McClellan Report - Sampling Devices
|
McClellan Report - Sampling Devices |
|
|
Report Cover Table of Contents Sec. 1 Sec. 2 Sec. 3 Sec. 4 Sec. 5 Sec. 6 Sec. 7 Sec. 8 SECTION 3 FIELD ACTIVITIES AND LABORATORY ANALYTICAL APPROACH 3.1 FIELD ACTIVITIES A total of 251 primary samples and 34 quality assurance/quality control (QA/QC) samples were collected from 20 wells at McClellan as part of this demonstration. Details of the field activities are discussed below. 3.1.1 SAMPLING STRATEGY Concurrent deployment of multiple types of samplers at the same depth in each well is desirable to obtain comparative data. However, the 4-inch well diameter imposed a physical limitation on the number of samplers that could be concurrently deployed at the same depth in each well. Therefore, sampling occurred in three phases as described below.
Table 3.1 is a summary of the types of sampling techniques that were used in each well. TABLE 3.1 3.1.2 FIELD MEASUREMENTS The depth to water was measured in each well prior to deployment during Phase 1, prior to retrievals during Phase 2, and prior to conventional sampling during Phase 3. Additionally, the total well depth was measured prior to deployment during Phase 1. Target sampler deployment depths were calculated after measuring the depth to water and the total well depth at the beginning of Phase 1, taking into consideration the reported screened interval of the well. Of the three sampling depths monitored per well, the intermediate interval was generally defined as the center of the saturated screened interval, the shallow interval was generally defined as being approximately 1 foot below the top of the saturated screened interval, and the deep interval was generally defined as being approximately 1 foot above the bottom of the open (i.e., non-buried) saturated screened interval. Table 3.2 is a summary of the depth to water measurements, the total depth measurements, the screened interval depths, and the sampling intervals for each well. TABLE 3.2 Measurements of traditional well stabilization parameters were made during conventional sampling. These parameters included groundwater temperature, pH, conductivity, DO, ORP, and turbidity. These measurements along with the total volume purged, the time spent purging, and the average pump rate for each well are summarized in Table 3.3. A maximum of three different types of diffusion samplers and one type of grab sampler were deployed in each well. The distribution of diffusion and grab samplers in each well was designed to facilitate inter-sampler comparisons while maintaining an overall deployment of RPPS in 20 wells; RCS, PsMS, HydraSleeve®, and Snap Samplers in 10 wells each; and PDBS in only those wells that were targeted for VOC analysis. Table 3.4 is a summary of the sample dates, deployment lengths, and time lags between all sampling events. 3.2 LABORATORY ANALYTICAL APPROACH 3.2.1 TARGET COMPOUNDS The following compounds were targeted for analysis in the priority listed below during the technology demonstration.
With the exception of VOCs, these compounds were targeted because they are not able to be monitored using the PDBS method, but are contaminants of concern at some DoD installations. VOCs were included in the target compound list to verify that all no-purge sampling devices also would be capable of accurately monitoring for these compounds. The final measurements of turbidity made during both types of conventional sampling were used to determine whether or not the samples should be field-filtered for TAL metals analysis using a 0.45-micron disposable filter. If the final turbidity measurement made immediately before sample collection was less than or equal to 5 Nephelometric Turbidity Units (NTUs), the samples were not filtered in the field and were submitted for total metals analysis. If the final turbidity measurement was greater than 5 NTUs, the samples were filtered according to procedures described in SOP #6 of the Work Plan (Parsons, 2004a), and were scheduled for dissolved metals analysis. All conventionally sampled wells that were analyzed for metals were field-filtered with the exception of well TABLE 3.3 TABLE 3.4 MW-400 where the measured turbidity was less than 5 NTUs. Additionally, all metals samples collected using the HydraSleeve® were field-filtered. Samples for hexavalent chromium analysis were not field-filtered. 3.2.2 LABORATORIES Two analytical laboratories were used during this demonstration to perform all of the required analyses. Columbia Analytical Services, Inc. (CAS) in Kelso, Washington performed the metals and 1,4 dioxane analyses. Sequoia Analytical (Sequoia), based in Sacramento, California performed the hexavalent chromium, anion, and VOC analyses. Sequoia used two different facilities to perform the requested analyses; hexavalent chromium and anions were analyzed in their Morgan Hill, California facility while VOCs were analyzed in their Petaluma, California facility. The maximum holding time permitted for hexavalent chromium is 24 hours. Therefore, samples were sent twice per day (once at approximately noon, and again at approximately 5 pm) to Sequoia using a hand-delivery courier. Samples were shipped daily each afternoon to CAS via overnight express courier. 