2017 Annual Groundwater Monitoring and Corrective Action Report

Transcription

1 Prepared For: Gavin Power, LLC Cheshire, Ohio Annual Groundwater Monitoring and Corrective Action Report Bottom Ash Complex Gavin Power Plant Cheshire, Ohio 31 January 2018 Environmental Resources Management 204 Chase Drive Hurricane, West Virginia The business of sustainability

2 Gavin Power, LLC Annual Groundwater Monitoring and Corrective Action Report Bottom Ash Complex at Gavin Power Plant in Cheshire, Ohio January 2018 Project No Leonard Rafalko Principal-in-Charge Joseph Robb, P.G. Project Manager Natasha Hausmann Senior Scientist Environmental Resources Management 204 Chase Drive Hurricane, West Virginia T: F:

3 TABLE OF CONTENTS 1.0 Introduction Program Status (e) Groundwater Flow Rate and Direction Sampling Summary Data Quality Statistical Analysis and Results Pooled vs Individual Well Comparisons Establishment of Upgradient Dataset Descriptive Statistics Outlier Determination Checking for Temporal Stability Establishing Upper Prediction Limits Final UPL Selection Conclusions Key Future Activities References 13 LIST OF TABLES TABLE 1 REGULATORY REQUIREMENT CROSS-REFERENCE TABLE 1 TABLE 2 SAMPLING DATES FOR EACH WELL 6 TABLE 3 FINAL UPLS FOR EACH ANALYTE 9 TABLE 4 DOWNGRADIENT MEASUREMENTS THAT EXCEED THE UPL 10 LIST OF FIGURES 2-1 EXISTING MONITORING WELL NETWORK MAPS 2-2 AUGUST 2016 GROUNDWATER FLOW DIRECTION MAP i

4 2-3 FEBRUARY GROUNDWATER FLOW DIRECTION MAP LIST OF APPENDICES APPENDIX A STATISTICAL SUPPORT APPENDIX B ANALYTICAL SUMMARY ii

5 1.0 INTRODUCTION The General James M. Gavin Power Plant (Plant) is a coal-fired generating station located in Gallia County in Cheshire, Ohio, along the Ohio River. The Plant consists of three regulated coal combustion residual (CCR) management units that are subject to regulation under Title 40, Code of Federal Regulations, Part 257 (40 CFR 257) (also known as the CCR Rule): the Residual Waste Landfill, the Fly Ash Reservoir, and the Bottom Ash Complex (BAC). This report was produced by Environmental Resource Management, Inc. (ERM), on behalf of Gavin Power, LLC, and focuses on the initial annual groundwater monitoring results for the BAC. The report summarizes the activity at the Site over the last year and provides a statistical summary of the findings of samples collected by October 17,, as required by 40 CFR Consistent with the notification requirements of the Rule, this annual groundwater monitoring report will be posted to the Plant s operating record and notification will be made to the State of Ohio no later than January 31, 2018, and the report will be placed on the publicly accessible internet site within thirty days thereafter (40 CFR (h), (h), (h)). Table 1 cross references the reporting requirements under the CCR Rule with the contents of this report. The BAC CCR unit is located adjacent to and immediately south of the main Plant area. The BAC consists of two ponds situated along the Ohio River. The larger pond is the Bottom Ash Pond (BAP) and the smaller pond is the Recirculation Pond or Reclaim Pond (RCP). The BAC receives bottom ash and miscellaneous Plant wastewaters including coal-pile runoff, cooling-tower blowdown, pyrites, and various Plant sump wastewaters. It has been operational since Table 1 Regulatory Requirement Cross-Reference Table Regulatory Citation in 40 CFR (e) (e) Requirement (paraphrased) Status of the groundwater monitoring program. Summarize key actions completed. Where Addressed in this Report Section 2.0 Section 2.2 ERM 1 GAVIN/

6 Regulatory Citation in 40 CFR (e) (e) (e)(1) (e)(2) (e)(3) (e)(4) Requirement (paraphrased) Describe any problems encountered. Key activities for upcoming year. Map, aerial image or diagram of CCR Unit and monitoring wells Identification of new monitoring wells installed or abandoned during the preceding year. Summary of groundwater data, wells sampled, date sampled, and whether sample was required under detection or assessment monitoring. Narrative Discussion of any transition between monitoring programs. Where Addressed in this Report Section 2.2 Section 4.0 Figure 2.1 There were no new monitoring wells installed or abandoned during the preceding year. Table 2, Appendix A Section 4.0 ERM 2 GAVIN/

