Study Question 1 Study Question 3 Study Question 4 Study Question 5 Study Question 6 Study Question 7 Study Question 8 Study Question 9 Study Question 10 Study Question 2

4. Statistical Analysis for Project Life Cycle Stages

This section explains the role of groundwater statistics in the activities of typical project life cycle stages. These project life cycle stages include release detection, site characterization, remediation, monitoring, and closure. The environmental projects may be cleanup projects, or compliance monitoring (for example Resource Conservation and Recovery Act (RCRA) facilities) projects. Study questions serve as a bridge connecting project life cycle stages with relevant statistical methods and were selected based on common project objectives that require statistical analyses. Ten common study questions and their associations with each of the life cycle stages are presented in Table 4-1. A more detailed discussion of the study questions is presented in Appendix C.

Table 4-1. Statistical Study Questions for life cycle stages

4.1 Considerations for Statistical Analysis

Exploratory data analysis (EDA) is a common step for all project life cycle stages or study questions. As discussed in Section 3.3.3, EDA is valuable for inspecting the quality and character of groundwater sample results. Data should be evaluated for frequency of detection, multiple detection limits, and outliersValues unusually discrepant from the rest of a series of observations (Unified Guidance). (either high or low values). Check the distributions of the sample data for normality or lognormality to select appropriate statistical methods. In general, do not make unverified assumptions regarding the statistical distribution of data. When the impact of nondetectsLaboratory analytical result known only to be below the method detection limit (MDL), or reporting limit (RL); see "censored data" (Unified Guidance). on the data set is combined with the impact of a small data set or data sets containing extreme values, distributional tests may fail to identify a known distribution and nonparametricStatistical test that does not depend on knowledge of the distribution of the sampled population (Unified Guidance). methods must be used.

4.2 Release Detection

At cleanup sites or at waste management facilities, groundwater monitoring may be used to determine if a release has occurred. A release may be detected by comparison of compliance well data to a criterionGeneral term used in this document to identify a groundwater concentration that is relevant to a project; used instead of designations such as Groundwater Protection Standard, clean-up standard, or clean-up level. or by detection of a trend in the compliance well data. Groundwater concentrations may be compared to a criterion to determine facility compliance. A release may also be detected when the concentration of a chemical in groundwater exceeds background. Accordingly, an important aspect of release detection monitoring is determining the background concentrations (either natural or anthropogenic) for chemicals. Natural background would be representative of pristine or pre-industrial conditions. Anthropogenic background refers to concentrations in the surrounding groundwater that may be impacted by human activity, but not by the site.

The study questions that are most applicable for release detection are shown here.

4.2.1 Background Conditions

4.2.2 Location or Selection of Background Wells

4.2.3 Monitoring for a Release

4.2.4 Statistical Methods for Release Detection Objectives

4.3 Site Characterization

Site characterization is typically the first phase of a cleanup project. Site characterization describes the physical conditions of the site such as soils, geology, hydrology, the presence of existing contamination, the potential for contamination to be released, and the actual and potential pathways and mechanisms for contamination transport. This stage of the project life cycle considers the chemical characteristics of the contaminants and their potential to be mobile in the environment. All of these aspects of site characterization are needed to develop an appropriate groundwater monitoring program, understand groundwater contaminant concentrations, and select and interpret groundwater statistical analyses.

Information collected during the characterization phase may support, refute, or provide additional details for the initial assumptions regarding the site. While some information may not be initially known, information must be collected to support the CSMconceptual site model. Development of the CSM starts at the beginning of the cleanup project and continues as additional information becomes available (see Section 3.2).

To ensure that the data collected will support the project goals, use a structured process such as the Triad approach or USEPA Data Quality Objectives (DQO) process to guide data acquisition. Developing, implementing, and optimizing groundwater monitoring all depend upon data of known quality and sufficient quantity. Section 3.3.1 includes more information about data quality.

The study questions that are most applicable for site characterization are shown here.

4.3.1 Physical Site Conditions

4.3.2 Existing Contamination and its Sources

4.3.3 Pathways and Mechanisms of Transport

4.3.4 Statistical Methods for Site Characterization Objectives

4.4 Remediation

Remediation of contaminated groundwater is a challenge. On average, attainment of regulatory closure at sites with contaminated groundwater takes significantly longer compared to sites that have contaminated soil but no groundwater impacts. At many sites, it is estimated that attainment of groundwater standards will take decades or more (See NAP 2012 for a discussion of the challenges of managing complex contaminated groundwater sites).

This section provides guidance on the use of statistics to support remedy selection and evaluation of remedy effectiveness. For remediation, statistical analyses are most useful for evaluating changes in concentration over time (trend analyses). An objective and accurate evaluation of changes in contaminant concentrations over time can help to resolve groundwater remediation issues.

