ML16141B276
| ML16141B276 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 05/19/2016 |
| From: | - No Known Affiliation |
| To: | Division of License Renewal |
| References | |
| Download: ML16141B276 (56) | |
Text
IPRenewal NPEmails From: Gray, Dara F <DGray@entergy.com>
Sent: Thursday, May 19, 2016 8:41 AM To: Wentzel, Michael
Subject:
[External_Sender] FW: Year Class Reports Attachments: 24100 Hudson River 2013 QC.pdf; 24200 Hudson River 2014 QC FINAL.pdf Importance: High Mike Here you go, and sorry again for the delay.
Dara Gray REM Chemistry/Environmental Indian Point Energy Center Dgray@entergy.com 914-254-8414 From: Wentzel, Michael [1]
Sent: Thursday, April 21, 2016 7:00 AM To: Gray, Dara F Cc: Louie, Richard
Subject:
Year Class Reports
- Dara, Thank you (and Rich) for sending the 2013 and 2014 Year Class Reports. We are having difficulty getting the files added to ADAMS because the files are password protected. Would it be possible to either get the password to unlock the files, or, unlocked versions of the files?
Please feel free to give me a call if you would like to discuss this further.
- Thanks, Mike Michael Wentzel Project Manager NRR/DLR/RERP (301) 415-6459 michael.wentzel@nrc.gov Please consider the environment before printing this e-mail.
1
Hearing Identifier: IndianPointUnits2and3NonPublic_EX Email Number: 6399 Mail Envelope Properties (DA94DFACF1201C4A91A21BD336C2520A23596363)
Subject:
[External_Sender] FW: Year Class Reports Sent Date: 5/19/2016 8:40:35 AM Received Date: 5/19/2016 8:40:50 AM From: Gray, Dara F Created By: DGray@entergy.com Recipients:
"Wentzel, Michael" <Michael.Wentzel@nrc.gov>
Tracking Status: None Post Office: LITXMETSP003.etrsouth.corp.entergy.com Files Size Date & Time MESSAGE 984 5/19/2016 8:40:50 AM 24100 Hudson River 2013 QC.pdf 394610 24200 Hudson River 2014 QC FINAL.pdf 434978 Options Priority: High Return Notification: No Reply Requested: No Sensitivity: Normal Expiration Date:
Recipients Received:
Quality Control Report for the 2013 Hudson River Ichthyoplankton Laboratory Program and 2013 Fall Juvenile Survey Prepared for:
Indian Point Energy Center 450 Broadway, Suite 1 Buchanan, New York 10511 Submitted:
July 2014 Prepared by:
Normandeau Associates, Inc.
25 Nashua Road Bedford, NH 03110 www.normandeau.com
2013 HUDSON RIVER QC REPORT Table of Contents Page
1.0 INTRODUCTION
................................................................................ 1 2.0 METHODS....................................................................................... 2 2.1 LABORATORY QUALITY CONTROL PROCEDURES ............................................. 2 2.1.1 Ichthyoplankton Survey........................................................ 2 2.1.2 Fall Juvenile Survey............................................................ 3 2.2 REPORTING PROCEDURES ................................................................... 4 2.2.1 Fraction Inspected ............................................................. 4 2.2.2 Percent Nonconforming ....................................................... 6 2.2.3 Percent Measurement Error................................................... 6 Sorting Task..................................................................... 6 Identification Task............................................................. 7 2.2.4 Average Outgoing Quality ..................................................... 8 2.2.5 Cumulative Error Rates ........................................................ 9 3.0 RESULTS .......................................................................................10 3.1 ICHTHYOPLANKTON LABORATORY PROGRAM ............................................... 10 3.2 FALL JUVENILE SURVEY .................................................................... 17 4.0 LITERATURE CITED ..........................................................................23 24100 Hudson River 2013 QC.docx 7/14/14 ii Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT List of Figures Page Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks. ...................................................................... 3 Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes. ........................................... 5 Figure 3. Example of percent measurement error calculations for individual taxa during the identification task. .......................................................... 7 24100 Hudson River 2013 QC.docx 7/14/14 iii Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT List of Tables Page Table 1. Task specific applications of continuous sampling plans for the 2013 Hudson River Ichthyoplankton Laboratory Program ................................. 2 Table 2. Task specific applications of continuous sampling plans for the 2013 Fall Juvenile Survey............................................................................ 4 Table 3. Fraction inspected, percent nonconforming, mean percent measurement error, and average outgoing quality of tasks performed by Normandeau for the 2013 Hudson River Ichthyoplankton Laboratory Program.................................................................................... 10 Table 4. Sample sorting fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program. ................................................ 11 Table 5. Sample sorting percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program. ................................................ 12 Table 6. Sample sorting mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program. ......................................... 13 Table 7. Sample identification fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program. ................................................ 14 Table 8. Sample identification percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program. ......................................... 15 Table 9. Sample identification mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program. ............................... 16 Table 10. Ranking of taxa missed during initial sort and found during sort QC. ........... 18 Table 11. Summary by life stage of the six highest ranked taxa missed during original sort and found during sort QC compared to total count. ................ 19 Table 12. Cumulative net and absolute error rates for commonly encountered taxa in samples selected for QC inspection of identification and counting process. ........................................................................ 20 Table 13. Fraction inspected, percent nonconforming, and average outgoing quality of laboratory tasks performed by Normandeau for the 2013 Fall Juvenile Survey........................................................................... 22 24100 Hudson River 2013 QC.docx 7/14/14 iv Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 1.0 Introduction This quality control report for the laboratory tasks of the 2013 Hudson River Ichthyoplankton Survey and the 2013 Fall Juvenile Survey was prepared for Entergys Indian Point Energy Center by Normandeau Associates Inc. (Normandeau).
To comply with Entergys requirements for valid and reliable data on the Hudson River Ichthyoplankton Laboratory Program and the Fall Juvenile Survey, Normandeau implemented a Quality Assurance Plan that provides a 10% Average Outgoing Quality Limit (AOQL) for all measurement parameters collected. The Quality Assurance Plan consists of two systems: a quality control (QC) system and a quality assurance (QA) system.
The QC system is managed by the program manager and conducted by operational personnel. The system monitors and documents the reliability and validity (accuracy, precision, completeness) of daily operations. The specific features of the QC system are determined by the Quality Assurance Department to insure that all procedures conform to Entergys data requirements. The QA system is managed by Normandeaus Quality Assurance Director and utilizes project independent personnel familiar with the work or activities under evaluation to conduct performance and systems audits. These audits are designed to provide objective evidence that the quality control program and technical requirements, methods, and procedures as outlined in the program Standard Operating Procedures are being implemented. The outcomes of the QA system activities are verification of the effectiveness of the QC system, assignment of corrective actions to resolve nonconforming procedures or data deficiencies, communication of audit results to project and staff managers for follow-up, and objective validation or improvement of project operations.
This report provides a compilation of QC system data verifying the results of the 2013 Hudson River Ichthyoplankton Laboratory Program and 2013 Fall Juvenile Survey activities. Determinations of the fraction inspected, percent nonconforming, and average outgoing quality are presented for both programs. In addition, for the 2013 Hudson River Ichthyoplankton Laboratory Program the results include percent measurement error, a summary of the number of each taxon-life stage found during sorting QC, and cumulative error rates for each taxon-life stage.
24100 Hudson River 2013 QC.docx 7/14/14 1 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 2.0 Methods 2.1 Laboratory Quality Control Procedures 2.1.1 Ichthyoplankton Survey For sorting and identification of samples from the 2013 Hudson River Ichthyoplankton Laboratory Program, Normandeau used a continuous sampling plan designed to provide a 10% AOQL (U.S. Department of Defense 1981). A flow diagram of how the sampling plan was applied is presented in Figure 1. A summary of the sampling plan, tolerances, and QC sample definitions used for each measurement parameter is presented in Table 1. Quality control inspection was applied on a laboratory-wide basis for the sorting task and to each individual processor for the identification task. Quality control samples were selected in a random manner utilizing random number tables. As determined from the sampling plan outlined in Table 1, a specified number of quality control samples were reprocessed by QC inspectors with expertise in the task being inspected. In cases where a sample was subdivided and counted, counts for all subdivisions were combined before calculating percent error for that sample. If the difference between the quality control value and the original value exceeded acceptable tolerances (Table 1), a third measurement could be obtained to verify one of the measurements. If a sample was found to have exceeded acceptable tolerances, all subsequent samples processed by the same technician were subjected to 100% quality control until an appropriate number of consecutive samples (i) were found within tolerance as determined by the continuous sampling plan (Table 1 and Figure 1). The standard operating procedures manual (Normandeau 2014) documents specific QA/QC methods utilized for this program.
Table 1. Task specific applications of continuous sampling plans for the 2013 Hudson River Ichthyoplankton Laboratory Program CSP-1 Laboratory AOQL 10% QC Sample Task i f Sample Tolerance Definition Sorting 8 1/7 +/- 2 if d20 organisms one sample
+/- 10% if >20 organisms Identification 8 1/7 +/- 2 if d20 one sample
+/- 10% if >20 for every taxon in the sample (in identifying, assigning a life stage, or counting any species, errors are cumulative by life stage within each taxon) 24100 Hudson River 2013 QC.docx 7/14/14 2 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks.
In some cases one of the taxonomists (either the original identifier or the QC inspector) was able to determine the taxon or life stage of damaged specimens when the other taxonomist recorded them as unknown life stage, unidentified taxon, or a higher level taxon (genus or family). If a more general taxon or life stage used by one taxonomist included the more specific category used by the other taxonomist, and that was the only reason for a count discrepancy, then that sample was not considered as failing the QC inspection. For example, damaged specimens recorded as Morone sp. by the original identifier and as Striped Bass by the QC inspector were considered to be in agreement because the category Morone sp. includes Striped Bass. In contrast, an original determination of unidentified gobiid would not be acceptable if the QC determination was Striped Bass, because Striped Bass is not included in the family Gobiidae.
2.1.2 Fall Juvenile Survey The Fall Juvenile Survey consisted of two types of collections, referred to as the Fall Shoals Survey (which used Tucker trawls) and the Beach Seine Survey. For laboratory identification and length measurements of young-of-the-year fishes in the 2013 Fall Juvenile Survey, Normandeau used a continuous sampling plan designed to provide a 10% Average Outgoing Quality Limit (U.S. Department of Defense, 1981). A flow diagram of how the 24100 Hudson River 2013 QC.docx 7/14/14 3 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT plan was applied is presented in Figure 2. A summary of the sampling plan, tolerances, and QC sample definitions used for each task is shown in Table 2. QC samples were selected as specified by the appropriate plan in Table 2, using random numbers, and reprocessed by QC inspectors. If the difference between original and QC values exceeded the acceptable tolerance, a third value was obtained as a resolution. The standard operating procedures manual (Normandeau 2013) documents specific QA/QC methods used for the 2013 Fall Juvenile Survey. Young-of-the-year fishes were identified in the laboratory for the first two Fall Shoals Survey river runs (sampling weeks) and the first three Beach Seine Survey river runs. Young-of-the-year fishes were identified in the field starting with Fall Shoals Survey river run 3 and Beach Seine Survey river run 4. The same quality control procedures applied to both field and laboratory identifications. All length measurements of young-of-the-year fishes occurred in the laboratory.
Table 2. Task specific applications of continuous sampling plans for the 2013 Fall Juvenile Survey.
QC QC Sample Task Plan AOQL i f x Tolerance Definition Identification CSP-V 7% 21 1/15 7 +/-10% of total count or +/-2 One taxon individuals when <25 fish
+/-1 mm when <34 mm TL Length CSP-V 7% 30 1/50 10 One fish
+/-3% when >34 mm TL 2.2 Reporting Procedures The 2013 Hudson River Ichthyoplankton Laboratory Program Sort and Identification Quality Control Logs were keyed, verified, and error-checked to produce SAS data sets.
From these data, fraction inspected, percent nonconforming, and percent measurement error (precision) were determined for each river run and for the entire study. For the 2013 Fall Juvenile Survey, QC data were used to determine fraction inspected and percent nonconforming for the entire study (combining all river runs processed in the laboratory for both the Fall Shoals Survey and the Beach Seine Survey).
2.2.1 Fraction Inspected Fraction Inspected Number of Samples Inspected x 100 (Equation 1)
Total Number of Samples River Run: Fraction inspected for a river run (Equation 1) was one hundred times the number of samples inspected divided by the total number of samples analyzed for that river run. For the ichthyoplankton sorting task, the number of samples inspected excludes training QC samples for new sorters, which do not represent the independent performance of the technician, as well as the samples inspected as part of the QC plan. For the identification task, the total number of samples identified excludes empty (no catch) samples, which did not require processing by an identifier.
24100 Hudson River 2013 QC.docx 7/14/14 4 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes.
24100 Hudson River 2013 QC.docx 7/14/14 5 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Entire Study: Fraction inspected for the entire study was one hundred times the number of samples inspected divided by the total number of samples analyzed during the study.
2.2.2 Percent Nonconforming Percent Nonconforming Number of Nonconforming Samples Inspected x 100 (Equation 2)
Number of Samples Inspected River Run: Percent nonconforming for a river run (Equation 2) was one hundred times the number of nonconforming quality control samples found for that river run divided by the total number of quality control samples inspected for that river run.
Entire Study: Percent nonconforming for the entire study was one hundred times the total number of nonconforming quality control samples for the study divided by the total number of quality control samples inspected for the study. The result of this analysis was a determination of the actual incoming quality level of each measurement parameter. (Note that because samples checked by QC found to be defective were rectified during QC, the average outgoing quality of the final data set was better than that indicated by the percent nonconforming.)
2.2.3 Percent Measurement Error Sorting Task Sorting Percent Measurement Error Quality Control Value x 100 (Equation 3)
(Original Value Quality Control Value)
Sample: Percent measurement error for a sorted sample (Equation 3) was one hundred times the quality control value divided by the sum of the original value and the quality control value. If the total count (original value plus quality control value) was less than or equal to 20, and the quality control value (i.e., the number of organisms missed by the sorter and found during sort QC inspection) was no more than two, the percent measurement error for the sorted sample was defined as zero.
River Run: Mean percent measurement error for sorted samples for a river run was the sum of the percent measurement errors for each sample inspected during the river run divided by the total number of samples inspected for the river run.
Entire Study: Mean percent measurement error for sorted samples for the entire study was the sum of the percent measurement errors for each sample inspected during the study divided by the total number of samples inspected for the study. (Note that this method of averaging gives equal weight to each sample, regardless of the number of organisms present).
24100 Hudson River 2013 QC.docx 7/14/14 6 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Identification Task Life Stage Percent Measurement Error (Original Value Quality Control Value) x 100 (Equation 4)
Quality Control Value Life Stage: Percent measurement error for a life stage (Equation 4) was one hundred times the difference between the original value and the quality control value divided by the quality control value. For life stages where the quality control value was 20 or less, if the original and quality control values differed by no more than two organisms, the percent measurement error was defined as zero. For life stages where the quality control value was 20 or less and the original and quality control values differed by more than two organisms, the percent measurement error was calculated utilizing Equation 4. If the quality control value was zero, the percent measurement error was calculated by multiplying the difference between the original and quality control values by 100. This can occasionally result in extremely large percent measurement error values (as much as several hundred percent for a life stage of a taxon in a sample), which are not truly indicative of the actual proportion of specimens misidentified, assigned to the wrong life stage, or miscounted in a sample. If the original count for a life stage was acceptably close to a resolution value but not to the quality control value, the percent measurement error was calculated as described above except that the resolution value was substituted for the quality control value.
Taxon: Percent measurement error for an identified taxon was the sum of the absolute values of percent measurement error for each life stage within the taxon. Refer to Figure 3 for an example of taxon percent measurement error calculations.
Post Yolk-Sac Young-of-Eggs Larvae the-Year Total Taxon 1 Original Value 103 176 25 Quality Control Value 100 194 26
% Measurement Error Life Stage 3.0 -9.3 -3.8 16.1 Taxon 2 Original Value 2 Quality Control Value 1
% Measurement Error Life Stage 0 0 Taxon 3 Original Value 8 Quality Control Value 2
% Measurement Error Life Stage 300 300 Figure 3. Example of percent measurement error calculations for individual taxa during the identification task.
24100 Hudson River 2013 QC.docx 7/14/14 7 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT River Run: Mean percent measurement error for the identification task for a river run was the sum of the percent measurement errors for all taxa inspected during the river run divided by the total number of taxa inspected for the river run. This statistic was computed by averaging taxa rather than samples because even though complete samples were inspected and reworked for identification quality control, the pass/fail criterion was whether any taxon in the sample individually exceeded the 10% tolerance.
Entire Study: Mean percent measurement error for identified taxa for the entire study was the sum of the percent measurement errors for all taxa inspected during the study divided by the total number of taxa inspected for the study.
2.2.4 Average Outgoing Quality At the completion of these studies, the Average Outgoing Quality (AOQ) was calculated for each measurement parameter inspected. Continuous sampling plans were used for all tasks. Continuous sampling plans are devised for processes involving a continuous or nearly continuous flow of products or other entities. For these types of processes, it is extremely difficult to organize units into discrete groups commonly referred to as lots. As a result, inspection must be performed on individual units drawn from a continuous flow of products and a decision made concerning the quality of units produced based on the inspection results. Rectification is performed on any nonconforming unit found during inspection, followed by 100% screening of a number of subsequent units depending on the sampling plan. Average Outgoing Quality for each laboratory task was calculated as a function of the percent nonconforming and the fraction of total units inspected (Stephens 1979). This calculation applies to continuous sampling plans when nonconforming units found are rectified:
p' (1 f )q i AOQ x 100 (Equation 5) f (1 f )q i where p' = Percent nonconforming as a decimal fraction f = Fraction of units inspected. This is a parameter of the sampling plan.
Q = 1-p' = Percent conforming as a decimal fraction I = Clearing interval. This is a parameter of the sampling plan.
Example:
p' = 0.0689 f = 1/7 = 0.1429 q = 1-0.0689 = 0.9311 i = 8 0.0689 (1 0.1429)(0.9311) 8 AOQ x 100 5.32%
0.1429 (1 0.1429)(0.9311) 8 The above equation for calculating AOQ was formulated specifically for CSP-1 sampling plans such as those used for the ichthyoplankton sorting and identification (Table 1). The 24100 Hudson River 2013 QC.docx 7/14/14 8 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT same equation was used to calculate AOQ for young-of-the-year identifications and measurements, which used CSP-V plans (Table 2). When Equation 5 is used for CSP-V plans, the calculated AOQ is conservatively high, because the equation does not take into account the times when the number of consecutive reinspections following a failure is x (which is smaller than i).