7 2.0 PROGRAM STATUS (E) Over the past 2 years, samples were collected from the certified federal monitoring-well network. The groundwater samples were collected as part of detection monitoring under 40 CFR The groundwater monitoring-well network consists of three upgradient monitoring wells (BAC-01, MW-1, and MW-6) and four downgradient monitoring wells (BAC-02, BAC-03, BAC-04, and BAC-05). All of the monitoring wells are screened in the uppermost aquifer around the BAP and RCP units. The uppermost aquifer is approximately 25 feet to 35 feet thick and is comprised of fine to coarse sand. The uppermost aquifer is located below an approximately 20-foot thick confining layer of silty clay with interbedded sand and silt, and above a shale bedrock unit. The monitoring well locations can be viewed on the site location map and aerial image provided in Figure 2-1. No new wells were installed or decommissioned after the certification of the well network (Geosyntec, 2016). 2.1 GROUNDWATER FLOW RATE AND DIRECTION Depth to groundwater measurements were made at each monitoring well prior to each sampling event. Groundwater elevations, calculated by subtracting the depth to groundwater from the surveyed reference elevation for each well, were reviewed for each sampling event. The hydraulic gradient for the eight sampling events was to the northeast, toward the Ohio River, except during the February event when the hydraulic gradient was to the northwest, away from the Ohio River. The reversal of the hydraulic gradient occurred after a period in January when the Ohio River reached an elevation of feet, based on records from the U.S. Geological Survey gauging station at Point Pleasant, West Virginia. The river stage decreased through February, and the principal groundwater flow in a northeasterly direction toward the river had resumed by the March gauging event. A potentiometric surface map for August 2016, which is similar to the other gauging events except February, is presented in Figure 2-2. A potentiometric surface map for February, for the period when the groundwater flow was away from the Ohio River, is presented in Figure 2-3. Measured hydraulic gradients ranged from to over the eight gauging events. Based on the measured hydraulic gradients, an ERM 3 GAVIN/

8 assumed porosity of 0.3, and a measured hydraulic conductivity of centimeters per second (Geosyntec 2012), the velocity of groundwater in the alluvial aquifer beneath the BAC varied between 300 and 2,400 feet per year. 2.2 SAMPLING SUMMARY A summary of the total number of samples collected for each well over the last 2 years, the well gradient designation (upgradient or downgradient of the CCR unit), and the geologic formation in which it is screened is provided in Table 2. Sampling occurred approximately every other month starting August 25, The results for metals represent total recoverable metals. Each of the BAC wells was sampled during each of the eight sampling events from August 2016 through July. With one exception, the wells were sampled for the 40 CFR 257 Appendix III and Appendix IV analytes. During the March sampling event, a pump malfunction resulted in collection of insufficient sample volume at well BAC-04, which prevented the laboratory from analyzing for combined radium, total dissolved solids, and certain anions. The pump was repaired and functioned properly in subsequent sampling events Data Quality As discussed below, ERM s data quality review found the laboratory analytical results to be valid, reliable, and useable for decision-making purposes with the listed qualifiers. No analytical results were rejected. ERM reviewed field and laboratory documentation to assess the validity, reliability, and usability of the analytical results. Samples from the first four sampling events (August 2016 through February ) were analyzed by American Electric Power (the former owner of the Gavin Plant) at the Dolan Chemical Laboratory located in Groveport Ohio. Available data quality information included field-sampling forms, chain-of-custody documentation, quantitation limits, and completeness of the analyses. Samples from the second four sampling events (March through July ) were analyzed by TestAmerica of North Canton, Ohio. Data quality information reviewed for these results included field sampling forms, chain-of-custody documentation, holding times, laboratory methods, ERM 4 GAVIN/

9 cooler temperatures, laboratory method blanks, laboratory control sample recoveries, field duplicate samples, matrix spikes/matrix spike duplicates, quantitation limits, and equipment blanks. Data qualifiers were appended to results in the project database as appropriate based on laboratory quality measurements (e.g., control sample recoveries) and field quality measurements (e.g., agreement between normal and field duplicate samples). ERM 5 GAVIN/

10 Table 2 Sampling Dates for Each Well Well BAC-01 BAC-02 BAC-03 BAC-04 BAC-05 MW-1 MW-6 Location Upgradient Downgradient Downgradient Downgradient Downgradient Upgradient Upgradient Geology Alluvium Alluvium Alluvium Alluvium Alluvium Alluvium Alluvium X X X X X X X X X X X X X X X X X X X X X Date Sampled X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Note: All samples summarized in this table were collected under Detection Monitoring as specified in ERM 6 GAVIN/

11 3.0 STATISTICAL ANALYSIS AND RESULTS Consistent with the CCR Rule and with the Statistical Analysis Plan (StAP, ERM ) in the operating record, a prediction limit approach (40 CFR (f)) was used to identify potential impacts to groundwater. The steps outlined in the decision framework in the StAP include: Pooled vs individual comparisons. Establishment of the upgradient dataset. Calculating prediction limits. Drawing conclusions. 3.1 POOLED VS INDIVIDUAL WELL COMPARISONS When multiple upgradient wells were available within the same geologic formation, concentrations were compared among these wells to determine if they could be pooled to create a single upgradient dataset, or alternately, if the background data set should be established for each individual upgradient well. For each analyte, Boxplots (see Appendix A, Figure 1) and Kruskal Wallis results (see Appendix A, Table 1) are provided for upgradient wells. The statistical test shows that an individual well data comparison is appropriate for all analytes. 3.2 ESTABLISHMENT OF UPGRADIENT DATASET When evaluating the concentrations of analytes in groundwater, USEPA guidance (2009) recommends performing a careful quality check of the data to identify any anomalies. In addition to the data validation that was performed (discussed above in Section 2.2.1), descriptive statistics, outlier testing, and checking for temporal stationarity were completed to finalize the upgradient dataset (the supporting documentation is found in Appendix A and discussed below) Descriptive Statistics Descriptive statistics were calculated for upgradient wells and analytes at the site (see Appendix A, Table 2). The descriptive statistics highlight a number of relevant characteristics about the alluvium upgradient datasets including: ERM 7 GAVIN/

