculated manually using the following equation: Where: As = Peak area for the analyte in the sample D = Dilution factor (dimensionless). If no dilution was necessary, D=1. CF = Mean calibration factor from the initial gas calibration (area per ppb) If a sample exhibits a concentration that potentially saturates the detector, dilution will be necessary

Soil AnalysisAt the beginning of each day of analysis, an aqueous calibration check and method blank analysis will be performed. (See Page 12, 13) Quality Control/Quality Assurance, for further discussion of calibration checks and method blanks.) Once it has been established that both the calibration check and method blank meets the acceptance criteria, field samples are analyzed. Following soil collection, 5 grams of soil are transferred into a separate 40-ml vial. The vial is then filled to the half-full mark with deionized water (approx. 20 mLs.). The vial is then shaken vigorously dissolving the soil into the DI water. The vial is then placed into a temperature-controlled oven at 90°C for approximately 15 minutes. Once heated, the vial is removed from the oven, then 0.20 cc of headspace from within the vial is collected with a disposable syringe, and the sample is immediately injected into the GC for analysis. After the analysis of 10 field samples, the method blank analysis is repeated. The calibration check is repeated at the end of each day.

Once the calibration and/or response factor (defined previously) and retention time for each individual compound has been entered into the ID table of the GC, the GC's Data Integrator identifies each compound and reports the concentrations in g/L. The concentration of the compounds within the soil is then converted to g/Kg by multiplying the concentration reported in g/L by the total volume of headspace (21 mL, actual) within the vial divided by the weight of the soil collected and analyzed (5 grams).

If a sample exhibits a high concentration that will potentially saturate the detector, dilution may be necessary. Using the "back-up" vial collected at the sampling location, dilution will be performed by transferring less than 5 grams of soil into a 40-mL vial and adding deionized water until the total volume reaches the half-full mark on the vial. The diluted sample is then analyzed as previously described. As an alternative to preparing a dilution, a smaller volume of headspace could be injected into the GC. In some cases, less soil may be used to increase dilution ranges.

Groundwater AnalysisAt the beginning of each day of analysis, an aqueous calibration check and method blank analysis will be performed. (See Page 12, 13) Quality Control/Quality Assurance, for further discussion of calibration checks and method blanks.) Once it has been established that both the calibration check and method blank meets the acceptance criteria, field samples are analyzed. Following groundwater collection, 20 ml of groundwater is transferred into a separate 40-ml vial. The vial is then placed into a temperature-controlled oven at 90 C for approximately 15 minutes. Once heated, the vial is removed from the oven, 0.20 cc of headspace from within the vial is collected with a disposable syringe, and the sample is immediately injected into the GC for analysis. After the analysis of 10 field samples, the method blank analysis is repeated. The calibration check is repeated at the end of each day. Once the calibration and/or response factor (defined previously) and retention time for each individual compound has been entered into the ID table of the GC, the GC's Data Integrator identifies each compound and reports the concentrations directly. The concentration of an analyte in the sample can also be calculated manually using the following equation: Where: As = Peak area for the analyte in the sample D = Dilution factor (dimensionless). If no dilution was necessary, D=1. CF = Mean calibration factor from the initial aqueous calibration (area per g/L) If a sample exhibits a concentration that may potentially saturate the detector, dilution will be necessary. Using the "back-up" vial collected at the sampling location, dilution will be performed by transferring less than 20 ml of groundwater into a 40-mL vial and adding deionized water until a total volume of 20 ml of liquid is in the vial. For example, a dilution factor of 2 would require 10 ml of groundwater sample and 10 ml of deionized water. Likewise, a dilution factor of 5 would require 4 ml of groundwater sample and 16 ml of deionized water. The diluted sample is then analyzed as previously described. As an alternative to preparing a dilution, a smaller volume of headspace could be injected into the GC, or these procedures may be used in combination. QUALITY CONTROL / QUALITY ASSURANCE

Calibration Check Calibration checks are performed to ensure that the relationship between instrument response and initial calibration is valid over the duration of the field effort. Calibration check samples will be analyzed at the beginning and end of each day of analysis. For the gaseous and aqueous calibration checks, a mid-point standard will be utilized. Aqueous standards are prepared on-site. The calibration checks regarding both aqueous and gas standards for each target analyte must fall within 25% of the true value. If this criterion is not met, then a second calibration check standard will be analyzed. If the second standard also falls outside of acceptance criteria, then the standards may be used to re-calibrate prior to further sample analysis. If field screened samples exhibit analyte concentrations greatly exceeding initial calibration ranges, calibration checks may include higher concentrations that reflect these levels.

Blank SamplesMethod Blanks Method blanks will be run at the beginning of each day and following every 10th sample to ensure a clean system. For soil gas samples, the method blank consists of ambient air collected in a disposable syringe. For groundwater samples, the method blank consists of reagent water placed in a 40-mL vial and heated in the temperature-controlled oven for approximately 15 minutes at 90°C. Once heated, 0.20cc of headspace from the vial is injected into the GC. If a detection is noted within a method blank, then the blank analysis will be repeated. If the associated field samples have similar detections, these samples will be reanalyzed if sufficient back-up volume was able to be collected. Rinsate Blanks Prior to the commencement of, or during sampling, rinsate blanks will be collected from syringes, soil/groundwater sampling tools, stirring rods, and/or sample vials to ensure that no cross-contamination has occurred. If compound detection is noted within a rinsate blank, then the equipment will be re-cleaned. If the associated field samples have similar detections, these samples will be reanalyzed if sufficient back-up volume was able to be collected.

Duplicate SamplesDuplicate samples are collected and analyzed to demonstrate the precision of the collection and analysis process. Duplicate samples are collected and analyzed at a frequency of 10% of the total samples collected. Since duplicate sample results are dependent on both collection and analysis, the optimal deviation for duplicate samples is a relative percent difference (RPD) of less than or equal to 25%. The RPD is calculated as follows: Where: C1 = Measured Concentration of the first sample C2 = Measured Concentration of the duplicate sample RPD = Relative Percent Difference If the RPD between duplicate samples dramatically exceeds 25% and sufficient "back-up" sample volume was able to be collected, the sample may be reanalyzed.