2025-03-11-minutes-solid-waste-and-maintenance

Home Browse Search

Ryegrass Limited Purpose Landfill and Baletill Closed Landfill Groundwater and Leachate Sampling and Analysis Plan Kittitas county Solid Waste T)Rt . Evall-lation 3.1 Statistical Analysis Ecology guidance (2028) recommends using the U.S. Environmental Protection Agency's (ETA's) Statistical Analysis of Groundwater Monitoring Data at RCRA Facilities - Unified Guidance (Unified Guidance) (EPA 2009) for guidance on statistical analysis of landfill groundwater monitoring data. 3.1.1 Approach Overview The statistical approach for evaluating data at the LPL and BalefilI Landfill will consist of establishing a subset of parameters for formal detection monitoring and conducting quarterly comparisons of data to intrawell upper prediction limits (UPLs) calculated for each well/parameter case. Control charts will be used to supplement the UPLs to evaluate whether any trends are occurring that could potentially be attributable to the landfill. Annually, the data will be evaluated for trends using the Mann-Kendall/Sen's Slope test, and the results will be used to support periodic updates of the control chart and prediction limit background data sets. The initial baseline data will consist of all historical data for existing wells, after conducting trend evaluation to confirm there are no upward trends. For new wells without historical data, the baseline data wili be established with the initial eight data points. The Unified Guidance provides a suggested rule for evaluating constituents that are "never - detected," including the VOCs. Any constituent that has never been previously detected should be evaluated by the following simple, quasi -statistical rule. A confirmed exceedance is registered if any well -constituent pair in the '100% non -detect' group exhibits quantified measurements (i.e., at or above the reporting limit [RQ) in two consecutive sample and resample events (EPA 2009). 3.i-MI !Upper Prediction Limits The statistical approach will use UPLs with retesting strategy for indicator parameters selected for formal detection monitoring. UPLs will be established from the distribution of the background data at each well, and future values will be evaluated against those UPLs using a 1-of-m scheme. As long as the compliance point measurements are similar to background, the prediction limit should contain all m of the future values or statistics with high probability (the level of confidence). For a 1-of-m test, all m values must be larger than the prediction limit to be declared a statistically significant increase (S%, as initial evidence that compliance point concentrations are higher than background. A 1-of-2 retesting scheme (EPA 2009) will be used to compare the new sample value(s) to the UPL. The I-of-2 scheme assumes that two samples will be collected for a particular constituent at a given well, including the Initial groundwater sample and one resample. The initial groundwater sample will be collected in the first and thlyd quarters, and the resamples will be collected in the second and fourth quarters. Statistical comparisons will be made in the first and third quarters and, if necessary, in the second and fourth quarters for retests. However, because groundwater is sampled quarterly, any measured concentrations above the UPL during the second and fourth quarters will also be confirmed or disconfirmed using samples collected in the third and first quarters, respectively. The Unified Guidance recommends that, in order to balance the site -wide false positive rate and the power of the statistical program, a subset of monitoring parameters should be selected for statistical comparisons. The statistically tested monitoring parameters should be limited to a few April 2024 1 553-1.763-010 3.1 R-9vised May 2024