12 There are a total of 21 well-analyte combinations for the upgradient dataset (three upgradient monitoring wells and seven constituents for Detection Monitoring (40 CFR 257 Appendix III)). Twenty-one well-analyte combinations have detection rates greater than or equal to 50 percent. Twenty well-analyte combinations have 100 percent detects. Fifteen well-analyte combinations follow a normal distribution (using Shapiro-Wilks Normality Test) and one well-analyte combination follows a log-normal distribution. The remaining well-analyte combinations have no discernible distribution. Although the StAP lists as eight detected values as the minimum data requirements, this first annual report uses five detected values as the cutoff for all of the statistical tests. The sample size requirements have been relaxed to allow for the analysis of most datasets at this initial stage of the monitoring program, while the sample sizes are still small. As additional samples are collected and the datasets updated, the minimum requirement will be increased as specified in the StAP Outlier Determination As discussed in the StAP, both statistical and visual outlier tests were performed on the upgradient datasets. No data points were identified as both a statistical and visual outliers (see Appendix A, Table 3 and Figure 2), so no values were excluded from upper prediction limit (UPL) calculations Checking for Temporal Stability A trend test was calculated for all detected values in the upgradient wells as long as they had at least five detected data points and at least 50 percent detection rate. A summary of the Mann Kendall trend test results and time series plots can be found in Appendix A, Table 4 and Figure 3. The following summarize the results of the trend analysis: There are a total of 21 well-analyte combinations in the upgradient dataset. Twenty-one well-analyte combinations meet the data requirements of the trend test. Six well-analyte combinations had a significant increasing trend. Three well-analyte combinations had a significant decreasing trend. ERM 8 GAVIN/

13 Twelve well-analyte combinations had no significant trend (i.e., concentrations were stable over time). 3.3 ESTABLISHING UPPER PREDICTION LIMITS As described in the StAP, a multi-part assessment of the monitoring wells was performed to determine what type of UPL should be used for the analysis. A complete table of UPLs and the methods used to calculate them can be found in Appendix A, Table 5. A total of nine well-analyte combinations were found to have either increasing or decreasing trends. For these well-analyte pairs, a bootstrapped UPL calculated around a Theil Sen trend was used to derive a more accurate UPL (ERM ). The remaining 12 well-analyte combinations were found to have no significant trend. Sanitas software was used to calculate static UPLs using an annual site-wide false positive rate of 0.1 and a 1-of-2 retesting approach as discussed in the StAP Final UPL Selection A final UPL is selected for each analyte and compared to the most recent sample in downgradient wells. The seven analytes had a UPL value calculated for each of the upgradient wells based on individual well data. For these wells and analytes, the maximum UPL was selected as the representative UPL for each analyte (ERM ). All final UPL values are shown in Table 3 below and Appendix A, Table 5. Table 3 Final UPLs for each Analyte UPL Type Analyte Geology LPL UPL Unit Individual well Boron Alluvium Individual well Calcium Alluvium 129 Individual well Chloride Alluvium 24.7 Individual well Fluoride Alluvium Individual well ph Alluvium SU Individual well Sulfate Alluvium 140. Individual well TDS Alluvium 505 LPL = lower prediction limit; = milligrams per liter; SU = standard units; TDS = total dissolved solids; UPL = upper prediction limit ERM 9 GAVIN/

14 3.4 CONCLUSIONS The downgradient samples collected during the July sampling event were used for compliance comparisons. All downgradient wells were below the UPLs with the following exceptions (see Table 4 below and Appendix A, Table 6). Table 4 Downgradient Measurements that Exceed the UPL Analyte Well Geology LPL UPL Sample Date Value Unit SSI Boron BAC-02 Alluvium Yes Boron BAC-03 Alluvium Yes Boron BAC-04 Alluvium Yes Boron BAC-05 Alluvium Yes Calcium BAC-02 Alluvium Yes Chloride BAC-02 Alluvium Yes Chloride BAC-03 Alluvium Yes Chloride BAC-04 Alluvium Yes Fluoride BAC-05 Alluvium Yes ph BAC-02 Alluvium SU Yes ph BAC-03 Alluvium SU Yes ph BAC-04 Alluvium SU Yes ph BAC-05 Alluvium SU Yes Sulfate BAC-02 Alluvium Yes Sulfate BAC-03 Alluvium Yes Sulfate BAC-04 Alluvium Yes Sulfate BAC-05 Alluvium Yes TDS BAC-02 Alluvium ,100 Yes TDS BAC-04 Alluvium Yes LPL = lower prediction limit; = milligrams per liter; SSI = statistically significant increase; SU = standard units; TDS = total dissolved solids; UPL = upper prediction limit The downgradient measurements that exceed the UPL are considered statistically significant increases (SSI) above background. The Unified Guidance (USEPA 2009) recommends re-testing as part of the UPL ERM 10 GAVIN/

15 method. Per the Unified Guidance and the 1-of-2 retesting scheme described in the StAP, the downgradient wells with SSIs may be resampled to support an alternate source demonstration ( (e)(2)). All downgradient wells with initial exceedances were examined for trends to assess the stability of concentrations. A summary of these trend test results can be found in Appendix A, Table 6. Of the wells with SSIs, the following wells had significantly increasing trends: BAC-02 for boron, calcium, chloride, and total dissolved solids BAC-05 for fluoride Additionally, the following wells had significantly decreasing trends: BAC-02 for ph BAC-03 for sulfate The decreasing trends for ph and sulfate are not consistent with an active release from the BAC. All trends will be monitored closely in future events. All wells with SSIs are plotted in Appendix A, Figure 4. A summary of all analytical results obtained from the BAC groundwater monitoring is provided in Appendix B. ERM 11 GAVIN/

