Inter-Well Comparisons
Inter-Well Comparisons
See Also:
With inter-well comparisons, data from the compliance well are compared to data pooled from one or more background wells. The background wells are required to be hydrogeologically up-gradient of the facility and should not be impacted by the facility or any other source of contamination.
One advantage of inter-well comparisons is that as few as four background samples are sufficient for most tests. Unlike intra-well comparisons, inter-well comparisons can evaluate a large number of compliance wells simultaneously. In some statistical methods, all compliance wells can be evaluated with a single test, although the test may not be appropriate for all situations.
The primary disadvantage of inter-well comparisons is that they are susceptible to spatial variability of natural ground water quality conditions, which can be significant. Spatial variability can cause compliance wells to show statistical significance that is not a result of facility activities, or it can mask existing contamination. Additionally, it is not unusual to have contamination in background wells from an adjacent facility, or as a result of fluctuating hydrogeologic gradients.
Compliance Well to Background Statistic
With compliance well to background statistic comparisons, the total set of samples from a compliance well is compared to a statistic calculated from the pooled background wells. The analysis method can indicate if the compliance well is contaminated, but can not discern which specific samples in the wells caused the statistically significant event.
- Parametric Analysis of Variance
- Kruskal-Wallis Non-Parametric Rank Analysis
- Wilcoxon Rank-Sum Non-Parametric Analysis
- Poisson Prediction Intervals
- Parametric t-test Comparison
- Quantile Test
Compliance Well Sample to Background Statistic
With compliance well sample to background statistic comparisons, each individual sample from each compliance well is compared to a statistic calculated from the pooled background wells. Such a method can determine which specific samples are statistically elevated with respect to background.
- Parametric Prediction Limits
- Parametric Prediction Limits with Resampling
- Non-Parametric Prediction Limits
- Parametric Tolerance Limits
- Poisson Tolerance Limits
- Non-Parametric Tolerance Limits
- Shewhart-CUSUM Control Chart (1989 Guidance)
- Shewhart-CUSUM Control Charts (Unified Guidance)
- EWMA Control Chart
Compliance Group to Background Group Comparisons
With compliance to background group comparisons, all compliance sampling points are compared as a group to all background sampling points. The test determines if compliance points as a group have elevated concentrations when compared to background. These tests are most appropriate for soil samples, as they can not determine which individual sampling points are elevated. These tests include: