Preparing a Groundwater Sampling Analysis Plan for a

  • Slides: 39
Download presentation
Preparing a Groundwater Sampling & Analysis Plan for a Type I Landfill Overview of

Preparing a Groundwater Sampling & Analysis Plan for a Type I Landfill Overview of the groundwater sampling protocol, constituents measurements, data evaluation, and reporting requirements.

Courtesy of Phys. org

Courtesy of Phys. org

(Preferred Method) (New addition)

(Preferred Method) (New addition)

Make sure to specify how purge water and excess sample water will be managed

Make sure to specify how purge water and excess sample water will be managed and disposed of.

Purge/ Sampling Water Disposal

Purge/ Sampling Water Disposal

Document, Document… Create a form for field personnel to record: (include a copy in

Document, Document… Create a form for field personnel to record: (include a copy in the GWSAP).

Field Quality Assurance (QA)/ Quality Control (QC) Samples

Field Quality Assurance (QA)/ Quality Control (QC) Samples

SPECIFY: SHOULD NOT COLD

SPECIFY: SHOULD NOT COLD

STATISTICS!!

STATISTICS!!

Statistical evaluation of monitoring data

Statistical evaluation of monitoring data

Edward R. Tufte W. Edwards Deming

Edward R. Tufte W. Edwards Deming

Background Wells GW flow Landfill Compliance Wells

Background Wells GW flow Landfill Compliance Wells

 • Collecting Data • Normality of Data Step 1 Step 2 • Outliers

• Collecting Data • Normality of Data Step 1 Step 2 • Outliers • Trends • Spatial Variation • Establishing Background • Updating Background Step 3

Limit - 380 Limit - 36

Limit - 380 Limit - 36

Prediction Limits (with Retesting) Control Chart Limits (with Retesting) Trend Test (Contaminated BG, Unstable

Prediction Limits (with Retesting) Control Chart Limits (with Retesting) Trend Test (Contaminated BG, Unstable Conditions) Confidence intervals (Assessment, Corrective Action) ANOVA for diagnostics (Identify Spatial Variation)

spatial variation lack of significant ANOVA

spatial variation lack of significant ANOVA

 • Parametric: • Non-parametric:

• Parametric: • Non-parametric:

Parametric Non-Parametric

Parametric Non-Parametric

should be normalized

should be normalized

Mann-Kendall • Key: trends evident in BG data Theil-Sen

Mann-Kendall • Key: trends evident in BG data Theil-Sen

Upper Limit Lower Limit • At least 4 values

Upper Limit Lower Limit • At least 4 values

Confidence Interval Trichloroethylene Data Set Limit = 5 273

Confidence Interval Trichloroethylene Data Set Limit = 5 273

 • Detection • Control Charts? Good choice • • • Combining Shewhart &

• Detection • Control Charts? Good choice • • • Combining Shewhart & CUSUM portions Prediction Limits? As Good Trend Tests? Best choice for unstable/trending data • • Prediction limits, control charts assume stable data Significant, upward slope implies change in GW quality • Assessment & Corrective Action • Confidence Intervals? The only choice

Power Curves Site-Wide False Positive Rates

Power Curves Site-Wide False Positive Rates

55% power at 3 80 -85% power at 4

55% power at 3 80 -85% power at 4

55%

55%

Sample size n=4 Sample size n=8

Sample size n=4 Sample size n=8

site-wide false positive rate

site-wide false positive rate

Double Quantification Rule (DQR) Not Historically Detected

Double Quantification Rule (DQR) Not Historically Detected

 • Any retesting scheme better than none

• Any retesting scheme better than none

Contact Information Texas Commission on Environmental Quality P. O. Box 13087, Austin, Texas 78711

Contact Information Texas Commission on Environmental Quality P. O. Box 13087, Austin, Texas 78711 MSW Permits Section Iryna Kushnirsky iryna. kushnirsky@tceq. texas. gov 512 -239 -1471 Eric Clegg, P. G. eric. clegg@tceq. texas. gov 512 -239 -1270