PCBs in Ambient Air – Method Evaluation and Background Monitoring – The Hudson River, NY Sediment Remediation Project

摘要

A pilot program was conducted in July-September 2005 to assist in the selection of sampling and analyses methods to be used for measurement of PCBs in ambient air during the Hudson River NY sediment remediation program. The field program evaluated use of EPA Method TO4A (high volume sampling) and EPA Method TO10A (low volume sampling) as well as aroclor pattern matching and congener specific analytical procedures. A six (6) week field monitoring program was also conducted to establish baseline concentrations for PCBs at a representative location riverside in the planned sediment dredging corridor. Mean concentrations for the two (2) types of sorbents used in low volume samplers (EPA Method 10A) were in good agreement (PUF alone mean =3.96 ng/m3 and PUF/XAD-2 mean = 4.30 ng/m3). As a result of these data it was concluded that routine sampling during remediation could take place with either PUF alone or with PUF/XAD-2. Results for high volume PUF samples (EPA Method TO4A) collocated with the low volume samples ranged from 3.8-5.74 ng/m3 with a mean concentration of 4.74 ng/m3. Reported concentrations for these same samples are in excellent agreement when comparing results for the two analytical methods examined. (Aroclor technique, mean = 4.74 ng/m3; congener method, mean = 4.72 ng/m3). These results were instrumental in the selection of aroclor or pattern matching as the method of choice for analysis of ambient air samples collected during the actual Hudson River Sediment Remedial Program. A classical aroclor 1242 pattern was observed in these samples suggesting that the PCBs may be associated with atmospheric particulate matter in the vicinity of the sampling location. The average background concentration at the Fort Edward site for all samples was 2.41 ng/m3 (n = 24) and ranged from 0.70 – 4.48 ng/m3. These concentrations are consistent with PCB levels found in rural settings. These results verified that both sampling methods were capable of meeting the project QOL sensitivity target of 30 ng/m3 on a routine basis.