16 4.0 KEY FUTURE ACTIVITIES Consistent with the 1-of-2 retesting approach described in the Unified Guidance (USEPA 2009) and the StAP (ERM ), initial exceedances will be retested as soon as practicable. Investigation of features adjacent to the BAC as well as additional upgradient wells will be undertaken to determine if alternate sources of SSIs are present. Assessment monitoring will be initiated unless an Alternate Source Demonstration can be successfully made by April 15, ERM 12 GAVIN/

17 5.0 REFERENCES Environmental Resources Management (ERM).. Groundwater Monitoring Plan. Bottom Ash Complex, Fly Ash Reservoir, and Residual Waste Landfill. Gavin Plan, Cheshire Ohio. Geosyntec Final Permit-To-Install Application. Expansion of the Gavin Plant Residual Waste Landfill. Hydrogeologic Study Report OAC (C)(4). Geosyntec Groundwater Monitoring Network Evaluation, Gavin Site Bottom Ash Complex, Cheshire, Ohio. USEPA Statistical Analysis of Groundwater Monitoring Data at RCRA Facilities. Unified Guidance. USEPA/530/R/09/007. Office of Resource Conservation and Recovery. Washington, D.C. ERM 13 GAVIN/

18 Figures

19 Main Plant Area &? A &? A BAC-02 Federal Sampling Program Groundwater Monitoring Well BAC-03 & A? Recirculation Pond & A? Legend BAC-04 p & A? BAC or iver Feet Ohi MW-1 Figure 1: Existing Monitoring Well Network Maps Bottom Ash Complex Gavin Power Plant Cheshire, Ohio Q:\Team\DMMV\Clients_F_K\Gavin\GavinPowerPlant\MXD\Figure1_MWNetwork_BAC_ mxd - Dana.Heusinkveld - 1/17/2018 Bottom Ash Pond &? A BAC-01 NOTES: 1. Locations are approximate 2. Aerial Imagery: ESRI World Imagery Reproduced under license in ArcGIS 10.4 & A? & A? MW-6

20 Main Plant Area Recirculation Pond &? A BAC &? A & A? Legend Groundwater Elevation (ft) & A? BAC MW Federal Sampling Program Groundwater Monitoring Well Potentiometric Contour (ft) $ Interpreted Groundwater Flow Direction D & A? BAC BAC Bottom Ash Pond p & A? &? A BAC or iver Feet Ohi MW Figure 2: August 2016 Groundwater Flow Direction Map Bottom Ash Complex Gavin Power Plant Cheshire, Ohio Q:\Team\DMMV\Clients_F_K\Gavin\GavinPowerPlant\MXD\Figure2_GWFlowDirection_BAC_ mxd - Dana.Heusinkveld - 1/17/2018 & A? NOTES: 1. Locations are approximate 2. Groundwater elevations based on measurements made on 8/25/ Aerial Imagery: ESRI World Imagery Reproduced under license in ArcGIS 10.4

21 Main Plant Area BAC &? A & A? Groundwater Elevation (ft) BAC MW Potentiometric Contour (ft) $ D Federal Sampling Program Groundwater Monitoring Well Interpreted Groundwater Flow Direction 543 & A? NOTES: 1. Locations are approximate 2. Groundwater elevations based on measurements made on 2/7/ 3. Aerial Imagery: ESRI World Imagery Reproduced under license in ArcGIS 10.4 & A? BAC BAC Bottom Ash Pond p & A? &? A BAC or iver Feet Ohi MW Figure 3: February Groundwater Flow Direction Map Bottom Ash Complex Gavin Power Plant Cheshire, Ohio Q:\Team\DMMV\Clients_F_K\Gavin\GavinPowerPlant\MXD\Figure3_GWFlowDirection_BAC_Feb_ mxd - Dana.Heusinkveld - 1/17/2018 Recirculation Pond &? A & A? Legend

22 Appendix A Statistical Support

23 Table A-1 Kruskal-Wallis Test Comparison of Upgradient Wells Gavin Power, LLC Bottom Ash Complex Analyte Geology N Num Detects Percent Detects DF KW Statistic p-value Conclusion UPL Type Boron Alluvium % Significant Difference Individual Calcium Alluvium % Significant Difference Individual Chloride Alluvium % <0.001 Significant Difference Individual Fluoride Alluvium % Significant Difference Individual ph Alluvium % <0.001 Significant Difference Individual Sulfate Alluvium % <0.001 Significant Difference Individual TDS Alluvium % <0.001 Significant Difference Individual Notes N: number of data points DF: Degrees of Freedom statistic: Kruskal Wallis test statistic p-value: P-values below 0.05 indicate that the median concentrations in the upgradient wells are significantly different from each other and the upgradient wells should not be pooled. p-value: P-values equal or above 0.05 indicate that the median concentrations in the upgradient wells are not significantly different from each other and the upgradient wells can be pooled. UPL: upper prediction limit TDS: total dissolved solids