2.2.5 Cumulative Error Rates Due to the non-independence of identification errors across taxa and life stages, and to the accumulation of errors within taxa, a relatively high fraction of samples may fail QC inspection even though only a small fraction of organisms are incorrectly identified or counted. In order to present the error frequencies more realistically for particular taxa-life stages, two additional statistics were calculated for each taxon-life stage for the identification/counting process.
Absolute Error Rate =
n n i 1 lIi Qi l / Q i 1 i Equation 6 Net Error Rate =
n n (I
i 1 i Qi ) / Q i 1 i Equation 7 where Ii = initial count for taxon-life stage in sample i Qi = QC count for taxon-life stage in sample i (or the resolution count, if Ii was acceptably close to it but not to the QC count) n = number of samples in the entire study If the sum of Qi for the entire study was zero for the taxon-life stage, then the sum of Qi was set equal to one for the purpose of calculating absolute and net error rate.
The absolute error rate is the approximate fraction of the taxon-life stage that was originally identified or counted incorrectly. This is an estimate of the fraction of erroneous countable items in the uninspected samples.
Net error rate is the approximate relative error in the total counts for the taxon-life stage.
For this index, positive (original count too high) and negative (original count too low) errors cancel each other so that the index reflects the relative net bias to the taxon-life stage abundance.
24100 Hudson River 2013 QC.docx 7/14/14 9 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 3.0 Results 3.1 Ichthyoplankton Laboratory Program The Average Outgoing Quality (AOQ) of the 2013 Hudson River Ichthyoplankton Laboratory Program was 6.41% for the sorting task and 0.52% for the identification task.
These AOQ levels represent the actual or achieved quality for measurement parameters and were within the 10% AOQL requirement of the study. The Average Fraction Inspected (AFI) was 35.81% for sorting and 15.34% for identification (Table 3).
Table 3. Fraction inspected, percent nonconforming, mean percent measurement error, and average outgoing quality of tasks performed by Normandeau for the 2013 Hudson River Ichthyoplankton Laboratory Program.
Mean Percent Fraction Percent Noncon- Measurement Error Task Inspected (%) forming (%) (%) AOQ (%)
Sorting 35.81 8.60 2.97 6.41 Identification 15.34 0.61 0.92 0.52 Sorting and identification tasks were also evaluated on the basis of river runs (sampling weeks). Sorted samples were inspected at a rate of 14.81% to 62.70% for individual river runs (Table 4). Nonconformance for the sorting task among the inspected samples ranged from 0% to 18.33% in the 23 river runs, and averaged 8.60% overall (Table 5). Sorting measurement error was between 0% and 6.10% and averaged 2.97% for the study (Table 6).
For the task of sample identification, 10.0% to 39.51% of samples were inspected from individual river runs (Table 7). Percent nonconforming for the identification task in individual river runs ranged from 0% to 11.11% (Table 8). Measurement error for individual river runs ranged from 0% to 6.25% and overall measurement error was 0.92%
for the identification task of this study (Table 9).
Measurement error results for the identification task are skewed towards high values as a result of the method of computation at the life stage level. In addition, measurement errors are summed over life stages within each taxon, which then amplifies the already skewed life stage values. These data are not indicative of actual measurement error and should only be compared to other measurement error results that are calculated using exactly the same methods. In all cases of failed QC samples, the data were corrected and the QC sample inspection frequency was maintained at 100% for that individual until acceptable results were demonstrated as determined by the QC sampling plan.
24100 Hudson River 2013 QC.docx 7/14/14 10 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 4. Sample sorting fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Total # of Fraction (Beginning Monday) Samples Inspected Samples Sorted Inspected (%)
11 Mar 13 13 74 17.57 18 Mar 13 37 74 50.00 25 Mar 13 34 74 45.95 1 Apr 13 54 126 42.86 8 Apr 13 39 126 30.95 15 Apr 13 36 126 28.57 22 Apr 13 31 135 22.96 29 Apr 13 41 135 30.37 6 May 13 61 135 45.19 13 May 13 79 126 62.70 20 May 13 55 126 43.65 27 May 13 60 126 47.62 3 Jun 13 49 123 39.84 10 Jun 13 58 123 47.15 17 Jun 13 49 123 39.84 24 Jun 13 21 122 17.21 8 Jul 13 24 81 29.63 22 Jul 13 22 75 29.33 5 Aug 13 12 81 14.81 19 Aug 13 42 81 51.85 2 Sep 13 19 81 23.46 16 Sep 13 16 81 19.75 30 Sep 13 20 81 24.69 Study 872 2,435 35.81 24100 Hudson River 2013 QC.docx 7/14/14 11 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 5. Sample sorting percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program.
- of Total # of Sampling Week Noncon- Samples % Non-(Beginning Monday) formities Inspected conformance 11 Mar 13 0 13 0.00 18 Mar 13 1 37 2.70 25 Mar 13 2 34 5.88 1 Apr 13 2 54 3.70 8 Apr 13 3 39 7.69 15 Apr 13 0 36 0.00 22 Apr 13 2 31 6.45 29 Apr 13 5 41 12.20 6 May 13 4 61 6.56 13 May 13 8 79 10.13 20 May 13 4 55 7.27 27 May 13 11 60 18.33 3 Jun 13 7 49 14.29 10 Jun 13 8 58 13.79 17 Jun 13 8 49 16.33 24 Jun 13 2 21 9.52 8 Jul 13 2 24 8.33 22 Jul 13 1 22 4.55 5 Aug 13 1 12 8.33 19 Aug 13 3 42 7.14 2 Sep 13 0 19 0.00 16 Sep 13 0 16 0.00 30 Sep 13 1 20 5.00 Study 75 872 8.60 24100 Hudson River 2013 QC.docx 7/14/14 12 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 6. Sample sorting mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Mean Percent (Beginning Monday) Samples Inspected Measurement Error 11 Mar 13 13 0.00 18 Mar 13 37 0.90 25 Mar 13 34 4.19 1 Apr 13 54 1.54 8 Apr 13 39 2.17 15 Apr 13 36 0.21 22 Apr 13 31 3.87 29 Apr 13 41 3.36 6 May 13 61 2.08 13 May 13 79 3.48 20 May 13 55 2.22 27 May 13 60 6.10 3 Jun 13 49 3.86 10 Jun 13 58 4.44 17 Jun 13 49 4.61 24 Jun 13 21 3.30 8 Jul 13 24 2.07 22 Jul 13 22 2.55 5 Aug 13 12 2.19 19 Aug 13 42 4.40 2 Sep 13 19 0.44 16 Sep 13 16 0.52 30 Sep 13 20 0.69 Study 872 2.97 24100 Hudson River 2013 QC.docx 7/14/14 13 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 7. Sample identification fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Total # of Fraction (Beginning Monday) Samples Inspected Samples Identified Inspected (%)
11 Mar 13 8 53 15.09 18 Mar 13 8 57 14.04 25 Mar 13 10 68 14.71 1 Apr 13 9 66 13.64 8 Apr 13 10 63 15.87 15 Apr 13 8 47 17.02 22 Apr 13 10 83 12.05 29 Apr 13 14 128 10.94 6 May 13 25 134 18.66 13 May 13 18 124 14.52 20 May 13 17 126 13.49 27 May 13 19 126 15.08 3 Jun 13 15 123 12.20 10 Jun 13 18 123 14.63 17 Jun 13 18 123 14.63 24 Jun 13 17 122 13.93 8 Jul 13 14 81 17.28 22 Jul 13 12 75 16.00 5 Aug 13 8 80 10.00 19 Aug 13 32 81 39.51 2 Sep 13 13 81 16.05 16 Sep 13 10 81 12.35 30 Sep 13 13 80 16.25 Study 326 2,125 15.34 24100 Hudson River 2013 QC.docx 7/14/14 14 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 8. Sample identification percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program.
- of Total # of Sampling Week Noncon- Samples % Non-(Beginning Monday) formities Inspected conformance 11 Mar 13 0 8 0.00 18 Mar 13 0 8 0.00 25 Mar 13 0 10 0.00 1 Apr 13 0 9 0.00 8 Apr 13 0 10 0.00 15 Apr 13 0 8 0.00 22 Apr 13 0 10 0.00 29 Apr 13 0 14 0.00 6 May 13 0 25 0.00 13 May 13 0 18 0.00 20 May 13 0 17 0.00 27 May 13 0 19 0.00 3 Jun 13 0 15 0.00 10 Jun 13 2 18 11.11 17 Jun 13 0 18 0.00 24 Jun 13 0 17 0.00 8 Jul 13 0 14 0.00 22 Jul 13 0 12 0.00 5 Aug 13 0 8 0.00 19 Aug 13 0 32 0.00 2 Sep 13 0 13 0.00 16 Sep 13 0 10 0.00 30 Sep 13 0 13 0.00 Study 2 326 0.61 24100 Hudson River 2013 QC.docx 7/14/14 15 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 9. Sample identification mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Mean Percent Number of (Beginning Monday) Samples Inspected Measurement Error Taxa Inspected 11 Mar 13 8 0.36 17 18 Mar 13 8 0.19 25 25 Mar 13 10 0.00 19 1 Apr 13 9 0.00 29 8 Apr 13 10 0.00 20 15 Apr 13 8 0.00 16 22 Apr 13 10 0.23 19 29 Apr 13 14 0.09 34 6 May 13 25 0.22 65 13 May 13 18 0.31 58 20 May 13 17 0.71 70 27 May 13 19 0.79 98 3 Jun 13 15 0.80 79 10 Jun 13 18 6.25 106 17 Jun 13 18 0.48 106 24 Jun 13 17 0.38 87 8 Jul 13 14 0.41 84 22 Jul 13 12 0.67 46 5 Aug 13 8 0.06 44 19 Aug 13 32 0.25 123 2 Sep 13 13 0.51 31 16 Sep 13 10 0.56 21 30 Sep 13 13 0.27 22 Study 326 0.92 1,219 24100 Hudson River 2013 QC.docx 7/14/14 16 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Additional organisms found during the sort QC were identified independently to determine the frequency of species and life stages missed during the initial sort. Six taxa accounted for 91% of the additional organisms found during sort QC: Striped Bass, White Perch, herring family, Bay Anchovy, Morone sp., and Freshwater Drum (Table 10). For these six taxa, the additional number found in the sort QC amounted to less than 2% of the total found during sample processing (Table 11). For most taxa-life stages the percentage missed by the original sorter was well under 5%.
The life stage most commonly missed by sorters was post yolk-sac larvae for White Perch, clupeids, and Morone sp.; eggs for Bay Anchovy and Freshwater Drum; and yolk-sac larvae for Striped Bass (Table 11). The life stage most frequently missed by sorters was usually the most abundant life stage.
During the identification process, absolute error rates for individual life stages of commonly encountered taxa ranged were usually less than 0.02. Generally, only those taxa-life stages with low total counts had absolute error rates above 0.05 (Table 12).
Net error rates were substantially lower than the absolute error rates in most cases, demonstrating that errors often tended to cancel each other out. This was especially noticeable for many of the most abundant taxa-life stages, such as Bay Anchovy eggs and Striped Bass yolk-sac larvae and post yolk-sac larvae.
3.2 Fall Juvenile Survey Results of the laboratory quality control program for the 2013 Fall Juvenile Survey (consisting of the Beach Seine Survey and the Fall Shoals Survey) were summarized by the same methods as the QC results for the 2013 Hudson River Ichthyoplankton Laboratory Program and are presented in Table 13.
There were 516 and 911 young-of-the-year fish identification records made in the laboratory for the Fall Shoals and Beach Seine surveys respectively and 4,084 and 3,536 young-of-the-year fish length measurement records were made for the Fall Shoals and Beach Seine surveys respectively.
24100 Hudson River 2013 QC.docx 7/14/14 17 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 10. Ranking of taxa missed during initial sort and found during sort QC.
Number of Organisms Taxon Found in Sort QC Percent Striped Bass 2,090 32.87 White Perch 1,888 29.69 Herring family 1,027 16.15 Bay Anchovy 485 7.63 Morone species 159 2.50 Freshwater Drum 158 2.48 Unidentified 108 1.70 Winter Flounder 90 1.42 Tautog 85 1.34 Cunner 74 1.16 Goby family 51 0.80 Gizzard Shad 32 0.50 Fourbeard Rockling 23 0.36 Atlantic Tomcod 16 0.25 Hogchoker 13 0.20 Atlantic Menhaden 12 0.19 Carp and minnow family 8 0.13 Windowpane 8 0.13 Weakfish 7 0.11 Atlantic Croaker 6 0.09 Common Carp 4 0.06 Fourspot Flounder 3 0.05 Inland Silverside 3 0.05 American Shad 2 0.03 Yellow Perch 2 0.03 Atlantic Cod 1 0.02 Butterfish 1 0.02 Menidia species 1 0.02 Northern Pipefish 1 0.02 Pollock 1 0.02 Total 6,359 100.00 24100 Hudson River 2013 QC.docx 7/14/14 18 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 11. Summary by life stage of the six highest ranked taxa missed during original sort and found during sort QC compared to total count.
Total Percent in Each Percent of Total Organisms Taxon Life Stage Number Stage Found Founda Bay Anchovy Eggs 391 80.62 1.06 36,727 Yolk-sac larvae 7 1.44 4.32 162 Post yolk-sac larvae 87 17.94 0.47 18,670 Young-of-the-year 0 0.00 0.00 4,628 Unidentified 0 0.00 0.00 16 Freshwater Eggs 114 72.15 2.41 4,737 Drum Yolk-sac larvae 30 18.99 1.25 2,406 Post yolk-sac larvae 13 8.23 0.91 1,428 Young-of-the-year 0 0.00 0.00 1 Unidentified 1 0.63 5.88 17 Herring family Eggs 193 18.79 1.63 11,813 Yolk-sac larvae 173 16.85 5.02 3,449 Post yolk-sac larvae 653 63.58 2.80 23,291 Unidentified 8 0.78 7.62 105 Morone species Yolk-sac larvae 13 8.18 8.13 160 Post yolk-sac larvae 91 57.23 2.47 3,677 Unidentified 55 34.59 2.48 2,219 Striped Bass Eggs 649 31.05 1.41 45,915 Yolk-sac larvae 786 37.61 1.71 45,927 Post yolk-sac larvae 639 30.57 0.79 81,178 Young-of-the-year 0 0.00 0.00 183 Unidentified 16 0.77 3.90 410 White Perch Eggs 344 18.22 3.97 8,664 Yolk-sac larvae 295 15.63 3.22 9,163 Post yolk-sac larvae 1,237 65.52 2.79 44,295 Young-of-the-year 0 0.00 0.00 42 Unidentified 12 0.64 5.36 224 a Includes both original count and additional organisms found during sort QC.
24100 Hudson River 2013 QC.docx 7/14/14 19 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 12. Cumulative net and absolute error rates for commonly encountered taxa in samples selected for QC inspection of identification and counting process.
Total Net Absolute Taxon Stage Count Error Error N Atlantic Menhaden Eggs 124 -0.00806 0.00806 7 Post Yolk Sac Larvae 879 -0.00228 0.00455 83 Atlantic Tomcod Young of the Year 168 -0.02381 0.03571 44 Yolk-Sac Larvae 3 0.00000 0.00000 2 Post Yolk Sac Larvae 138 0.00000 0.00000 28 Bay Anchovy Young of the Year 222 0.00000 0.00000 48 Unidentified 0 1.00000 1.00000 1 Eggs 4,569 -0.00263 0.00875 57 Yolk-Sac Larvae 73 0.04110 0.04110 4 Post Yolk Sac Larvae 3,963 -0.00606 0.01060 101 Young of the Year 840 0.01429 0.01905 61 Freshwater Drum Unidentified 2 0.00000 0.00000 1 Eggs 156 0.00000 0.00000 5 Yolk-Sac Larvae 204 0.00980 0.03922 20 Post Yolk Sac Larvae 255 -0.05098 0.05098 24 Young of the Year 1 0.00000 0.00000 1 Goby family Unidentified 1 0.00000 0.00000 1 Post Yolk Sac Larvae 1,007 -0.00695 0.01291 68 Herring family Unidentified 9 0.11111 0.11111 5 Eggs 1,340 -0.00896 0.00896 16 Yolk-Sac Larvae 310 -0.00323 0.01613 33 Post Yolk Sac Larvae 2,558 -0.00156 0.01407 101 Hogchoker Eggs 692 -0.00289 0.00867 16 Yolk-Sac Larvae 30 -0.03333 0.03333 5 Post Yolk Sac Larvae 4 0.25000 0.25000 4 Young Of The Year 5 0.00000 0.00000 3 Morone species Unidentified 263 -0.00760 0.09125 28 Yolk-Sac Larvae 6 0.00000 0.33333 4 Post Yolk Sac Larvae 646 0.00464 0.03870 41 Striped Bass Unidentified 28 0.10714 0.17857 7 Eggs 4,207 0.00071 0.00071 59 Yolk-Sac Larvae 4,889 0.00900 0.02168 78 Post Yolk Sac Larvae 10,275 -0.00467 0.01071 114 Young of the Year 41 0.00000 0.00000 14 Weakfish Eggs 647 0.00155 0.01700 9 Post Yolk Sac Larvae 30 0.00000 0.00000 14 Young Of The Year 22 0.00000 0.00000 10 (continued) 24100 Hudson River 2013 QC.docx 7/14/14 20 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 12. (Continued)
Total Net Absolute Taxon Stage Count Error Error N White Perch Unidentified 8 0.12500 0.12500 3 Eggs 936 -0.00641 0.00641 27 Yolk-Sac Larvae 586 0.02218 0.03242 49 Post Yolk Sac Larvae 4,015 -0.00897 0.02092 88 Young of the Year 3 0.00000 0.00000 2 Winter Flounder Unidentified 2 0.00000 0.00000 1 Yolk-Sac Larvae 16 0.00000 0.00000 8 Post Yolk Sac Larvae 170 0.00000 0.01176 27 Young of the Year 13 0.00000 0.00000 5 24100 Hudson River 2013 QC.docx 7/14/14 21 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 13. Fraction inspected, percent nonconforming, and average outgoing quality of laboratory tasks performed by Normandeau for the 2013 Fall Juvenile Survey.
Average Fraction Percent Average Outgoing Task Inspected (%) Nonconforming (%) Quality (%)
Identification 6.66 0.00 0.00 Measurement 2.03 0.00 0.00 24100 Hudson River 2013 QC.docx 7/14/14 22 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 4.0 Literature Cited Normandeau Associates Inc. 2013. 2013 Hudson River fall juvenile and beach seine surveys standard operating procedures. June 2013. Prepared for Indian Point Energy Center.
__________. 2014. Quality assurance plan for the Hudson River ichthyoplankton laboratory program. Rev. 2, April 2014. Prepared for Entergy Nuclear Operations, Inc.
Stephens, K.S. 1979. Volume 2: How to perform continuous sampling (CSP). American Society for Quality Control. 70 pp.
