Parametric Prediction Limit
Parametric Prediction Limit
See Also:
Non-Parametric Prediction Limit
Parametric Prediction Limit with Resampling
Inter-Well Comparison:
Description:
The inter-well parametric prediction limit is used to compare samples from background wells to a selected number of recent sampling dates for from compliance wells. If there are more than one sample per date, the samples for that date are averaged, and the average is compared to the prediction limit.
Verification Resampling
The default number of recent sampling dates is specified from Options | Default Options | Comparison or the prediction limit right-click menu. In most cases, a value of 1 is appropriate. However, you can specify different “recent date” values for each individual well. When the parametric prediction interval report is displayed, the bottom portion of the window will display a list of all compliance wells included in the statistical analysis. Select a well or wells. Click the right mouse button to display the context menu, and select the desired number of recent dates for the selected well(s).
Use:
As an inter-well comparison, the parametric prediction limit is useful for comparing background measurements to compliance wells in situations where there are multiple samples from compliance wells for a single sampling event. The method also uses only recent samples from compliance wells, and is therefore useful for compliance wells which may have been contaminated in the past and an assessment of current conditions without regard to past contaminant levels is required.
A major advantage of the parametric prediction limit is that only one sample from any compliance well is required. This enables it to be used on newly installed wells to asses contamination.
The prediction limit analysis is performed for all compliance wells for the specified parameter.
The parametric prediction limit requires that data be normally distributed.
Implementation:
ChemStat provides six algorithms for calculating the parametric prediction limit.
- USEPA 1992 Guidance 99% Confidence
- USEPA 1992 Guidance 95% Confidence
- USEPA Draft Unified Guidance (2004) 99% Confidence One-Sided
- USEPA Draft Unified Guidance (2004) 95% Confidence One-Sided
- USEPA Draft Unified Guidance (2004) 99% Confidence Two-Sided
- USEPA Draft Unified Guidance (2004) 95% Confidence Two-Sided
Select the desired calculation method from the right-click menu accessed over the prediction limit report.
Remarks:
There is a considerable difference in the implementation of this method as provided in the USEPA guidance documents. Examples of the USEPA implementations of this method are provided in both the 1989 and 1992 guidance documents and in the Draft Unified Guidance document.
Intra-Well Comparison:
Description:
The intra-well parametric prediction limit is used to compare the samples from a selected number of recent sampling dates from a well to historical samples from that same well.
If there is more than one measurement per date, the measurements for that date are averaged. The number of recent sampling dates, and the number of baseline samples for comparison are specified from Select | Recent Dates or the prediction limit right-click menu. This displays the box shown right. In most cases, a value of 1 is appropriate for recent dates. At least 5 baseline samples known to not be impacted by the facility, should be available for the selected well.
Use:
As an intra-well comparison, the parametric prediction limit can be used to test for contamination in a compliance well without the risk of false positives due to natural variability of ground water. It is important that the baseline samples used to calculate the prediction limit have not be impacted by the facility. It is preferable that baseline samples be only those samples collected before waste placement at the facility.
The prediction limit analysis maintains a minimum error rate of 5% for the well, and a minimum error rate of 1% for each comparison. The parametric prediction limit requires that data be normally distributed.
Implementation:
ChemStat provides six algorithms for calculating the parametric prediction limit.
- USEPA 1992 Guidance 99% Confidence
- USEPA 1992 Guidance 95% Confidence
- USEPA Draft Unified Guidance (2004) 99% Confidence One-Sided
- USEPA Draft Unified Guidance (2004) 95% Confidence One-Sided
- USEPA Draft Unified Guidance (2004) 99% Confidence Two-Sided
- USEPA Draft Unified Guidance (2004) 95% Confidence Two-Sided
Select the desired calculation method from the right-click menu accessed over the prediction limit report.
Remarks:
Intra-well comparisons with the prediction limit are not documented in the USEPA 1989 and 1992 statistical guidance documents.