24 Table A-2 Descriptive Statistics for Upgradient Wells Gavin Power, LLC Bottom Ash Complex Geology Analyte Well Units N Num Detects Percent Detects Min ND Max ND Min Detect Median Mean Max Detect SD CV Distribution UPL Type Alluvium Boron BAC % % Normal Individual Alluvium Boron MW % % Normal Individual Alluvium Boron MW % % Lognormal Individual Alluvium Calcium BAC % % Normal Individual Alluvium Calcium MW % % Normal Individual Alluvium Calcium MW % % NDD Individual Alluvium Chloride BAC % % Normal Individual Alluvium Chloride MW % % Normal Individual Alluvium Chloride MW % % Normal Individual Alluvium Fluoride BAC % % NDD Individual Alluvium Fluoride MW % % NDD Individual Alluvium Fluoride MW % % NDD Individual Alluvium ph BAC-01 SU % % Normal Individual Alluvium ph MW-1 SU % % Normal Individual Alluvium ph MW-6 SU % % Normal Individual Alluvium Sulfate BAC % % NDD Individual Alluvium Sulfate MW % % Normal Individual Alluvium Sulfate MW % % Normal Individual Alluvium TDS BAC % % Normal Individual Alluvium TDS MW % % Normal Individual Alluvium TDS MW % % Normal Individual Notes : milligrams per liter N: number of data points Min ND: The minimum non-detected value Max ND: The maximum non-detected value SD: Standard Deviation CV: Coefficient of Variation (standard deviation divided by the mean) Normal: the data fit a normal distribution Lognormal: The data fit a lognormal distribution NDD: no discernable distribution SU: standard units TDS:Total dissolved solids

25 Table A-3 Potential Outliers in Upgradient Wells Gavin Power, LLC Bottom Ash Complex Well Sample Geology Date Analyte RL Units Detect Concentration UPL Type Distribution Statistical Outlier Visual Outlier Normal Outlier Log Statistical Outlier Log Visual Outlier Lognormal Outlier Statistical and Visual Outlier Final Outlier Determination MW-6 MW Alluvium Boron TRUE Individual Lognormal X X X X Not an outlier MW-1 MW Alluvium Calcium 124 TRUE 124 Individual Normal X X Not an outlier MW-6 MW Alluvium Calcium 123 TRUE 123 Individual NDD X X Not an outlier MW-1 MW Alluvium Chloride 1 TRUE 22 Individual Normal X X Not an outlier BAC-01 BAC Alluvium Fluoride 0.05 TRUE 0.14 Individual NDD X X Not an outlier BAC-01 BAC Alluvium Fluoride 0.05 TRUE 0.14 Individual NDD X X Not an outlier BAC-01 BAC Alluvium Fluoride 0.05 TRUE 0.14 Individual NDD X X Not an outlier BAC-01 BAC Alluvium Fluoride 0.05 TRUE 0.14 Individual NDD X X Not an outlier Notes RL Reporting limit UPL: upper prediction limit NDD: No Discernible Distribution SU: Standard units : milligrams per liter Outlier tests were performed on detected data only. Statistical outliers were determined using a Dixon's test for N < 25 and with Rosner's test for N > 25. Visual outliers were identified if they fall above the confidence envelope on the QQ plot. Data points were considered potential outliers if they were both statistical and visual outliers. NDD wells had data points considered as potential outliers if they were either a normal or lognormal outlier. [Blank] data distribution indicates that the well data did not have enough detected data points for outlier analysis. Lognormally distributed data was first log-transformed before visual and statistical outlier tests were performed. Normal data distribution indicates that the well data was directly used for statistical and visual outlier tests. NDD indicates that the botrh untransformed and transformed data were examined with statistical and visual outlier tests.

26 Table A-4 Mann Kendall Test for Trends in Upgradient Wells Gavin Power, LLC Bottom Ash Complex Analyte Geology UPL Type Well N Num Detects Percent Detects p-value Tau Conclusion Boron Alluvium Individual BAC % Increasing Trend Boron Alluvium Individual MW % Stable, No Trend Boron Alluvium Individual MW % Stable, No Trend Calcium Alluvium Individual BAC % Stable, No Trend Calcium Alluvium Individual MW % Stable, No Trend Calcium Alluvium Individual MW % Stable, No Trend Chloride Alluvium Individual BAC % Stable, No Trend Chloride Alluvium Individual MW % Increasing Trend Chloride Alluvium Individual MW % Increasing Trend Fluoride Alluvium Individual BAC % Increasing Trend Fluoride Alluvium Individual MW % Increasing Trend Fluoride Alluvium Individual MW % Stable, No Trend ph Alluvium Individual BAC % Stable, No Trend ph Alluvium Individual MW % Decreasing Trend ph Alluvium Individual MW % Decreasing Trend Sulfate Alluvium Individual BAC % Decreasing Trend Sulfate Alluvium Individual MW % Stable, No Trend Sulfate Alluvium Individual MW % Stable, No Trend TDS Alluvium Individual BAC % Stable, No Trend TDS Alluvium Individual MW % Increasing Trend TDS Alluvium Individual MW % Stable, No Trend Notes UPL: upper prediction limit N: number of data points p-value: A two-sided p-value describing the probability of the H0 being true (a=0.05) tau: Kendall's tau statistic Trend tests were performed only if the upgradient dataset met the minium data quality criteria (ERM ). TDS: Total dissolved solids