3.2.3 SAMPLE VOLUME As described in the Work Plan (Parsons, 2004a), the diffusion and grab samplers do not collect large volumes of groundwater (relative to conventional sampling methods), and the available sample volume does not always fulfill normal laboratory and/or analytical method recommendations. This characteristic is not necessarily a critical limitation since most analytical methods do not actually require the larger sample volumes recommended in standard analytical procedures. An ITRC Diffusion Sampler subteam has estimated the minimum sample volumes required for common environmental analytical methods; details are available on the ITRC diffusion sampling website at http://64.203.146.40/news.asp#41. Prior coordination with the analytical laboratories enabled use of smaller sample volumes to perform the required analytical methods while still maintaining required detection limits. Table 3.5 is a summary of the approximate maximum volume capacities of each type of no-purge sampling device used in this study per sample depth (some sampling devices required more than one sampler per depth interval). The volumes listed in Table 3.5 are the maximum obtainable with the configuration used at McClellan; larger volumes can potentially be obtained in some cases by reconfiguring the samplers (e.g., using more PsMS canisters). It should be noted that a larger-volume Snap Sampler™ and HydraSleeve® are now available. Table 3.6 summarizes the minimum sample volume requirements (per analysis) specified by the analytical laboratories. Groundwater samples from each well were analyzed for only a subset of the target analyte list. The minimum sample volumes shown in Table 3.6 were used for diffusion, grab, and low-flow samples to maintain consistency and to facilitate comparison of the results. However, in order to maintain consistency between the three-volume purge method historically used for these wells as part of LTM and the conventional samples collected as part of this demonstration, normal sample volumes specified in the McClellan QAPP (URS, 2003) were collected for the three-volume purge method. TABLE 3.5 TABLE 3.6 One or more additional sets of sample bottles were filled and submitted to the analytical laboratory along with the primary sample whenever sufficient sample volume was available. This practice allowed the laboratory to reanalyze samples as necessary due to the need for sample dilution or other circumstances. 3.3 DEVIATIONS FROM WORK PLAN The field activities generally occurred in accordance with the Work Plan (Parsons, 2004a). However, the following notable deviations occurred during this evaluation.
3.4 QA/QC SAMPLE COLLECTION A total of 34 samples were collected for QA/QC purposes. The number and type of each of these samples is summarized in Table 3.7. Generally QA/QC sample collection followed the schedule described in the Work Plan (Parsons, 2004a). However, some variances did occur as described below. Sequoia did not provide trip blank samples as part of the Phase 2 bottle order. However, one trip blank sample was provided via courier by Sequoia on June 9, 2004. This was the only trip blank sample collected during the Phase 2 activities. This sample was sent to Sequoia along with the daily shipment of VOC samples on June 9, 2004. However, Sequoia did not analyze this sample. No explanation was available from Sequoia as to why this sample was not analyzed. Trip blank samples were provided by Sequoia for the Phase 3 activities, and one of these samples was shipped along with each TABLE 3.7 a/ Although four samples were collected with the intention of being used as field duplicates, a fifth field duplicate sample was available for the analyses performed by Sequoia (see Note b/ below). b/ These samples were designated for MS/MSD analyses on the chains of custody. However, Sequoia treated them as primary samples and did not spike them. They therefore are considered field duplicate samples for analyses performed by Sequoia only. Although no other samples were designated by the field scientists as MS/MSD samples, both Sequoia and CAS chose other samples at random upon which to perform MS/MSD analyses (see Appendix A). c/ Source water blank was comprised of the water used to fill the diffusion samplers prior to deployment. d/ NA = not applicable. e/ Purified water blank was comprised of the water used for decontamination. cooler containing samples intended for VOC analysis. As a result of the lack of trip blanks during Phase 2, the degree to which low-level VOC detections may be attributable to cross-contamination during sample shipping and handling cannot be fully confirmed. Two of the samples collected with the intent of being used by the laboratories as matrix spike/matrix spike duplicate (MS/MSD) samples were not treated as MS/MSD samples by Sequoia although they were by CAS. Instead, Sequoia analyzed these samples as primary samples. They are therefore considered duplicate samples for QA/QC purposes. These samples were MW173-3VOL-MS/MSD and MW225-MICROMS/ MSD. Despite this oversight, other samples were selected at random by Sequoia for MS/MSD analysis (see Appendix A). In the instances where field samples designated as MS/MSDs were not analyzed as such, measurements of accuracy and analytical precision based on MS/MSD results were not developed for samples collected using a given sampling method. In the Work Plan (Parsons, 2004a), two field duplicates and two MS/MSD samples were scheduled for collection with the HydraSleeve®. However, due to an oversight, no field duplicates or MS/MSD samples for this sampler type were collected. Therefore, information regarding precision of the HydraSleeve® sampling process based on MS/MSD results and the impact of potential matrix effects on the analytical testing is not available. A total of four field duplicate samples were collected for both the low-flow and threevolume purge sampling methods while only two were scheduled according to the Work Plan (Parsons, 2004a). Although only one equipment rinseate was scheduled for the three-volume purge method (Parsons, 2004a), two were actually collected; one from the bailer only, and another from both the bailer and the in-line filter. Report Cover Table of Contents Sec. 1 Sec. 2 Sec. 3 Sec. 4 Sec. 5 Sec. 6 Sec. 7 Sec. 8 Copyright © 2008 • GeoInsight • 800-996-2225 |
