INVESTIGATION OF THE TOTAL ORGANIC HALOGEN ANALYTICAL METHOD AT THE WASTE SAMPLING CHARACTERIZATION FACILITY (WSCF)

JG, DOUGLAS ; MEZNARICH HD, PHD ; JR, OLSEN ; GA, ROSS ; M, STAUFFER

Hanford Site (Wash.)

关键词: Sample Preparation;    Organic Halogen Compounds;    Radioactive Waste Facilities;    Ground Water;    Errors;   

DOI  :  10.2172/940020
RP-ID  :  HNF-39194 Rev 0
RP-ID  :  DE-AC06-96RL13200
RP-ID  :  940020

美国|英语

来源: UNT Digital Library

【 摘 要 】

Total organic halogen (TOX) is used as a parameter to screen groundwater samples at the Hanford Site. Trending is done for each groundwater well, and changes in TOX and other screening parameters can lead to costly changes in the monitoring protocol. The Waste Sampling and Characterization Facility (WSCF) analyzes groundwater samples for TOX using the United States Environmental Protection Agency (EPA) SW-846 method 9020B (EPA 1996a). Samples from the Soil and Groundwater Remediation Project (S&GRP) are submitted to the WSCF for analysis without information regarding the source of the sample; each sample is in essence a 'blind' sample to the laboratory. Feedback from the S&GRP indicated that some of the WSCF-generated TOX data from groundwater wells had a number of outlier values based on the historical trends (Anastos 2008a). Additionally, analysts at WSCF observed inconsistent TOX results among field sample replicates. Therefore, the WSCF lab performed an investigation of the TOX analysis to determine the cause of the outlier data points. Two causes were found that contributed to generating out-of-trend TOX data: (1) The presence of inorganic chloride in the groundwater samples: at inorganic chloride concentrations greater than about 10 parts per million (ppm), apparent TOX values increase with increasing chloride concentration. A parallel observation is the increase in apparent breakthrough of TOX from the first to the second activated-carbon adsorption tubes with increasing inorganic chloride concentration. (2) During the sample preparation step, excessive purging of the adsorption tubes with oxygen pressurization gas after sample loading may cause channeling in the activated-carbon bed. This channeling leads to poor removal of inorganic chloride during the subsequent wash step with aqueous potassium nitrate. The presence of this residual inorganic chloride then produces erroneously high TOX values. Changes in sample preparation were studied to more effectively remove inorganic chloride from the activated carbon adsorption tubes. With the TOX sample preparation equipment and TOX analyzers at WSCF, the nitrate wash recommended by EPA SW-846 method 9020B was found to be inadequate to remove inorganic chloride interference. Increasing the nitrate wash concentration from 10 grams per liter (g/L) to 100 g/L potassium nitrate and increasing the nitrate wash volume from 3 milliliters (mL) to 10 mL effectively removed the inorganic chloride up to at least 100 ppm chloride in the sample matrix. Excessive purging of the adsorption tubes during sample preparation was eliminated. These changes in sample preparation have been incorporated in the analytical procedure. The results using the revised sample preparation procedure show better agreement of TOX values both for replicate analyses of single samples and for the analysis of replicate samples acquired from the same groundwater well. Furthermore, less apparent column breakthrough now occurs with the revised procedure. One additional modification made to sample preparation was to discontinue the treatment of groundwater samples with sodium bisulfite. Sodium bisulfite is used to remove inorganic chlorine from the sample; inorganic chlorine is not expected to be a constituent in these groundwater samples. Several other factors were also investigated as possible sources of anomalous TOX results: (1) Instrument instability: examination of the history of results for TOX laboratory control samples and initial calibration verification standards indicate good long-term precision for the method and instrument. Determination of a method detection limit of 2.3 ppb in a deionized water matrix indicates the method and instrumentation have good stability and repeatability. (2) Non-linear instrument response: the instrument is shown to have good linear response from zero to 200 parts per billion (ppb) TOX. This concentration range encompasses the majority of samples received at WSCF for TOX analysis. (3) Improper sample preservation: ion-chromatographic analysis of several samples with anomalous TOX results revealed that the samples were properly preserved with sulfuric acid and not hydrochloric acid.