U.S. Department of Defense. 1981. Military standard. Single- and multi-level continuous sampling procedures and table for inspection by attributes. MIL-STD-1235B.
24100 Hudson River 2013 QC.docx 7/14/14 23 Normandeau Associates, Inc.
Quality Control Report for the 2014 Hudson River Ichthyoplankton Laboratory Program and 2014 Fall Juvenile Survey Prepared for:
Indian Point Energy Center 450 Broadway, Suite 1 Buchanan, New York 10511 Submitted:
October 2015 Prepared by:
Normandeau Associates, Inc.
25 Nashua Road Bedford, NH 03110 www.normandeau.com
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table of Contents Page
1.0 INTRODUCTION
................................................................................ 1 2.0 METHODS....................................................................................... 2 2.1 LABORATORY QUALITY CONTROL PROCEDURES ............................................. 2 2.1.1 Ichthyoplankton Survey........................................................ 2 2.1.2 Fall Juvenile and Beach Seine Surveys ...................................... 3 2.2 REPORTING PROCEDURES ................................................................... 5 2.2.1 Percent Inspected .............................................................. 5 2.2.2 Percent Nonconforming ....................................................... 6 2.2.3 Percent Error.................................................................... 6 2.2.4 Average Outgoing Quality ..................................................... 8 2.2.5 Cumulative Error Rates ........................................................ 9 3.0 RESULTS .......................................................................................10 3.1 ICHTHYOPLANKTON LABORATORY PROGRAM ............................................... 10 3.2 FALL JUVENILE SURVEY .................................................................... 18 4.0 LITERATURE CITED ..........................................................................23 24200 Hudson River 2014 QC FINAL.docx 11/18/15 ii Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM List of Figures Page Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks. ...................................................................... 3 Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes. ........................................... 4 24200 Hudson River 2014 QC FINAL.docx 11/18/15 iii Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM List of Tables Page Table 1. Task specific applications of continuous sampling plans for the 2014 Hudson River Ichthyoplankton Laboratory Program. ................................ 2 Table 2. Task specific applications of continuous sampling plans for the 2014 Fall Juvenile Survey............................................................................ 5 Table 3. Examples of percent error calculations for individual taxa and samples during both the identification and sorting tasks. .................................... 8 Table 4. Average percent inspected, percent nonconforming, mean percent error, and average outgoing quality of tasks performed by Normandeau for the 2014 Hudson River Ichthyoplankton Laboratory Program................. 11 Table 5. Sample sorting percent inspected results, 2014 Hudson River Ichthyoplankton Laboratory Program. ................................................ 12 Table 6. Percent nonconformance of samples inspected during the sorting procedure, 2014 Hudson River Ichthyoplankton Laboratory Program............ 13 Table 7. Mean percent error for the sorting procedure by sampling week, 2014 Hudson River Ichthyoplankton Laboratory Program. ............................... 14 Table 8. Percent of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program. ............................... 15 Table 9. Sample identification percent nonconformance results, 2014 Hudson River Ichthyoplankton Laboratory Program. ......................................... 16 Table 10. Mean percent error of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program............ 17 Table 11. Ranking of ichthyoplankton taxa missed during initial sort and found during sort QC, 2014 Hudson River Ichthyoplankton Laboratory Program....... 19 Table 12. Summary of ichthyoplankton by life stage for the six highest ranked taxa missed during original sort and found during sort QC compared to total count, 2014 Hudson River Ichthyoplankton Laboratory Program........... 20 Table 13. Cumulative net and absolute error rates for commonly encountered taxa in ichthyoplankton samples selected for QC inspection of identification process, 2014 Hudson River Ichthyoplankton Laboratory Program.................................................................................... 21 Table 14. Percent inspected, percent nonconforming, and average outgoing quality of laboratory tasks for the 2014 Fall Juvenile and Beach Seine Surveys..................................................................................... 22 24200 Hudson River 2014 QC FINAL.docx 11/18/15 iv Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM 1.0 Introduction This Quality Control (QC) report for the laboratory tasks of the 2014 Hudson River Ichthyoplankton Survey, 2014 Fall Juvenile Survey and, 2014 Beach Seine Survey was prepared for Entergys Indian Point Energy Center by Normandeau Associates Inc.
(Normandeau).
To comply with Entergys requirements for valid and reliable data on the Hudson River Ichthyoplankton Laboratory Program and the Fall Juvenile and Beach Seine Surveys, Normandeau implemented a Quality Assurance Plan that provides a 10% Average Outgoing Quality Limit (AOQL) for all measurement parameters collected. The Quality Assurance Plan consists of two systems: a Quality Control (QC) system and a Quality Assurance (QA) system. The QC system is managed by the program manager and conducted by operational personnel. The system monitors and documents the reliability and validity (accuracy, precision, completeness) of daily operations. The specific features of the QC system are determined by the Quality Assurance Department to insure that all procedures conform to Entergys data requirements. The QA system is managed by Normandeaus Quality Assurance Director and utilizes project independent personnel familiar with the work or activities under evaluation to conduct performance and systems audits. These audits are designed to provide objective evidence that the QC program and technical requirements, methods, and procedures as outlined in the program Standard Operating Procedures are being implemented. The outcomes of the QA system activities are verification of the effectiveness of the QC system, assignment of corrective actions to resolve nonconforming procedures or data deficiencies, communication of audit results to project and staff managers for follow-up, and objective validation or improvement of project operations.
This report provides a compilation of QC system data verifying the results of the 2014 Hudson River Ichthyoplankton Laboratory Program and 2014 Fall Juvenile and Beach Seine Surveys activities. Determinations of the percent inspected, percent nonconforming, and average outgoing quality are presented for both programs. In addition, for the 2014 Hudson River Ichthyoplankton Laboratory Program the results include percent error, a summary of the number of each taxon-life stage found during sorting QC, and cumulative error rates for each taxon-life stage.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 1 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM 2.0 Methods 2.1 Laboratory Quality Control Procedures 2.1.1 Ichthyoplankton Survey The Ichthyoplankton Survey collected samples by a 1-m wide epibenthic sled and a 1-m2 Tucker trawl, each with a 500 m mesh net. For sorting and identification of samples from the 2014 Hudson River Ichthyoplankton Laboratory Program, Normandeau used a continuous sampling plan designed to provide a 10% AOQL (U.S. Department of Defense 1981). A flow diagram of how the sampling plan was applied is presented in Figure 1. A summary of the sampling plan, tolerances, and QC sample definitions used for each measurement parameter is presented in Table 1. QC inspection was applied on a laboratory-wide basis for the sorting task and to each individual processor for the identification task. Samples were selected in a random manner utilizing random number tables. As determined from the sampling plan outlined in Table 1, a specified number of samples were reprocessed by QC inspectors with expertise in the task being inspected. In cases where a sample was subdivided and counted, counts for all subdivisions were combined before calculating percent error for that sample. If the difference between the QC value and the original value exceeded acceptable tolerances (Table 1), a third measurement could be obtained to verify one of the measurements. If a sample from the block of 1 out of 7 samples (f) was found to have exceeded acceptable tolerances, all samples within that block were subjected to review to ensure that the quality meets acceptable tolerance levels.
Furthermore, all subsequent samples processed by the same technician were subjected to 100% QC until 8 consecutive samples (i) were found within acceptable tolerance limits as determined by the continuous sampling plan (Table 1 and Figure 1). The Quality Assurance plan (Normandeau 2014) documents specific QA/QC methods utilized for this program.
Table 1. Task specific applications of continuous sampling plans for the 2014 Hudson River Ichthyoplankton Laboratory Program.
CSP-1 QC Laboratory AOQL 10% Sample Task i* f** Sample Tolerance Definition Sorting 8 1/7 +/- 2 organisms if d20 organisms one
+/- 10% if >20 organisms sample Identification 8 1/7 +/- 2 organisms if d20 organisms one
+/- 10% if >20 organisms sample for every taxon in the sample (in identifying, assigning a life stage, or counting any species, errors are cumulative by life stage within each taxon)
- i = number of samples within the 100% sample inspection training period
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks.
In some cases, one of the taxonomists (either the original identifier or the QC inspector) was able to determine the taxon or life stage of damaged specimens when the other taxonomist recorded them as unknown life stage, unidentified taxon, or a higher level taxon (genus or family). If a more general taxon or life stage used by one taxonomist included the more specific category used by the other taxonomist and was only separated by one consecutive taxonomic level such as Family-Genus or Genus-Species, and that was the only reason for a count discrepancy, then that sample was not considered failing the QC inspection. For example, damaged specimens recorded as Morone sp. (Genus level) by the original identifier and as Morone saxatilis (Species level) by the QC inspector were considered to be in agreement because the category Morone sp. includes Morone saxatilis. In contrast, an original determination of unidentified Gobiidae (Family) would not be acceptable if the QC determination was Morone saxatilis (Species), because 1) taxonomically, Morone saxatilis is not in the family Gobiidae and 2) M. saxatilis is more than one taxonomic level separated from the family Gobiidae.
2.1.2 Fall Juvenile and Beach Seine Surveys The Fall Juvenile Survey, also historically referred as Fall Shoals Survey, consisted of collections from two gear types, a 3-m wide x 0.9 m high beam trawl (3.2 cm stretch mesh codend and 1.3 cm stretch mesh liner) and 1-m2 Tucker trawl (2.0 mm mesh net). Juvenile fish are also sampled along the shore by the Beach Seine Survey, which uses a 2.4-m high 24200 Hudson River 2014 QC FINAL.docx 11/18/15 3 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM seine with two 12-m long wings of 1.0-cm bar mesh and 0.5-cm bar mesh bag. For laboratory identification and length measurements of young-of-the-year fishes in the 2014 Fall Juvenile Survey, Normandeau used a continuous sampling plan designed to provide a 10% Average Outgoing Quality Limit (U.S. Department of Defense, 1981). A flow diagram of how the plan was applied is presented in Figure 2.
Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes.
A summary of the sampling plan, tolerances, and QC sample definitions used for each task is shown in Table 2. QC samples were selected as specified by the appropriate plan in Table 2, using random numbers, and reprocessed by QC inspectors. If the difference between original and QC values exceeded the acceptable tolerance, a third value was obtained as a resolution. The standard operating procedures manual (Normandeau 2013) documents specific QA/QC methods used for the 2014 Fall Juvenile and Beach Seine Surveys. Young-24200 Hudson River 2014 QC FINAL.docx 11/18/15 4 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM of-the-year fishes were identified in the laboratory for the first two Fall Juvenile Survey River Runs (sampling weeks) and the first three Beach Seine Survey River Runs. Young-of-the-year fishes were identified in the field starting with Fall Juvenile Survey River Run 3 and Beach Seine Survey River Run 4. The same QC procedures applied to both field and laboratory identifications. All length measurements of young-of-the-year fishes occurred in the laboratory.
Table 2. Task specific applications of continuous sampling plans for the 2014 Fall Juvenile Survey.
QC QC Sample Task Plan AOQL i* f** x Tolerance Definition Identification CSP-V 7% 21 1/15 7 +/-10% of total count or +/-2 One taxon individuals when <25 fish
+/-1 mm when <34 mm TL Length CSP-V 7% 30 1/50 10 One fish
+/-3% when >34 mm TL
- i = number of samples within the 100% sample inspection training period
- f = fraction of samples randomly selected for QC 2.2 Reporting Procedures The 2014 Hudson River Ichthyoplankton Laboratory Program Sort and Identification QC Logs were keyed, verified, and error-checked to produce SAS data sets. From these data, percent inspected, percent nonconforming, and percent error (precision) were determined for each River Run and for the entire study. For the 2014 Fall Juvenile Survey, QC data were used to determine percent inspected and percent nonconforming for the entire study (combining all River Runs processed in the laboratory for both the Fall Juvenile Survey and the Beach Seine Survey).
2.2.1 Percent Inspected Percent Inspected Number of Samples Inspected x 100 (Equation 1)
Total Number of Samples Analyzed River Run: Percent inspected (%) for a River Run (Equation 1) was the number of samples inspected divided by the total number of samples analyzed for that River Run times one hundred. For the ichthyoplankton sorting task, the number of samples inspected excludes training QC samples for new sorters, which do not represent the independent performance of the technician, as well as the samples inspected as part of the QC plan. For the identification task, the total number of samples identified excludes empty (no catch) samples, which did not require processing by an identifier.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 5 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Entire Study: Percent inspected for the entire study was the number of samples inspected divided by the total number of samples analyzed times one hundred during the study.
2.2.2 Percent Nonconforming Percent Nonconforming Number of Nonconforming Quality Control Samples Inspected x 100 Total Number of Quality Control Samples Inspected (Equation 2)
River Run: Percent nonconforming for a River Run (Equation 2) was the number of nonconforming QC samples found for that River Run divided by the total number of QC samples inspected times one hundred for that River Run.
Entire Study: Percent nonconforming for the entire study was the total number of nonconforming QC samples for the study divided by the total number of QC samples inspected times one hundred for the study. The result of this analysis was a determination of the actual incoming quality level of each measurement parameter. (Note that because samples checked by QC found to be defective were rectified during QC, the average outgoing quality of the final data set was better than that indicated by the percent nonconforming.)
2.2.3 Percent Error Sorting Task Sorting Percent Error Quality Control Count x 100 (Equation 3)
(Original Count Quality Control Count )
Sample: Percent error for a sorted sample (Equation 3) was the QC count divided by the sum of the original count and the QC count (adjusted total) times one hundred. The denominator is an adjusted total because the sorting procedure is the initial process by which all countable organisms are removed from the field sample. Therefore, if QC personnel have found additional organisms that the original sorter missed, then the total count must reflect this addition to the sample total. The sorting task is not life stage or taxon specific, it is the total count of all countable organisms sorted. Refer to Table 3 for an example of this sample percent error calculation.
River Run: Mean percent error for sorted samples for a River Run was the sum of the percent errors for each sample inspected during the River Run divided by the total number of samples inspected for the River Run times one hundred.
Entire Study: Mean percent error for sorted samples for the entire study was the sum of the percent measurement errors for each sample inspected during the study divided by the total number of samples inspected for the study times one hundred. (Note that this method of 24200 Hudson River 2014 QC FINAL.docx 11/18/15 6 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM averaging gives equal weight to each sample, regardless of the number of organisms present).
Identification Task Life Stage Percent Error (Original Count Quality Control Count )
x 100 (Equation 4)
Quality Control Count Life Stage: Percent error for a life stage (Equation 4) was the difference between the original life stage count and the QC life stage count divided by the QC life stage count times one hundred.
Exceptions to Equation 4:
- 1) For life stages where the QC count was 20 individuals or less and the original and QC counts differed by no more than two organisms, the percent error was defined as
+/- 1 or 2. For life stages where the QC value was 20 or less and the original and QC values differed by more than two organisms, the percent error was calculated utilizing Equation 4. Refer to Table 3 for an example of this life stage percent error calculation exception.
- 2) If the QC count was zero, the percent error will be 100%. This error was calculated by dividing the difference between the original count and the QC count by the original count times one hundred. Refer to Table 3 for an example of this QC=0 percent error calculation exception.
If the sample fails then a resolution by a third party must occur. Percent error will then be recalculated for the QC counts and the original counts but by using Equation 4 but using the Resolution Counts as the denominator. Then, if the original count for a life stage was acceptably close (within 10%) to a resolution value but not to the QC value, the original counts are not to be changed and the samples passes. If the original counts are not within 10% from the resolution count then the sample fails and the original data counts are replaced by the QC counts. During the rare occurrence that the original values are greater than 10% from the resolution and the QC counts, then the sample will be reanalyzed by all three people and the identifiers sample processing will not continue until agreement can be reached on the identification of the sample.
Taxon: Percent error for an identified taxon was the sum of the absolute values of percent error for each life stage within the taxon. Refer to Table 3 for an example of taxon percent error calculations.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 7 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 3. Examples of percent error calculations for individual taxa and samples during both the identification and sorting tasks.
Yolk-Sac Post Yolk-Eggs Larvae Sac Larvae Total Sorting Task Example:
Sample Error Original Count 250 52 351 653 Quality Control Count 3 0 5 8
% Error by Sample 1.2%
Identification Task Examples:
Taxon 1 Original Count 103 186 31 Quality Control Count 100 194 26
% Error by Life Stage 3.0% 4.1% 5.2% 12.3%
Taxon 2 Original Count 15 Quality Control Count 13
% Error by Life Stage +/- 2 +/- 2 Taxon 3 Original Count 8 Quality Control Count 2
% Error by Life Stage 300% 300%
Taxon 4 Original Count 25 20 2 Quality Control Count 25 0 22
% Error by Life Stage 100% 10% 110%
River Run: Mean percent error for the identification task for a River Run was the sum of the percent measurement errors for all taxa inspected during the River Run divided by the total number of taxa inspected for the River Run. This statistic was computed by averaging taxa rather than samples because even though complete samples were inspected and reworked for identification QC, the pass/fail criterion was whether any taxon in the sample individually exceeded the 10% tolerance.
Entire Study: Mean percent error for identified taxa for the entire study was the sum of the percent errors for all taxa inspected during the study divided by the total number of taxa inspected for the study.
2.2.4 Average Outgoing Quality At the completion of these studies, the Average Outgoing Quality (AOQ) was calculated for each measurement parameter inspected. Continuous sampling plans were used for all tasks. Continuous sampling plans are devised for processes involving a continuous or nearly continuous flow of products or other entities. For these types of processes, it is extremely difficult to organize units into discrete groups commonly referred to as lots. As a 24200 Hudson River 2014 QC FINAL.docx 11/18/15 8 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM result, inspection must be performed on individual units drawn from a continuous flow of products and a decision made concerning the quality of units produced based on the inspection results. Rectification is performed on any nonconforming unit found during inspection, followed by 100% screening of a number of subsequent units depending on the sampling plan. Average Outgoing Quality for each laboratory task was calculated as a function of the percent nonconforming and the fraction of total units inspected (Stephens 1979). This calculation applies to continuous sampling plans when nonconforming units found are rectified:
p' (1 f )q i AOQ x 100 (Equation 5) f (1 f )q i where p' = Percent nonconforming as a decimal fraction f = Proportion of units inspected. This is a parameter of the sampling plan.
q = 1-p' = Percent conforming as a decimal fraction i = Clearing interval. This is a parameter of the sampling plan.
Example:
p' = 0.0689 f = 1/7 = 0.1429 q = 1-0.0689 = 0.9311 i = 8 0.0689 (1 0.1429)(0.9311) 8 AOQ x 100 5.32%
0.1429 (1 0.1429)(0.9311) 8 The above equation for calculating AOQ was formulated specifically for CSP-1 sampling plans such as those used for the ichthyoplankton sorting and identification (Table 1). The same equation was used to calculate AOQ for young-of-the-year identifications and measurements, which used CSP-V plans (Table 2). When Equation 5 is used for CSP-V plans, the calculated AOQ is conservatively high, because the equation does not take into account the times when the number of consecutive re-inspections following a failure is x (which is smaller than i).