The study questions that are most applicable for remediation are shown here.

4.4.1 Remedy Selection

4.4.2 Remedy Effectiveness

4.5 Monitoring

Groundwater monitoring is conducted to observe and assess characteristics of interest at cleanup, RCRAResource Conservation and Recovery Act facility, or waste disposal sites. Often, monitoring is conducted on a long-term basis, sometimes for decades. Monitoring may be required even after closure of a site during post-closure monitoring. As such, monitoring may be conducted to describe characteristics at a specific location or point in time or to show how these characteristics change over time or space.

Changes in groundwater quality may have either natural or human causes. Proper evaluation of groundwater data helps you understand whether the criteria or goals of the monitoring program are met or if significant, adverse changes in groundwater concentrations have occurred. Proper design of the monitoring network depends upon the type of the site, the contaminants present, and the regulatory program. Prior to implementing a monitoring program, review well placement, parameter selection, sampling frequency, and whether or not a release has been identified. Typical activities in the monitoring stage include observing changes in concentration levels over time and space, and comparing concentrations to numerical criteria. After sufficient data are collected, it may also be possible to optimize sampling locations and sampling frequencies to improve and streamline the monitoring program (see Section 4.5.3). 

The study questions that are most applicable for monitoring are shown here.

4.5.1 Monitoring for Concentration Changes

4.5.2 Compliance with Criteria

4.5.3 Optimization of Long-term Monitoring Networks

4.5.4 Statistical Methods for Monitoring Objectives

4.6 Closure

Closure is the final stage of the project life cycle and therefore is subject to extra scrutiny. At this stage in the process, data planning and collection should have been managed through a systematic planning process, and the CSMconceptual site model is assumed to be complete for the purposes of making a final determination on whether monitoring may be permanently discontinued and the site closed. This decision point may be reached at any time during the life cycle process (for example, during site characterization, remediation, or monitoring). Significant variation occurs across regulatory programs, but in general, when contaminants are no longer detected in any wells for several sampling events or over a specified period of time, the remedial goals are deemed complete and the groundwater is no longer considered contaminated. However, in instances where contaminants decrease but remain measurably present, or are present in natural or anthropogenic background, statistics can support a closure decision.

Given the importance of this decision, managers must have a high degree of confidence that the data fully support closure. Closure should verify that site contaminants are no longer present in the groundwater, or are not present at concentrations that pose an unacceptable risk to human health or the environment. In cases where concentrations of contaminants are allowed to remain (such as under institutional or engineering control scenarios), trend analysis results may be used to show that contaminants will not migrate or increase concentrations outside of the defined boundaries of the controlled area.

Following remediation, formal statistical testing will usually involve an upper confidence limit around the mean or an upper percentile compared against a criterion. The overriding concern in corrective action is that remediation efforts must have sufficient statistical proof to be declared successful. Since groundwater is now presumed to be contaminated, a facility should not exit corrective action until there is sufficient evidence that contamination has been abated (see Chapter 7.2, Unified Guidance).

By the time a site reaches the closure life cycle stage, the evaluation assumes that contamination exists. Therefore, the statistical approach may involve comparing an upper confidence limit of the data to a criterion. The upper confidence limit (UCL)The upper value on a range of values around the statistic (for example, mean) where the population statistic (for example, mean) is expected to be located with a given level of certainty, such as 95% ( 2013). should lie below the criterion to accept the hypothesis that concentration levels support closure. If the entire confidence interval (considering both the lower and upper confidence limits) lies below the criterion, there is statistically significant evidence that the true value of the parameter (for instance, the mean) is less than the criterion. When the confidence interval straddles the criterion, the correct decision is uncertain within the stated confidence levelDegree of confidence associated with a statistical estimate or test, denoted as (1 – alpha) (Unified Guidance).. The true value of the parameter might be less than or greater than the criterion and no clear decision with high statistical confidence is possible (see Chapter 5, Unified Guidance).

The study questions that are most applicable for closure are shown here.

4.6.1 Compliance with Criteria

4.6.2 Trends Toward Compliance Criteria

4.6.3 Statistical Methods for Closure Objectives

Publication Date: December 2013

Permission is granted to refer to or quote from this publication with the customary acknowledgment of the source (see suggested citation and disclaimer).


This web site is owned by ITRC.

1250 H Street, NW • Suite 850 • Washington, DC 20005

(202) 266-4933 • Email: [email protected]

Terms of Service, Privacy Policy, and Usage Policy


ITRC is sponsored by the Environmental Council of the States.