27 Table A-5 Calculated UPLs for Upgradient Datasets Gavin Power, LLC Bottom Ash Complex Analyte Geology UPL Type Trend Well N Num Detects Percent Detects LPL UPL Units ND Adjustment Transformation Alpha Method Final LPL Final UPL Boron Alluvium Individual Increasing Trend BAC % None No NP Detrended UPL X Boron Alluvium Individual Stable, No Trend MW % None No Param Intra 1 of 2 Boron Alluvium Individual Stable, No Trend MW % None ln(x) Param Intra 1 of 2 Calcium Alluvium Individual Stable, No Trend BAC % 120 None No Param Intra 1 of 2 Calcium Alluvium Individual Stable, No Trend MW % 129 None No Param Intra 1 of 2 X Calcium Alluvium Individual Stable, No Trend MW % 123 None No NP Intra (normality) 1 of 2 Chloride Alluvium Individual Stable, No Trend BAC % 24.7 None No Param Intra 1 of 2 X Chloride Alluvium Individual Increasing Trend MW % 22.9 None No NP Detrended UPL Chloride Alluvium Individual Increasing Trend MW % 20.6 None No NP Detrended UPL Fluoride Alluvium Individual Increasing Trend BAC % None No NP Detrended UPL X Fluoride Alluvium Individual Increasing Trend MW % None No NP Detrended UPL Fluoride Alluvium Individual Stable, No Trend MW % None x^ Param Intra 1 of 2 ph Alluvium Individual Stable, No Trend BAC % SU None No Param Intra 1 of 2 X ph Alluvium Individual Decreasing Trend MW % SU None No NP Detrended UPL X ph Alluvium Individual Decreasing Trend MW % SU None No NP Detrended UPL Sulfate Alluvium Individual Decreasing Trend BAC % 101 None No NP Detrended UPL Sulfate Alluvium Individual Stable, No Trend MW % 137 None No Param Intra 1 of 2 Sulfate Alluvium Individual Stable, No Trend MW % 140 None No Param Intra 1 of 2 X TDS Alluvium Individual Stable, No Trend BAC % 469 None No Param Intra 1 of 2 TDS Alluvium Individual Increasing Trend MW % 499 None No NP Detrended UPL TDS Alluvium Individual Stable, No Trend MW % 505 None No Param Intra 1 of 2 X Notes N: number of data points LPL: lower prediction limit. These were only calculated for ph UPL: upper prediction limit. UPLs were constructed with a site wide false positive rate of 0.1 and a 1 of 2 retesting. UPLs were calculated using Sanitas Software. ND: Non-dectect : milligrams per liter SU: Standard units NP: non parametric RL: Reporting Limit Intra: indicates an individual UPL was used Inter: indicates a pooled UPL was used In the case where multiple UPLs were calculated for an analyte, the maximum UPL was used as the final UPL. In the case where multiple LPLs were calculated for an ph the minimum LPL was used as the final LPL.

28 Table A-6 Comparison of downgradient wells to UPLs Gavin Power, LLC Bottom Ash Complex Analyte Well Geology LPL UPL Units Recent Date Observation Obs > UPL Notes Mann Kendall P-value Mann Kendall tau Boron BAC-02 Alluvium /19/ 2.7 X Trend Test: Increasing Trend Boron BAC-03 Alluvium /14/ 2 X Trend Test: Stable, No Trend Boron BAC-04 Alluvium /19/ 2.5 X Trend Test: Stable, No Trend 1 0 Boron BAC-05 Alluvium /19/ 4.3 X Trend Test: Stable, No Trend Calcium BAC-02 Alluvium 129 7/19/ 190 X Trend Test: Increasing Trend Calcium BAC-03 Alluvium 129 7/14/ 88 Calcium BAC-04 Alluvium 129 7/19/ 86 Calcium BAC-05 Alluvium 129 7/19/ 87 Chloride BAC-02 Alluvium /19/ 110 X Trend Test: Increasing Trend Chloride BAC-03 Alluvium /14/ 61 X Trend Test: Stable, No Trend Chloride BAC-04 Alluvium /19/ 49 X Trend Test: Stable, No Trend Chloride BAC-05 Alluvium /19/ 21 Fluoride BAC-02 Alluvium /19/ 0.16 Fluoride BAC-03 Alluvium /14/ 0.07 Fluoride BAC-04 Alluvium /19/ Fluoride BAC-05 Alluvium /19/ 0.21 X Trend Test: Increasing Trend ph BAC-02 Alluvium SU 7/19/ 6.02 X Trend Test: Decreasing Trend ph BAC-03 Alluvium SU 7/14/ 5.93 X Trend Test: Stable, No Trend ph BAC-04 Alluvium SU 7/19/ 5.94 X Trend Test: Stable, No Trend ph BAC-05 Alluvium SU 7/19/ 6.53 X Trend Test: Stable, No Trend Sulfate BAC-02 Alluvium 140 7/19/ 440 X Trend Test: Stable, No Trend Sulfate BAC-03 Alluvium 140 7/14/ 190 X Trend Test: Decreasing Trend Sulfate BAC-04 Alluvium 140 7/19/ 220 X Trend Test: Stable, No Trend Sulfate BAC-05 Alluvium 140 7/19/ 160 X Trend Test: Stable, No Trend TDS BAC-02 Alluvium 505 7/19/ 1100 X Trend Test: Increasing Trend TDS BAC-03 Alluvium 505 7/14/ 500 TDS BAC-04 Alluvium 505 7/19/ 520 X Trend Test: Stable, No Trend TDS BAC-05 Alluvium 505 7/19/ 460 Notes LPL: lower prediction limit UPL: upper prediction limit : milligrams per liter Obs > UPL:The observation is greater than the UPL SU: Standard units tau: Kendall's tau statistic p-value: A two-sided p-value describing the probability of the H0 being true (a=0.05) "Exceed 'X' indicates that the most recent observed value is higher than the UPL (or out of range of the LPL and UPL in the case of ph.)" "Exceed 'X0' indicates that the two most recent values are higher than the UPL, but the upgradient well is 100% ND." "Exceed '0' indicated that the most recent observed value is higher than the UPL, but is not scored as an SSI due to Double Quantification Rule (ERM )."