2.2.5 Cumulative Error Rates Due to the non-independence of identification errors across taxa and life stages, and to the accumulation of errors within taxa, a relatively high fraction of samples may fail QC inspection even though only a small fraction of organisms are incorrectly identified or counted. In order to present the error frequencies more realistically for particular taxa-life stages, two additional statistics were calculated for each taxon-life stage for the identification/counting process.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 9 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Absolute Error Rate =
n n i 1 lIi Qi l / Q i 1 i Equation 6 Net Error Rate =
n n i 1 (I i Q i ) / Q i 1 i Equation 7 where Ii = initial count for taxon-life stage in sample i Qi = QC count for taxon-life stage in sample i (or the resolution count, if Ii was acceptably close to it but not to the QC count) n = number of samples in the entire study If the sum of Qi for the entire study was zero for the taxon-life stage, then the sum of Qi was set equal to one for the purpose of calculating absolute and net error rate.
The absolute error rate is the approximate fraction of the taxon-life stage that was originally identified or counted incorrectly. This is an estimate of the fraction of erroneous countable items in the uninspected samples.
Net error rate is the approximate relative error in the total counts for the taxon-life stage.
For this index, positive (original count too high) and negative (original count too low) errors cancel each other so that the index reflects the relative net bias to the taxon-life stage abundance.
3.0 Results 3.1 Ichthyoplankton Laboratory Program The Average Outgoing Quality (AOQ) of the 2014 Hudson River Ichthyoplankton Laboratory Program was 4.41% for the sorting task and 0.58% for the identification task.
These AOQ levels represent the actual or achieved quality for measurement parameters and were within the 10% AOQL requirement of the study. The average of the Percent Inspected among River Runs was 27.60% for the sorting task and 15.69% for the identification task (Table 4).
24200 Hudson River 2014 QC FINAL.docx 11/18/15 10 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 4. Average percent inspected, percent nonconforming, mean percent error, and average outgoing quality (AOQ) of tasks performed by Normandeau for the 2014 Hudson River Ichthyoplankton Laboratory Program.
Percent Percent Inspected Nonconforming Mean Percent Task (%) (%) Error (%) AOQ (%)
Sorting 27.60 5.57 1.95 4.41 Identification 15.69 0.68 4.65 0.58 Sorting and identification tasks were also evaluated on the basis of River Runs (sampling weeks). Sorted samples were inspected at a rate of 11.11% to 45.24% for individual River Runs (Table 5). Nonconformance for the sorting task among the inspected samples ranged from 0% to 21.05% in the 20 River Runs (Table 6). Sorting error ranged between 0.14% and 6.79% (Table 7). For the task of sample identification, 11.54% to 32.10% of samples were inspected from individual River Runs (Table 8). Percent nonconforming for the identification task in individual River Runs ranged from 0% to 3.85% (Table 9). Percent error for individual River Runs ranged from 0% to 30.28% (Table 10).
Percent error results for the identification task are skewed towards high values as a result of the method of computation at the life stage level. In addition, errors are summed over life stages within each taxon, which then amplifies the already skewed life stage values. These data are not indicative of actual error and should only be compared to other error results that are calculated using exactly the same methods. In all cases of failed QC samples, the data were corrected and the QC sample inspection frequency was maintained at 100% for that individual until acceptable results were demonstrated as determined by the QC sampling plan.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 11 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 5. Sample sorting percent inspected results, 2014 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Total # of Percent (Beginning Monday) Samples Inspected Samples Sorted Inspected (%)
31 Mar 14 53 126 42.06 07 Apr 14 52 126 41.27 14 Apr 14 41 126 32.54 21 Apr 14 29 135 21.48 28 Apr 14 23 135 17.04 05 May 14 30 135 22.22 12 May 14 38 126 30.16 19 May 14 57 126 45.24 26 May 14 52 126 41.27 02 Jun 14 43 122 35.25 09 Jun 14 35 123 28.46 16 Jun 14 32 123 26.02 23 Jun 14 27 123 21.95 07 Jul 14 12 81 14.81 21 Jul 14 9 75 12.00 04 Aug 14 12 78 15.38 18 Aug 14 9 81 11.11 01 Sep 14 27 81 33.33 15 Sep 14 11 81 13.58 29 Sep 14 18 81 22.22 Study 610 2210 27.60 24200 Hudson River 2014 QC FINAL.docx 11/18/15 12 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 6. Percent nonconformance of samples inspected during the sorting procedure, 2014 Hudson River Ichthyoplankton Laboratory Program.
Sampling Total No. of Week Number of Samples Percent (Beginning Monday) Nonconformities Inspected Nonconformance 31 Mar 14 2 53 3.77 07 Apr 14 1 52 1.92 14 Apr 14 1 41 2.44 21 Apr 14 0 29 0.00 28 Apr 14 0 23 0.00 05 May 14 2 30 6.67 12 May 14 3 38 7.89 19 May 14 12 57 21.05 26 May 14 1 52 1.92 02 Jun 14 2 43 4.65 09 Jun 14 5 35 14.29 16 Jun 14 0 32 0.00 23 Jun 14 2 27 7.41 07 Jul 14 0 12 0.00 21 Jul 14 0 9 0.00 04 Aug 14 0 12 0.00 18 Aug 14 0 9 0.00 01 Sep 14 3 27 11.11 15 Sep 14 0 11 0.00 29 Sep 14 0 18 0.00 Study 34 610 5.57 24200 Hudson River 2014 QC FINAL.docx 11/18/15 13 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 7. Mean percent error for the sorting procedure by sampling week, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Sampling Week Mean Percent Samples (Beginning Monday) Error Inspected 31 Mar 14 53 1.58 07 Apr 14 52 0.30 14 Apr 14 41 0.80 21 Apr 14 29 0.58 28 Apr 14 23 0.19 05 May 14 30 1.86 12 May 14 38 2.45 19 May 14 57 6.79 26 May 14 52 1.78 02 Jun 14 43 1.17 09 Jun 14 35 2.89 16 Jun 14 32 0.85 23 Jun 14 27 3.04 07 Jul 14 12 2.27 21 Jul 14 9 0.78 04 Aug 14 12 0.85 18 Aug 14 9 2.49 01 Sep 14 27 2.36 15 Sep 14 11 1.33 29 Sep 14 18 0.14 Study 610 1.95 24200 Hudson River 2014 QC FINAL.docx 11/18/15 14 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 8. Percent of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Sampling Week Total No. of Samples Percent (Beginning Monday) Samples Inspected Identified Inspected 31 Mar 14 10 61 16.39 07 Apr 14 8 52 15.38 14 Apr 14 7 52 13.46 21 Apr 14 10 67 14.93 28 Apr 14 14 98 14.29 05 May 14 18 125 14.40 12 May 14 18 124 14.52 19 May 14 18 126 14.29 26 May 14 17 126 13.49 02 Jun 14 17 122 13.93 09 Jun 14 21 123 17.07 16 Jun 14 26 123 21.14 23 Jun 14 16 123 13.01 07 Jul 14 12 81 14.81 21 Jul 14 11 75 14.67 04 Aug 14 9 78 11.54 18 Aug 14 13 81 16.05 01 Sep 14 11 81 13.58 15 Sep14 13 81 16.05 29 Sep 14 26 81 32.10 Study 295 1880 15.69 24200 Hudson River 2014 QC FINAL.docx 11/18/15 15 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 9. Sample identification percent nonconformance results, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Percent Sampling Week Number of Samples Nonconformance (Beginning Monday) Nonconformities Inspected (Week) 31 Mar 14 0 10 0.00 07 Apr 14 0 8 0.00 14 Apr 14 0 7 0.00 21 Apr 14 0 10 0.00 28 Apr 14 0 14 0.00 05 May 14 0 18 0.00 12 May 14 0 18 0.00 19 May 14 0 18 0.00 26 May 14 0 17 0.00 02 Jun 14 0 17 0.00 09 Jun 14 0 21 0.00 16 Jun 14 1 26 3.85 23 Jun 14 0 16 0.00 07 Jul 14 0 12 0.00 21 Jul 14 0 11 0.00 04 Aug 14 0 9 0.00 18 Aug 14 0 13 0.00 01 Sep 14 0 11 0.00 15 Sep 14 0 13 0.00 29 Sep 14 1 26 3.85 Study 2 295 0.68 24200 Hudson River 2014 QC FINAL.docx 11/18/15 16 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 10. Mean percent error of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Number of Sampling Week Mean Percent Samples Taxa (Beginning Monday) Error Inspected Inspected 31 Mar 14 10 0.00 12 07 Apr 14 8 0.00 15 14 Apr 14 7 0.00 11 21 Apr 14 10 0.15 19 28 Apr 14 14 0.02 27 05 May 14 18 0.18 35 12 May 14 18 0.11 63 19 May 14 18 0.47 76 26 May 14 17 0.76 74 02 Jun 14 17 0.71 95 09 Jun 14 21 0.43 108 16 Jun 14 26 30.28 163 23 Jun 14 16 0.45 88 07 Jul 14 12 0.38 63 21 Jul 14 11 0.31 54 04 Aug 14 9 0.11 52 18 Aug 14 13 0.13 49 01 Sep 14 11 0.02 37 15 Sep 14 13 0.00 47 29 Sep 14 26 1.36 59 Study 295 4.65 1147 Additional organisms found during the sort QC were identified independently to determine the frequency of species and life stages missed during the initial sort. Six taxa accounted for 92% of the additional organisms found during sort QC: Striped Bass, White Perch, Herring family, Bay Anchovy, Morone sp., and Goby family (Table 11). For these six taxa, the additional number of organisms per stage found in the sort QC amounted to less than 2% of the total found during sample processing. Exceptions of >2% were within the most abundant life stage: Bay Anchovy yolk-sac larvae (2.83%), Goby family eggs (100%) and yolk-sac larvae (25%), Herring unidentifiable stage (16.13%), Morone species unidentifiable stage (2.59%), Striped Bass eggs (2.93%) and unidentifiable stage (7.62%) and, White Perch eggs (6.84) and yolk-sac larvae (2.24%; Table 12). For most taxa by life stage, the percentage missed by the original sorter was well under 5%.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 17 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM During the identification process, absolute error rates for individual life stages of commonly encountered taxa were usually less than 0.02. Generally, only those taxa-life stages with low total counts had absolute error rates above 0.05 (Table 13).
Net error rates were substantially lower than the absolute error rates in most cases, demonstrating that errors often tended to cancel each other out. This was especially noticeable for many of the most abundant taxa-life stages, such as Bay Anchovy eggs and Striped Bass yolk-sac and post yolk-sac larvae.
3.2 Fall Juvenile Survey Results of the laboratory QC program for the 2014 Fall Juvenile Survey (consisting of the Beach Seine Survey and the Fall Shoals Survey) were summarized by the same methods as the QC results for the 2014 Hudson River Ichthyoplankton Laboratory Program and are presented in Table 14.
There were 1,213 and 847 young-of-the-year fish identification records made in the laboratory for the Fall Juvenile and Beach Seine surveys respectively and 7,588 and 4,923 young-of-the-year fish length measurement records were made for the Fall Juvenile and Beach Seine surveys respectively.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 18 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 11. Ranking of ichthyoplankton taxa missed during initial sort and found during sort Quality Control (QC), 2014 Hudson River Ichthyoplankton Laboratory Program.
Number of Organisms Taxon found in Sort QC Percent Striped Bass 2196 47.55 White Perch 904 19.58 Herring family 627 13.58 Bay Anchovy 305 6.60 Morone species 111 2.40 Unidentified 110 2.38 Goby family 96 2.08 Atlantic Tomcod 64 1.39 Winter Flounder 58 1.26 Cunner 37 0.80 Tautog 24 0.52 Freshwater Drum 22 0.48 Hogchoker 14 0.30 Windowpane 11 0.24 Atlantic Croaker 9 0.19 Atlantic Menhaden 9 0.19 Gizzard Shad 6 0.13 American Shad 3 0.06 American Sand Lance 2 0.04 Inland Silverside 2 0.04 American Eel 1 0.02 Carp and Minnow family 1 0.02 Common Carp 1 0.02 Summer Flounder 1 0.02 Sunfish family 1 0.02 Tessellated Darter 1 0.02 Weakfish 1 0.02 Yellow Perch 1 0.02 Total 4618 100.00 24200 Hudson River 2014 QC FINAL.docx 11/18/15 19 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 12. Summary of ichthyoplankton by life stage for the six highest ranked taxa missed during original sort and found during sort QC compared to total count, 2014 Hudson River Ichthyoplankton Laboratory Program.
Percent of Number Abundance Missed in Total Percent of Total Each by Life Original Organisms Found During Taxon Life Stage Stage Sort Found a QC Eggs 44.26 135 35421 0.38 Yolk-sac larvae 0.98 3 106 2.83 Bay Post yolk-sac larvae 54.75 167 19551 0.85 Anchovy Young-of-the-year 0 0 9436 0 Unidentified 0 0 27 0 Eggs 4.17 4 4 100 Goby Yolk-sac larvae 2.08 2 8 25 family Post yolk-sac larvae 93.75 90 8095 1.11 Unidentified 0 0 1 0 Eggs 13.08 82 12980 0.63 Herring Yolk-sac larvae 7.97 50 6075 0.82 family Post yolk-sac larvae 78.15 490 36086 1.36 Unidentified 0.8 5 31 16.13 Yolk-sac larvae 0 0 11 0 Morone Post yolk-sac larvae 82.88 92 4940 1.86 species Unidentified 17.12 19 733 2.59 Eggs 57.7 1267 43186 2.93 Yolk-sac larvae 22.22 488 36314 1.34 Striped Post yolk-sac larvae 19.31 424 181809 0.23 Bass Young-of-the-year 0.05 1 984 0.1 Unidentified 0.73 16 210 7.62 Eggs 33.08 299 4374 6.84 Yolk-sac larvae 16.92 153 6840 2.24 White Post yolk-sac larvae 50 452 28617 1.58 Perch Young-of-the-year 0 0 127 0 Unidentified 0 0 13 0 a Includes both original count and additional organisms found during sort QC.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 20 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 13. Cumulative net and absolute error rates for commonly encountered taxa in ichthyoplankton samples selected for QC inspection of identification process, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total Net Absolute No.
Taxon Stage Count Error Error Samples Atlantic Croaker Post Yolk Sac Larvae 1029 0.00389 0.00389 31 Young Of The Year 54 -0.07407 0.07407 9 Atlantic Menhaden Eggs 185 0.00000 0.00000 8 Yolk-Sac Larvae 7 0.00000 0.00000 4 Post Yolk Sac Larvae 35 1.34286 1.34286 18 Young Of The Year 6 0.16667 0.16667 5 Atlantic Tomcod Unidentified 12 0.00000 0.00000 3 Yolk-Sac Larvae 27 0.03704 0.03704 8 Post Yolk Sac Larvae 851 0.00118 0.00353 29 Young Of The Year 1118 -0.00089 0.00089 74 Bay Anchovy Unidentified 6 0.00000 0.00000 5 Eggs 7364 0.00326 0.00598 52 Yolk-Sac Larvae 18 0.05556 0.05556 6 Post Yolk Sac Larvae 3208 -0.01527 0.02276 113 Young Of The Year 1150 0.00348 0.00522 69 Blueback Herring Young Of The Year 736 0.00000 0.00272 25 Freshwater Drum Eggs 1209 0.00083 0.00083 10 Yolk-Sac Larvae 4 0.25000 0.25000 5 Post Yolk Sac Larvae 42 0.02381 0.02381 11 Goby family Unidentified 1 0.00000 0.00000 1 Post Yolk Sac Larvae 1357 -0.00221 0.00958 56 Herring family Unidentified 1 0.00000 0.00000 1 Eggs 1812 0.00386 0.00497 28 Yolk-Sac Larvae 1002 0.00399 0.01198 44 Post Yolk Sac Larvae 5001 -0.00100 0.01380 130 Hogchoker Eggs 744 0.00134 0.00941 15 Yolk-Sac Larvae 13 -0.07692 0.07692 8 Post Yolk Sac Larvae 15 -0.06667 0.06667 8 Young Of The Year 6 0.50000 0.50000 5 Morone species Unidentified 36 0.08333 0.13889 11 Post Yolk Sac Larvae 361 0.06925 0.08033 30 Striped Bass Unidentified 44 -0.02273 0.11364 8 Eggs 6499 -0.00015 0.00077 52 Yolk-Sac Larvae 4757 0.00399 0.01577 67 Post Yolk Sac Larvae 19207 -0.00193 0.00869 109 Young Of The Year 91 -0.01099 0.01099 28 Tautog Eggs 604 -0.00331 0.00993 11 Post Yolk Sac Larvae 3 0.33333 0.33333 3 (continued) 24200 Hudson River 2014 QC FINAL.docx 11/18/15 21 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 13. (Continued)
Total Net Absolute No.
Taxon Stage Count Error Error Samples White Perch Eggs 170 -0.01176 0.02353 22 Yolk-Sac Larvae 643 0.00467 0.01089 41 Post Yolk Sac Larvae 2182 0.00962 0.03437 80 Young Of The Year 42 0.00000 0.00000 7 Winter Flounder Yolk-Sac Larvae 37 0.00000 0.00000 6 Post Yolk Sac Larvae 265 0.00000 0.00000 16 Young Of The Year 2 0.00000 0.00000 2 Table 14. Percent inspected, percent nonconforming, and average outgoing quality of laboratory tasks for the 2014 Fall Juvenile and Beach Seine Surveys.
Percent Average Outgoing Task Percent Inspected (%) Nonconforming (%) Quality (%)
Identification 6.26 0.00 0.00 Measurement 1.92 0.00 0.00 24200 Hudson River 2014 QC FINAL.docx 11/18/15 22 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM 4.0 Literature Cited Normandeau Associates Inc. 2013. 2013 Hudson River fall juvenile and beach seine surveys standard operating procedures. June 2013. Prepared for Indian Point Energy Center.
__________. 2014. Quality assurance plan for the Hudson River ichthyoplankton laboratory program. Rev. 2, April 2014. Prepared for Entergy Nuclear Operations, Inc.
Stephens, K.S. 1979. Volume 2: How to perform continuous sampling (CSP). American Society for Quality Control. 70 pp.
U.S. Department of Defense. 1981. Military standard. Single- and multi-level continuous sampling procedures and table for inspection by attributes. MIL-STD-1235B.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 23 Normandeau Associates, Inc.
IPRenewal NPEmails From: Gray, Dara F <DGray@entergy.com>
Sent: Thursday, May 19, 2016 8:41 AM To: Wentzel, Michael
Subject:
[External_Sender] FW: Year Class Reports Attachments: 24100 Hudson River 2013 QC.pdf; 24200 Hudson River 2014 QC FINAL.pdf Importance: High Mike Here you go, and sorry again for the delay.
