38 



Matrix spike samples (MS) and matrix spike duplicates (MSD) are prepared by 

 dividing a sample into multiple aliquots, spiking an aliquot with a known concentration of 

 analyte, and proceeding with the analysis as though the spike was a sample. In 1997, 

 samples NLDS QCA and NLDS 10A 0.75-1.0 m were selected for matrix spike and matrix 

 spike duplicate analysis. In 1998, samples NL-14B and NL-15A were selected for matrix 

 spike and matrix spike duplicate analysis. 



Matrix spike recovery for metals should yield 75 to 125% recovery of the known 

 value, as stated in EPA Method 6010. In 1997, the laboratory reported matrix spike 

 recoveries for As, Cd, Cr, Cu, Pb, and Hg in NLDS QCA ranging between 97 and 122%, 

 within Method 6010 QC acceptance criteria. However, Ni and Zn had recoveries of 138% 

 and 203%, respectively, exceeding the criteria limits. Recovery of Zn in NLDS 10A was 

 156.6% of the matrix spike sample. Because the laboratory control sample recoveries for all 

 analytes were within control limits, the elevated spike recoveries were attributed to sample 

 non-homogeneity (typical for dredged material) or matrix effect. A laboratory QC sample 

 was prepared with each sample batch. The recovery for the QC sample for eight metals 

 analyzed ranged from 95 to 103%. A second QC sample for Zn had a recovery of 102%. 

 Considering all of the QC information provided for metals, the recoveries for all metals 

 indicated acceptable accuracy. 



In 1998, WHG reported matrix spike recoveries for As, Cd, Cr, Cu, Pb, Hg, Ni, and 

 Zn in NLDS NL-14B ranging between 87 and 120%, within Method 6010 QC acceptance 

 criteria. A laboratory QC sample (spiked blank/laboratory control spikes) was prepared with 

 each sample batch. The recovery for the QC sample for eight metals analyzed ranged from 

 71 to 97%. The recoveries for all metals indicated acceptable accuracy. 



For TOC, the laboratory standard operating procedure targets a range of 75-125% 

 recovery. In 1997, two samples were spiked for TOC (NLDS QCA and NLDS 10A 0.75- 

 1.0 m). Recoveries were 104% and 112%, indicating acceptable accuracy for TOC data. In 

 1998, sample NLDS 14B was spiked for TOC and had recovery of 84%, indicating 

 acceptable accuracy for TOC data. Prior to the analysis of the samples, the laboratory QC 

 sample was analyzed and 99% was recovered. After the samples were analyzed, the 

 laboratory QC sample had a recovery of 106%, within the range of acceptable recovery 

 percentages. 



Recoveries of PAH matrix spike and matrix spike duplicates must fall within the 

 range of 35-125% as stated in EPA SW-846 Method 8270 (USEPA 1997). Using NLDS 

 1 A for the 1997 MS/MSD analyses, recoveries ranged from 49 to 156%. Fluoranthene and 



Monitoring Cruise at the New London Disposal Site, Seawolf Mound 1995 - 1998 