29 Unit: Bottom Ash Complex Figure A-1: Boxplots of Upgradient Wells Analyte: Boron; Geology: Alluvium Significant Difference N data: Analyte: Calcium; Geology: Alluvium Significant Difference N data: Concentration () Concentration () BAC 01 MW 1 MW 6 BAC 01 MW 1 MW 6 26 Analyte: Chloride; Geology: Alluvium Significant Difference N data: Analyte: Fluoride; Geology: Alluvium Significant Difference N data: Concentration () BAC 01 MW 1 MW 6 18 BAC 01 MW 1 MW 6 Concentration ()

30 Unit: Bottom Ash Complex Figure A-1: Boxplots of Upgradient Wells Analyte: ph; Geology: Alluvium Significant Difference N data: Analyte: Sulfate; Geology: Alluvium Significant Difference N data: Concentration (SU) Concentration () BAC 01 MW 1 MW 6 BAC 01 MW 1 MW 6 Analyte: TDS; Geology: Alluvium Significant Difference N data: Concentration () 450 BAC 01 MW 1 MW 6 400

31 Analyte: Boron Wells: BAC 01 Detect Identified Outlier Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles 0.08 Analyte: Boron Wells: MW 1 Detect Identified Outlier Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

32 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells 2.2 Analyte: Boron Wells: MW 6 Detect Identified Outlier Individual Analysis Lognormal Distribution Intentionally left blank, not Normal/NDD distribution. Log Concentration () Log Quantiles Analyte: Calcium Wells: BAC 01 Detect Identified Outlier Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

33 Analyte: Calcium Wells: MW 1 Detect Identified Outlier Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles Analyte: Calcium Wells: MW 6 Individual Analysis NDD Distribution Analyte: Calcium Wells: MW 6 Individual Analysis NDD Distribution Detect Identified Outlier 4.80 Detect Identified Outlier 120 Concentration () Log Concentration () Normal Quantiles Log Quantiles

34 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Analyte: Chloride Wells: BAC 01 Detect Identified Outlier Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles Analyte: Chloride Wells: MW 1 Detect Identified Outlier Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

35 Analyte: Chloride Wells: MW 6 Detect Identified Outlier Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles 0.14 Analyte: Fluoride Wells: BAC 01 Detect Identified Outlier Individual Analysis NDD Distribution 2.00 Analyte: Fluoride Wells: BAC 01 Detect Identified Outlier Individual Analysis NDD Distribution Concentration () Log Concentration () Normal Quantiles Log Quantiles

36 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Analyte: Fluoride Wells: MW 1 Individual Analysis NDD Distribution Analyte: Fluoride Wells: MW 1 Individual Analysis NDD Distribution 0.10 Detect Identified Outlier 2.5 Detect Identified Outlier Concentration () Log Concentration () Normal Quantiles Log Quantiles Analyte: Fluoride Wells: MW 6 Individual Analysis NDD Distribution Analyte: Fluoride Wells: MW 6 Individual Analysis NDD Distribution Detect Identified Outlier 2.35 Detect Identified Outlier Concentration () Log Concentration () Normal Quantiles Log Quantiles

37 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells 6.85 Analyte: ph Wells: BAC 01 Detect Identified Outlier Individual Analysis Normal Distribution Concentration (ph units) Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles 7.20 Analyte: ph Wells: MW 1 Detect Identified Outlier Individual Analysis Normal Distribution Concentration (ph units) Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

38 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Analyte: ph Wells: MW 6 Detect Identified Outlier Individual Analysis Normal Distribution Concentration (ph units) Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles Analyte: Sulfate Wells: BAC 01 Individual Analysis NDD Distribution Analyte: Sulfate Wells: BAC 01 Individual Analysis NDD Distribution 110 Detect Identified Outlier 4.70 Detect Identified Outlier Concentration () Log Concentration () Normal Quantiles Log Quantiles

39 Analyte: Sulfate Wells: MW 1 Detect Identified Outlier Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles 130 Analyte: Sulfate Wells: MW 6 Detect Identified Outlier Individual Analysis Normal Distribution 128 Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

40 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells Analyte: TDS Wells: BAC 01 Detect Identified Outlier Individual Analysis Normal Distribution 420 Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles Analyte: TDS Wells: MW 1 Detect Identified Outlier Individual Analysis Normal Distribution Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

41 Unit: Bottom Ash Complex Figure A-2: QQ Plots of Upgradient Wells 480 Analyte: TDS Wells: MW 6 Detect Identified Outlier Individual Analysis Normal Distribution 470 Concentration () Intentionally left blank, not Lognormal/NDD distribution Normal Quantiles

42 0.15 Unit: Bottom Ash Complex Alluvium Figure A-3: Timeseries of Upgradient Wells Chemical: Boron Significant Difference (Individual Analysis) Wells BAC 01 MW 1 MW 6 Concentration () Oct 2016 Jan Chemical: Calcium Significant Difference (Individual Analysis) Apr Jul Symbols Detect Wells BAC 01 MW 1 MW 6 Concentration () Symbols Detect Oct 2016 Jan Apr Jul