Dara Gray REM Chemistry/Environmental Indian Point Energy Center Dgray@entergy.com 914-254-8414 From: Wentzel, Michael [2]
Sent: Thursday, April 21, 2016 7:00 AM To: Gray, Dara F Cc: Louie, Richard
Subject:
Year Class Reports
- Dara, Thank you (and Rich) for sending the 2013 and 2014 Year Class Reports. We are having difficulty getting the files added to ADAMS because the files are password protected. Would it be possible to either get the password to unlock the files, or, unlocked versions of the files?
Please feel free to give me a call if you would like to discuss this further.
- Thanks, Mike Michael Wentzel Project Manager NRR/DLR/RERP (301) 415-6459 michael.wentzel@nrc.gov Please consider the environment before printing this e-mail.
1
Hearing Identifier: IndianPointUnits2and3NonPublic_EX Email Number: 6399 Mail Envelope Properties (DA94DFACF1201C4A91A21BD336C2520A23596363)
Subject:
[External_Sender] FW: Year Class Reports Sent Date: 5/19/2016 8:40:35 AM Received Date: 5/19/2016 8:40:50 AM From: Gray, Dara F Created By: DGray@entergy.com Recipients:
"Wentzel, Michael" <Michael.Wentzel@nrc.gov>
Tracking Status: None Post Office: LITXMETSP003.etrsouth.corp.entergy.com Files Size Date & Time MESSAGE 984 5/19/2016 8:40:50 AM 24100 Hudson River 2013 QC.pdf 394610 24200 Hudson River 2014 QC FINAL.pdf 434978 Options Priority: High Return Notification: No Reply Requested: No Sensitivity: Normal Expiration Date:
Recipients Received:
Quality Control Report for the 2013 Hudson River Ichthyoplankton Laboratory Program and 2013 Fall Juvenile Survey Prepared for:
Indian Point Energy Center 450 Broadway, Suite 1 Buchanan, New York 10511 Submitted:
July 2014 Prepared by:
Normandeau Associates, Inc.
25 Nashua Road Bedford, NH 03110 www.normandeau.com
2013 HUDSON RIVER QC REPORT Table of Contents Page
1.0 INTRODUCTION
................................................................................ 1 2.0 METHODS....................................................................................... 2 2.1 LABORATORY QUALITY CONTROL PROCEDURES ............................................. 2 2.1.1 Ichthyoplankton Survey........................................................ 2 2.1.2 Fall Juvenile Survey............................................................ 3 2.2 REPORTING PROCEDURES ................................................................... 4 2.2.1 Fraction Inspected ............................................................. 4 2.2.2 Percent Nonconforming ....................................................... 6 2.2.3 Percent Measurement Error................................................... 6 Sorting Task..................................................................... 6 Identification Task............................................................. 7 2.2.4 Average Outgoing Quality ..................................................... 8 2.2.5 Cumulative Error Rates ........................................................ 9 3.0 RESULTS .......................................................................................10 3.1 ICHTHYOPLANKTON LABORATORY PROGRAM ............................................... 10 3.2 FALL JUVENILE SURVEY .................................................................... 17 4.0 LITERATURE CITED ..........................................................................23 24100 Hudson River 2013 QC.docx 7/14/14 ii Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT List of Figures Page Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks. ...................................................................... 3 Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes. ........................................... 5 Figure 3. Example of percent measurement error calculations for individual taxa during the identification task. .......................................................... 7 24100 Hudson River 2013 QC.docx 7/14/14 iii Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT List of Tables Page Table 1. Task specific applications of continuous sampling plans for the 2013 Hudson River Ichthyoplankton Laboratory Program ................................. 2 Table 2. Task specific applications of continuous sampling plans for the 2013 Fall Juvenile Survey............................................................................ 4 Table 3. Fraction inspected, percent nonconforming, mean percent measurement error, and average outgoing quality of tasks performed by Normandeau for the 2013 Hudson River Ichthyoplankton Laboratory Program.................................................................................... 10 Table 4. Sample sorting fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program. ................................................ 11 Table 5. Sample sorting percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program. ................................................ 12 Table 6. Sample sorting mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program. ......................................... 13 Table 7. Sample identification fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program. ................................................ 14 Table 8. Sample identification percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program. ......................................... 15 Table 9. Sample identification mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program. ............................... 16 Table 10. Ranking of taxa missed during initial sort and found during sort QC. ........... 18 Table 11. Summary by life stage of the six highest ranked taxa missed during original sort and found during sort QC compared to total count. ................ 19 Table 12. Cumulative net and absolute error rates for commonly encountered taxa in samples selected for QC inspection of identification and counting process. ........................................................................ 20 Table 13. Fraction inspected, percent nonconforming, and average outgoing quality of laboratory tasks performed by Normandeau for the 2013 Fall Juvenile Survey........................................................................... 22 24100 Hudson River 2013 QC.docx 7/14/14 iv Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 1.0 Introduction This quality control report for the laboratory tasks of the 2013 Hudson River Ichthyoplankton Survey and the 2013 Fall Juvenile Survey was prepared for Entergys Indian Point Energy Center by Normandeau Associates Inc. (Normandeau).
To comply with Entergys requirements for valid and reliable data on the Hudson River Ichthyoplankton Laboratory Program and the Fall Juvenile Survey, Normandeau implemented a Quality Assurance Plan that provides a 10% Average Outgoing Quality Limit (AOQL) for all measurement parameters collected. The Quality Assurance Plan consists of two systems: a quality control (QC) system and a quality assurance (QA) system.
The QC system is managed by the program manager and conducted by operational personnel. The system monitors and documents the reliability and validity (accuracy, precision, completeness) of daily operations. The specific features of the QC system are determined by the Quality Assurance Department to insure that all procedures conform to Entergys data requirements. The QA system is managed by Normandeaus Quality Assurance Director and utilizes project independent personnel familiar with the work or activities under evaluation to conduct performance and systems audits. These audits are designed to provide objective evidence that the quality control program and technical requirements, methods, and procedures as outlined in the program Standard Operating Procedures are being implemented. The outcomes of the QA system activities are verification of the effectiveness of the QC system, assignment of corrective actions to resolve nonconforming procedures or data deficiencies, communication of audit results to project and staff managers for follow-up, and objective validation or improvement of project operations.
This report provides a compilation of QC system data verifying the results of the 2013 Hudson River Ichthyoplankton Laboratory Program and 2013 Fall Juvenile Survey activities. Determinations of the fraction inspected, percent nonconforming, and average outgoing quality are presented for both programs. In addition, for the 2013 Hudson River Ichthyoplankton Laboratory Program the results include percent measurement error, a summary of the number of each taxon-life stage found during sorting QC, and cumulative error rates for each taxon-life stage.
24100 Hudson River 2013 QC.docx 7/14/14 1 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 2.0 Methods 2.1 Laboratory Quality Control Procedures 2.1.1 Ichthyoplankton Survey For sorting and identification of samples from the 2013 Hudson River Ichthyoplankton Laboratory Program, Normandeau used a continuous sampling plan designed to provide a 10% AOQL (U.S. Department of Defense 1981). A flow diagram of how the sampling plan was applied is presented in Figure 1. A summary of the sampling plan, tolerances, and QC sample definitions used for each measurement parameter is presented in Table 1. Quality control inspection was applied on a laboratory-wide basis for the sorting task and to each individual processor for the identification task. Quality control samples were selected in a random manner utilizing random number tables. As determined from the sampling plan outlined in Table 1, a specified number of quality control samples were reprocessed by QC inspectors with expertise in the task being inspected. In cases where a sample was subdivided and counted, counts for all subdivisions were combined before calculating percent error for that sample. If the difference between the quality control value and the original value exceeded acceptable tolerances (Table 1), a third measurement could be obtained to verify one of the measurements. If a sample was found to have exceeded acceptable tolerances, all subsequent samples processed by the same technician were subjected to 100% quality control until an appropriate number of consecutive samples (i) were found within tolerance as determined by the continuous sampling plan (Table 1 and Figure 1). The standard operating procedures manual (Normandeau 2014) documents specific QA/QC methods utilized for this program.
Table 1. Task specific applications of continuous sampling plans for the 2013 Hudson River Ichthyoplankton Laboratory Program CSP-1 Laboratory AOQL 10% QC Sample Task i f Sample Tolerance Definition Sorting 8 1/7 +/- 2 if d20 organisms one sample
+/- 10% if >20 organisms Identification 8 1/7 +/- 2 if d20 one sample
+/- 10% if >20 for every taxon in the sample (in identifying, assigning a life stage, or counting any species, errors are cumulative by life stage within each taxon) 24100 Hudson River 2013 QC.docx 7/14/14 2 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks.
In some cases one of the taxonomists (either the original identifier or the QC inspector) was able to determine the taxon or life stage of damaged specimens when the other taxonomist recorded them as unknown life stage, unidentified taxon, or a higher level taxon (genus or family). If a more general taxon or life stage used by one taxonomist included the more specific category used by the other taxonomist, and that was the only reason for a count discrepancy, then that sample was not considered as failing the QC inspection. For example, damaged specimens recorded as Morone sp. by the original identifier and as Striped Bass by the QC inspector were considered to be in agreement because the category Morone sp. includes Striped Bass. In contrast, an original determination of unidentified gobiid would not be acceptable if the QC determination was Striped Bass, because Striped Bass is not included in the family Gobiidae.
2.1.2 Fall Juvenile Survey The Fall Juvenile Survey consisted of two types of collections, referred to as the Fall Shoals Survey (which used Tucker trawls) and the Beach Seine Survey. For laboratory identification and length measurements of young-of-the-year fishes in the 2013 Fall Juvenile Survey, Normandeau used a continuous sampling plan designed to provide a 10% Average Outgoing Quality Limit (U.S. Department of Defense, 1981). A flow diagram of how the 24100 Hudson River 2013 QC.docx 7/14/14 3 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT plan was applied is presented in Figure 2. A summary of the sampling plan, tolerances, and QC sample definitions used for each task is shown in Table 2. QC samples were selected as specified by the appropriate plan in Table 2, using random numbers, and reprocessed by QC inspectors. If the difference between original and QC values exceeded the acceptable tolerance, a third value was obtained as a resolution. The standard operating procedures manual (Normandeau 2013) documents specific QA/QC methods used for the 2013 Fall Juvenile Survey. Young-of-the-year fishes were identified in the laboratory for the first two Fall Shoals Survey river runs (sampling weeks) and the first three Beach Seine Survey river runs. Young-of-the-year fishes were identified in the field starting with Fall Shoals Survey river run 3 and Beach Seine Survey river run 4. The same quality control procedures applied to both field and laboratory identifications. All length measurements of young-of-the-year fishes occurred in the laboratory.
Table 2. Task specific applications of continuous sampling plans for the 2013 Fall Juvenile Survey.
QC QC Sample Task Plan AOQL i f x Tolerance Definition Identification CSP-V 7% 21 1/15 7 +/-10% of total count or +/-2 One taxon individuals when <25 fish
+/-1 mm when <34 mm TL Length CSP-V 7% 30 1/50 10 One fish
+/-3% when >34 mm TL 2.2 Reporting Procedures The 2013 Hudson River Ichthyoplankton Laboratory Program Sort and Identification Quality Control Logs were keyed, verified, and error-checked to produce SAS data sets.
From these data, fraction inspected, percent nonconforming, and percent measurement error (precision) were determined for each river run and for the entire study. For the 2013 Fall Juvenile Survey, QC data were used to determine fraction inspected and percent nonconforming for the entire study (combining all river runs processed in the laboratory for both the Fall Shoals Survey and the Beach Seine Survey).
2.2.1 Fraction Inspected Fraction Inspected Number of Samples Inspected x 100 (Equation 1)
Total Number of Samples River Run: Fraction inspected for a river run (Equation 1) was one hundred times the number of samples inspected divided by the total number of samples analyzed for that river run. For the ichthyoplankton sorting task, the number of samples inspected excludes training QC samples for new sorters, which do not represent the independent performance of the technician, as well as the samples inspected as part of the QC plan. For the identification task, the total number of samples identified excludes empty (no catch) samples, which did not require processing by an identifier.
24100 Hudson River 2013 QC.docx 7/14/14 4 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes.
24100 Hudson River 2013 QC.docx 7/14/14 5 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Entire Study: Fraction inspected for the entire study was one hundred times the number of samples inspected divided by the total number of samples analyzed during the study.
2.2.2 Percent Nonconforming Percent Nonconforming Number of Nonconforming Samples Inspected x 100 (Equation 2)
Number of Samples Inspected River Run: Percent nonconforming for a river run (Equation 2) was one hundred times the number of nonconforming quality control samples found for that river run divided by the total number of quality control samples inspected for that river run.
Entire Study: Percent nonconforming for the entire study was one hundred times the total number of nonconforming quality control samples for the study divided by the total number of quality control samples inspected for the study. The result of this analysis was a determination of the actual incoming quality level of each measurement parameter. (Note that because samples checked by QC found to be defective were rectified during QC, the average outgoing quality of the final data set was better than that indicated by the percent nonconforming.)
2.2.3 Percent Measurement Error Sorting Task Sorting Percent Measurement Error Quality Control Value x 100 (Equation 3)
(Original Value Quality Control Value)
Sample: Percent measurement error for a sorted sample (Equation 3) was one hundred times the quality control value divided by the sum of the original value and the quality control value. If the total count (original value plus quality control value) was less than or equal to 20, and the quality control value (i.e., the number of organisms missed by the sorter and found during sort QC inspection) was no more than two, the percent measurement error for the sorted sample was defined as zero.
River Run: Mean percent measurement error for sorted samples for a river run was the sum of the percent measurement errors for each sample inspected during the river run divided by the total number of samples inspected for the river run.
Entire Study: Mean percent measurement error for sorted samples for the entire study was the sum of the percent measurement errors for each sample inspected during the study divided by the total number of samples inspected for the study. (Note that this method of averaging gives equal weight to each sample, regardless of the number of organisms present).
24100 Hudson River 2013 QC.docx 7/14/14 6 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Identification Task Life Stage Percent Measurement Error (Original Value Quality Control Value) x 100 (Equation 4)
Quality Control Value Life Stage: Percent measurement error for a life stage (Equation 4) was one hundred times the difference between the original value and the quality control value divided by the quality control value. For life stages where the quality control value was 20 or less, if the original and quality control values differed by no more than two organisms, the percent measurement error was defined as zero. For life stages where the quality control value was 20 or less and the original and quality control values differed by more than two organisms, the percent measurement error was calculated utilizing Equation 4. If the quality control value was zero, the percent measurement error was calculated by multiplying the difference between the original and quality control values by 100. This can occasionally result in extremely large percent measurement error values (as much as several hundred percent for a life stage of a taxon in a sample), which are not truly indicative of the actual proportion of specimens misidentified, assigned to the wrong life stage, or miscounted in a sample. If the original count for a life stage was acceptably close to a resolution value but not to the quality control value, the percent measurement error was calculated as described above except that the resolution value was substituted for the quality control value.
Taxon: Percent measurement error for an identified taxon was the sum of the absolute values of percent measurement error for each life stage within the taxon. Refer to Figure 3 for an example of taxon percent measurement error calculations.
Post Yolk-Sac Young-of-Eggs Larvae the-Year Total Taxon 1 Original Value 103 176 25 Quality Control Value 100 194 26
% Measurement Error Life Stage 3.0 -9.3 -3.8 16.1 Taxon 2 Original Value 2 Quality Control Value 1
% Measurement Error Life Stage 0 0 Taxon 3 Original Value 8 Quality Control Value 2
% Measurement Error Life Stage 300 300 Figure 3. Example of percent measurement error calculations for individual taxa during the identification task.
24100 Hudson River 2013 QC.docx 7/14/14 7 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT River Run: Mean percent measurement error for the identification task for a river run was the sum of the percent measurement errors for all taxa inspected during the river run divided by the total number of taxa inspected for the river run. This statistic was computed by averaging taxa rather than samples because even though complete samples were inspected and reworked for identification quality control, the pass/fail criterion was whether any taxon in the sample individually exceeded the 10% tolerance.
Entire Study: Mean percent measurement error for identified taxa for the entire study was the sum of the percent measurement errors for all taxa inspected during the study divided by the total number of taxa inspected for the study.
2.2.4 Average Outgoing Quality At the completion of these studies, the Average Outgoing Quality (AOQ) was calculated for each measurement parameter inspected. Continuous sampling plans were used for all tasks. Continuous sampling plans are devised for processes involving a continuous or nearly continuous flow of products or other entities. For these types of processes, it is extremely difficult to organize units into discrete groups commonly referred to as lots. As a result, inspection must be performed on individual units drawn from a continuous flow of products and a decision made concerning the quality of units produced based on the inspection results. Rectification is performed on any nonconforming unit found during inspection, followed by 100% screening of a number of subsequent units depending on the sampling plan. Average Outgoing Quality for each laboratory task was calculated as a function of the percent nonconforming and the fraction of total units inspected (Stephens 1979). This calculation applies to continuous sampling plans when nonconforming units found are rectified:
p' (1 f )q i AOQ x 100 (Equation 5) f (1 f )q i where p' = Percent nonconforming as a decimal fraction f = Fraction of units inspected. This is a parameter of the sampling plan.
Q = 1-p' = Percent conforming as a decimal fraction I = Clearing interval. This is a parameter of the sampling plan.
Example:
p' = 0.0689 f = 1/7 = 0.1429 q = 1-0.0689 = 0.9311 i = 8 0.0689 (1 0.1429)(0.9311) 8 AOQ x 100 5.32%
0.1429 (1 0.1429)(0.9311) 8 The above equation for calculating AOQ was formulated specifically for CSP-1 sampling plans such as those used for the ichthyoplankton sorting and identification (Table 1). The 24100 Hudson River 2013 QC.docx 7/14/14 8 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT same equation was used to calculate AOQ for young-of-the-year identifications and measurements, which used CSP-V plans (Table 2). When Equation 5 is used for CSP-V plans, the calculated AOQ is conservatively high, because the equation does not take into account the times when the number of consecutive reinspections following a failure is x (which is smaller than i).
2.2.5 Cumulative Error Rates Due to the non-independence of identification errors across taxa and life stages, and to the accumulation of errors within taxa, a relatively high fraction of samples may fail QC inspection even though only a small fraction of organisms are incorrectly identified or counted. In order to present the error frequencies more realistically for particular taxa-life stages, two additional statistics were calculated for each taxon-life stage for the identification/counting process.
Absolute Error Rate =
n n i 1 lIi Qi l / Q i 1 i Equation 6 Net Error Rate =
n n (I
i 1 i Qi ) / Q i 1 i Equation 7 where Ii = initial count for taxon-life stage in sample i Qi = QC count for taxon-life stage in sample i (or the resolution count, if Ii was acceptably close to it but not to the QC count) n = number of samples in the entire study If the sum of Qi for the entire study was zero for the taxon-life stage, then the sum of Qi was set equal to one for the purpose of calculating absolute and net error rate.
The absolute error rate is the approximate fraction of the taxon-life stage that was originally identified or counted incorrectly. This is an estimate of the fraction of erroneous countable items in the uninspected samples.
Net error rate is the approximate relative error in the total counts for the taxon-life stage.
For this index, positive (original count too high) and negative (original count too low) errors cancel each other so that the index reflects the relative net bias to the taxon-life stage abundance.
24100 Hudson River 2013 QC.docx 7/14/14 9 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 3.0 Results 3.1 Ichthyoplankton Laboratory Program The Average Outgoing Quality (AOQ) of the 2013 Hudson River Ichthyoplankton Laboratory Program was 6.41% for the sorting task and 0.52% for the identification task.
These AOQ levels represent the actual or achieved quality for measurement parameters and were within the 10% AOQL requirement of the study. The Average Fraction Inspected (AFI) was 35.81% for sorting and 15.34% for identification (Table 3).
Table 3. Fraction inspected, percent nonconforming, mean percent measurement error, and average outgoing quality of tasks performed by Normandeau for the 2013 Hudson River Ichthyoplankton Laboratory Program.
Mean Percent Fraction Percent Noncon- Measurement Error Task Inspected (%) forming (%) (%) AOQ (%)
Sorting 35.81 8.60 2.97 6.41 Identification 15.34 0.61 0.92 0.52 Sorting and identification tasks were also evaluated on the basis of river runs (sampling weeks). Sorted samples were inspected at a rate of 14.81% to 62.70% for individual river runs (Table 4). Nonconformance for the sorting task among the inspected samples ranged from 0% to 18.33% in the 23 river runs, and averaged 8.60% overall (Table 5). Sorting measurement error was between 0% and 6.10% and averaged 2.97% for the study (Table 6).
For the task of sample identification, 10.0% to 39.51% of samples were inspected from individual river runs (Table 7). Percent nonconforming for the identification task in individual river runs ranged from 0% to 11.11% (Table 8). Measurement error for individual river runs ranged from 0% to 6.25% and overall measurement error was 0.92%
for the identification task of this study (Table 9).
Measurement error results for the identification task are skewed towards high values as a result of the method of computation at the life stage level. In addition, measurement errors are summed over life stages within each taxon, which then amplifies the already skewed life stage values. These data are not indicative of actual measurement error and should only be compared to other measurement error results that are calculated using exactly the same methods. In all cases of failed QC samples, the data were corrected and the QC sample inspection frequency was maintained at 100% for that individual until acceptable results were demonstrated as determined by the QC sampling plan.
24100 Hudson River 2013 QC.docx 7/14/14 10 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 4. Sample sorting fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Total # of Fraction (Beginning Monday) Samples Inspected Samples Sorted Inspected (%)
11 Mar 13 13 74 17.57 18 Mar 13 37 74 50.00 25 Mar 13 34 74 45.95 1 Apr 13 54 126 42.86 8 Apr 13 39 126 30.95 15 Apr 13 36 126 28.57 22 Apr 13 31 135 22.96 29 Apr 13 41 135 30.37 6 May 13 61 135 45.19 13 May 13 79 126 62.70 20 May 13 55 126 43.65 27 May 13 60 126 47.62 3 Jun 13 49 123 39.84 10 Jun 13 58 123 47.15 17 Jun 13 49 123 39.84 24 Jun 13 21 122 17.21 8 Jul 13 24 81 29.63 22 Jul 13 22 75 29.33 5 Aug 13 12 81 14.81 19 Aug 13 42 81 51.85 2 Sep 13 19 81 23.46 16 Sep 13 16 81 19.75 30 Sep 13 20 81 24.69 Study 872 2,435 35.81 24100 Hudson River 2013 QC.docx 7/14/14 11 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 5. Sample sorting percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program.
- of Total # of Sampling Week Noncon- Samples % Non-(Beginning Monday) formities Inspected conformance 11 Mar 13 0 13 0.00 18 Mar 13 1 37 2.70 25 Mar 13 2 34 5.88 1 Apr 13 2 54 3.70 8 Apr 13 3 39 7.69 15 Apr 13 0 36 0.00 22 Apr 13 2 31 6.45 29 Apr 13 5 41 12.20 6 May 13 4 61 6.56 13 May 13 8 79 10.13 20 May 13 4 55 7.27 27 May 13 11 60 18.33 3 Jun 13 7 49 14.29 10 Jun 13 8 58 13.79 17 Jun 13 8 49 16.33 24 Jun 13 2 21 9.52 8 Jul 13 2 24 8.33 22 Jul 13 1 22 4.55 5 Aug 13 1 12 8.33 19 Aug 13 3 42 7.14 2 Sep 13 0 19 0.00 16 Sep 13 0 16 0.00 30 Sep 13 1 20 5.00 Study 75 872 8.60 24100 Hudson River 2013 QC.docx 7/14/14 12 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 6. Sample sorting mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Mean Percent (Beginning Monday) Samples Inspected Measurement Error 11 Mar 13 13 0.00 18 Mar 13 37 0.90 25 Mar 13 34 4.19 1 Apr 13 54 1.54 8 Apr 13 39 2.17 15 Apr 13 36 0.21 22 Apr 13 31 3.87 29 Apr 13 41 3.36 6 May 13 61 2.08 13 May 13 79 3.48 20 May 13 55 2.22 27 May 13 60 6.10 3 Jun 13 49 3.86 10 Jun 13 58 4.44 17 Jun 13 49 4.61 24 Jun 13 21 3.30 8 Jul 13 24 2.07 22 Jul 13 22 2.55 5 Aug 13 12 2.19 19 Aug 13 42 4.40 2 Sep 13 19 0.44 16 Sep 13 16 0.52 30 Sep 13 20 0.69 Study 872 2.97 24100 Hudson River 2013 QC.docx 7/14/14 13 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 7. Sample identification fraction inspected results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Total # of Fraction (Beginning Monday) Samples Inspected Samples Identified Inspected (%)
11 Mar 13 8 53 15.09 18 Mar 13 8 57 14.04 25 Mar 13 10 68 14.71 1 Apr 13 9 66 13.64 8 Apr 13 10 63 15.87 15 Apr 13 8 47 17.02 22 Apr 13 10 83 12.05 29 Apr 13 14 128 10.94 6 May 13 25 134 18.66 13 May 13 18 124 14.52 20 May 13 17 126 13.49 27 May 13 19 126 15.08 3 Jun 13 15 123 12.20 10 Jun 13 18 123 14.63 17 Jun 13 18 123 14.63 24 Jun 13 17 122 13.93 8 Jul 13 14 81 17.28 22 Jul 13 12 75 16.00 5 Aug 13 8 80 10.00 19 Aug 13 32 81 39.51 2 Sep 13 13 81 16.05 16 Sep 13 10 81 12.35 30 Sep 13 13 80 16.25 Study 326 2,125 15.34 24100 Hudson River 2013 QC.docx 7/14/14 14 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 8. Sample identification percent nonconformance results, 2013 Hudson River Ichthyoplankton Laboratory Program.
- of Total # of Sampling Week Noncon- Samples % Non-(Beginning Monday) formities Inspected conformance 11 Mar 13 0 8 0.00 18 Mar 13 0 8 0.00 25 Mar 13 0 10 0.00 1 Apr 13 0 9 0.00 8 Apr 13 0 10 0.00 15 Apr 13 0 8 0.00 22 Apr 13 0 10 0.00 29 Apr 13 0 14 0.00 6 May 13 0 25 0.00 13 May 13 0 18 0.00 20 May 13 0 17 0.00 27 May 13 0 19 0.00 3 Jun 13 0 15 0.00 10 Jun 13 2 18 11.11 17 Jun 13 0 18 0.00 24 Jun 13 0 17 0.00 8 Jul 13 0 14 0.00 22 Jul 13 0 12 0.00 5 Aug 13 0 8 0.00 19 Aug 13 0 32 0.00 2 Sep 13 0 13 0.00 16 Sep 13 0 10 0.00 30 Sep 13 0 13 0.00 Study 2 326 0.61 24100 Hudson River 2013 QC.docx 7/14/14 15 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 9. Sample identification mean percent measurement error results, 2013 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Mean Percent Number of (Beginning Monday) Samples Inspected Measurement Error Taxa Inspected 11 Mar 13 8 0.36 17 18 Mar 13 8 0.19 25 25 Mar 13 10 0.00 19 1 Apr 13 9 0.00 29 8 Apr 13 10 0.00 20 15 Apr 13 8 0.00 16 22 Apr 13 10 0.23 19 29 Apr 13 14 0.09 34 6 May 13 25 0.22 65 13 May 13 18 0.31 58 20 May 13 17 0.71 70 27 May 13 19 0.79 98 3 Jun 13 15 0.80 79 10 Jun 13 18 6.25 106 17 Jun 13 18 0.48 106 24 Jun 13 17 0.38 87 8 Jul 13 14 0.41 84 22 Jul 13 12 0.67 46 5 Aug 13 8 0.06 44 19 Aug 13 32 0.25 123 2 Sep 13 13 0.51 31 16 Sep 13 10 0.56 21 30 Sep 13 13 0.27 22 Study 326 0.92 1,219 24100 Hudson River 2013 QC.docx 7/14/14 16 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Additional organisms found during the sort QC were identified independently to determine the frequency of species and life stages missed during the initial sort. Six taxa accounted for 91% of the additional organisms found during sort QC: Striped Bass, White Perch, herring family, Bay Anchovy, Morone sp., and Freshwater Drum (Table 10). For these six taxa, the additional number found in the sort QC amounted to less than 2% of the total found during sample processing (Table 11). For most taxa-life stages the percentage missed by the original sorter was well under 5%.
The life stage most commonly missed by sorters was post yolk-sac larvae for White Perch, clupeids, and Morone sp.; eggs for Bay Anchovy and Freshwater Drum; and yolk-sac larvae for Striped Bass (Table 11). The life stage most frequently missed by sorters was usually the most abundant life stage.
During the identification process, absolute error rates for individual life stages of commonly encountered taxa ranged were usually less than 0.02. Generally, only those taxa-life stages with low total counts had absolute error rates above 0.05 (Table 12).
Net error rates were substantially lower than the absolute error rates in most cases, demonstrating that errors often tended to cancel each other out. This was especially noticeable for many of the most abundant taxa-life stages, such as Bay Anchovy eggs and Striped Bass yolk-sac larvae and post yolk-sac larvae.
3.2 Fall Juvenile Survey Results of the laboratory quality control program for the 2013 Fall Juvenile Survey (consisting of the Beach Seine Survey and the Fall Shoals Survey) were summarized by the same methods as the QC results for the 2013 Hudson River Ichthyoplankton Laboratory Program and are presented in Table 13.
There were 516 and 911 young-of-the-year fish identification records made in the laboratory for the Fall Shoals and Beach Seine surveys respectively and 4,084 and 3,536 young-of-the-year fish length measurement records were made for the Fall Shoals and Beach Seine surveys respectively.
24100 Hudson River 2013 QC.docx 7/14/14 17 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 10. Ranking of taxa missed during initial sort and found during sort QC.
Number of Organisms Taxon Found in Sort QC Percent Striped Bass 2,090 32.87 White Perch 1,888 29.69 Herring family 1,027 16.15 Bay Anchovy 485 7.63 Morone species 159 2.50 Freshwater Drum 158 2.48 Unidentified 108 1.70 Winter Flounder 90 1.42 Tautog 85 1.34 Cunner 74 1.16 Goby family 51 0.80 Gizzard Shad 32 0.50 Fourbeard Rockling 23 0.36 Atlantic Tomcod 16 0.25 Hogchoker 13 0.20 Atlantic Menhaden 12 0.19 Carp and minnow family 8 0.13 Windowpane 8 0.13 Weakfish 7 0.11 Atlantic Croaker 6 0.09 Common Carp 4 0.06 Fourspot Flounder 3 0.05 Inland Silverside 3 0.05 American Shad 2 0.03 Yellow Perch 2 0.03 Atlantic Cod 1 0.02 Butterfish 1 0.02 Menidia species 1 0.02 Northern Pipefish 1 0.02 Pollock 1 0.02 Total 6,359 100.00 24100 Hudson River 2013 QC.docx 7/14/14 18 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 11. Summary by life stage of the six highest ranked taxa missed during original sort and found during sort QC compared to total count.
Total Percent in Each Percent of Total Organisms Taxon Life Stage Number Stage Found Founda Bay Anchovy Eggs 391 80.62 1.06 36,727 Yolk-sac larvae 7 1.44 4.32 162 Post yolk-sac larvae 87 17.94 0.47 18,670 Young-of-the-year 0 0.00 0.00 4,628 Unidentified 0 0.00 0.00 16 Freshwater Eggs 114 72.15 2.41 4,737 Drum Yolk-sac larvae 30 18.99 1.25 2,406 Post yolk-sac larvae 13 8.23 0.91 1,428 Young-of-the-year 0 0.00 0.00 1 Unidentified 1 0.63 5.88 17 Herring family Eggs 193 18.79 1.63 11,813 Yolk-sac larvae 173 16.85 5.02 3,449 Post yolk-sac larvae 653 63.58 2.80 23,291 Unidentified 8 0.78 7.62 105 Morone species Yolk-sac larvae 13 8.18 8.13 160 Post yolk-sac larvae 91 57.23 2.47 3,677 Unidentified 55 34.59 2.48 2,219 Striped Bass Eggs 649 31.05 1.41 45,915 Yolk-sac larvae 786 37.61 1.71 45,927 Post yolk-sac larvae 639 30.57 0.79 81,178 Young-of-the-year 0 0.00 0.00 183 Unidentified 16 0.77 3.90 410 White Perch Eggs 344 18.22 3.97 8,664 Yolk-sac larvae 295 15.63 3.22 9,163 Post yolk-sac larvae 1,237 65.52 2.79 44,295 Young-of-the-year 0 0.00 0.00 42 Unidentified 12 0.64 5.36 224 a Includes both original count and additional organisms found during sort QC.
24100 Hudson River 2013 QC.docx 7/14/14 19 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 12. Cumulative net and absolute error rates for commonly encountered taxa in samples selected for QC inspection of identification and counting process.
Total Net Absolute Taxon Stage Count Error Error N Atlantic Menhaden Eggs 124 -0.00806 0.00806 7 Post Yolk Sac Larvae 879 -0.00228 0.00455 83 Atlantic Tomcod Young of the Year 168 -0.02381 0.03571 44 Yolk-Sac Larvae 3 0.00000 0.00000 2 Post Yolk Sac Larvae 138 0.00000 0.00000 28 Bay Anchovy Young of the Year 222 0.00000 0.00000 48 Unidentified 0 1.00000 1.00000 1 Eggs 4,569 -0.00263 0.00875 57 Yolk-Sac Larvae 73 0.04110 0.04110 4 Post Yolk Sac Larvae 3,963 -0.00606 0.01060 101 Young of the Year 840 0.01429 0.01905 61 Freshwater Drum Unidentified 2 0.00000 0.00000 1 Eggs 156 0.00000 0.00000 5 Yolk-Sac Larvae 204 0.00980 0.03922 20 Post Yolk Sac Larvae 255 -0.05098 0.05098 24 Young of the Year 1 0.00000 0.00000 1 Goby family Unidentified 1 0.00000 0.00000 1 Post Yolk Sac Larvae 1,007 -0.00695 0.01291 68 Herring family Unidentified 9 0.11111 0.11111 5 Eggs 1,340 -0.00896 0.00896 16 Yolk-Sac Larvae 310 -0.00323 0.01613 33 Post Yolk Sac Larvae 2,558 -0.00156 0.01407 101 Hogchoker Eggs 692 -0.00289 0.00867 16 Yolk-Sac Larvae 30 -0.03333 0.03333 5 Post Yolk Sac Larvae 4 0.25000 0.25000 4 Young Of The Year 5 0.00000 0.00000 3 Morone species Unidentified 263 -0.00760 0.09125 28 Yolk-Sac Larvae 6 0.00000 0.33333 4 Post Yolk Sac Larvae 646 0.00464 0.03870 41 Striped Bass Unidentified 28 0.10714 0.17857 7 Eggs 4,207 0.00071 0.00071 59 Yolk-Sac Larvae 4,889 0.00900 0.02168 78 Post Yolk Sac Larvae 10,275 -0.00467 0.01071 114 Young of the Year 41 0.00000 0.00000 14 Weakfish Eggs 647 0.00155 0.01700 9 Post Yolk Sac Larvae 30 0.00000 0.00000 14 Young Of The Year 22 0.00000 0.00000 10 (continued) 24100 Hudson River 2013 QC.docx 7/14/14 20 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 12. (Continued)
Total Net Absolute Taxon Stage Count Error Error N White Perch Unidentified 8 0.12500 0.12500 3 Eggs 936 -0.00641 0.00641 27 Yolk-Sac Larvae 586 0.02218 0.03242 49 Post Yolk Sac Larvae 4,015 -0.00897 0.02092 88 Young of the Year 3 0.00000 0.00000 2 Winter Flounder Unidentified 2 0.00000 0.00000 1 Yolk-Sac Larvae 16 0.00000 0.00000 8 Post Yolk Sac Larvae 170 0.00000 0.01176 27 Young of the Year 13 0.00000 0.00000 5 24100 Hudson River 2013 QC.docx 7/14/14 21 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT Table 13. Fraction inspected, percent nonconforming, and average outgoing quality of laboratory tasks performed by Normandeau for the 2013 Fall Juvenile Survey.
Average Fraction Percent Average Outgoing Task Inspected (%) Nonconforming (%) Quality (%)
Identification 6.66 0.00 0.00 Measurement 2.03 0.00 0.00 24100 Hudson River 2013 QC.docx 7/14/14 22 Normandeau Associates, Inc.
2013 HUDSON RIVER QC REPORT 4.0 Literature Cited Normandeau Associates Inc. 2013. 2013 Hudson River fall juvenile and beach seine surveys standard operating procedures. June 2013. Prepared for Indian Point Energy Center.
__________. 2014. Quality assurance plan for the Hudson River ichthyoplankton laboratory program. Rev. 2, April 2014. Prepared for Entergy Nuclear Operations, Inc.
Stephens, K.S. 1979. Volume 2: How to perform continuous sampling (CSP). American Society for Quality Control. 70 pp.
U.S. Department of Defense. 1981. Military standard. Single- and multi-level continuous sampling procedures and table for inspection by attributes. MIL-STD-1235B.