43 Unit: Bottom Ash Complex Alluvium Figure A-3: Timeseries of Upgradient Wells 20 Chemical: Chloride Significant Difference (Individual Analysis) Wells BAC 01 MW 1 MW 6 Concentration () Symbols Detect Oct 2016 Jan Apr Jul Chemical: Fluoride Significant Difference (Individual Analysis) Wells BAC 01 MW 1 MW 6 Concentration () Symbols Detect Oct 2016 Jan Apr Jul

44 6 Chemical: ph Significant Difference (Individual Analysis) Unit: Bottom Ash Complex Alluvium Figure A-3: Timeseries of Upgradient Wells Wells BAC 01 MW 1 MW 6 Concentration (SU) Oct 2016 Jan Chemical: Sulfate Significant Difference (Individual Analysis) Apr Jul Symbols Detect Wells BAC 01 MW 1 MW 6 Concentration () Symbols Detect Oct 2016 Jan Apr Jul

45 Chemical: TDS Significant Difference (Individual Analysis) Unit: Bottom Ash Complex Alluvium Figure A-3: Timeseries of Upgradient Wells Wells BAC 01 MW 1 MW 6 Concentration () Symbols Detect Oct 2016 Jan Apr Jul

46 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Concentration () Chemical: Boron, Well: BAC 02 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (0.177) 0.5 Stats Increasing Trend 0.0 N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul 2.0 Chemical: Boron, Well: BAC 03 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (0.177) Concentration () Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

47 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Chemical: Boron, Well: BAC 04 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (0.177) Concentration () Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul 4 Chemical: Boron, Well: BAC 05 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (0.177) Concentration () Stats Stable, No Trend 0 N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

48 150 Chemical: Calcium, Well: BAC 02 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (129) Concentration () Stats Increasing Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul Chemical: Chloride, Well: BAC 02 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (24.7) Concentration () Stats Increasing Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

49 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Chemical: Chloride, Well: BAC 03 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (24.7) Concentration () Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul Chemical: Chloride, Well: BAC 04 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (24.7) Concentration () Stats Stable, No Trend N Total: 7 N Detect: 7 % Detect: 100 Oct 2016 Jan Apr Jul

50 0.20 Chemical: Fluoride, Well: BAC 05 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (0.161) Concentration () Stats Increasing Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul 6 Chemical: ph, Well: BAC 02 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit LPL (6.63) UPL (7.22) Concentration (SU) Stats Decreasing Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

51 6 Chemical: ph, Well: BAC 03 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit LPL (6.63) UPL (7.22) Concentration (SU) Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul 6 Chemical: ph, Well: BAC 04 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit LPL (6.63) UPL (7.22) Concentration (SU) Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

52 6 Chemical: ph, Well: BAC 05 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit LPL (6.63) UPL (7.22) Concentration (SU) Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul 400 Chemical: Sulfate, Well: BAC 02 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (140) Concentration () Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

53 Chemical: Sulfate, Well: BAC 03 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (140) Concentration () Stats Decreasing Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul 200 Chemical: Sulfate, Well: BAC 04 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (140) Concentration () Stats Stable, No Trend N Total: 7 N Detect: 7 % Detect: 100 Oct 2016 Jan Apr Jul

54 Concentration () Chemical: Sulfate, Well: BAC 05 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (140) 50 0 Stats Stable, No Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul Chemical: TDS, Well: BAC 02 Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (505) Concentration () Stats Increasing Trend N Total: 8 N Detect: 8 % Detect: 100 Oct 2016 Jan Apr Jul

55 Chemical: TDS, Well: BAC 04 Unit: Bottom Ash Complex Figure A-4: Trend Analysis of Downgradient Wells with Exceedances Geology: Alluvium Symbols Detect Detect Limit Theil Sens Fit UPL (505) Concentration () Stats Stable, No Trend N Total: 7 N Detect: 7 % Detect: 100 Oct 2016 Jan Apr Jul

56 Appendix B Analytical Summary

57 Appendix B Analytical Data Summary Gavin Power, LLC Bottom Ash Complex FEDERAL FEDERAL FEDERAL FEDERAL FEDERAL FEDERAL Sample Date 8/25/2016 8/25/2016 8/26/2016 8/26/2016 8/26/2016 8/26/2016 Sample Type N N N N N N Location ID BAC-02 MW-1 BAC-01 BAC-03 BAC-04 BAC-05 Sample ID BAC MW BAC BAC BAC BAC Analyte Unit Antimony 6E-05 2E-05 2E-05 5E-05 9E Arsenic Barium Beryllium 3.5E-05 2E-05 1E-05 1E-05 2E Boron Cadmium E-05 2E Calcium Chloride Chromium Cobalt Combined Radium pci/l Fluoride Lead Lithium Mercury 3E-06 2E-06 2E-06 2E-06 2E-06 3E-06 Molybdenum ph, Field ph units ph, Field SU Selenium E Sulfate Thallium E-05 1E-05 3E E E-05 Total dissolved solids Notes: FD = Field Duplicate Sample N = Normal Sample If a sample was analyzed for mercury by both Method 7470 and low-level Method 1631, the low-level Method 1631 result is shown. ERM Page 1 of 15 GAVIN/