24100 Hudson River 2013 QC.docx 7/14/14 23 Normandeau Associates, Inc.
Quality Control Report for the 2014 Hudson River Ichthyoplankton Laboratory Program and 2014 Fall Juvenile Survey Prepared for:
Indian Point Energy Center 450 Broadway, Suite 1 Buchanan, New York 10511 Submitted:
October 2015 Prepared by:
Normandeau Associates, Inc.
25 Nashua Road Bedford, NH 03110 www.normandeau.com
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table of Contents Page
1.0 INTRODUCTION
................................................................................ 1 2.0 METHODS....................................................................................... 2 2.1 LABORATORY QUALITY CONTROL PROCEDURES ............................................. 2 2.1.1 Ichthyoplankton Survey........................................................ 2 2.1.2 Fall Juvenile and Beach Seine Surveys ...................................... 3 2.2 REPORTING PROCEDURES ................................................................... 5 2.2.1 Percent Inspected .............................................................. 5 2.2.2 Percent Nonconforming ....................................................... 6 2.2.3 Percent Error.................................................................... 6 2.2.4 Average Outgoing Quality ..................................................... 8 2.2.5 Cumulative Error Rates ........................................................ 9 3.0 RESULTS .......................................................................................10 3.1 ICHTHYOPLANKTON LABORATORY PROGRAM ............................................... 10 3.2 FALL JUVENILE SURVEY .................................................................... 18 4.0 LITERATURE CITED ..........................................................................23 24200 Hudson River 2014 QC FINAL.docx 11/18/15 ii Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM List of Figures Page Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks. ...................................................................... 3 Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes. ........................................... 4 24200 Hudson River 2014 QC FINAL.docx 11/18/15 iii Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM List of Tables Page Table 1. Task specific applications of continuous sampling plans for the 2014 Hudson River Ichthyoplankton Laboratory Program. ................................ 2 Table 2. Task specific applications of continuous sampling plans for the 2014 Fall Juvenile Survey............................................................................ 5 Table 3. Examples of percent error calculations for individual taxa and samples during both the identification and sorting tasks. .................................... 8 Table 4. Average percent inspected, percent nonconforming, mean percent error, and average outgoing quality of tasks performed by Normandeau for the 2014 Hudson River Ichthyoplankton Laboratory Program................. 11 Table 5. Sample sorting percent inspected results, 2014 Hudson River Ichthyoplankton Laboratory Program. ................................................ 12 Table 6. Percent nonconformance of samples inspected during the sorting procedure, 2014 Hudson River Ichthyoplankton Laboratory Program............ 13 Table 7. Mean percent error for the sorting procedure by sampling week, 2014 Hudson River Ichthyoplankton Laboratory Program. ............................... 14 Table 8. Percent of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program. ............................... 15 Table 9. Sample identification percent nonconformance results, 2014 Hudson River Ichthyoplankton Laboratory Program. ......................................... 16 Table 10. Mean percent error of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program............ 17 Table 11. Ranking of ichthyoplankton taxa missed during initial sort and found during sort QC, 2014 Hudson River Ichthyoplankton Laboratory Program....... 19 Table 12. Summary of ichthyoplankton by life stage for the six highest ranked taxa missed during original sort and found during sort QC compared to total count, 2014 Hudson River Ichthyoplankton Laboratory Program........... 20 Table 13. Cumulative net and absolute error rates for commonly encountered taxa in ichthyoplankton samples selected for QC inspection of identification process, 2014 Hudson River Ichthyoplankton Laboratory Program.................................................................................... 21 Table 14. Percent inspected, percent nonconforming, and average outgoing quality of laboratory tasks for the 2014 Fall Juvenile and Beach Seine Surveys..................................................................................... 22 24200 Hudson River 2014 QC FINAL.docx 11/18/15 iv Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM 1.0 Introduction This Quality Control (QC) report for the laboratory tasks of the 2014 Hudson River Ichthyoplankton Survey, 2014 Fall Juvenile Survey and, 2014 Beach Seine Survey was prepared for Entergys Indian Point Energy Center by Normandeau Associates Inc.
(Normandeau).
To comply with Entergys requirements for valid and reliable data on the Hudson River Ichthyoplankton Laboratory Program and the Fall Juvenile and Beach Seine Surveys, Normandeau implemented a Quality Assurance Plan that provides a 10% Average Outgoing Quality Limit (AOQL) for all measurement parameters collected. The Quality Assurance Plan consists of two systems: a Quality Control (QC) system and a Quality Assurance (QA) system. The QC system is managed by the program manager and conducted by operational personnel. The system monitors and documents the reliability and validity (accuracy, precision, completeness) of daily operations. The specific features of the QC system are determined by the Quality Assurance Department to insure that all procedures conform to Entergys data requirements. The QA system is managed by Normandeaus Quality Assurance Director and utilizes project independent personnel familiar with the work or activities under evaluation to conduct performance and systems audits. These audits are designed to provide objective evidence that the QC program and technical requirements, methods, and procedures as outlined in the program Standard Operating Procedures are being implemented. The outcomes of the QA system activities are verification of the effectiveness of the QC system, assignment of corrective actions to resolve nonconforming procedures or data deficiencies, communication of audit results to project and staff managers for follow-up, and objective validation or improvement of project operations.
This report provides a compilation of QC system data verifying the results of the 2014 Hudson River Ichthyoplankton Laboratory Program and 2014 Fall Juvenile and Beach Seine Surveys activities. Determinations of the percent inspected, percent nonconforming, and average outgoing quality are presented for both programs. In addition, for the 2014 Hudson River Ichthyoplankton Laboratory Program the results include percent error, a summary of the number of each taxon-life stage found during sorting QC, and cumulative error rates for each taxon-life stage.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 1 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM 2.0 Methods 2.1 Laboratory Quality Control Procedures 2.1.1 Ichthyoplankton Survey The Ichthyoplankton Survey collected samples by a 1-m wide epibenthic sled and a 1-m2 Tucker trawl, each with a 500 m mesh net. For sorting and identification of samples from the 2014 Hudson River Ichthyoplankton Laboratory Program, Normandeau used a continuous sampling plan designed to provide a 10% AOQL (U.S. Department of Defense 1981). A flow diagram of how the sampling plan was applied is presented in Figure 1. A summary of the sampling plan, tolerances, and QC sample definitions used for each measurement parameter is presented in Table 1. QC inspection was applied on a laboratory-wide basis for the sorting task and to each individual processor for the identification task. Samples were selected in a random manner utilizing random number tables. As determined from the sampling plan outlined in Table 1, a specified number of samples were reprocessed by QC inspectors with expertise in the task being inspected. In cases where a sample was subdivided and counted, counts for all subdivisions were combined before calculating percent error for that sample. If the difference between the QC value and the original value exceeded acceptable tolerances (Table 1), a third measurement could be obtained to verify one of the measurements. If a sample from the block of 1 out of 7 samples (f) was found to have exceeded acceptable tolerances, all samples within that block were subjected to review to ensure that the quality meets acceptable tolerance levels.
Furthermore, all subsequent samples processed by the same technician were subjected to 100% QC until 8 consecutive samples (i) were found within acceptable tolerance limits as determined by the continuous sampling plan (Table 1 and Figure 1). The Quality Assurance plan (Normandeau 2014) documents specific QA/QC methods utilized for this program.
Table 1. Task specific applications of continuous sampling plans for the 2014 Hudson River Ichthyoplankton Laboratory Program.
CSP-1 QC Laboratory AOQL 10% Sample Task i* f** Sample Tolerance Definition Sorting 8 1/7 +/- 2 organisms if d20 organisms one
+/- 10% if >20 organisms sample Identification 8 1/7 +/- 2 organisms if d20 organisms one
+/- 10% if >20 organisms sample for every taxon in the sample (in identifying, assigning a life stage, or counting any species, errors are cumulative by life stage within each taxon)
- i = number of samples within the 100% sample inspection training period
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Figure 1. Quality control inspection plan for ichthyoplankton sorting and identification tasks.
In some cases, one of the taxonomists (either the original identifier or the QC inspector) was able to determine the taxon or life stage of damaged specimens when the other taxonomist recorded them as unknown life stage, unidentified taxon, or a higher level taxon (genus or family). If a more general taxon or life stage used by one taxonomist included the more specific category used by the other taxonomist and was only separated by one consecutive taxonomic level such as Family-Genus or Genus-Species, and that was the only reason for a count discrepancy, then that sample was not considered failing the QC inspection. For example, damaged specimens recorded as Morone sp. (Genus level) by the original identifier and as Morone saxatilis (Species level) by the QC inspector were considered to be in agreement because the category Morone sp. includes Morone saxatilis. In contrast, an original determination of unidentified Gobiidae (Family) would not be acceptable if the QC determination was Morone saxatilis (Species), because 1) taxonomically, Morone saxatilis is not in the family Gobiidae and 2) M. saxatilis is more than one taxonomic level separated from the family Gobiidae.
2.1.2 Fall Juvenile and Beach Seine Surveys The Fall Juvenile Survey, also historically referred as Fall Shoals Survey, consisted of collections from two gear types, a 3-m wide x 0.9 m high beam trawl (3.2 cm stretch mesh codend and 1.3 cm stretch mesh liner) and 1-m2 Tucker trawl (2.0 mm mesh net). Juvenile fish are also sampled along the shore by the Beach Seine Survey, which uses a 2.4-m high 24200 Hudson River 2014 QC FINAL.docx 11/18/15 3 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM seine with two 12-m long wings of 1.0-cm bar mesh and 0.5-cm bar mesh bag. For laboratory identification and length measurements of young-of-the-year fishes in the 2014 Fall Juvenile Survey, Normandeau used a continuous sampling plan designed to provide a 10% Average Outgoing Quality Limit (U.S. Department of Defense, 1981). A flow diagram of how the plan was applied is presented in Figure 2.
Figure 2. Quality control inspection plan for identification and length measurement of young-of-the-year fishes.
A summary of the sampling plan, tolerances, and QC sample definitions used for each task is shown in Table 2. QC samples were selected as specified by the appropriate plan in Table 2, using random numbers, and reprocessed by QC inspectors. If the difference between original and QC values exceeded the acceptable tolerance, a third value was obtained as a resolution. The standard operating procedures manual (Normandeau 2013) documents specific QA/QC methods used for the 2014 Fall Juvenile and Beach Seine Surveys. Young-24200 Hudson River 2014 QC FINAL.docx 11/18/15 4 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM of-the-year fishes were identified in the laboratory for the first two Fall Juvenile Survey River Runs (sampling weeks) and the first three Beach Seine Survey River Runs. Young-of-the-year fishes were identified in the field starting with Fall Juvenile Survey River Run 3 and Beach Seine Survey River Run 4. The same QC procedures applied to both field and laboratory identifications. All length measurements of young-of-the-year fishes occurred in the laboratory.
Table 2. Task specific applications of continuous sampling plans for the 2014 Fall Juvenile Survey.
QC QC Sample Task Plan AOQL i* f** x Tolerance Definition Identification CSP-V 7% 21 1/15 7 +/-10% of total count or +/-2 One taxon individuals when <25 fish
+/-1 mm when <34 mm TL Length CSP-V 7% 30 1/50 10 One fish
+/-3% when >34 mm TL
- i = number of samples within the 100% sample inspection training period
- f = fraction of samples randomly selected for QC 2.2 Reporting Procedures The 2014 Hudson River Ichthyoplankton Laboratory Program Sort and Identification QC Logs were keyed, verified, and error-checked to produce SAS data sets. From these data, percent inspected, percent nonconforming, and percent error (precision) were determined for each River Run and for the entire study. For the 2014 Fall Juvenile Survey, QC data were used to determine percent inspected and percent nonconforming for the entire study (combining all River Runs processed in the laboratory for both the Fall Juvenile Survey and the Beach Seine Survey).
2.2.1 Percent Inspected Percent Inspected Number of Samples Inspected x 100 (Equation 1)
Total Number of Samples Analyzed River Run: Percent inspected (%) for a River Run (Equation 1) was the number of samples inspected divided by the total number of samples analyzed for that River Run times one hundred. For the ichthyoplankton sorting task, the number of samples inspected excludes training QC samples for new sorters, which do not represent the independent performance of the technician, as well as the samples inspected as part of the QC plan. For the identification task, the total number of samples identified excludes empty (no catch) samples, which did not require processing by an identifier.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 5 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Entire Study: Percent inspected for the entire study was the number of samples inspected divided by the total number of samples analyzed times one hundred during the study.
2.2.2 Percent Nonconforming Percent Nonconforming Number of Nonconforming Quality Control Samples Inspected x 100 Total Number of Quality Control Samples Inspected (Equation 2)
River Run: Percent nonconforming for a River Run (Equation 2) was the number of nonconforming QC samples found for that River Run divided by the total number of QC samples inspected times one hundred for that River Run.
Entire Study: Percent nonconforming for the entire study was the total number of nonconforming QC samples for the study divided by the total number of QC samples inspected times one hundred for the study. The result of this analysis was a determination of the actual incoming quality level of each measurement parameter. (Note that because samples checked by QC found to be defective were rectified during QC, the average outgoing quality of the final data set was better than that indicated by the percent nonconforming.)
2.2.3 Percent Error Sorting Task Sorting Percent Error Quality Control Count x 100 (Equation 3)
(Original Count Quality Control Count )
Sample: Percent error for a sorted sample (Equation 3) was the QC count divided by the sum of the original count and the QC count (adjusted total) times one hundred. The denominator is an adjusted total because the sorting procedure is the initial process by which all countable organisms are removed from the field sample. Therefore, if QC personnel have found additional organisms that the original sorter missed, then the total count must reflect this addition to the sample total. The sorting task is not life stage or taxon specific, it is the total count of all countable organisms sorted. Refer to Table 3 for an example of this sample percent error calculation.
River Run: Mean percent error for sorted samples for a River Run was the sum of the percent errors for each sample inspected during the River Run divided by the total number of samples inspected for the River Run times one hundred.
Entire Study: Mean percent error for sorted samples for the entire study was the sum of the percent measurement errors for each sample inspected during the study divided by the total number of samples inspected for the study times one hundred. (Note that this method of 24200 Hudson River 2014 QC FINAL.docx 11/18/15 6 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM averaging gives equal weight to each sample, regardless of the number of organisms present).
Identification Task Life Stage Percent Error (Original Count Quality Control Count )
x 100 (Equation 4)
Quality Control Count Life Stage: Percent error for a life stage (Equation 4) was the difference between the original life stage count and the QC life stage count divided by the QC life stage count times one hundred.
Exceptions to Equation 4:
- 1) For life stages where the QC count was 20 individuals or less and the original and QC counts differed by no more than two organisms, the percent error was defined as
+/- 1 or 2. For life stages where the QC value was 20 or less and the original and QC values differed by more than two organisms, the percent error was calculated utilizing Equation 4. Refer to Table 3 for an example of this life stage percent error calculation exception.
- 2) If the QC count was zero, the percent error will be 100%. This error was calculated by dividing the difference between the original count and the QC count by the original count times one hundred. Refer to Table 3 for an example of this QC=0 percent error calculation exception.
If the sample fails then a resolution by a third party must occur. Percent error will then be recalculated for the QC counts and the original counts but by using Equation 4 but using the Resolution Counts as the denominator. Then, if the original count for a life stage was acceptably close (within 10%) to a resolution value but not to the QC value, the original counts are not to be changed and the samples passes. If the original counts are not within 10% from the resolution count then the sample fails and the original data counts are replaced by the QC counts. During the rare occurrence that the original values are greater than 10% from the resolution and the QC counts, then the sample will be reanalyzed by all three people and the identifiers sample processing will not continue until agreement can be reached on the identification of the sample.
Taxon: Percent error for an identified taxon was the sum of the absolute values of percent error for each life stage within the taxon. Refer to Table 3 for an example of taxon percent error calculations.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 7 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 3. Examples of percent error calculations for individual taxa and samples during both the identification and sorting tasks.
Yolk-Sac Post Yolk-Eggs Larvae Sac Larvae Total Sorting Task Example:
Sample Error Original Count 250 52 351 653 Quality Control Count 3 0 5 8
% Error by Sample 1.2%
Identification Task Examples:
Taxon 1 Original Count 103 186 31 Quality Control Count 100 194 26
% Error by Life Stage 3.0% 4.1% 5.2% 12.3%
Taxon 2 Original Count 15 Quality Control Count 13
% Error by Life Stage +/- 2 +/- 2 Taxon 3 Original Count 8 Quality Control Count 2
% Error by Life Stage 300% 300%
Taxon 4 Original Count 25 20 2 Quality Control Count 25 0 22
% Error by Life Stage 100% 10% 110%
River Run: Mean percent error for the identification task for a River Run was the sum of the percent measurement errors for all taxa inspected during the River Run divided by the total number of taxa inspected for the River Run. This statistic was computed by averaging taxa rather than samples because even though complete samples were inspected and reworked for identification QC, the pass/fail criterion was whether any taxon in the sample individually exceeded the 10% tolerance.
Entire Study: Mean percent error for identified taxa for the entire study was the sum of the percent errors for all taxa inspected during the study divided by the total number of taxa inspected for the study.
2.2.4 Average Outgoing Quality At the completion of these studies, the Average Outgoing Quality (AOQ) was calculated for each measurement parameter inspected. Continuous sampling plans were used for all tasks. Continuous sampling plans are devised for processes involving a continuous or nearly continuous flow of products or other entities. For these types of processes, it is extremely difficult to organize units into discrete groups commonly referred to as lots. As a 24200 Hudson River 2014 QC FINAL.docx 11/18/15 8 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM result, inspection must be performed on individual units drawn from a continuous flow of products and a decision made concerning the quality of units produced based on the inspection results. Rectification is performed on any nonconforming unit found during inspection, followed by 100% screening of a number of subsequent units depending on the sampling plan. Average Outgoing Quality for each laboratory task was calculated as a function of the percent nonconforming and the fraction of total units inspected (Stephens 1979). This calculation applies to continuous sampling plans when nonconforming units found are rectified:
p' (1 f )q i AOQ x 100 (Equation 5) f (1 f )q i where p' = Percent nonconforming as a decimal fraction f = Proportion of units inspected. This is a parameter of the sampling plan.
q = 1-p' = Percent conforming as a decimal fraction i = Clearing interval. This is a parameter of the sampling plan.
Example:
p' = 0.0689 f = 1/7 = 0.1429 q = 1-0.0689 = 0.9311 i = 8 0.0689 (1 0.1429)(0.9311) 8 AOQ x 100 5.32%
0.1429 (1 0.1429)(0.9311) 8 The above equation for calculating AOQ was formulated specifically for CSP-1 sampling plans such as those used for the ichthyoplankton sorting and identification (Table 1). The same equation was used to calculate AOQ for young-of-the-year identifications and measurements, which used CSP-V plans (Table 2). When Equation 5 is used for CSP-V plans, the calculated AOQ is conservatively high, because the equation does not take into account the times when the number of consecutive re-inspections following a failure is x (which is smaller than i).
2.2.5 Cumulative Error Rates Due to the non-independence of identification errors across taxa and life stages, and to the accumulation of errors within taxa, a relatively high fraction of samples may fail QC inspection even though only a small fraction of organisms are incorrectly identified or counted. In order to present the error frequencies more realistically for particular taxa-life stages, two additional statistics were calculated for each taxon-life stage for the identification/counting process.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 9 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Absolute Error Rate =
n n i 1 lIi Qi l / Q i 1 i Equation 6 Net Error Rate =
n n i 1 (I i Q i ) / Q i 1 i Equation 7 where Ii = initial count for taxon-life stage in sample i Qi = QC count for taxon-life stage in sample i (or the resolution count, if Ii was acceptably close to it but not to the QC count) n = number of samples in the entire study If the sum of Qi for the entire study was zero for the taxon-life stage, then the sum of Qi was set equal to one for the purpose of calculating absolute and net error rate.
The absolute error rate is the approximate fraction of the taxon-life stage that was originally identified or counted incorrectly. This is an estimate of the fraction of erroneous countable items in the uninspected samples.
Net error rate is the approximate relative error in the total counts for the taxon-life stage.
For this index, positive (original count too high) and negative (original count too low) errors cancel each other so that the index reflects the relative net bias to the taxon-life stage abundance.
3.0 Results 3.1 Ichthyoplankton Laboratory Program The Average Outgoing Quality (AOQ) of the 2014 Hudson River Ichthyoplankton Laboratory Program was 4.41% for the sorting task and 0.58% for the identification task.
These AOQ levels represent the actual or achieved quality for measurement parameters and were within the 10% AOQL requirement of the study. The average of the Percent Inspected among River Runs was 27.60% for the sorting task and 15.69% for the identification task (Table 4).
24200 Hudson River 2014 QC FINAL.docx 11/18/15 10 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 4. Average percent inspected, percent nonconforming, mean percent error, and average outgoing quality (AOQ) of tasks performed by Normandeau for the 2014 Hudson River Ichthyoplankton Laboratory Program.
Percent Percent Inspected Nonconforming Mean Percent Task (%) (%) Error (%) AOQ (%)
Sorting 27.60 5.57 1.95 4.41 Identification 15.69 0.68 4.65 0.58 Sorting and identification tasks were also evaluated on the basis of River Runs (sampling weeks). Sorted samples were inspected at a rate of 11.11% to 45.24% for individual River Runs (Table 5). Nonconformance for the sorting task among the inspected samples ranged from 0% to 21.05% in the 20 River Runs (Table 6). Sorting error ranged between 0.14% and 6.79% (Table 7). For the task of sample identification, 11.54% to 32.10% of samples were inspected from individual River Runs (Table 8). Percent nonconforming for the identification task in individual River Runs ranged from 0% to 3.85% (Table 9). Percent error for individual River Runs ranged from 0% to 30.28% (Table 10).
Percent error results for the identification task are skewed towards high values as a result of the method of computation at the life stage level. In addition, errors are summed over life stages within each taxon, which then amplifies the already skewed life stage values. These data are not indicative of actual error and should only be compared to other error results that are calculated using exactly the same methods. In all cases of failed QC samples, the data were corrected and the QC sample inspection frequency was maintained at 100% for that individual until acceptable results were demonstrated as determined by the QC sampling plan.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 11 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 5. Sample sorting percent inspected results, 2014 Hudson River Ichthyoplankton Laboratory Program.
Sampling Week Total # of Total # of Percent (Beginning Monday) Samples Inspected Samples Sorted Inspected (%)
31 Mar 14 53 126 42.06 07 Apr 14 52 126 41.27 14 Apr 14 41 126 32.54 21 Apr 14 29 135 21.48 28 Apr 14 23 135 17.04 05 May 14 30 135 22.22 12 May 14 38 126 30.16 19 May 14 57 126 45.24 26 May 14 52 126 41.27 02 Jun 14 43 122 35.25 09 Jun 14 35 123 28.46 16 Jun 14 32 123 26.02 23 Jun 14 27 123 21.95 07 Jul 14 12 81 14.81 21 Jul 14 9 75 12.00 04 Aug 14 12 78 15.38 18 Aug 14 9 81 11.11 01 Sep 14 27 81 33.33 15 Sep 14 11 81 13.58 29 Sep 14 18 81 22.22 Study 610 2210 27.60 24200 Hudson River 2014 QC FINAL.docx 11/18/15 12 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 6. Percent nonconformance of samples inspected during the sorting procedure, 2014 Hudson River Ichthyoplankton Laboratory Program.
Sampling Total No. of Week Number of Samples Percent (Beginning Monday) Nonconformities Inspected Nonconformance 31 Mar 14 2 53 3.77 07 Apr 14 1 52 1.92 14 Apr 14 1 41 2.44 21 Apr 14 0 29 0.00 28 Apr 14 0 23 0.00 05 May 14 2 30 6.67 12 May 14 3 38 7.89 19 May 14 12 57 21.05 26 May 14 1 52 1.92 02 Jun 14 2 43 4.65 09 Jun 14 5 35 14.29 16 Jun 14 0 32 0.00 23 Jun 14 2 27 7.41 07 Jul 14 0 12 0.00 21 Jul 14 0 9 0.00 04 Aug 14 0 12 0.00 18 Aug 14 0 9 0.00 01 Sep 14 3 27 11.11 15 Sep 14 0 11 0.00 29 Sep 14 0 18 0.00 Study 34 610 5.57 24200 Hudson River 2014 QC FINAL.docx 11/18/15 13 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 7. Mean percent error for the sorting procedure by sampling week, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Sampling Week Mean Percent Samples (Beginning Monday) Error Inspected 31 Mar 14 53 1.58 07 Apr 14 52 0.30 14 Apr 14 41 0.80 21 Apr 14 29 0.58 28 Apr 14 23 0.19 05 May 14 30 1.86 12 May 14 38 2.45 19 May 14 57 6.79 26 May 14 52 1.78 02 Jun 14 43 1.17 09 Jun 14 35 2.89 16 Jun 14 32 0.85 23 Jun 14 27 3.04 07 Jul 14 12 2.27 21 Jul 14 9 0.78 04 Aug 14 12 0.85 18 Aug 14 9 2.49 01 Sep 14 27 2.36 15 Sep 14 11 1.33 29 Sep 14 18 0.14 Study 610 1.95 24200 Hudson River 2014 QC FINAL.docx 11/18/15 14 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 8. Percent of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Sampling Week Total No. of Samples Percent (Beginning Monday) Samples Inspected Identified Inspected 31 Mar 14 10 61 16.39 07 Apr 14 8 52 15.38 14 Apr 14 7 52 13.46 21 Apr 14 10 67 14.93 28 Apr 14 14 98 14.29 05 May 14 18 125 14.40 12 May 14 18 124 14.52 19 May 14 18 126 14.29 26 May 14 17 126 13.49 02 Jun 14 17 122 13.93 09 Jun 14 21 123 17.07 16 Jun 14 26 123 21.14 23 Jun 14 16 123 13.01 07 Jul 14 12 81 14.81 21 Jul 14 11 75 14.67 04 Aug 14 9 78 11.54 18 Aug 14 13 81 16.05 01 Sep 14 11 81 13.58 15 Sep14 13 81 16.05 29 Sep 14 26 81 32.10 Study 295 1880 15.69 24200 Hudson River 2014 QC FINAL.docx 11/18/15 15 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 9. Sample identification percent nonconformance results, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Percent Sampling Week Number of Samples Nonconformance (Beginning Monday) Nonconformities Inspected (Week) 31 Mar 14 0 10 0.00 07 Apr 14 0 8 0.00 14 Apr 14 0 7 0.00 21 Apr 14 0 10 0.00 28 Apr 14 0 14 0.00 05 May 14 0 18 0.00 12 May 14 0 18 0.00 19 May 14 0 18 0.00 26 May 14 0 17 0.00 02 Jun 14 0 17 0.00 09 Jun 14 0 21 0.00 16 Jun 14 1 26 3.85 23 Jun 14 0 16 0.00 07 Jul 14 0 12 0.00 21 Jul 14 0 11 0.00 04 Aug 14 0 9 0.00 18 Aug 14 0 13 0.00 01 Sep 14 0 11 0.00 15 Sep 14 0 13 0.00 29 Sep 14 1 26 3.85 Study 2 295 0.68 24200 Hudson River 2014 QC FINAL.docx 11/18/15 16 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 10. Mean percent error of samples inspected during the identification procedure, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total No. of Number of Sampling Week Mean Percent Samples Taxa (Beginning Monday) Error Inspected Inspected 31 Mar 14 10 0.00 12 07 Apr 14 8 0.00 15 14 Apr 14 7 0.00 11 21 Apr 14 10 0.15 19 28 Apr 14 14 0.02 27 05 May 14 18 0.18 35 12 May 14 18 0.11 63 19 May 14 18 0.47 76 26 May 14 17 0.76 74 02 Jun 14 17 0.71 95 09 Jun 14 21 0.43 108 16 Jun 14 26 30.28 163 23 Jun 14 16 0.45 88 07 Jul 14 12 0.38 63 21 Jul 14 11 0.31 54 04 Aug 14 9 0.11 52 18 Aug 14 13 0.13 49 01 Sep 14 11 0.02 37 15 Sep 14 13 0.00 47 29 Sep 14 26 1.36 59 Study 295 4.65 1147 Additional organisms found during the sort QC were identified independently to determine the frequency of species and life stages missed during the initial sort. Six taxa accounted for 92% of the additional organisms found during sort QC: Striped Bass, White Perch, Herring family, Bay Anchovy, Morone sp., and Goby family (Table 11). For these six taxa, the additional number of organisms per stage found in the sort QC amounted to less than 2% of the total found during sample processing. Exceptions of >2% were within the most abundant life stage: Bay Anchovy yolk-sac larvae (2.83%), Goby family eggs (100%) and yolk-sac larvae (25%), Herring unidentifiable stage (16.13%), Morone species unidentifiable stage (2.59%), Striped Bass eggs (2.93%) and unidentifiable stage (7.62%) and, White Perch eggs (6.84) and yolk-sac larvae (2.24%; Table 12). For most taxa by life stage, the percentage missed by the original sorter was well under 5%.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 17 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM During the identification process, absolute error rates for individual life stages of commonly encountered taxa were usually less than 0.02. Generally, only those taxa-life stages with low total counts had absolute error rates above 0.05 (Table 13).
Net error rates were substantially lower than the absolute error rates in most cases, demonstrating that errors often tended to cancel each other out. This was especially noticeable for many of the most abundant taxa-life stages, such as Bay Anchovy eggs and Striped Bass yolk-sac and post yolk-sac larvae.
3.2 Fall Juvenile Survey Results of the laboratory QC program for the 2014 Fall Juvenile Survey (consisting of the Beach Seine Survey and the Fall Shoals Survey) were summarized by the same methods as the QC results for the 2014 Hudson River Ichthyoplankton Laboratory Program and are presented in Table 14.
There were 1,213 and 847 young-of-the-year fish identification records made in the laboratory for the Fall Juvenile and Beach Seine surveys respectively and 7,588 and 4,923 young-of-the-year fish length measurement records were made for the Fall Juvenile and Beach Seine surveys respectively.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 18 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 11. Ranking of ichthyoplankton taxa missed during initial sort and found during sort Quality Control (QC), 2014 Hudson River Ichthyoplankton Laboratory Program.
Number of Organisms Taxon found in Sort QC Percent Striped Bass 2196 47.55 White Perch 904 19.58 Herring family 627 13.58 Bay Anchovy 305 6.60 Morone species 111 2.40 Unidentified 110 2.38 Goby family 96 2.08 Atlantic Tomcod 64 1.39 Winter Flounder 58 1.26 Cunner 37 0.80 Tautog 24 0.52 Freshwater Drum 22 0.48 Hogchoker 14 0.30 Windowpane 11 0.24 Atlantic Croaker 9 0.19 Atlantic Menhaden 9 0.19 Gizzard Shad 6 0.13 American Shad 3 0.06 American Sand Lance 2 0.04 Inland Silverside 2 0.04 American Eel 1 0.02 Carp and Minnow family 1 0.02 Common Carp 1 0.02 Summer Flounder 1 0.02 Sunfish family 1 0.02 Tessellated Darter 1 0.02 Weakfish 1 0.02 Yellow Perch 1 0.02 Total 4618 100.00 24200 Hudson River 2014 QC FINAL.docx 11/18/15 19 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 12. Summary of ichthyoplankton by life stage for the six highest ranked taxa missed during original sort and found during sort QC compared to total count, 2014 Hudson River Ichthyoplankton Laboratory Program.
Percent of Number Abundance Missed in Total Percent of Total Each by Life Original Organisms Found During Taxon Life Stage Stage Sort Found a QC Eggs 44.26 135 35421 0.38 Yolk-sac larvae 0.98 3 106 2.83 Bay Post yolk-sac larvae 54.75 167 19551 0.85 Anchovy Young-of-the-year 0 0 9436 0 Unidentified 0 0 27 0 Eggs 4.17 4 4 100 Goby Yolk-sac larvae 2.08 2 8 25 family Post yolk-sac larvae 93.75 90 8095 1.11 Unidentified 0 0 1 0 Eggs 13.08 82 12980 0.63 Herring Yolk-sac larvae 7.97 50 6075 0.82 family Post yolk-sac larvae 78.15 490 36086 1.36 Unidentified 0.8 5 31 16.13 Yolk-sac larvae 0 0 11 0 Morone Post yolk-sac larvae 82.88 92 4940 1.86 species Unidentified 17.12 19 733 2.59 Eggs 57.7 1267 43186 2.93 Yolk-sac larvae 22.22 488 36314 1.34 Striped Post yolk-sac larvae 19.31 424 181809 0.23 Bass Young-of-the-year 0.05 1 984 0.1 Unidentified 0.73 16 210 7.62 Eggs 33.08 299 4374 6.84 Yolk-sac larvae 16.92 153 6840 2.24 White Post yolk-sac larvae 50 452 28617 1.58 Perch Young-of-the-year 0 0 127 0 Unidentified 0 0 13 0 a Includes both original count and additional organisms found during sort QC.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 20 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 13. Cumulative net and absolute error rates for commonly encountered taxa in ichthyoplankton samples selected for QC inspection of identification process, 2014 Hudson River Ichthyoplankton Laboratory Program.
Total Net Absolute No.
Taxon Stage Count Error Error Samples Atlantic Croaker Post Yolk Sac Larvae 1029 0.00389 0.00389 31 Young Of The Year 54 -0.07407 0.07407 9 Atlantic Menhaden Eggs 185 0.00000 0.00000 8 Yolk-Sac Larvae 7 0.00000 0.00000 4 Post Yolk Sac Larvae 35 1.34286 1.34286 18 Young Of The Year 6 0.16667 0.16667 5 Atlantic Tomcod Unidentified 12 0.00000 0.00000 3 Yolk-Sac Larvae 27 0.03704 0.03704 8 Post Yolk Sac Larvae 851 0.00118 0.00353 29 Young Of The Year 1118 -0.00089 0.00089 74 Bay Anchovy Unidentified 6 0.00000 0.00000 5 Eggs 7364 0.00326 0.00598 52 Yolk-Sac Larvae 18 0.05556 0.05556 6 Post Yolk Sac Larvae 3208 -0.01527 0.02276 113 Young Of The Year 1150 0.00348 0.00522 69 Blueback Herring Young Of The Year 736 0.00000 0.00272 25 Freshwater Drum Eggs 1209 0.00083 0.00083 10 Yolk-Sac Larvae 4 0.25000 0.25000 5 Post Yolk Sac Larvae 42 0.02381 0.02381 11 Goby family Unidentified 1 0.00000 0.00000 1 Post Yolk Sac Larvae 1357 -0.00221 0.00958 56 Herring family Unidentified 1 0.00000 0.00000 1 Eggs 1812 0.00386 0.00497 28 Yolk-Sac Larvae 1002 0.00399 0.01198 44 Post Yolk Sac Larvae 5001 -0.00100 0.01380 130 Hogchoker Eggs 744 0.00134 0.00941 15 Yolk-Sac Larvae 13 -0.07692 0.07692 8 Post Yolk Sac Larvae 15 -0.06667 0.06667 8 Young Of The Year 6 0.50000 0.50000 5 Morone species Unidentified 36 0.08333 0.13889 11 Post Yolk Sac Larvae 361 0.06925 0.08033 30 Striped Bass Unidentified 44 -0.02273 0.11364 8 Eggs 6499 -0.00015 0.00077 52 Yolk-Sac Larvae 4757 0.00399 0.01577 67 Post Yolk Sac Larvae 19207 -0.00193 0.00869 109 Young Of The Year 91 -0.01099 0.01099 28 Tautog Eggs 604 -0.00331 0.00993 11 Post Yolk Sac Larvae 3 0.33333 0.33333 3 (continued) 24200 Hudson River 2014 QC FINAL.docx 11/18/15 21 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM Table 13. (Continued)
Total Net Absolute No.
Taxon Stage Count Error Error Samples White Perch Eggs 170 -0.01176 0.02353 22 Yolk-Sac Larvae 643 0.00467 0.01089 41 Post Yolk Sac Larvae 2182 0.00962 0.03437 80 Young Of The Year 42 0.00000 0.00000 7 Winter Flounder Yolk-Sac Larvae 37 0.00000 0.00000 6 Post Yolk Sac Larvae 265 0.00000 0.00000 16 Young Of The Year 2 0.00000 0.00000 2 Table 14. Percent inspected, percent nonconforming, and average outgoing quality of laboratory tasks for the 2014 Fall Juvenile and Beach Seine Surveys.
Percent Average Outgoing Task Percent Inspected (%) Nonconforming (%) Quality (%)
Identification 6.26 0.00 0.00 Measurement 1.92 0.00 0.00 24200 Hudson River 2014 QC FINAL.docx 11/18/15 22 Normandeau Associates, Inc.
2014 LABORATORY QC REPORT FOR THE HUDSON RIVER BIOLOGICAL MONITORING PROGRAM 4.0 Literature Cited Normandeau Associates Inc. 2013. 2013 Hudson River fall juvenile and beach seine surveys standard operating procedures. June 2013. Prepared for Indian Point Energy Center.
__________. 2014. Quality assurance plan for the Hudson River ichthyoplankton laboratory program. Rev. 2, April 2014. Prepared for Entergy Nuclear Operations, Inc.
Stephens, K.S. 1979. Volume 2: How to perform continuous sampling (CSP). American Society for Quality Control. 70 pp.
U.S. Department of Defense. 1981. Military standard. Single- and multi-level continuous sampling procedures and table for inspection by attributes. MIL-STD-1235B.
24200 Hudson River 2014 QC FINAL.docx 11/18/15 23 Normandeau Associates, Inc.